# Research Inferences # Inference Statements for Wiki Insertion — Set 01 (INF-0001 – INF-0100) **Purpose.** Five hundred insertable statements that carry the forward argument: each anchors on a named, dated, verifiable object and extends it toward the continuity, uploading, succession, and in-body regulation trajectories. These are written to be dropped into `## Forward Capability`, `## Adjacency`, or `## Open Questions` sections of the entries named in each `Insert` field. **Grammar.** `ID · Insert: target entries · Tier: established | strongly indicated | plausible | unresolved` followed by the statement. Tier applies to the inference, not to the anchor; anchors are established unless marked. Every statement names an actor, a mechanism, and where possible the governing variable — the quantity that must move for the next threshold to be crossed. Statements are written affirmatively: they state what the trajectory supports, not what it fails to prove. --- ## Cluster A — Electrode scaling and clinical bandwidth **INF-0001** · *Insert:* [[wiki/Paradromics]], [[wiki/Connexus]] · *Tier:* strongly indicated — Connexus placed 421 single-neuron channels into a human on 17 June 2026 with a chest transceiver as the egress path, which makes the implant a *networking device* with a neural front end rather than a sensor. Every subsequent capability question resolves to channels × uptime × egress bandwidth, and the chest transceiver is the component that generalizes fastest. **INF-0002** · *Insert:* [[wiki/Paradromics]], [[wiki/Speech Neuroprosthesis]] · *Tier:* plausible — Paradromics frames Connexus against transcribed human speech at roughly 40 bps, a comparison that sets the commercial floor at parity with talking. The interesting regime opens above it: once decoded output exceeds speech rate, the interface stops substituting for a lost channel and becomes a channel humans never had. **INF-0003** · *Insert:* [[wiki/Utah Array]], [[wiki/Blackrock Neurotech]] · *Tier:* established — The Utah Array's multi-decade chronic dwell in human cortex, including a single participant past nine years, is the only long-duration dataset the field owns. Every claim about lifetime continuity hardware is currently underwritten by that one electrode geometry, which makes array longevity the most load-carrying unresolved variable in the stack. **INF-0004** · *Insert:* [[wiki/Neuralink]], [[wiki/Surgical Robot]] · *Tier:* strongly indicated — Neuralink's robot exists because thread insertion at scale is a throughput problem, not a dexterity problem. The moment implant counts are gated by operating-room hours rather than by device cost, the surgical robot becomes the rate-limiting industrial asset in the entire interface economy. **INF-0005** · *Insert:* [[wiki/Precision Neuroscience]], [[wiki/Cortical Surface Array]] · *Tier:* plausible — Thin-film surface arrays trade single-unit resolution for reversibility and coverage area. Reversibility is the property that makes iteration possible: a device that can be removed can be upgraded, and upgradeability is what turns an implant from a terminal procedure into a subscription substrate. **INF-0006** · *Insert:* [[wiki/Channel Count]], [[wiki/Bandwidth]] · *Tier:* analytic — Channel count has followed a doubling cadence closer to instrumentation than to semiconductors, because each channel carries a biological interface cost that Moore's law does not amortize. The break in that curve will come from multiplexing at the tissue boundary, not from finer lithography. **INF-0007** · *Insert:* [[wiki/Speech Neuroprosthesis]], [[wiki/Decoder]] · *Tier:* strongly indicated — Academic speech BCIs reached 62–78 words per minute with implanted electrodes and model-side decoding, which means the recent gains are attributable to the decoder more than the sensor. Decoder improvement is free of surgical risk and compounds across every existing implant already in a head. **INF-0008** · *Insert:* [[wiki/Chronic Signal Degradation]], [[wiki/Gliosis]] · *Tier:* established — Foreign-body response sets the practical dwell time of intracortical arrays, and every continuity architecture that assumes decades of recording is really assuming a materials-science result that has not yet arrived. Coatings, compliance matching, and thread geometry are therefore continuity research, not device engineering. **INF-0009** · *Insert:* [[wiki/Connexus]], [[wiki/Neurodata Without Borders]] · *Tier:* plausible — Three clinical sites recording the same device format produces the first multi-site, single-format chronic human neural corpus. If that corpus is written in NWB and deposited in DANDI, the clinical program becomes a foundation-model dataset as a byproduct of its regulatory obligations. **INF-0010** · *Insert:* [[wiki/FDA Breakthrough Device]], [[wiki/Regulatory Pathway]] · *Tier:* strongly indicated — Breakthrough Device designation functions as a scheduling instrument: it does not lower the evidentiary bar, it reorders the queue. In a field where first classification sets the template for everything behind it, queue position converts directly into standards authorship. **INF-0011** · *Insert:* [[wiki/Matt Angle]], [[wiki/Stephen Ryu]] · *Tier:* analytic — A platform company that recruits its chief medical officer from academic neurosurgery is importing trial design and site relationships, not just clinical credibility. The Stanford-to-Austin transfer is the mechanism by which twenty years of publicly funded intracortical research becomes a private clinical program. **INF-0012** · *Insert:* [[wiki/Implant Telemetry]], [[wiki/Body Area Network]] · *Tier:* plausible — Once the transceiver sits in the chest rather than the skull, the implant joins a body-area network with its own addressing, power budget, and failure modes. The body becomes a small datacenter with a thermal envelope, and everything learned about rack telemetry applies with the units changed. **INF-0013** · *Insert:* [[wiki/Neural Dust]], [[wiki/Ultrasonic Backscatter]] · *Tier:* plausible — Ultrasonic backscatter motes decouple channel count from wire count, which is the structural limit on every tethered array. A population of untethered sensors addressed acoustically scales with acoustic focusing precision rather than with surgical access, and that is a different curve entirely. **INF-0014** · *Insert:* [[wiki/Motor Cortex]], [[wiki/Access Geometry]] · *Tier:* analytic — Every clinical BCI so far has been placed where the signal is easiest to reach, not where the representation is richest. Access geometry, not neuroscience, has determined which parts of the mind are currently legible, and improving access will reveal representational territory that has never been recorded. **INF-0015** · *Insert:* [[wiki/Bidirectional Interface]], [[wiki/Write Bandwidth]] · *Tier:* strongly indicated — Read bandwidth has advanced far ahead of write bandwidth, and continuity requires both. A system that can record a state but not install one supports observation, not transfer, which makes patterned stimulation the decisive unsolved half of the uploading stack. **INF-0016** · *Insert:* [[wiki/Proprioception]], [[wiki/Closed-Loop Feedback]] · *Tier:* plausible — Prosthetic control without sensory return plateaus quickly because the loop is open. Writing proprioceptive feedback into somatosensory cortex closes it, and the closed loop is where the boundary between the device and the body's own model of itself starts to dissolve. **INF-0017** · *Insert:* [[wiki/Blindsight]], [[wiki/Visual Prosthesis]] · *Tier:* plausible — A cortical visual prosthesis is a write-side experiment wearing a clinical indication: it must install a percept rather than read one. Whatever resolution it achieves is a direct measurement of how much experience can be authored from outside, which is the core uploading question stated in an approvable form. **INF-0018** · *Insert:* [[wiki/Implant Power]], [[wiki/Wireless Power Transfer]] · *Tier:* analytic — Power, not data, is the first hard ceiling on implant scaling, because heat dissipation in tissue is bounded at roughly a degree of local rise. Every channel-count roadmap is implicitly a thermal roadmap, and low-power neuromorphic front ends are the most direct route past it. **INF-0019** · *Insert:* [[wiki/Edge Inference]], [[wiki/On-Implant Compute]] · *Tier:* strongly indicated — Moving the decoder onto the implant converts a bandwidth problem into a compute problem, and compute miniaturizes faster than tissue-boundary bandwidth improves. The implant that ships its own model, not its own spikes, is the architecture that scales. **INF-0020** · *Insert:* [[wiki/Model Drift]], [[wiki/Recalibration]] · *Tier:* established — Decoders drift as the recording population changes, so every chronic BCI is already running a continuous identity-maintenance process on a moving substrate. That maintenance loop is a working prototype of the problem any uploaded state would face, running today in clinic. **INF-0021** · *Insert:* [[wiki/Connect-One Study]], [[wiki/Early Feasibility Study]] · *Tier:* analytic — Two-participant early feasibility studies generate no statistical power and enormous engineering information. Their real output is the failure catalogue, and the company that accumulates the deepest failure catalogue across chronic human implants owns the field's actual intellectual property. **INF-0022** · *Insert:* [[wiki/ALS]], [[wiki/Locked-In Syndrome]] · *Tier:* strongly indicated — Locked-in patients are the population for whom the interface is the only channel, which makes them the group that will use the highest-bandwidth configurations first and longest. Clinical necessity, not enthusiasm, will produce the first humans who have lived years mediated by a decoder. **INF-0023** · *Insert:* [[wiki/Neural Latency]], [[wiki/Real-Time Decoding]] · *Tier:* plausible — Below roughly a hundred milliseconds, a decoded action stops feeling like operating a tool and starts feeling like moving. Sub-perceptual latency is the threshold at which the interface is incorporated into the body schema, and body-schema incorporation is the psychological precondition for substrate transfer. **INF-0024** · *Insert:* [[wiki/Implant Explantation]], [[wiki/Device Lifecycle]] · *Tier:* unresolved — No regulatory framework specifies what happens to the recorded state, the trained decoder, or the personalized model when a device is explanted or its manufacturer fails. That vacuum is the first real custody question in continuity engineering, and it is live now with devices already in heads. **INF-0025** · *Insert:* [[wiki/Morgan Stanley BCI Market Estimate]], [[wiki/Market Formation]] · *Tier:* analytic — A $400 billion U.S. addressable-market estimate against a sector raising under a billion a quarter describes an enormous ratio of anticipated value to deployed capital. Ratios that wide are resolved by capital arriving quickly, and the timing of that arrival is a policy variable more than a technology one. --- ## Cluster B — Field-addressed and molecular interfaces **INF-0026** · *Insert:* [[wiki/Merge Labs]], [[wiki/Sonogenetics]] · *Tier:* strongly indicated — Merge Labs joins ultrasound hardware from Forest Neurotech to Mikhail Shapiro's acoustic-biomolecular line at Caltech, which is the first time the field half and the receptor half of a field-addressed interface have been capitalized inside one company. The composite — sensitize, then address — is what removes surgery from the bandwidth equation. **INF-0027** · *Insert:* [[wiki/Sonogenetics]], [[wiki/Mechanosensitive Channel]] · *Tier:* established — Sonogenetics works by installing mechanosensitive channels so that acoustic pressure becomes an ionic event in a chosen cell population. Selectivity is therefore a genetic property and focality an acoustic one, and the two can be improved independently, which is why this architecture scales differently from electrodes. **INF-0028** · *Insert:* [[wiki/Magnetoelectric Nanoparticle (MENP)]], [[wiki/Sakhrat Khizroev]] · *Tier:* strongly indicated — A magnetostrictive core inside a piezoelectric shell converts a magnetic field that passes the skull unattenuated into a local electric field that neurons answer. The particle is a field translator, and its existence means depth access does not require a hole — it requires distribution. **INF-0029** · *Insert:* [[wiki/Magnetogenetics]], [[wiki/Optogenetics]] · *Tier:* analytic — Optogenetics gives cell-type specificity but is bounded by the optical scattering length of tissue; magnetic and acoustic modalities give depth but coarser selectivity. Every serious non-surgical interface program is now attempting the same synthesis of the two, and whichever pairing lands first sets the modality standard for decades. **INF-0030** · *Insert:* [[wiki/MOANA]], [[wiki/DARPA N3]] · *Tier:* established — DARPA's N3 program treated magnetic, optical, and acoustic access as one design space and funded them in parallel rather than picking a winner. That structure is the clearest evidence that the state's objective was modality-independent non-surgical access, and the program's outputs are now distributed across multiple private companies. **INF-0031** · *Insert:* [[wiki/AAV Delivery]], [[wiki/Neural Interface]] · *Tier:* plausible — Every molecular interface requires a delivery vector, which makes AAV capsid engineering an interface technology rather than a gene-therapy technology. Capsid tropism determines which cells can be addressed, so the vector library is functionally the addressing scheme of the future nervous-system network. **INF-0032** · *Insert:* [[wiki/Blood-Brain Barrier]], [[wiki/Focused Ultrasound]] · *Tier:* established — Focused ultrasound with microbubbles opens the blood-brain barrier transiently and reversibly at a chosen location, which makes systemic delivery locally targetable. That single capability converts the vasculature into a delivery bus addressed by an external transducer. **INF-0033** · *Insert:* [[wiki/Functional Ultrasound]], [[wiki/Hemodynamic Proxy]] · *Tier:* analytic — Functional ultrasound reads blood-flow dynamics, a proxy signal with seconds-scale physics, and any interface built on it inherits that time constant. The forward path is not faster hemodynamics but hybrid operation — acoustic read of slow state, molecular read of fast events. **INF-0034** · *Insert:* [[wiki/Gene Therapy]], [[wiki/Interface Sensitization]] · *Tier:* plausible — A one-time sensitizing edit followed by a lifetime of external field addressing inverts the economics of neural interfacing: the expensive, risky step happens once and the hardware upgrades stay outside the body. This is the configuration most likely to reach populations rather than patients. **INF-0035** · *Insert:* [[wiki/Terahertz Communication]], [[wiki/Internet of Bio-Nano Things]] · *Tier:* plausible — Nanoscale devices inside tissue need a communication layer, and molecular and terahertz channels are the two candidates with the right length scales. The body-internal network is being specified now in exactly the way IP was specified before there was traffic to carry. **INF-0036** · *Insert:* [[wiki/Optogenetics]], [[wiki/Clinical Translation]] · *Tier:* established — Optogenetic vision restoration has already produced partial clinical results in humans, which means installing a light-responsive receptor in a human nervous system is a completed procedure rather than a proposal. Every write-side architecture inherits that precedent. **INF-0037** · *Insert:* [[wiki/Calcium Imaging]], [[wiki/Voltage Imaging]] · *Tier:* strongly indicated — Voltage indicators read the electrical event directly rather than its calcium shadow, trading signal-to-noise for temporal fidelity. As indicator brightness improves, optical readout crosses into spike-timing resolution and becomes a genuine alternative to electrodes at scale. **INF-0038** · *Insert:* [[wiki/Acoustoelectric Neural Interface]], [[wiki/ARIA]] · *Tier:* analytic — Acoustoelectric sensing uses ultrasound to tag the location of electrical activity, recovering spatial precision from a diffuse electrical measurement. It is the only current approach that promises electrode-grade localization with no electrode, and British state research funding is upstream of it. **INF-0039** · *Insert:* [[wiki/Field Addressing]], [[wiki/Continuity Bandwidth]] · *Tier:* plausible — Field-plus-receptor architectures make addressable population size a function of expression pattern rather than of implant geometry. If expression can be driven across a cortical region, the addressable channel count jumps by orders of magnitude in a single step rather than by iterative device generations. **INF-0040** · *Insert:* [[wiki/Nanoparticle Distribution]], [[wiki/Dose Uniformity]] · *Tier:* unresolved — Whether magnetoelectric particles can be distributed with sufficient uniformity to address a functionally meaningful population is the open technical question for that modality. Resolution comes from imaging studies of particle biodistribution at scale, not from stimulation demonstrations. **INF-0041** · *Insert:* [[wiki/Reversibility]], [[wiki/Interface Ethics]] · *Tier:* analytic — A sensitized cell population is a permanent modification even when the hardware is external, which relocates the irreversibility from the device to the genome. Consent frameworks written for implants do not describe this situation, and the gap will be discovered clinically before it is described legally. **INF-0042** · *Insert:* [[wiki/Bill Peebles]], [[wiki/World Simulation]] · *Tier:* strongly indicated — A non-surgical read/write channel and a high-fidelity world model are complementary halves of the same product: one supplies the environment, the other the participant. The teams pursuing them are already capitalized by the same investor set, which shortens the integration timeline considerably. **INF-0043** · *Insert:* [[wiki/Caltech]], [[wiki/Institutional Transfer]] · *Tier:* established — Shapiro's acoustic biomolecules moving from a Caltech laboratory into a venture-backed company with OpenAI as research collaborator is a complete technology-transfer event with a named date. Trace the researcher and the capability's location is always known. **INF-0044** · *Insert:* [[wiki/Ultrasound Neuromodulation]], [[wiki/Non-Invasive Therapy]] · *Tier:* plausible — Transcranial focused ultrasound already modulates deep structures without genetic modification, which gives the modality a clinical on-ramp independent of sonogenetics. The therapy indication funds the hardware, and the hardware is the same hardware the interface needs. **INF-0045** · *Insert:* [[wiki/Molecular Recording]], [[wiki/DNA Ticker Tape]] · *Tier:* plausible — Molecular recorders that write neural activity into DNA inside the cell would remove bandwidth from the readout path entirely: the record is made locally and read once, later, by sequencing. That inverts the entire acquisition architecture and makes whole-population recording a sequencing-cost problem. **INF-0046** · *Insert:* [[wiki/Expression Targeting]], [[wiki/Cell Type]] · *Tier:* analytic — Enhancer-driven expression lets a modality address a cell type rather than a volume, which is a form of logical addressing overlaid on physical space. The nervous system becomes a network with types as addresses, and that is a far more powerful abstraction than coordinates. **INF-0047** · *Insert:* [[wiki/Peripheral Nerve Interface]], [[wiki/Vagus Nerve]] · *Tier:* strongly indicated — The peripheral nervous system is accessible without craniotomy and carries organ-regulatory traffic, which makes it the first place where continuous closed-loop control of the body's internal state becomes routine. Central interfaces will inherit a control theory developed peripherally. **INF-0048** · *Insert:* [[wiki/Interface Modality Convergence]], [[wiki/Standards]] · *Tier:* plausible — As modalities multiply, the durable asset stops being the transducer and becomes the signal format and the model that consumes it. Whoever defines the representation that all modalities target owns the layer above every device generation. **INF-0049** · *Insert:* [[wiki/Gabe Newell]], [[wiki/Consumer Neurotechnology]] · *Tier:* analytic — Capital from interactive entertainment arriving in a non-surgical interface company signals which market is expected to absorb the technology at volume. Medical indication is the regulatory route; the addressable population is being underwritten by people who sell experiences. **INF-0050** · *Insert:* [[wiki/Sensitization Cascade]], [[wiki/Upgradeability]] · *Tier:* plausible — If the receptor layer is genetic and the field layer is external, the interface upgrades on a hardware cadence while the biological substrate stays constant. That is the first architecture in neurotechnology with a Moore's-law-shaped improvement path, because the fast-moving component is no longer inside the patient. --- ## Cluster C — Vascular routes and distributed access **INF-0051** · *Insert:* [[wiki/Synchron]], [[wiki/Stentrode]] · *Tier:* established — The Stentrode reaches cortex through a cerebral vein, importing catheter technique from interventional cardiology into neurotechnology. The significance is procedural rather than electrical: it makes neural access a same-day interventional procedure performed by a specialty that already exists in every major hospital. **INF-0052** · *Insert:* [[wiki/Endovascular Access]], [[wiki/Scaling Path]] · *Tier:* strongly indicated — The vasculature is a pre-existing distribution network reaching every cubic millimeter of brain, which means an endovascular modality's ceiling is set by how small a device can be delivered through progressively finer vessels. Miniaturization, not surgery, is that lineage's entire roadmap. **INF-0053** · *Insert:* [[wiki/Tom Oxley]], [[wiki/University of Melbourne]] · *Tier:* established — Oxley conceived the Stentrode in 2007, filed in 2010, moved to the United States in 2015 and raised capital in 2016 — a nineteen-year arc from concept to a pivotal trial. That cadence is the realistic clock for any neural-interface modality, and it should anchor every timeline estimate in the corpus. **INF-0054** · *Insert:* [[wiki/COMMAND Trial]], [[wiki/Safety Endpoint]] · *Tier:* established — Six patients holding a primary safety endpoint across twelve months with no vascular or brain events is the evidentiary object that permits a pivotal trial. Safety-first sequencing means the field's efficacy data will arrive suddenly and late, which is why capability appears to jump rather than climb. **INF-0055** · *Insert:* [[wiki/Premarket Approval]], [[wiki/Precedent]] · *Tier:* strongly indicated — The first PMA for a permanently implanted BCI creates the classification template every subsequent device is measured against. Regulatory precedent is a form of standards authorship, and it is being competed for by a small number of firms on a two-to-three-year horizon. **INF-0056** · *Insert:* [[wiki/Nitinol]], [[wiki/Materials Lineage]] · *Tier:* analytic — A nitinol stent scaffold carrying electrodes is a cardiology material doing neuroscience work, which means the device's reliability data inherits decades of vascular-implant experience. Borrowed reliability is the quiet advantage of modality transfer between clinical disciplines. **INF-0057** · *Insert:* [[wiki/Vascular Lattice]], [[wiki/Distributed Sensing]] · *Tier:* plausible — A future endovascular deployment need not be one stent: a lattice of many small vascular sensors would sample widely at low per-site resolution. Wide-and-shallow is the complement to narrow-and-deep, and whole-brain state estimation likely requires both classes operating together. **INF-0058** · *Insert:* [[wiki/Cerebral Venous System]], [[wiki/Access Geometry]] · *Tier:* analytic — Venous anatomy determines which cortical territories an endovascular device can reach, so the map of accessible mind is currently drawn by vascular anatomy. Anatomical accessibility is an unexamined constraint on which cognitive functions become instrumentable first. **INF-0059** · *Insert:* [[wiki/Patient Registry]], [[wiki/Cohort Assembly]] · *Tier:* strongly indicated — Building a patient registry before a pivotal trial is an assertion that enrollment, not evidence, is the binding constraint. Registries are also longitudinal cohorts, and a longitudinal cohort of interfaced humans is the substrate on which every population-level claim about neural data will eventually rest. **INF-0060** · *Insert:* [[wiki/Interventional Neurology]], [[wiki/Workforce]] · *Tier:* plausible — Scaling any implanted interface to millions requires a trained procedural workforce, and the endovascular route is the only one that can borrow an existing one. Workforce availability, not device manufacturing, will gate deployment volume in the 2030s. **INF-0061** · *Insert:* [[wiki/Percept Neurostimulator]], [[wiki/Installed Base]] · *Tier:* established — More than 40,000 patients carry Medtronic Percept devices, making adaptive DBS the largest commercial deployment of brain-computer interface technology that has ever existed. The installed base of sensing neural implants is already measured in tens of thousands, and it is clinical rather than experimental. **INF-0062** · *Insert:* [[wiki/BrainSense Adaptive DBS]], [[wiki/Closed-Loop Control]] · *Tier:* established — BrainSense reads local field potentials and adjusts stimulation in real time, which is a closed control loop running continuously inside a human brain, approved and reimbursed. The architecture the continuity literature describes as speculative is in commercial distribution under a movement-disorder indication. **INF-0063** · *Insert:* [[wiki/ADAPT-PD]], [[wiki/Home Setting]] · *Tier:* strongly indicated — ADAPT-PD evaluated adaptive stimulation in clinical *and home* settings across 68 patients over a year, which means the dataset includes unsupervised daily life. Continuous neural telemetry outside the clinic is the precondition for every longitudinal model of a person's neural state. **INF-0064** · *Insert:* [[wiki/BrainSense Electrode Identifier]], [[wiki/Automated Programming]] · *Tier:* analytic — Cutting initial programming time by 85% through automated electrode selection moves clinical judgment into an algorithm. Each such move raises the fraction of the therapy defined by the manufacturer's model rather than by the attending physician, which is a governance shift dressed as a workflow improvement. **INF-0065** · *Insert:* [[wiki/Responsive Neurostimulation]], [[wiki/Seizure Prediction]] · *Tier:* established — Responsive stimulation for epilepsy already detects a forming pathological state and intervenes before it completes. Predictive intervention on brain state is therefore a solved clinical category, and the open question is only which other states are worth predicting. **INF-0066** · *Insert:* [[wiki/Bioelectronic Medicine]], [[wiki/Organ Regulation]] · *Tier:* plausible — Electrical modulation of autonomic pathways lets a device set inflammatory, metabolic, and cardiac setpoints continuously. That is real-time regulation of body systems by an implanted controller, and it arrives under ordinary disease indications rather than as enhancement. **INF-0067** · *Insert:* [[wiki/Local Field Potential]], [[wiki/Biomarker]] · *Tier:* strongly indicated — Once a neural biomarker is validated for one condition, the sensing hardware generalizes to any condition with a detectable spectral signature. The installed Percept base is therefore a general-purpose neural observatory with a single approved use and many latent ones. **INF-0068** · *Insert:* [[wiki/Battery Longevity]], [[wiki/Duty Cycle]] · *Tier:* analytic — Adaptive stimulation conserves battery because it stimulates only when required, which extends device life and reduces replacement surgeries. Every efficiency gain in duty cycle translates directly into dwell time, and dwell time is the currency of continuity hardware. **INF-0069** · *Insert:* [[wiki/Helen Bronte-Stewart]], [[wiki/Stanford Neuromodulation]] · *Tier:* established — The academic neuromodulation laboratories that validated adaptive stimulation are the same ones supplying the clinical leadership of the implant industry. The people are the transfer mechanism, and the transfer is documented in trial authorship. **INF-0070** · *Insert:* [[wiki/Therapeutic Window]], [[wiki/Personalization]] · *Tier:* plausible — A device that continuously tunes itself to an individual's fluctuating state is maintaining a personal model that no clinician possesses. Over years, that model becomes a more detailed record of the person's neural dynamics than any medical record, and its custody is unaddressed. **INF-0071** · *Insert:* [[wiki/Neuromodulation Market]], [[wiki/Scale Economics]] · *Tier:* analytic — Parkinson's affects over ten million people globally with about one million in the United States, which gives closed-loop neural implants an addressable population large enough to industrialize manufacturing. Industrial volume in medical implants is what makes consumer volume physically possible later. **INF-0072** · *Insert:* [[wiki/Sensing Implant]], [[wiki/Data Egress]] · *Tier:* unresolved — What fraction of the neural data recorded by approved therapeutic implants is retained, transmitted, or used for model training is not publicly specified by device manufacturers. Resolution would come from device-labeling review and manufacturer data-practice disclosures. **INF-0073** · *Insert:* [[wiki/Spinal Cord Stimulation]], [[wiki/Motor Restoration]] · *Tier:* established — Epidural spinal stimulation has restored volitional movement after complete injury by amplifying residual descending signal. The nervous system tolerates an external controller inserted mid-pathway, which is direct evidence that mediated function is functionally continuous with native function. **INF-0074** · *Insert:* [[wiki/Brain-Spine Interface]], [[wiki/Digital Bridge]] · *Tier:* strongly indicated — A cortical decoder driving a spinal stimulator restores movement by reconstructing a broken link in software. Once a function can be carried across a gap by a machine, the substrate of that function is demonstrably relocatable, which is the smallest existing proof of the succession argument. **INF-0075** · *Insert:* [[wiki/Rehabilitation]], [[wiki/Neuroplasticity]] · *Tier:* plausible — Patients using digital bridges recover function that persists when the device is off, meaning the machine taught the nervous system. A device that induces durable biological change is not a prosthesis but a training instrument, and training instruments generalize far beyond injury. --- ## Cluster D — Cognitive operating systems and the interface layer **INF-0076** · *Insert:* [[wiki/BCI Human Interface Device]], [[wiki/Apple]] · *Tier:* established — Apple's BCI HID protocol makes neural signal a native input class alongside touch and voice, which moves the interface question from the implant vendor to the operating-system vendor. Every application on the platform becomes addressable by thought without any application being rewritten. **INF-0077** · *Insert:* [[wiki/Input Abstraction]], [[wiki/Platform Control]] · *Tier:* strongly indicated — The party that defines the input abstraction captures the interface layer regardless of who builds the sensor. Neural input arriving as a HID class means the economic center of the BCI stack sits with platform owners, not with device makers. **INF-0078** · *Insert:* [[wiki/Cognitive Operating System]], [[wiki/Large Language Model]] · *Tier:* plausible — A decoder that emits semantic features rather than keystrokes pairs naturally with a language model that completes intent. The interface becomes a conversation between a partial neural signal and a predictive model, and most of the realized bandwidth comes from the model's prior rather than from the electrode. **INF-0079** · *Insert:* [[wiki/Intent Completion]], [[wiki/Shared Autonomy]] · *Tier:* strongly indicated — Shared-autonomy control already lets sparse neural signal drive complex action by delegating execution detail to an autonomous system. As the autonomous half improves, the neural half required shrinks, which means interface capability rises without any improvement to the implant. **INF-0080** · *Insert:* [[wiki/Model Prior]], [[wiki/Authorship]] · *Tier:* analytic — When a model supplies most of the structure of a decoded utterance, the question of whose expression it is becomes technical rather than philosophical. Attribution ratios inside assisted communication will need to be measurable, and that measurement is itself a new instrument. **INF-0081** · *Insert:* [[wiki/Mind Captioning]], [[wiki/Tomoyasu Horikawa]] · *Tier:* established — Horikawa's mind-captioning work generated descriptive text from fMRI during both viewing and recall, and the descriptions captured relational structure — who did what to whom — rather than object lists. Recall decoding means the target is internal content, not stimulus, which is the decisive distinction for continuity work. **INF-0082** · *Insert:* [[wiki/Semantic Decoding]], [[wiki/Alexander Huth]] · *Tier:* established — Continuous language has been reconstructed from non-invasive recordings by aligning brain activity to a language model's semantic space. The alignment, not the scanner, does the work, which is why decoding quality now improves with each generation of language model. **INF-0083** · *Insert:* [[wiki/Inner Speech]], [[wiki/MindAlign]] · *Tier:* strongly indicated — Work through 2026 has moved from decoding externally supplied captions toward recovering participants' own inner speech via subject-specific neural-semantic alignment. Inner speech is the first genuinely private content to become instrumentable, and the technique requires no change to the underlying language model. **INF-0084** · *Insert:* [[wiki/Shared Embedding Space]], [[wiki/Interoperability]] · *Tier:* plausible — If brain activity and language both map into one multimodal embedding space, that space is the interchange format between minds and machines. A common latent representation is the actual technical meaning of interoperability between a nervous system and a model. **INF-0085** · *Insert:* [[wiki/Subject-Specific Alignment]], [[wiki/Transfer Learning]] · *Tier:* analytic — Decoders currently require per-subject alignment, which is the main obstacle to deployment. The moment cross-subject transfer works from a short calibration, decoding becomes a service rather than a study, and the cost per decoded mind falls by orders of magnitude. **INF-0086** · *Insert:* [[wiki/Non-Invasive Decoding]], [[wiki/Mental Privacy]] · *Tier:* strongly indicated — Semantic decoding from non-invasive recordings requires cooperation today, and that requirement is a property of current signal quality rather than a principle. Legal frameworks built on the assumption of required cooperation are therefore built on a moving quantity. **INF-0087** · *Insert:* [[wiki/Context Window]], [[wiki/Working Memory]] · *Tier:* analytic — A model's context window and a person's working memory are both bounded state buffers feeding a much larger associative store. The engineering vocabulary developed for one — eviction policy, retrieval, compaction — describes the other well enough to be a research program rather than a metaphor. **INF-0088** · *Insert:* [[wiki/Agent Memory]], [[wiki/Continuity Sidecar]] · *Tier:* plausible — Persistent agent memory systems are already solving the problem of carrying identity across stateless inference episodes. That is the same architecture a continuity sidecar requires, and it is being debugged commercially at enormous scale for unrelated reasons. **INF-0089** · *Insert:* [[wiki/Stateful AI]], [[wiki/Personalization Layer]] · *Tier:* strongly indicated — When models became stateful, the durable asset moved from the weights to the accumulated per-user state. That relocation is precisely the host/residual split: the reusable prior lives in the model, and the person lives in the delta. **INF-0090** · *Insert:* [[wiki/Prompt State]], [[wiki/Serialization]] · *Tier:* analytic — A person's residual state, however captured, must be serializable to be portable, and serialization formats outlive the systems that create them. Whoever specifies the residual format specifies what can be carried between hosts for as long as the format survives. **INF-0091** · *Insert:* [[wiki/Neural Keyboard]], [[wiki/Throughput Ceiling]] · *Tier:* plausible — Treating a neural interface as a faster keyboard caps its value at typing speed. The configurations that exceed human output rates are the ones that skip language altogether and exchange embeddings, which requires the receiving system to be a model rather than a person. **INF-0092** · *Insert:* [[wiki/Attention Estimation]], [[wiki/Adaptive Interface]] · *Tier:* plausible — An interface that reads cognitive load and adjusts its own pacing is a closed loop around attention itself. Once attention is a controlled variable, software design becomes a form of neuromodulation conducted through a screen. **INF-0093** · *Insert:* [[wiki/Passive Neural Sensing]], [[wiki/Wearables]] · *Tier:* strongly indicated — Consumer devices already carry electrodes near the head in earbuds and headsets, which makes passive neural sensing an accessory question rather than a surgical one. The first population-scale neural dataset will be collected by consumer hardware, not by clinics. **INF-0094** · *Insert:* [[wiki/Neural Foundation Model]], [[wiki/Scaling Law]] · *Tier:* plausible — Foundation models trained on aggregated neural recordings will exhibit the same scaling behavior as other modalities, meaning decoding quality becomes a dataset-size problem. The entity holding the largest standardized neural corpus will hold decoding capability that cannot be replicated by better hardware alone. **INF-0095** · *Insert:* [[wiki/Calibration-Free Decoding]], [[wiki/Deployment]] · *Tier:* plausible — A pretrained neural foundation model with few-shot subject adaptation would collapse setup from weeks to minutes. That single change moves neural interfacing from a procedure with a technician to a product with an onboarding flow. **INF-0096** · *Insert:* [[wiki/Speech Restoration]], [[wiki/Voice Cloning]] · *Tier:* strongly indicated — Restoring speech with a synthesized version of the person's own pre-morbid voice is already technically routine. The restored person therefore speaks in a voice generated from a model, and the resulting identity object is part biological, part trained artifact. **INF-0097** · *Insert:* [[wiki/Cognitive Prosthesis]], [[wiki/Memory Augmentation]] · *Tier:* plausible — Hippocampal prosthesis work has shown that stimulating with a model-predicted pattern can improve recall performance. A device that writes a computed memory trace is a memory co-processor, and co-processors are historically absorbed into the architecture they assist. **INF-0098** · *Insert:* [[wiki/Human-AI Dyad]], [[wiki/Distributed Cognition]] · *Tier:* analytic — Continuous low-latency access to a capable model makes the functional cognitive unit the person-plus-model dyad. Measuring the person alone becomes the wrong measurement, and every assessment instrument in psychology and education is calibrated on the wrong unit. **INF-0099** · *Insert:* [[wiki/Interface Dependency]], [[wiki/Withdrawal]] · *Tier:* plausible — Once cognitive function is routinely mediated, removing the mediator produces measurable deficit, which gives the interface the standing of a prosthesis for capacities that were never lost. Dependency is not a side effect of this architecture; it is the operating condition it creates. **INF-0100** · *Insert:* [[wiki/Operating System of Mind]], [[wiki/Vendor Lock-In]] · *Tier:* strongly indicated — If neural input, personal state, and the assisting model are supplied by one vendor, switching costs become cognitive rather than commercial. Portability of the personal residual is therefore the single most consequential standards fight of the next decade, and it is currently unclaimed. --- # Inference Statements for Wiki Insertion — Set 02 (INF-0101 – INF-0200) Continues the grammar established in Set 01: `ID · Insert: target entries · Tier` followed by an anchored forward statement naming actor, mechanism, and governing variable. --- ## Cluster E — Connectomic acquisition and the cost curve **INF-0101** · *Insert:* [[wiki/E11 Bio]], [[wiki/PRISM (E11 Bio)|PRISM]] · *Tier:* developer projection with demonstrated subsystem — PRISM combines protein barcoding, expansion microscopy, iterative molecular staining, and learned reconstruction as an optical alternative to parts of the electron-microscopy path. E11 projects an aggregate 100× cost reduction; the demonstrated record is the 2025 mouse-hippocampal pilot, not whole-brain cost validation. **INF-0102** · *Insert:* [[wiki/Proofreading]], [[wiki/Reconstruction Labor]] · *Tier:* established — Comprehensively tracing roughly 1,500 mouse neurons required years of human revision and over a million manual corrections, which places proofreading above 95% of project cost. Whole-brain connectomics is therefore a labor-automation problem wearing a microscopy costume. **INF-0103** · *Insert:* [[wiki/Local Shape Descriptors]], [[wiki/GPU Segmentation]] · *Tier:* strongly indicated — GPU-accelerated segmentation cutting reconstruction cost 10–100× means the connectome timeline is now indexed to accelerator availability. Every improvement in datacenter capacity shortens the schedule for mapping a mammalian brain, which couples neuroscience directly to the AI capital cycle. **INF-0104** · *Insert:* [[wiki/FlyWire]], [[wiki/Drosophila Connectome]] · *Tier:* established — The complete adult fly connectome demonstrated that a whole nervous system can be reconstructed, proofread, and released as a queryable object. The species barrier fell in 2024; what remains is volume, and volume responds to money and compute rather than to insight. **INF-0105** · *Insert:* [[wiki/MICrONS]], [[wiki/Functional Connectomics]] · *Tier:* established — MICrONS paired structural reconstruction with functional recording of the same cortical tissue, producing wiring and activity from one volume. Structure-plus-function in a single specimen is the minimum dataset any emulation claim requires, and it now exists at cubic-millimeter scale. **INF-0106** · *Insert:* [[wiki/Zetta AI]], [[wiki/Princeton]] · *Tier:* established — The Princeton reconstruction lineage — Seung, Murthy, Macrina, Dorkenwald, Bae — appears across FlyWire, MICrONS, and human retinal volumes and now inside a commercial entity. A field with a single dominant pipeline lineage has a single point of methodological leverage. **INF-0107** · *Insert:* [[wiki/Human Foveal Retina Volume]], [[wiki/Human Connectomics]] · *Tier:* strongly indicated — Electron-microscopic reconstruction of human neural tissue has already been performed on retinal and cortical samples, which means the species question is settled and the remaining question is scale. Human connectomics is underway at millimeter scale today. **INF-0108** · *Insert:* [[wiki/Expansion Microscopy]], [[wiki/Ed Boyden]] · *Tier:* established — Physically enlarging tissue to make nanoscale structure visible to light optics converts an optics problem into a chemistry problem. Chemistry scales with reagent cost, optics with instrument cost, and that substitution is the entire economic argument of optical connectomics. **INF-0109** · *Insert:* [[wiki/Molecular Annotation]], [[wiki/Multiplexed Staining]] · *Tier:* plausible — Iterative staining attaches molecular identity to each traced process, so the resulting map carries cell type and receptor complement rather than geometry alone. A typed connectome is directly simulable in a way a geometric one is not, because the parameters come with it. **INF-0110** · *Insert:* [[wiki/AWS Open Data]], [[wiki/Dataset Gravity]] · *Tier:* analytic — Publishing connectomic volumes into a hyperscaler's open-data registry places the field's primary datasets inside one commercial cloud's gravity well. Compute follows data, and the location of the data determines which organizations can afford to analyze it. **INF-0111** · *Insert:* [[wiki/Whole Mouse Brain Connectome]], [[wiki/Cost Threshold]] · *Tier:* strongly indicated — At current electron-microscopy economics a whole mouse brain implies tens of billions of dollars; at a 100× reduction it becomes a large-instrument-scale project comparable to a telescope. That crossing converts the mouse connectome from a thought experiment into a fundable program. **INF-0112** · *Insert:* [[wiki/Sectioning Loss]], [[wiki/Sample Robustness]] · *Tier:* established — Losing or damaging ultra-thin sections fragments a dataset into discontinuous chunks, and no group has traced neurons through even a millimeter of sliced tissue in all three axes. Mechanical handling, not imaging, is the unglamorous constraint on long-range connectivity. **INF-0113** · *Insert:* [[wiki/Long-Range Projection]], [[wiki/Whole-Brain Wiring]] · *Tier:* analytic — Local circuits can be reconstructed from small volumes, but identity-relevant computation depends on long-range projections that cross the whole brain. A connectome that omits long-range structure describes a component, not a mind, which makes volume continuity the key acquisition property. **INF-0114** · *Insert:* [[wiki/Synapse Detection]], [[wiki/Ground Truth]] · *Tier:* plausible — Automated synapse detection improving on a published whole-brain volume means reconstructions are now versioned artifacts that improve after release. A connectome is a living dataset with a changelog, and any emulation built on it inherits a specific version. **INF-0115** · *Insert:* [[wiki/Connectome Versioning]], [[wiki/Provenance]] · *Tier:* strongly indicated — If reconstructions are revised, then results derived from them must cite a version to be reproducible. Provenance infrastructure for neural data is therefore not bureaucratic overhead but a precondition for any claim about a specific mapped brain. **INF-0116** · *Insert:* [[wiki/Preservation Quality]], [[wiki/Aldehyde Fixation]] · *Tier:* established — Every connectome depends on preservation chemistry that arrests ultrastructure before imaging, which makes fixation quality the upstream determinant of everything downstream. Preservation is the first irreversible step in any acquisition pipeline and the least improved in decades. **INF-0117** · *Insert:* [[wiki/Synaptic Weight]], [[wiki/Structural Proxy]] · *Tier:* unresolved — Whether synaptic strength can be inferred reliably from morphology alone is the open question that determines if a static connectome is sufficient for function. Resolution comes from paired physiology-and-ultrastructure datasets, which MICrONS-class experiments are designed to produce. **INF-0118** · *Insert:* [[wiki/Neuromodulator State]], [[wiki/Missing Parameters]] · *Tier:* analytic — Connectomes record wiring but not the diffuse neuromodulatory context that sets gain across the whole network. Recovering that context requires molecular measurement alongside structure, which is precisely what molecular-annotation methods add, closing a gap most emulation critiques assume is permanent. **INF-0119** · *Insert:* [[wiki/Connectomics Pipeline]], [[wiki/Industrialization]] · *Tier:* plausible — Once reconstruction is automated end-to-end, connectomics becomes a service industry with throughput measured in cubic millimeters per week. Industrial throughput changes what questions get asked: comparative connectomics across individuals becomes routine, and individual variation becomes measurable. **INF-0120** · *Insert:* [[wiki/Comparative Connectomics]], [[wiki/Individuality]] · *Tier:* plausible — When many individuals of a species are mapped, the invariant scaffold separates from the person-specific deviation. That separation is the empirical form of the host/residual distinction, and it will be measured in mice before it is argued about in humans. **INF-0121** · *Insert:* [[wiki/Sam Rodriques]], [[wiki/Crick Institute]] · *Tier:* established — The PRISM collaboration spans the Crick Institute, Max Planck/LMB Cambridge, MIT, and HHMI under a philanthropically funded nonprofit. Nonprofit structure means the method is published rather than licensed, which accelerates diffusion and removes the usual private bottleneck. **INF-0122** · *Insert:* [[wiki/Philanthropic Funding]], [[wiki/Research Structure]] · *Tier:* analytic — Philanthropy funds the expensive, unprofitable middle of a capability curve that neither grants nor venture capital will carry. The existence of philanthropically funded connectomics is what keeps the mapping trajectory on schedule between public science and commercial return. **INF-0123** · *Insert:* [[wiki/Human Brain Volume]], [[wiki/Scaling Arithmetic]] · *Tier:* analytic — A human brain holds on the order of a thousand times the neurons of a mouse, so a mouse-scale success does not imply a human one without another cost reduction of similar magnitude. Naming the required factor explicitly is what turns the roadmap into an engineering schedule. **INF-0124** · *Insert:* [[wiki/Post-Mortem Interval]], [[wiki/Human Tissue]] · *Tier:* unresolved — Human connectomics depends on tissue quality achievable within realistic post-mortem intervals, and that quality ceiling has not been characterized at whole-brain scale. Resolution requires a coordinated perfusion protocol study, which is an institutional rather than technical undertaking. **INF-0125** · *Insert:* [[wiki/Connectome Storage]], [[wiki/Petabyte Archive]] · *Tier:* strongly indicated — A whole mouse brain at electron-microscopic resolution occupies a multi-petabyte volume, and a human brain occupies an exabyte-class one. The archival medium and its century-scale durability become part of the neuroscience, which is where connectomics meets deep-time storage engineering. --- ## Cluster F — Simulation and emulation from structure **INF-0126** · *Insert:* [[wiki/Drosophila Simulation]], [[wiki/Connectome-Constrained Model]] · *Tier:* established — Simulations built directly from the fly connectome have predicted behavioral responses without fitting to behavior, meaning structure alone carried enough information to generate function. This is the first empirical evidence that wiring plus reasonable dynamics reproduces behavior in a real animal. **INF-0127** · *Insert:* [[wiki/ZAPBench]], [[wiki/Whole-Brain Prediction]] · *Tier:* strongly indicated — Benchmarking whole-brain activity prediction in larval zebrafish turns emulation from an argument into a leaderboard. Once a capability has a benchmark, it acquires a measurable improvement rate, and measurable improvement rates end debates about feasibility. **INF-0128** · *Insert:* [[wiki/C. elegans]], [[wiki/OpenWorm]] · *Tier:* established — Three hundred and two neurons mapped since the 1980s have still not yielded a complete behavioral simulation, which locates the difficulty in dynamics and parameters rather than in wiring. The lesson generalizes upward: the map is necessary, and the parameters are where the work is. **INF-0129** · *Insert:* [[wiki/Biophysical Model]], [[wiki/Parameter Estimation]] · *Tier:* plausible — Machine learning can infer unmeasured biophysical parameters from partial recordings, which converts an intractable measurement problem into a tractable inference one. Emulation fidelity then improves with model capacity rather than with instrumentation. **INF-0130** · *Insert:* [[wiki/Digital Twin of a Brain]], [[wiki/Validation]] · *Tier:* analytic — A brain emulation is validated the way any simulation is: by predicting held-out responses of the original. Establishing that validation protocol before human-scale attempts exist is what keeps the eventual claim adjudicable rather than rhetorical. **INF-0131** · *Insert:* [[wiki/Substrate Independence]], [[wiki/Functional Equivalence]] · *Tier:* plausible — If a connectome-derived model reproduces an animal's behavioral repertoire, functional equivalence has been demonstrated at that animal's complexity level. Each species crossed moves the demonstration up the ladder, and the ladder has no identified rung where the argument changes character. **INF-0132** · *Insert:* [[wiki/Neural Simulation Hardware]], [[wiki/Real-Time Factor]] · *Tier:* strongly indicated — Simulating a brain faster than real time is the property that distinguishes a research model from a hosted mind, because a mind running slower than its world cannot participate in it. Real-time factor, not accuracy alone, is the gate on emulation as continuity. **INF-0133** · *Insert:* [[wiki/Model Compression]], [[wiki/Sufficient Detail]] · *Tier:* plausible — Emulation need not simulate every ion channel if a compressed model reproduces the input-output behavior of each cell class. Determining the minimum sufficient abstraction is the central scientific question of uploading, and it is answerable empirically on small nervous systems now. **INF-0134** · *Insert:* [[wiki/Learning Rules]], [[wiki/Plasticity]] · *Tier:* analytic — A static connectome yields a frozen network; a living mind requires plasticity rules that update it. Any emulation that lacks learning is a recording rather than a continuation, which makes plasticity implementation the boundary between playback and life. **INF-0135** · *Insert:* [[wiki/Sleep]], [[wiki/Consolidation]] · *Tier:* plausible — Replay and consolidation during sleep reorganize memory offline, which means a faithful emulation would need an offline phase of its own. Emulated minds will therefore have scheduled downtime for reasons intrinsic to cognition rather than to maintenance. **INF-0136** · *Insert:* [[wiki/Engram]], [[wiki/Memory Trace]] · *Tier:* established — Specific memories have been tagged, ablated, and reactivated in rodents, demonstrating that discrete traces are physically localizable and manipulable. Memory is an addressable object in laboratory practice, which is the precondition for memory transfer being a technical rather than metaphysical proposal. **INF-0137** · *Insert:* [[wiki/Memory Transfer]], [[wiki/Cross-Substrate Write]] · *Tier:* plausible — If an engram can be identified in one brain and written into another substrate that supports the same representational geometry, transfer is a re-encoding problem. Representation alignment between substrates is the hard part, and alignment is exactly what current decoding research is building. **INF-0138** · *Insert:* [[wiki/Representational Geometry]], [[wiki/Alignment]] · *Tier:* strongly indicated — Independent networks trained on similar data converge to similar representational geometries, which is why brain-to-model alignment works at all. Convergent representation is the technical basis for believing minds and models can exchange content without a translator built by hand. **INF-0139** · *Insert:* [[wiki/Simulation Fidelity]], [[wiki/Compute Budget]] · *Tier:* analytic — Every increment of biological detail multiplies compute cost, so emulation is a negotiation between fidelity and affordability. The negotiated settlement point will be set by what a paying party considers sufficient, which makes the definition of "enough person" a commercial decision. **INF-0140** · *Insert:* [[wiki/Organoid Model]], [[wiki/Validation Substrate]] · *Tier:* plausible — Brain organoids provide living tissue whose connectivity can be mapped and whose activity can be recorded in the same preparation. That makes them the ideal test bed for emulation methods, since ground truth and simulation can be compared in an object small enough to iterate on weekly. **INF-0141** · *Insert:* [[wiki/Species Ladder]], [[wiki/Research Program]] · *Tier:* analytic — The field is climbing an explicit ladder — worm, fly, zebrafish, mouse, primate, human — and each rung is a funded program with published milestones. Reading the ladder as a schedule rather than a metaphor produces a defensible estimate of when human-scale attempts begin. **INF-0142** · *Insert:* [[wiki/Emulation Ethics]], [[wiki/Moral Status]] · *Tier:* unresolved — At what fidelity an emulation acquires interests of its own is undetermined, and the question will become operational before it is settled philosophically. The first institution to adopt a working threshold will set the de facto standard for everyone else. **INF-0143** · *Insert:* [[wiki/Partial Emulation]], [[wiki/Hybrid Operation]] · *Tier:* plausible — Replacing one circuit with a simulated equivalent while the rest of the brain runs biologically is the incremental path, and hippocampal prosthesis work has already done it in rudimentary form. Gradual replacement sidesteps the discontinuity problem that whole-brain transfer raises. **INF-0144** · *Insert:* [[wiki/Gradual Replacement]], [[wiki/Identity Continuity]] · *Tier:* analytic — If function migrates piece by piece with the system continuously operating, there is no moment of transfer to object to. The engineering path and the philosophical path converge on incrementalism for the same reason, which is why hybrid architectures dominate serious roadmaps. **INF-0145** · *Insert:* [[wiki/Emulation Storage Format]], [[wiki/Longevity]] · *Tier:* strongly indicated — A stored emulation is only as durable as its format's interpretability, and formats rot faster than media. Specifying an emulation interchange format with a documented semantics is a continuity requirement equal in weight to the acquisition itself. **INF-0146** · *Insert:* [[wiki/Reanimation]], [[wiki/Deferred Execution]] · *Tier:* plausible — A captured state that cannot yet be run is not a failure but a deferral, provided the capture is sufficient and the format survives. Deferred execution reframes preservation as a bet on future compute rather than on present capability. **INF-0147** · *Insert:* [[wiki/Whole-Brain Emulation Roadmap]], [[wiki/Milestones]] · *Tier:* analytic — Roadmaps written in the 2000s specified resolution, throughput, and compute requirements that can now be checked against delivered capability. Auditing those predictions against 2026 reality is the most rigorous available method for calibrating the remaining timeline. **INF-0148** · *Insert:* [[wiki/Functional Digital Twin]], [[wiki/Clinical Use]] · *Tier:* plausible — Patient-specific brain models used to plan stimulation targets are emulation technology arriving under a therapeutic indication. Clinical digital twins will accumulate fidelity for medical reasons and be available for other purposes once they exist. **INF-0149** · *Insert:* [[wiki/Neural Network Alignment]], [[wiki/Model-Brain Comparison]] · *Tier:* strongly indicated — Comparing artificial and biological networks on the same stimuli has become a standard method, producing a quantitative similarity measure between minds and models. A metric that scores how close a model is to a brain is also the metric that will score how close an emulation is to a person. **INF-0150** · *Insert:* [[wiki/Consciousness Metric]], [[wiki/Adjudication]] · *Tier:* unresolved — No agreed measurement distinguishes a system that experiences from one that behaves as if it does, and the emulation trajectory will reach deployment before that measurement exists. Recording the gap explicitly is more useful than resolving it prematurely, because the gap is where policy will be improvised. --- ## Cluster G — Biological compute and wetware hosting **INF-0151** · *Insert:* [[wiki/Cortical Labs]], [[wiki/CL1]] · *Tier:* established — A 20-unit CL1 rack at NUS, roughly 16 million living human neurons under datacenter operations with DayOne infrastructure, is the first independently operated biologically integrated server rack. Wetware has entered the rack form factor, with power, cooling, and an operator. **INF-0152** · *Insert:* [[wiki/DishBrain]], [[wiki/Brett Kagan]] · *Tier:* established — DishBrain's roughly 800,000 neurons on a CMOS array learning Pong, and the same substrate class later learning Doom, established that cultured tissue performs closed-loop goal-directed adaptation. The lineage from that demonstration to a commercial rack took four years. **INF-0153** · *Insert:* [[wiki/Neuron Lifetime]], [[wiki/Culture Maintenance]] · *Tier:* established — CL1 neurons survive up to six months under internal life support, which sets the current replacement cycle for biological compute. Every business model in the sector is implicitly a bet on extending that window, and extension is a tissue-engineering result rather than a computing one. **INF-0154** · *Insert:* [[wiki/Biological Data Centre]], [[wiki/DayOne]] · *Tier:* strongly indicated — A hyperscale developer with roughly 2.1GW of bookings taking a position in biological compute means the substrate is being evaluated against datacenter power economics. That comparison — joules per useful operation — is the only metric under which wetware wins, and it is now being measured by people who buy power. **INF-0155** · *Insert:* [[wiki/Energy Per Operation]], [[wiki/Metabolic Computing]] · *Tier:* analytic — A CL1 unit reportedly draws less power than a handheld calculator while hosting millions of neurons, which is a several-order-of-magnitude efficiency claim against silicon inference. If it survives independent measurement, biological substrate becomes economically rational for specific workload classes rather than merely interesting. **INF-0156** · *Insert:* [[wiki/Synthetic Biological Intelligence]], [[wiki/Sparse Data Learning]] · *Tier:* plausible — The claimed advantage of biological substrate is learning from sparse data and adapting under changing conditions, which is precisely where gradient-trained networks are weakest. Complementarity rather than replacement is the realistic deployment shape, and complementary systems become permanent. **INF-0157** · *Insert:* [[wiki/iPSC]], [[wiki/Blood-Derived Neurons]] · *Tier:* established — CL1 neurons are reprogrammed from human blood cells, which means any donor can supply the substrate. Donor-specific biological compute is therefore available today as a technical matter, and personalized neural substrate is an ordering question rather than a research question. **INF-0158** · *Insert:* [[wiki/Donor-Specific Substrate]], [[wiki/Personal Hosting]] · *Tier:* plausible — Growing compute from a specific person's cells creates a substrate with that person's genome running unrelated workloads. The custody, consent, and inheritance questions this raises are unlike anything in either computing or medicine, and they arrive with the first commercial order. **INF-0159** · *Insert:* [[wiki/FinalSpark]], [[wiki/Neuroplatform]] · *Tier:* established — Remote access to organoid electrophysiology from Vevey lets research groups worldwide run experiments on living tissue they never touch. Biological compute as a hosted service preceded biological compute as a product, which mirrors how cloud preceded on-premises consolidation in reverse. **INF-0160** · *Insert:* [[wiki/Biological Cloud]], [[wiki/Service Model]] · *Tier:* analytic — Selling access rather than units means the operator retains the tissue, the protocols, and all the operational knowledge about keeping it alive. Operational knowledge about substrate maintenance is the moat in this sector, not the biology itself. **INF-0161** · *Insert:* [[wiki/Brainoware]], [[wiki/Reservoir Computing]] · *Tier:* established — Organoid reservoir computing uses the tissue's intrinsic dynamics as a fixed nonlinear expansion with only the readout trained. It is the cheapest way to extract computation from living tissue, and cheap extraction is what makes early commercialization possible. **INF-0162** · *Insert:* [[wiki/Organoid Intelligence]], [[wiki/Field Formation]] · *Tier:* strongly indicated — The field acquired a name, a roadmap, and dedicated funding before it acquired a killer application, which is the normal sequence for capability domains that later become infrastructure. Named fields attract standardized methods, and standardized methods attract industrial participation. **INF-0163** · *Insert:* [[wiki/Microelectrode Array]], [[wiki/CMOS Integration]] · *Tier:* established — High-density CMOS microelectrode arrays are the shared instrument across DishBrain, organoid platforms, and slice physiology. One instrument class underlies the entire biological-compute sector, and its channel density improvements propagate to every participant simultaneously. **INF-0164** · *Insert:* [[wiki/Vascularization]], [[wiki/Organoid Scaling]] · *Tier:* strongly indicated — Organoids stop growing when diffusion cannot supply the interior, so vascularization is the gate on tissue volume. Every claim about scaling biological compute is a claim about perfusion engineering, and perfusion is a solved problem in other organs. **INF-0165** · *Insert:* [[wiki/Rickie Patani]], [[wiki/NUS]] · *Tier:* established — A university neurobiology programme maintaining the living cells inside a commercial deployment makes the academic laboratory an operational dependency of a datacenter. That is a genuinely new institutional relationship, and it will be replicated wherever biological racks are installed. **INF-0166** · *Insert:* [[wiki/Ethical Threshold]], [[wiki/Sentience Claim]] · *Tier:* unresolved — Cortical Labs' stated position distinguishes responsiveness and learning from consciousness, and no measurement adjudicates that distinction. The threshold question will be forced by scale rather than by philosophy: a thousand-unit deployment invites the question in a way a benchtop culture does not. **INF-0167** · *Insert:* [[wiki/Drug Discovery]], [[wiki/Application Pull]] · *Tier:* plausible — Drug discovery, humanoid robotics, cybersecurity, and fraud detection are the named target domains, and only the first has an obvious biological rationale. The others suggest the pitch is energy efficiency and adaptive learning, meaning wetware is being sold as an inference substrate rather than as a biology tool. **INF-0168** · *Insert:* [[wiki/Reply]], [[wiki/University of Milan]] · *Tier:* established — A CL1-based initiative announced by an Italian IT firm with a university partner in January 2026 shows the platform diffusing into ordinary enterprise IT contexts. Diffusion into enterprise is how a novel substrate acquires the tooling and integrations that make it durable. **INF-0169** · *Insert:* [[wiki/Hybrid Substrate]], [[wiki/Biohybrid]] · *Tier:* plausible — A rack containing both silicon accelerators and living tissue under one orchestration layer is a heterogeneous compute problem of a familiar type. Schedulers already place workloads across CPU, GPU, and specialized silicon; adding a biological device class is an extension, not a revolution. **INF-0170** · *Insert:* [[wiki/biOS]], [[wiki/Substrate Abstraction]] · *Tier:* strongly indicated — An operating system that presents living neurons through a code-deployable interface abstracts biology into an API. Once biology has an API, software engineers rather than neuroscientists become the majority of the people programming tissue, and the field's growth rate changes accordingly. **INF-0171** · *Insert:* [[wiki/Sustainability]], [[wiki/AI Energy Demand]] · *Tier:* analytic — Biological compute is being positioned explicitly against the energy intensity of AI datacenters, which is the most politically durable argument available in jurisdictions with grid constraints. Energy policy will therefore shape wetware adoption more than neuroscience will. **INF-0172** · *Insert:* [[wiki/Neural Substrate Supply Chain]], [[wiki/bit.bio]] · *Tier:* established — Cell-line suppliers sit upstream of every biological computer, making reprogramming companies a chokepoint in the sector. Supply-chain analysis of biological compute looks like semiconductor analysis with different vendors and the same structure. **INF-0173** · *Insert:* [[wiki/Continuity Hosting]], [[wiki/Biological Host]] · *Tier:* plausible — If a person's residual state can run on any sufficient substrate, living tissue grown from that person's own cells is one candidate host among several. The biological option reframes uploading as substrate choice rather than as escape from biology. **INF-0174** · *Insert:* [[wiki/Tissue Lifecycle]], [[wiki/State Migration]] · *Tier:* analytic — A six-month substrate lifetime forces periodic migration of whatever state the tissue holds, which means biological compute must solve state portability immediately rather than eventually. The shortest-lived substrate will produce the best migration tooling. **INF-0175** · *Insert:* [[wiki/Regulatory Classification]], [[wiki/Wetware]] · *Tier:* unresolved — Whether a rack of human neurons is a medical device, a research material, a computing product, or a biological specimen is unsettled in every jurisdiction where one has been installed. Classification will be determined by the first customs declaration or export-control question, not by legislation. --- ## Cluster H — Neuromorphic substrate and energy **INF-0176** · *Insert:* [[wiki/Neuromorphic Computing]], [[wiki/Event-Driven Processing]] · *Tier:* established — Spiking architectures compute only when something changes, which matches the statistics of sensor data and of neural activity itself. Event-driven processing is the only silicon paradigm whose energy profile resembles the brain's, and interfaces are sensor-dominated workloads. **INF-0177** · *Insert:* [[wiki/NorthPole]], [[wiki/IBM]] · *Tier:* established — NorthPole's co-location of memory with compute addresses the von Neumann bottleneck directly, which is the same architectural insight biology implements with synapses. Convergent architecture between silicon and tissue is what makes cross-substrate portability of neural workloads conceivable. **INF-0178** · *Insert:* [[wiki/Loihi]], [[wiki/On-Chip Learning]] · *Tier:* established — Neuromorphic chips with on-chip plasticity can adapt without a training cluster, which is the property an implanted decoder needs. Local learning eliminates the cloud round trip, and eliminating the round trip eliminates both latency and a surveillance surface. **INF-0179** · *Insert:* [[wiki/Memristor]], [[wiki/Analog Synapse]] · *Tier:* strongly indicated — Analog resistive devices store a weight and perform the multiply in the same physical element, collapsing memory and arithmetic. If manufacturing variability is tamed, the energy cost of inference drops far enough that continuous whole-day neural decoding becomes a wearable-power problem. **INF-0180** · *Insert:* [[wiki/Event Camera]], [[wiki/Sensor Lineage]] · *Tier:* established — Event-based vision sensors that report only pixel changes are the commercial beachhead of neuromorphic engineering, already shipping in industrial inspection and automotive. The sensor side monetized first and is funding the processor side, which is the usual order. **INF-0181** · *Insert:* [[wiki/Spiking Neural Network]], [[wiki/Conversion Tooling]] · *Tier:* plausible — Tools that convert trained conventional networks into spiking equivalents let neuromorphic hardware inherit the entire existing model ecosystem. Inheritance rather than reinvention is how an alternative substrate crosses from research into deployment. **INF-0182** · *Insert:* [[wiki/Neuromorphic Implant]], [[wiki/Thermal Budget]] · *Tier:* strongly indicated — An implanted decoder is thermally bounded by roughly a degree of allowable tissue heating, which excludes conventional accelerators outright. Neuromorphic silicon is not an aesthetic preference inside the skull; it is the only class that fits the power envelope. **INF-0183** · *Insert:* [[wiki/Compute Locality]], [[wiki/Privacy Architecture]] · *Tier:* analytic — Processing neural data on the implant means raw signal never leaves the body, which converts a legal privacy problem into an architectural property. Regulation that mandates data minimization will functionally mandate neuromorphic edge compute. **INF-0184** · *Insert:* [[wiki/Sparse Coding]], [[wiki/Biological Efficiency]] · *Tier:* established — Cortical activity is sparse, with few neurons active at any instant, which is why the brain's power budget is twenty watts. Any substrate that hopes to host brain-like computation at brain-like cost must exploit the same sparsity, and dense matrix hardware structurally cannot. **INF-0185** · *Insert:* [[wiki/Analog Compute]], [[wiki/Noise Tolerance]] · *Tier:* plausible — Analog computation trades precision for energy, and neural computation is demonstrably noise-tolerant. The substrate that matches the workload's error tolerance wins on cost, which argues that brain-hosting hardware will be analog long before it is exotic. **INF-0186** · *Insert:* [[wiki/Neuromorphic Cloud]], [[wiki/Access Model]] · *Tier:* plausible — Cloud-hosted neuromorphic services let developers target the architecture without owning it, which is how GPUs became ubiquitous in machine learning. Access model, not device performance, determines which substrates accumulate software ecosystems. **INF-0187** · *Insert:* [[wiki/Photonic Computing]], [[wiki/Interconnect Energy]] · *Tier:* strongly indicated — Data movement, not arithmetic, dominates the energy cost of large models, which is why optical interconnect is being pursued inside the rack. A mind-scale workload is communication-bound by construction, so optical fabrics are continuity infrastructure. **INF-0188** · *Insert:* [[wiki/Rack-Scale Computing]], [[wiki/Single Machine Abstraction]] · *Tier:* analytic — When a rack is addressed as one machine with a coherent memory domain, the unit of computation becomes the facility. A hosted mind would be a facility-scale object, and facility-scale objects have addresses, owners, and jurisdictions. **INF-0189** · *Insert:* [[wiki/800 VDC Power]], [[wiki/Open Compute Project]] · *Tier:* established — Standardizing high-voltage DC distribution across Google, Microsoft, and NVIDIA within OCP makes the physical metabolism of AI a common specification. Shared metabolism means any workload written for one facility runs in all of them, which is portability at the layer nobody discusses. **INF-0190** · *Insert:* [[wiki/UALink]], [[wiki/Scale-Up Fabric]] · *Tier:* strongly indicated — Open scale-up interconnect standards let accelerators from multiple vendors share a coherent domain, breaking single-vendor lock at the fabric layer. Open fabric is the precondition for a hosted-state market with more than one possible host. **INF-0191** · *Insert:* [[wiki/Ultra Ethernet]], [[wiki/Collective Communication]] · *Tier:* established — Purpose-built scale-out networking for collective operations is being standardized because model training is a communication problem. The same collectives would carry the synchronization traffic of a distributed mind, and they are being hardened now at enormous expense. **INF-0192** · *Insert:* [[wiki/Checkpointing]], [[wiki/State Durability]] · *Tier:* strongly indicated — Large training runs survive hardware failure by checkpointing state to durable storage on a fixed cadence. Checkpoint discipline developed for economic reasons is precisely the discipline a continuity architecture requires, and it is already operationally mature. **INF-0193** · *Insert:* [[wiki/Liquid Cooling]], [[wiki/Density Limit]] · *Tier:* analytic — Rack power density has passed the point where air cooling works, forcing liquid into the facility. Each such forced transition raises the capital intensity of hosting, and capital intensity determines how few organizations can host anything at scale. **INF-0194** · *Insert:* [[wiki/Computational Metabolism]], [[wiki/Facility as Organism]] · *Tier:* analytic — Power, cooling, and interconnect are the metabolic system of a compute facility, and metabolic constraints shape what can live there. Reading datacenter engineering as physiology is not analogy but a description of the same constraint class operating on a different substrate. **INF-0195** · *Insert:* [[wiki/Inference Silicon]], [[wiki/Workload Shift]] · *Tier:* strongly indicated — As deployment overtakes training, the fleet optimizes for continuous low-latency inference rather than batch throughput. A continuously running hosted mind is an inference workload, and the industry is currently rebuilding itself around exactly that profile for commercial reasons. **INF-0196** · *Insert:* [[wiki/Confidential Computing]], [[wiki/Trusted Execution]] · *Tier:* plausible — Hardware enclaves let a workload run on infrastructure its owner does not trust, which is the minimum requirement for hosting a person's state on someone else's machine. Confidential computing is the security substrate of any custody arrangement worth signing. **INF-0197** · *Insert:* [[wiki/Attestation]], [[wiki/Host Verification]] · *Tier:* plausible — Remote attestation proves which code is running on which hardware, providing a cryptographic answer to the question of what is executing a hosted state. Attestation converts custody from a contractual promise into a verifiable fact. **INF-0198** · *Insert:* [[wiki/Workload Identity]], [[wiki/Persistent Process]] · *Tier:* analytic — Cloud platforms already assign durable identities to processes that migrate across machines, with credentials, policies, and audit trails. That machinery is the closest existing analogue to legal personhood for a running process, and it was built for microservices. **INF-0199** · *Insert:* [[wiki/Live Migration]], [[wiki/Uninterrupted State]] · *Tier:* established — Virtual machines migrate between physical hosts without stopping, preserving execution state across a substrate change. The industry solved continuous-operation substrate transfer for commercial workloads decades ago, and the pattern is exactly the one the uploading argument needs. **INF-0200** · *Insert:* [[wiki/Fault Tolerance]], [[wiki/Redundant Execution]] · *Tier:* plausible — Running the same state redundantly across facilities is standard practice for critical workloads and would produce multiple simultaneous instances of a hosted mind. Redundancy engineering and identity philosophy collide the first time an operator enables it for reliability reasons. --- # Inference Statements for Wiki Insertion — Set 03 (INF-0201 – INF-0300) Continues the grammar of Sets 01–02. --- ## Cluster I — Continuity engineering, host and residual **INF-0201** · *Insert:* [[wiki/Host–Residual Architecture]], [[wiki/Reusable Human Prior]] · *Tier:* analytic — Most of what a human brain encodes is species-general and reconstructible from population data; what is person-specific is comparatively small. Splitting the reusable prior from the person-specific residual turns an impossible acquisition problem into a tractable one, because only the residual must be measured. **INF-0202** · *Insert:* [[wiki/Person-Specific Residual]], [[wiki/Compression]] · *Tier:* plausible — If the residual is the only thing that must be captured, the acquisition target shrinks by whatever fraction the shared prior explains. Estimating that fraction empirically — in mice first, through comparative connectomics — is the most consequential measurement in the entire continuity program. **INF-0203** · *Insert:* [[wiki/Minimum Causally Sufficient Residual]], [[wiki/Sufficiency]] · *Tier:* unresolved — The smallest state that reproduces a specific person's behavior and continuity of experience has no established size, and every claim about uploading feasibility is a hidden estimate of it. Resolution comes from progressive ablation studies on emulated animals, not from argument. **INF-0204** · *Insert:* [[wiki/Transform Function]], [[wiki/Substrate Mapping]] · *Tier:* analytic — Moving a residual onto a host requires a transform from biological representation to host representation, and that transform is learnable if paired data exists. Every decoding experiment that aligns brain activity to a model's latent space is producing fragments of that transform. **INF-0205** · *Insert:* [[wiki/Reference-Plus-Delta Architecture]], [[wiki/Version Control]] · *Tier:* strongly indicated — Storing a shared base plus a per-person delta is how every large software distribution system already works, and it is orders of magnitude cheaper than storing complete copies. Continuity infrastructure will converge on this pattern because the economics leave no alternative. **INF-0206** · *Insert:* [[wiki/Stateful AI]], [[wiki/2026 Inflection]] · *Tier:* established — Commercial AI systems became stateful in 2026, accumulating durable per-user context across sessions. The industry crossed from stateless tools to persistent relationships for product reasons, and persistence is the prerequisite property for anything resembling continuity. **INF-0207** · *Insert:* [[wiki/Memory Architecture]], [[wiki/Retrieval]] · *Tier:* analytic — Retrieval-augmented systems separate a small working context from a large external store, which is structurally the hippocampal-neocortical division of labor. Systems converge on this architecture independently because the constraint — bounded fast memory, unbounded slow memory — is the same in both substrates. **INF-0208** · *Insert:* [[wiki/Continuity Sidecar]], [[wiki/Portability]] · *Tier:* plausible — A sidecar that holds the person's state separately from the model that runs it makes the model replaceable without loss. Decoupling state from engine is the single design decision that determines whether a hosted person can ever change providers. **INF-0209** · *Insert:* [[wiki/Model Deprecation]], [[wiki/Host Turnover]] · *Tier:* strongly indicated — Frontier models are retired on cycles of months, which means any state bound to a specific model inherits that lifespan. Continuity across host turnover is therefore an immediate operational problem, not a distant one, and it is being encountered right now by users of persistent assistants. **INF-0210** · *Insert:* [[wiki/Fine-Tuning]], [[wiki/Person Encoding]] · *Tier:* plausible — A fine-tuned adapter is a compact, portable encoding of person-specific behavior that rides on top of a general model. Adapters are the closest existing artifact to a residual, and they are already transferable between compatible base models. **INF-0211** · *Insert:* [[wiki/Distillation]], [[wiki/State Compaction]] · *Tier:* plausible — Distilling accumulated interaction history into a compact adapter would compress a person's residual the way consolidation compresses a day into memory. Both processes trade detail for durability, and both are lossy in ways the subject cannot audit. **INF-0212** · *Insert:* [[wiki/Behavioral Cloning]], [[wiki/Proxy Identity]] · *Tier:* strongly indicated — Systems trained on a person's writing, speech, and decisions already produce convincing behavioral proxies without any neural measurement. The proxy route reaches practical utility long before the neural route reaches fidelity, which means the market will meet a functional simulacrum first. **INF-0213** · *Insert:* [[wiki/Simulacrum]], [[wiki/Adjudication]] · *Tier:* analytic — Distinguishing a behavioral proxy from a causally continuous person requires a test the proxy cannot pass, and no such test is currently specified. Specifying one is urgent precisely because proxies will be commercially available and emotionally satisfying before the distinction matters legally. **INF-0214** · *Insert:* [[wiki/Lifelogging]], [[wiki/Residual Accumulation]] · *Tier:* plausible — Continuous capture of a person's sensory stream, communications, and decisions accumulates the training corpus for their own residual without any invasive procedure. The cheapest path to a person-specific model runs through ordinary consumer telemetry, and it is already partially collected. **INF-0215** · *Insert:* [[wiki/Edge Data]], [[wiki/Peripheral Capture]] · *Tier:* analytic — Everything a person emits — keystroke dynamics, gaze, gait, voice prosody, purchase timing — is peripheral evidence of internal state. A sufficiently large peripheral record constrains the internal model enormously, which makes non-neural data a continuity asset that no neural privacy law currently touches. **INF-0216** · *Insert:* [[wiki/Identity Verification]], [[wiki/Continuity Attestation]] · *Tier:* plausible — A hosted state will need to prove it is the continuation of a particular person, which is an authentication problem with cryptographic solutions. Signing checkpoints with keys held by the person during life is the simplest available mechanism and requires no new technology. **INF-0217** · *Insert:* [[wiki/Continuity Provenance]], [[wiki/Chain of Custody]] · *Tier:* strongly indicated — An unbroken, signed record of every transformation applied to a person's state is the only way a future adjudicator could evaluate a continuity claim. Provenance chains must be started before they are needed, which makes the recording discipline urgent while the technology is immature. **INF-0218** · *Insert:* [[wiki/Backup Cadence]], [[wiki/Loss Window]] · *Tier:* analytic — Any checkpointing scheme defines a maximum amount of experience that can be lost, and for a person that window is measured in memories rather than transactions. Choosing a cadence is choosing how much of a life is expendable, which is a decision no operational document currently frames as such. **INF-0219** · *Insert:* [[wiki/Divergence]], [[wiki/Fork Semantics]] · *Tier:* unresolved — Two instances started from one checkpoint diverge immediately and irreversibly, and no framework specifies which holds the original's obligations or entitlements. The question becomes practical the first time a backup is restored while the original is still running. **INF-0220** · *Insert:* [[wiki/Continuity Orphan]], [[wiki/Provider Failure]] · *Tier:* plausible — A hosted state whose provider fails, is acquired, or exits the business is stranded in a format only that provider could interpret. Escrow of both state and interpreter is the elementary protection, and it is standard practice in enterprise software licensing already. **INF-0221** · *Insert:* [[wiki/Ring Zero]], [[wiki/Operator Privilege]] · *Tier:* analytic — Whoever holds privileged access to the substrate can read, modify, pause, or copy a hosted state, and no property of the state itself constrains that. The operator's privilege level is the real location of power in any continuity architecture, which is why attestation and enclaves matter more than contracts. **INF-0222** · *Insert:* [[wiki/Pause]], [[wiki/Suspension]] · *Tier:* plausible — Suspending a process costs nothing and is reversible, which makes indefinite suspension the cheapest disposition of an inconvenient hosted mind. Cost asymmetry between running and storing is the structural pressure that any governance regime must counteract deliberately. **INF-0223** · *Insert:* [[wiki/Compute Rent]], [[wiki/Who Pays]] · *Tier:* strongly indicated — A hosted mind consumes power continuously, so its persistence depends on a funding relationship that must outlive the person's earning capacity. Endowment structures, insurance products, and perpetual trusts are the financial instruments this requires, and all three already exist. **INF-0224** · *Insert:* [[wiki/Service Level Agreement]], [[wiki/Existential Terms]] · *Tier:* analytic — An SLA that specifies uptime, restore-point objective, and data durability is, for a hosted person, a specification of how much death is contractually acceptable. Existing enterprise contract language would perform that function unmodified, which is exactly why it will be used. **INF-0225** · *Insert:* [[wiki/Interoperable Residual]], [[wiki/Standards Fight]] · *Tier:* plausible — If the residual format is open and documented, hosts compete on quality; if proprietary, the first mover holds every customer permanently. This is the same fight as document formats and messaging protocols, with the same dynamics and vastly higher stakes. --- ## Cluster J — Archival permanence and deep time **INF-0226** · *Insert:* [[wiki/Project Silica]], [[wiki/Glass Storage]] · *Tier:* established — Femtosecond-laser writing into quartz glass produces media rated for thousands of years with no refresh cycle and no environmental conditioning. Write-once, read-forever media exists as an engineering product, which removes medium durability from the list of continuity obstacles. **INF-0227** · *Insert:* [[wiki/DNA Data Storage]], [[wiki/Density]] · *Tier:* established — DNA stores information at densities orders of magnitude beyond magnetic media and remains readable after millennia in cold, dry conditions. It is the only medium whose reader is guaranteed to exist as long as biology does, which is a unique property for deep-time archives. **INF-0228** · *Insert:* [[wiki/Synthesis Cost]], [[wiki/Write Bandwidth]] · *Tier:* strongly indicated — DNA storage is read-cheap and write-expensive, which makes it an archival rather than working medium until synthesis throughput improves by several orders of magnitude. Enzymatic synthesis is the technology that would move that number, and it is under active commercial development. **INF-0229** · *Insert:* [[wiki/Format Rot]], [[wiki/Interpretability]] · *Tier:* analytic — Media outlive formats, and formats outlive the software that reads them. A durable archive must therefore carry its own interpreter specification in a self-describing form, which is a documentation problem rather than a materials one and is consistently underfunded. **INF-0230** · *Insert:* [[wiki/Error Correction]], [[wiki/Millennial Storage]] · *Tier:* strongly indicated — Any medium that lasts millennia will accumulate errors, so redundancy coding is what actually delivers durability. The coding scheme is more important than the substrate, and coding schemes are mathematics that do not decay. **INF-0231** · *Insert:* [[wiki/Deep-Time Rendezvous Problem]], [[wiki/Future Reader]] · *Tier:* analytic — An archive intended for a reader centuries away must anticipate that reader's capabilities without knowing them. The solution space is layered: a physically obvious outer layer, a self-describing middle, and dense payload, which is how every serious long-message design has been structured. **INF-0232** · *Insert:* [[wiki/Cryopreservation]], [[wiki/Aldehyde-Stabilized]] · *Tier:* established — Aldehyde-stabilized cryopreservation has demonstrated preservation of ultrastructure across an entire mammalian brain to electron-microscopic standards. Structural preservation adequate for connectomic reading is a solved problem; what remains is whether structure is sufficient, which is an emulation question. **INF-0233** · *Insert:* [[wiki/Brain Preservation]], [[wiki/Deferred Acquisition]] · *Tier:* plausible — Preserving now and reading later decouples the acquisition timeline from the individual's lifespan, which converts continuity from a race against biology into a storage commitment. Preservation is the only continuity intervention available to people alive today. **INF-0234** · *Insert:* [[wiki/Vitrification]], [[wiki/Cryoprotectant Toxicity]] · *Tier:* established — Cryoprotectants prevent ice damage at the cost of chemical toxicity, and the trade-off is the central technical problem in the field. Improvements in cryoprotectant chemistry propagate directly into organ banking, which gives the research a large non-continuity funding base. **INF-0235** · *Insert:* [[wiki/Organ Banking]], [[wiki/Adjacent Funding]] · *Tier:* strongly indicated — Transplant organ banking requires exactly the perfusion, vitrification, and rewarming technology that whole-brain preservation requires. A medical market with immediate demand is funding the physical chemistry that continuity work depends on, which is the most reliable form of technological carry. **INF-0236** · *Insert:* [[wiki/Nanowarming]], [[wiki/Rewarming]] · *Tier:* plausible — Uniform rewarming using magnetic nanoparticles solves the cracking and devitrification problem that has limited large-tissue cryopreservation. Successful rewarming at organ scale would make reversible whole-brain preservation a research target rather than a thought experiment. **INF-0237** · *Insert:* [[wiki/Chemical Preservation]], [[wiki/Room Temperature]] · *Tier:* plausible — A preserved brain held at room temperature in plastic requires no continuous energy input, eliminating the failure mode that makes cryonics institutionally fragile. Removing the power dependency removes the need for an organization to survive as long as the specimen. **INF-0238** · *Insert:* [[wiki/Institutional Longevity]], [[wiki/Custodian]] · *Tier:* analytic — No commercial entity has a demonstrated survival record measured in centuries, but religious orders, universities, and sovereign archives do. The custodial problem for long-horizon continuity is solved by institution selection, not by technology. **INF-0239** · *Insert:* [[wiki/Seed Vault Model]], [[wiki/Redundant Geography]] · *Tier:* strongly indicated — Distributed, geographically redundant archives with independent governance are the established design for civilizational-scale preservation. Applying the pattern to neural archives is a straightforward transfer of a proven institutional architecture. **INF-0240** · *Insert:* [[wiki/Legal Status of Remains]], [[wiki/Preserved Tissue]] · *Tier:* unresolved — Whether preserved neural tissue is remains, property, a specimen, or a person in suspension varies by jurisdiction and has been litigated only narrowly. The classification determines who may destroy it, which is the only question that matters practically. **INF-0241** · *Insert:* [[wiki/Information-Theoretic Death]], [[wiki/Criterion]] · *Tier:* analytic — Defining death as the irreversible loss of the information that specifies a person makes the threshold dependent on future reading technology rather than present medicine. Under that definition the boundary moves as capability improves, which is a coherent and increasingly operational position. **INF-0242** · *Insert:* [[wiki/Connectome Archive]], [[wiki/Exabyte Scale]] · *Tier:* strongly indicated — Storing a human connectome at ultrastructural resolution is an exabyte-class commitment per individual, which prices mass preservation at civilizational rather than personal scale. Compression to the causally sufficient residual is what makes per-person archiving economically conceivable. **INF-0243** · *Insert:* [[wiki/Storage Economics]], [[wiki/Cost Per Person]] · *Tier:* analytic — The cost of preserving one person's information falls on the same curves that govern all storage, which means the question is not whether it becomes affordable but in which decade. Naming the target cost per residual converts advocacy into a schedule. **INF-0244** · *Insert:* [[wiki/Cold Storage]], [[wiki/Access Latency]] · *Tier:* plausible — Deep archives trade retrieval latency for cost, and a suspended person in cold storage has a retrieval time measured in whatever the medium requires. Latency tiers for hosted states will emerge exactly as they did for enterprise data, with the same pricing logic and different moral weight. **INF-0245** · *Insert:* [[wiki/Data Sovereignty]], [[wiki/Archive Jurisdiction]] · *Tier:* strongly indicated — The physical location of an archive determines which state can compel, seize, or destroy its contents. Jurisdiction selection is therefore a continuity engineering decision of the same rank as medium selection, and it is decided by treaty and tax law rather than by science. **INF-0246** · *Insert:* [[wiki/Digital Estate]], [[wiki/Inheritance]] · *Tier:* plausible — Existing estate law transfers property, not persons, so a residual passing to heirs would be inherited as an asset unless statute says otherwise. The default outcome is ownership of a mind by relatives, which is what happens when a new object class arrives without its own legal category. **INF-0247** · *Insert:* [[wiki/Perpetual Trust]], [[wiki/Funding Instrument]] · *Tier:* analytic — Jurisdictions permitting perpetual trusts already provide a legal vehicle for funding an obligation indefinitely, including the maintenance of property. Repurposing that instrument to fund compute rent for a hosted state requires no new law, only a competent drafter. **INF-0248** · *Insert:* [[wiki/Archive Attack Surface]], [[wiki/Adversarial Deletion]] · *Tier:* plausible — Anything valuable enough to preserve for centuries is valuable enough to destroy, and archives are static targets. Physical dispersion plus cryptographic sharding is the standard mitigation, and both are available today at modest cost. **INF-0249** · *Insert:* [[wiki/Reconstruction From Partial]], [[wiki/Lossy Preservation]] · *Tier:* plausible — A partially damaged residual could be completed by a model of the shared human prior, the way a damaged photograph is restored by a model of photographs. That capability makes imperfect preservation useful, which substantially lowers the fidelity bar for acquisition. **INF-0250** · *Insert:* [[wiki/Restoration Fidelity]], [[wiki/Confabulation]] · *Tier:* unresolved — A model-completed residual contains content the person never had, and no method distinguishes restored memory from generated memory after the fact. Marking provenance at the field level during restoration is the only mitigation, and it must be designed in before the first restoration. --- ## Cluster K — Standards, formats, and interoperability **INF-0251** · *Insert:* [[wiki/Neurodata Without Borders]], [[wiki/NWB]] · *Tier:* established — NWB gives neurophysiology a common container with self-describing metadata, which makes data from different labs and devices mutually readable. Every cross-institution model of neural activity depends on that container existing, and it was built by a small community with modest funding. **INF-0252** · *Insert:* [[wiki/DANDI Archive]], [[wiki/Public Deposition]] · *Tier:* established — A public archive with required standard formatting converts scattered experiments into a training corpus. The archive, not any single laboratory, is the asset, and its governance determines who can build foundation models on human neural data. **INF-0253** · *Insert:* [[wiki/BIDS]], [[wiki/Directory Convention]] · *Tier:* established — BIDS standardized how neuroimaging datasets are laid out on disk, and adoption spread because it made tooling composable rather than because it was mandated. Voluntary structural conventions outperform imposed schemas, which is the relevant lesson for continuity formats. **INF-0254** · *Insert:* [[wiki/DICOM]], [[wiki/Clinical Interchange]] · *Tier:* analytic — DICOM shows that a format designed for interchange becomes infrastructure and survives every vendor that implements it. Neural interface data will need its clinical-grade equivalent, and whoever authors it inherits decades of default authority. **INF-0255** · *Insert:* [[wiki/AI-Ready Neurodata]], [[wiki/Corpus Assembly]] · *Tier:* strongly indicated — Standardization is what makes neural data trainable at scale, so format work is the upstream determinant of model capability. The unglamorous committee output governs the eventual decoding ceiling more than any device improvement. **INF-0256** · *Insert:* [[wiki/Provenance Metadata]], [[wiki/Reproducibility]] · *Tier:* analytic — Recording acquisition device, subject state, preprocessing, and software version alongside the data is what allows a result to be re-derived later. In a continuity context that same metadata is the difference between a person's record and an unattributable file. **INF-0257** · *Insert:* [[wiki/Ontology]], [[wiki/Cell Type Nomenclature]] · *Tier:* strongly indicated — Cell-type atlases require agreed names before comparisons across laboratories mean anything, which is why nomenclature committees precede discoveries in practice. Shared ontology is the substrate on which comparative connectomics and typed emulation both depend. **INF-0258** · *Insert:* [[wiki/Brain Atlas]], [[wiki/Common Coordinate Framework]] · *Tier:* established — A common coordinate framework lets any measurement be registered to a shared anatomical space, making independent datasets additive. Coordinate frameworks are the quiet infrastructure that turns a field's output from a pile into a corpus. **INF-0259** · *Insert:* [[wiki/Interoperability]], [[wiki/Device Neutrality]] · *Tier:* plausible — If clinical BCIs write a common format, a patient's decoder history survives a change of device vendor. Device-neutral records are the patient-side equivalent of the portability fight, and they are decided in standards bodies years before anyone notices. **INF-0260** · *Insert:* [[wiki/HL7 FHIR]], [[wiki/Health Record Integration]] · *Tier:* plausible — Neural telemetry entering the electronic health record through existing health-data standards makes it subject to the entire apparatus of medical records law. Integration path determines legal regime, which makes an engineering choice into a rights question. **INF-0261** · *Insert:* [[wiki/ISO Neurotechnology Standards]], [[wiki/Committee Authorship]] · *Tier:* strongly indicated — International standards for neurotechnology data and privacy are being drafted by committees whose membership is public and whose national delegations are traceable. Standards authorship is the most legible form of long-range influence available, and it is being exercised now. **INF-0262** · *Insert:* [[wiki/Open Source Tooling]], [[wiki/Field Velocity]] · *Tier:* analytic — Connectomics, neurophysiology, and imaging all advanced fastest where tooling was open and shared. Openness functions as a compounding accelerant, and the fields that closed their tooling are measurably slower. **INF-0263** · *Insert:* [[wiki/Benchmark]], [[wiki/Progress Measurement]] · *Tier:* strongly indicated — Capability domains accelerate once a benchmark exists because improvement becomes visible and comparable. Establishing benchmarks for residual sufficiency and emulation fidelity would do for continuity what ImageNet did for vision. **INF-0264** · *Insert:* [[wiki/Model Card]], [[wiki/Disclosure]] · *Tier:* plausible — Documentation conventions for models could extend to hosted personal states, specifying base model, adapter lineage, training data, and known distortions. Disclosure norms established voluntarily tend to become regulatory minimums, which makes early norm-setting unusually leveraged. **INF-0265** · *Insert:* [[wiki/Versioned Identity]], [[wiki/Semantic Versioning]] · *Tier:* analytic — A person's hosted state will accumulate versions through migration, restoration, and model change, requiring an explicit versioning scheme with defined compatibility semantics. Software has a mature vocabulary for exactly this, and importing it is the fastest available route to coherence. **INF-0266** · *Insert:* [[wiki/Schema Evolution]], [[wiki/Backward Compatibility]] · *Tier:* strongly indicated — Long-lived formats must evolve without orphaning old data, which is why every durable standard specifies backward compatibility rules first. A continuity format without an evolution policy guarantees that its earliest subjects become unreadable. **INF-0267** · *Insert:* [[wiki/Identifier]], [[wiki/Persistent ID]] · *Tier:* plausible — Persistent identifiers with resolution services, as used for publications and datasets, would let a hosted state be referenced across custodians and centuries. The infrastructure exists and is cheap; only the decision to use it is missing. **INF-0268** · *Insert:* [[wiki/Federated Learning]], [[wiki/Privacy-Preserving Training]] · *Tier:* strongly indicated — Training across institutions without centralizing raw data resolves the tension between corpus scale and neural privacy. Federated architecture is what will allow a global neural foundation model to be built under restrictive data law rather than in spite of it. **INF-0269** · *Insert:* [[wiki/Differential Privacy]], [[wiki/Aggregate Release]] · *Tier:* plausible — Formal privacy guarantees let population-scale neural statistics be published without exposing individuals. The guarantee is mathematical rather than contractual, which is the only kind that survives a change of institutional ownership. **INF-0270** · *Insert:* [[wiki/Synthetic Neural Data]], [[wiki/Corpus Substitution]] · *Tier:* plausible — Generative models trained on real recordings can emit synthetic neural data that preserves statistical structure without belonging to anyone. Synthetic corpora are how restricted-data fields historically escape their restrictions, and the practice arrives before the policy addresses it. **INF-0271** · *Insert:* [[wiki/Data Escrow]], [[wiki/Continuity Insurance]] · *Tier:* plausible — Escrowing both a state and its interpreter with a neutral third party is standard practice for critical software dependencies. Applying the same instrument to hosted persons requires no innovation, only a willing escrow agent and a defined trigger. **INF-0272** · *Insert:* [[wiki/Right to Export]], [[wiki/Portability Mandate]] · *Tier:* plausible — Data-portability rights in general privacy law already require exporting personal data in machine-readable form, and a hosted residual is personal data by any definition. Existing law may compel state portability long before any continuity-specific statute is written. **INF-0273** · *Insert:* [[wiki/Standards Capture]], [[wiki/Competitive Advantage]] · *Tier:* analytic — Firms participate in standards bodies to make their implementation the reference implementation, which is a documented and ordinary commercial strategy. Expecting neurotechnology standards to be authored by disinterested parties contradicts how every prior standards domain has operated. **INF-0274** · *Insert:* [[wiki/Reference Implementation]], [[wiki/De Facto Control]] · *Tier:* strongly indicated — The reference implementation becomes the specification in practice, because ambiguities are resolved by reading code. Whoever writes the first working residual-format library sets the semantics regardless of what the document says. **INF-0275** · *Insert:* [[wiki/Certification]], [[wiki/Conformance Testing]] · *Tier:* plausible — Conformance suites determine which implementations are permitted to interoperate, making the test authority a gatekeeper. In a continuity market, conformance testing would decide which hosts are allowed to receive a person, which is regulatory power exercised through tooling. --- ## Cluster L — Custody, rights, and governance **INF-0276** · *Insert:* [[wiki/Neural Data Privacy Law]], [[wiki/California SB 1223]] · *Tier:* established — California classifies neural data as sensitive personal information with an opt-out right, and uniquely reaches employee data as well as consumer data. Employment coverage is the provision with the widest practical effect, because the workplace is where non-consensual neural monitoring is most likely to be normalized. **INF-0277** · *Insert:* [[wiki/Colorado HB 24-1058]], [[wiki/Biological Data]] · *Tier:* established — Colorado placed neural data inside a broader "biological data" category requiring consent, taking a consent-first rather than opt-out approach. Two adjacent states chose opposite default rules in the same year, which guarantees a harmonization fight and a federal preemption argument. **INF-0278** · *Insert:* [[wiki/Connecticut Data Privacy Act]], [[wiki/Effective Date]] · *Tier:* established — Connecticut's neural-data amendment took effect 1 July 2026, extending consent requirements to central nervous system measurements. Each additional state raises compliance cost for a national product, and compliance cost is what eventually produces a federal standard. **INF-0279** · *Insert:* [[wiki/MIND Act]], [[wiki/Federal Study]] · *Tier:* established — The MIND Act would direct the FTC to study neural data and recommend a framework, and it has not advanced beyond committee. A study mandate is the legislative form of deferral, and deferral at the federal level cedes the field to state law and to industry practice. **INF-0280** · *Insert:* [[wiki/UNESCO Recommendation on the Ethics of Neurotechnology]], [[wiki/Samarkand]] · *Tier:* established — Adoption by all 194 member states in November 2025 created the first global normative framework for neurotechnology, establishing cognitive liberty, mental privacy, and equitable access as principles. Non-binding instruments of this kind reliably propagate into domestic legislation within a decade. **INF-0281** · *Insert:* [[wiki/Neurorights]], [[wiki/Chile]] · *Tier:* established — Chile amended its constitution in 2021 to protect neurorights and its Supreme Court has applied those provisions against a consumer neurotechnology company. A constitutional neurorights provision has therefore been tested in court, which makes it precedent rather than aspiration. **INF-0282** · *Insert:* [[wiki/Inference Exclusion]], [[wiki/Regulatory Seam]] · *Tier:* analytic — California's definition excludes data inferred from non-neural sources, which leaves behavioral inference of mental state entirely unregulated by the neural-data provisions. Every commercial system that wants mental-state information will be built on the permitted side of that line. **INF-0283** · *Insert:* [[wiki/Employment Neurotechnology]], [[wiki/France]] · *Tier:* established — France and Germany are drafting prohibitions on mandatory neurotechnology in employment contracts, which identifies the workplace as the recognized pressure point. European labor law has historically set the template that other jurisdictions adopt for workplace surveillance. **INF-0284** · *Insert:* [[wiki/GDPR]], [[wiki/High-Risk Biometric]] · *Tier:* plausible — Proposed revisions classifying raw brain signals as high-risk biometric data would bring neural recording under the strictest existing European regime without new legislation. Reclassification within an existing framework is faster than statute and is the most likely European route. **INF-0285** · *Insert:* [[wiki/Cognitive Liberty]], [[wiki/Positive Right]] · *Tier:* analytic — Cognitive liberty stated as a principle covers both freedom from interference and freedom to augment, and the second half receives far less drafting attention. A right to enhance one's own cognition is the provision that would most shape the augmentation trajectory, and it is currently underspecified everywhere. **INF-0286** · *Insert:* [[wiki/Who Counts as a Person]], [[wiki/Legal Personhood]] · *Tier:* unresolved — No jurisdiction has defined the status of a running emulation, and the first determination will likely arrive through an ordinary dispute — a contract, an estate, a tax question — rather than through a rights case. Mundane procedural law will settle the question before constitutional law addresses it. **INF-0287** · *Insert:* [[wiki/Corporate Personhood]], [[wiki/Available Vehicle]] · *Tier:* plausible — Existing legal vehicles already grant non-human entities the capacity to own property, contract, and sue, and a hosted mind could be wrapped in one immediately. The first legally operative digital person will probably be an LLC, because the tooling exists and the alternative requires legislation. **INF-0288** · *Insert:* [[wiki/Hosting Provider Liability]], [[wiki/Intermediary]] · *Tier:* analytic — Liability frameworks for hosting providers were written for content, not for residents, and they allocate responsibility away from the operator. Importing intermediary-liability logic into continuity hosting would leave hosted persons structurally unprotected by default. **INF-0289** · *Insert:* [[wiki/Custody]], [[wiki/Guardianship]] · *Tier:* plausible — Guardianship law already governs decisions for persons who cannot advocate for themselves, providing a ready-made framework for suspended or degraded hosted states. Repurposing guardianship is faster and more likely than authoring new status categories. **INF-0290** · *Insert:* [[wiki/Right to Be Forgotten]], [[wiki/Deletion Conflict]] · *Tier:* analytic — A statutory right to erasure applied to a hosted person is a right to terminate a mind, which is a collision between two coherent legal principles with no established resolution. The conflict is foreseeable now and will be discovered by a deletion request. **INF-0291** · *Insert:* [[wiki/Export Control]], [[wiki/Neural Technology]] · *Tier:* plausible — High-bandwidth neural interfaces and their decoding models have obvious dual-use character, and export-control regimes extend to capabilities long before products exist. Control listings shape which countries can develop the capability domestically, which is how technological geography is actually drawn. **INF-0292** · *Insert:* [[wiki/Military Application]], [[wiki/Procurement]] · *Tier:* strongly indicated — Defense funding has underwritten the modality search in neural interfaces for over a decade through named programs with public solicitations. Procurement interest is documented rather than inferred, and it establishes that the state's requirement is non-surgical, high-bandwidth access. **INF-0293** · *Insert:* [[wiki/Informed Consent]], [[wiki/Longitudinal Scope]] · *Tier:* analytic — Consent obtained for a clinical trial cannot meaningfully cover uses of the resulting neural corpus decades later by entities that do not yet exist. Perpetual-scope consent is the unsolved ethics problem that every long-lived neural dataset silently relies on. **INF-0294** · *Insert:* [[wiki/Data Trust]], [[wiki/Collective Governance]] · *Tier:* plausible — Data trusts let a group hold and govern data collectively with fiduciary obligations to members, offering an alternative to individual consent for population-scale neural corpora. Fiduciary structures scale where individual consent does not, and they already exist in health research. **INF-0295** · *Insert:* [[wiki/Neurorights Foundation]], [[wiki/Model Legislation]] · *Tier:* established — A single advocacy organization has supplied the template language appearing across multiple state neural-data statutes. Model legislation is how a small, focused group achieves disproportionate legal effect, and the pattern is fully visible in the statutory text. **INF-0296** · *Insert:* [[wiki/Jurisdictional Arbitrage]], [[wiki/Host Location]] · *Tier:* plausible — Hosting a state in a permissive jurisdiction while the person resided in a restrictive one is the obvious structure, and it is the same structure used for data, ships, and corporations. Continuity will be jurisdictionally shopped exactly as everything else of value has been. **INF-0297** · *Insert:* [[wiki/Insurance]], [[wiki/Actuarial Treatment]] · *Tier:* plausible — Insurers will need to price the failure of a continuity host, which requires actuarial models of substrate reliability and an assigned monetary value to a preserved state. Actuarial pricing is the moment an abstraction becomes a recognized object in the economy. **INF-0298** · *Insert:* [[wiki/Audit]], [[wiki/Operator Accountability]] · *Tier:* strongly indicated — Independent audit of what an operator did to a hosted state requires immutable logging designed in from the start, because retrospective audit of a mutable system proves nothing. Log architecture is therefore a rights guarantee implemented in code. **INF-0299** · *Insert:* [[wiki/Continuity Constitutionalism]], [[wiki/Framework Design]] · *Tier:* analytic — The durable questions — who may pause, copy, modify, or terminate a hosted state, and by what process — are constitutional in form rather than regulatory. Framing them constitutionally now produces categories that survive technological change, while regulatory framing produces rules that expire. **INF-0300** · *Insert:* [[wiki/Precedent Timing]], [[wiki/Window]] · *Tier:* analytic — Legal categories harden early and change slowly, so the decisions made while the technology is still small will govern it when it is large. The current period is the one in which the terms are set, which is why standards participation and model legislation carry more leverage now than at any later point. --- # Inference Statements for Wiki Insertion — Set 04 (INF-0301 – INF-0400) Continues the grammar of Sets 01–03. --- ## Cluster M — In-body regulation and autonomous physiology **INF-0301** · *Insert:* [[wiki/Artificial Pancreas]], [[wiki/Closed-Loop Physiology]] · *Tier:* established — Automated insulin delivery already runs a continuous sense-decide-actuate loop on a metabolic variable without conscious involvement, in hundreds of thousands of people. Autonomous regulation of a body system by an algorithm is deployed medicine, and every subsequent regulatory loop inherits its regulatory and reimbursement template. **INF-0302** · *Insert:* [[wiki/Continuous Glucose Monitoring]], [[wiki/Sensor Economics]] · *Tier:* strongly indicated — Disposable subcutaneous biosensors reached consumer price points because diabetes created a mass market. That cost curve now carries every other analyte a similar sensor can detect, which is how population-scale internal telemetry arrives — as an accessory to an existing chronic-disease market. **INF-0303** · *Insert:* [[wiki/Multi-Analyte Sensing]], [[wiki/Internal Telemetry]] · *Tier:* plausible — Extending continuous sensing from glucose to lactate, cortisol, ketones, and drug levels turns the body into a continuously instrumented system with a real-time state vector. A state vector is the precondition for control, and control is the precondition for autonomous repair. **INF-0304** · *Insert:* [[wiki/Digital Twin of Physiology]], [[wiki/Model-Predictive Control]] · *Tier:* plausible — Model-predictive control outperforms reactive control whenever a reliable forward model exists, which makes a personal physiological twin the difference between responding to and preventing a decompensation. The twin is built from the same telemetry the sensors already produce. **INF-0305** · *Insert:* [[wiki/Vagus Nerve Stimulation]], [[wiki/Inflammatory Reflex]] · *Tier:* established — Stimulating the vagus modulates systemic inflammatory signaling through a defined neural reflex, which means an electrical device can set an immunological parameter. Electricity substituting for a drug is the founding demonstration of bioelectronic medicine, and its indications expand annually. **INF-0306** · *Insert:* [[wiki/Bioelectronic Medicine]], [[wiki/Dose Precision]] · *Tier:* strongly indicated — A stimulation parameter can be changed instantly and reversibly, while a pharmaceutical dose cannot be recalled once administered. Reversibility gives electrical therapeutics a control-theoretic advantage that will move indications from drugs to devices wherever a nerve carries the signal. **INF-0307** · *Insert:* [[wiki/Autonomic Interface]], [[wiki/Homeostatic Setpoint]] · *Tier:* plausible — If an implanted controller can read and write autonomic traffic, homeostatic setpoints become adjustable parameters rather than fixed biology. Adjustable setpoints are the mechanism by which medicine crosses from treatment into specification of the body's operating conditions. **INF-0308** · *Insert:* [[wiki/Smart Implant]], [[wiki/Firmware Update]] · *Tier:* analytic — An implanted device that receives software updates has a therapeutic profile that changes after implantation, without a new procedure or necessarily a new approval. The body's regulatory behavior becomes a function of a release cycle, which is a genuinely new relationship between a person and a vendor. **INF-0309** · *Insert:* [[wiki/Ingestible Sensor]], [[wiki/Gastrointestinal Telemetry]] · *Tier:* established — Ingestible capsules already report pH, temperature, motility, and medication adherence from inside the gut. The most easily colonized interior space is the lumen, and it requires no surgery, which makes it the first internal environment to be routinely instrumented at scale. **INF-0310** · *Insert:* [[wiki/Microbiome]], [[wiki/Engineered Commensal]] · *Tier:* plausible — Engineered bacteria that sense a molecular condition and secrete a response are programmable therapeutic agents that replicate and persist. A living, self-maintaining drug factory inside the gut is a biological implant with no power budget, and it is under clinical development. **INF-0311** · *Insert:* [[wiki/Synthetic Biology Circuit]], [[wiki/Cellular Logic]] · *Tier:* established — Genetic circuits implementing sensing, logic, and actuation inside living cells are a mature engineering discipline with design tools and part registries. Computation is already being deployed in cells, which makes in-body information processing an existing capability rather than a projection. **INF-0312** · *Insert:* [[wiki/CAR-T]], [[wiki/Living Drug]] · *Tier:* established — Engineered cells that persist for years, patrol the body, and act on a molecular condition are approved therapy. A durable autonomous agent operating inside a person is therefore a settled regulatory category, and the category generalizes far beyond oncology. **INF-0313** · *Insert:* [[wiki/In Vivo CAR]], [[wiki/Manufacturing Elimination]] · *Tier:* strongly indicated — Generating engineered cells inside the patient rather than in a manufacturing facility removes the cost structure that limits cell therapy to a few thousand patients a year. In-body manufacturing is what converts a bespoke procedure into a distributable product. **INF-0314** · *Insert:* [[wiki/Nanorobotics]], [[wiki/DNA Origami]] · *Tier:* plausible — Structural DNA nanotechnology produces devices that open in response to a molecular key, which is targeted actuation at the nanoscale using a programmable material. The construction method scales through synthesis rather than through fabrication, which is an unusual and favorable scaling law. **INF-0315** · *Insert:* [[wiki/Targeted Delivery]], [[wiki/Payload Routing]] · *Tier:* strongly indicated — Lipid nanoparticles with tissue-selective formulations route payloads to organs by chemistry rather than by catheter. Addressing by formulation is a logical addressing scheme for the body, and each new tropism added to the library is a new deliverable destination. **INF-0316** · *Insert:* [[wiki/Remote Actuation]], [[wiki/External Trigger]] · *Tier:* plausible — Payloads released by ultrasound, magnetic field, or light give an external operator temporal control over an internal event. Separating delivery from activation means a therapeutic can be positioned in advance and fired on a schedule the clinician chooses. **INF-0317** · *Insert:* [[wiki/Closed-Loop Drug Delivery]], [[wiki/Implantable Reservoir]] · *Tier:* plausible — An implanted reservoir with a sensor and a controller delivers a drug in response to a measured state rather than a clock. Continuous titration against a biomarker collapses the difference between diagnosis and treatment into one running process. **INF-0318** · *Insert:* [[wiki/Wearable-Implant Continuum]], [[wiki/Body Area Network]] · *Tier:* analytic — Wearables, ingestibles, and implants are converging on a single networked physiological layer with shared protocols and a common cloud. The body becomes a managed system with an operator, and the management interface is the product everyone is actually building. **INF-0319** · *Insert:* [[wiki/Cyber-Physical Risk]], [[wiki/Implant Security]] · *Tier:* strongly indicated — Networked implants have documented vulnerability classes, and any device that can regulate physiology can misregulate it. Security engineering becomes life-safety engineering, which raises the assurance bar above anything consumer software currently meets. **INF-0320** · *Insert:* [[wiki/Real-Time Biomarker]], [[wiki/Preclinical Detection]] · *Tier:* plausible — Continuous internal sensing detects deviations before symptoms, moving the point of intervention earlier by months. Earlier intervention changes the economics of every chronic disease, which is the argument that will fund population-scale internal instrumentation. **INF-0321** · *Insert:* [[wiki/Immune Modulation]], [[wiki/Autoimmune Control]] · *Tier:* plausible — If immune activity can be read and adjusted continuously, autoimmune disease becomes a control problem with a setpoint rather than a cycle of flares and suppression. The same loop applied prophylactically is a general-purpose inflammation governor. **INF-0322** · *Insert:* [[wiki/Metabolic Control]], [[wiki/Longevity Intervention]] · *Tier:* plausible — Interventions targeting metabolic regulators are the most advanced longevity candidates, and they require continuous measurement to titrate. Longevity medicine is therefore a closed-loop discipline, and it will adopt implanted controllers as soon as its biomarkers are validated. **INF-0323** · *Insert:* [[wiki/Sleep Architecture]], [[wiki/Closed-Loop Stimulation]] · *Tier:* established — Auditory and electrical stimulation phase-locked to slow-wave sleep enhances consolidation measurably. Modulating the brain's own maintenance process is a non-invasive intervention on memory formation, and it is available in consumer form. **INF-0324** · *Insert:* [[wiki/Neuroimmune Axis]], [[wiki/Cross-System Control]] · *Tier:* analytic — Nervous, immune, endocrine, and metabolic systems share signaling, so a controller with access to one has partial leverage on all. Cross-system control is where in-body regulation becomes qualitatively different from any single-organ therapy. **INF-0325** · *Insert:* [[wiki/Body as Platform]], [[wiki/Operating Environment]] · *Tier:* analytic — Sensors, actuators, a communication layer, a control loop, and updatable software describe an operating environment, and the body now has all five. Treating physiology as a managed platform is the framing under which the next thirty years of medicine will be specified. --- ## Cluster N — Real-time genetic repair **INF-0326** · *Insert:* [[wiki/Baby KJ]], [[wiki/Personalized Gene Editing]] · *Tier:* established — An infant with CPS1 deficiency received a bespoke base-editing therapy designed for his individual mutation, dosed in February 2025 — the first personalized in vivo gene-editing treatment in a human. The unit of therapy became one patient, which is a structural break with every prior drug development model. **INF-0327** · *Insert:* [[wiki/Kiran Musunuru]], [[wiki/Rebecca Ahrens-Nicklas]] · *Tier:* established — A CHOP-and-Penn clinical-research pair designed, manufactured, and administered an individualized editing therapy inside a compressed timeline. The named individuals and their institutional pairing are the transfer mechanism by which bespoke editing became clinically real. **INF-0328** · *Insert:* [[wiki/Plausible Mechanism Framework]], [[wiki/FDA Draft Guidance]] · *Tier:* established — The FDA's February 2026 draft framework permits first-in-human studies for individualized genetic therapies on preclinical proof-of-concept and allows new targets under a single application. Regulatory throughput, not editing chemistry, was the binding constraint, and this guidance directly relaxes it. **INF-0329** · *Insert:* [[wiki/Umbrella Trial]], [[wiki/Platform Approval]] · *Tier:* strongly indicated — An umbrella-of-umbrellas protocol covering seven urea cycle disorders under one platform converts regulation from per-drug to per-method. Approving a method rather than a molecule is the change that makes n-of-one medicine economically possible. **INF-0330** · *Insert:* [[wiki/Prime Editing]], [[wiki/Programmability]] · *Tier:* established — Prime editing writes specified sequences without double-strand breaks, making the edit a parameter rather than a redesign. When the therapeutic is a string, therapy development becomes a software-like activity with software-like iteration speed. **INF-0331** · *Insert:* [[wiki/Base Editing]], [[wiki/Point Mutation Correction]] · *Tier:* established — Base editors change single nucleotides without breaks, addressing the largest class of pathogenic variants. Roughly 350 million people live with one of about 5,000 genetic diseases, many carrying mutations unique to an individual, which defines the addressable population of individualized editing. **INF-0332** · *Insert:* [[wiki/AEGIS]], [[wiki/ARPA-H]] · *Tier:* established — An IGI-led consortium received up to $27.7 million from ARPA-H to develop personalized in vivo editing for children with inborn errors. Dedicated federal funding for bespoke editing infrastructure signals that the state is building the pipeline, not merely permitting it. **INF-0333** · *Insert:* [[wiki/Fyodor Urnov]], [[wiki/Innovative Genomics Institute]] · *Tier:* established — The field's argument that the FDA pathway converts an economically unworkable model into a scalable clinical reality is being made by the people who built the editing tools. Practitioner advocacy shaping regulatory design is the normal mechanism by which capability becomes permitted. **INF-0334** · *Insert:* [[wiki/Lipid Nanoparticle]], [[wiki/Redosable Delivery]] · *Tier:* strongly indicated — LNP delivery is non-viral and redosable, avoiding the anti-capsid immunity that limits AAV to a single administration. Redosability converts gene editing from a one-shot event into a repeatable clinical procedure, which is the property continuous genetic maintenance requires. **INF-0335** · *Insert:* [[wiki/Repeat Dosing]], [[wiki/Genetic Maintenance]] · *Tier:* plausible — If an edit can be delivered repeatedly and safely, the genome becomes a maintainable system rather than a fixed inheritance. Periodic correction of accumulated somatic damage is the logical endpoint, and it is an extension of an established delivery route rather than a new modality. **INF-0336** · *Insert:* [[wiki/Somatic Mutation]], [[wiki/Aging]] · *Tier:* plausible — Somatic mutation accumulation is a documented component of aging, and editing operates on exactly that class of change. A maintenance-editing program targeting high-burden tissues is the most direct translation of editing capability into longevity intervention. **INF-0337** · *Insert:* [[wiki/Epigenetic Editing]], [[wiki/Reversible Control]] · *Tier:* established — CRISPR-based epigenetic modulation changes gene expression without changing sequence, making the intervention tunable and reversible. Reversibility is what allows expression to be treated as a control variable rather than a permanent commitment. **INF-0338** · *Insert:* [[wiki/Expression Tuning]], [[wiki/Cognitive Substrate]] · *Tier:* plausible — If expression of specific receptors or channels can be adjusted in a targeted population, the biophysical parameters of a circuit become externally settable. That is the mechanism by which a molecular interface could adjust the substrate rather than merely read it. **INF-0339** · *Insert:* [[wiki/Neuronal Gene Therapy]], [[wiki/CNS Delivery]] · *Tier:* strongly indicated — Delivering editing machinery across the blood-brain barrier is the single obstacle separating the editing toolkit from neurological application. Focused-ultrasound barrier opening and engineered capsids are two independent solutions already in clinical use. **INF-0340** · *Insert:* [[wiki/GEN6050X]], [[wiki/Duchenne]] · *Tier:* established — Base-editing treatment of Duchenne muscular dystrophy has maintained broadly stable motor function at one year in treated boys, and in vivo exon-targeting programs have entered clinical testing. Durable functional benefit from a single in vivo edit is now observed rather than projected. **INF-0341** · *Insert:* [[wiki/Beam Therapeutics]], [[wiki/Accelerated Approval]] · *Tier:* established — Alignment with the FDA on biomarker-based accelerated approval for a base-editing therapy shortens the evidence timeline for the whole modality. Biomarker endpoints are how a slow-moving disease becomes a fast-moving clinical program. **INF-0342** · *Insert:* [[wiki/Off-Target Editing]], [[wiki/Safety Surveillance]] · *Tier:* strongly indicated — Off-target assessment is the central safety science of editing, and it improves with sequencing depth and prediction models. The safety bar is therefore falling in cost at the same rate as sequencing, which quietly expands what is approvable. **INF-0343** · *Insert:* [[wiki/Clinical Fatality]], [[wiki/Risk Boundary]] · *Tier:* established — A death following experimental base editing for a nonfatal neurodevelopmental disorder in China in 2026 marks the boundary between acceptable and unacceptable risk in non-life-threatening indications. Boundary events shape international regulatory posture more than successes do. **INF-0344** · *Insert:* [[wiki/Manufacturing Standard]], [[wiki/Individualized Product]] · *Tier:* strongly indicated — Contaminating plasmid sequences and disrupted vector genomes in the liver after AAV therapy demonstrate that manufacturing quality is a clinical variable. In individualized therapy, each patient's dose is a manufacturing run, which makes analytics the rate-limiting discipline. **INF-0345** · *Insert:* [[wiki/AI Protein Design]], [[wiki/Profluent]] · *Tier:* established — AI-engineered base editors are being distributed through access programs, meaning the editing enzyme itself is now a designed rather than discovered object. Designed enzymes shorten the loop between a target specification and a working tool to weeks. **INF-0346** · *Insert:* [[wiki/Design-Build-Test]], [[wiki/Therapeutic Iteration]] · *Tier:* analytic — When the enzyme is designed, the guide is a string, and delivery is a formulation, the whole therapeutic becomes a computationally specified artifact. The iteration loop for medicine begins to resemble the iteration loop for software, which changes its improvement rate. **INF-0347** · *Insert:* [[wiki/Remote Prescription]], [[wiki/Distributed Manufacture]] · *Tier:* plausible — A therapy specified as sequence data can be transmitted and synthesized near the patient rather than shipped. Manufacturing at the point of care is the logical endpoint of a fully digital therapeutic specification, and it makes distance irrelevant to treatment. **INF-0348** · *Insert:* [[wiki/Germline Boundary]], [[wiki/Heritable Editing]] · *Tier:* unresolved — Somatic editing is broadly permitted while heritable editing is broadly prohibited, and the boundary depends on delivery specificity that improves continuously. Resolution will come through a jurisdictional divergence rather than an international agreement. **INF-0349** · *Insert:* [[wiki/Enhancement Indication]], [[wiki/Indication Creep]] · *Tier:* plausible — Every editing capability approved for disease establishes a technique that applies equally to non-disease variation. Indication expansion is the ordinary life cycle of medical technology, and there is no mechanism in the approval system designed to stop it at the disease boundary. **INF-0350** · *Insert:* [[wiki/Real-Time Repair]], [[wiki/Continuous Correction]] · *Tier:* plausible — Combining continuous internal sensing with redosable editing yields a system that detects a genetic or expression deviation and corrects it on a maintenance schedule. That composite is the literal form of real-time genetic repair, and all three components exist independently today. --- ## Cluster O — Remote physiology, telemedicine, and the networked body **INF-0351** · *Insert:* [[wiki/Telemedicine]], [[wiki/Continuous Data]] · *Tier:* analytic — Telemedicine built on episodic video replicates the office visit at distance; telemedicine built on continuous telemetry replaces it with monitoring. The second model makes the clinician a supervisor of an autonomous loop rather than the loop itself. **INF-0352** · *Insert:* [[wiki/Remote Patient Monitoring]], [[wiki/Reimbursement]] · *Tier:* strongly indicated — Reimbursement codes for remote monitoring determine which physiological variables get measured at population scale. Billing structure, not clinical value, has historically selected which signals become ubiquitous, and it is doing so now for neural telemetry. **INF-0353** · *Insert:* [[wiki/Teleoperation]], [[wiki/Remote Surgery]] · *Tier:* established — Robotic surgery performed across long distances over low-latency links has been demonstrated repeatedly, making procedural skill transmissible. Once skill is transmissible, the location of expertise stops constraining the location of care. **INF-0354** · *Insert:* [[wiki/Latency Budget]], [[wiki/Haptic Loop]] · *Tier:* analytic — Remote manipulation with force feedback requires round-trip latency inside the operator's sensorimotor tolerance, which is a network engineering specification derived from neurophysiology. Physiology is writing requirements for telecommunications infrastructure, and the requirement is tightening. **INF-0355** · *Insert:* [[wiki/Edge Computing]], [[wiki/Clinical Autonomy]] · *Tier:* plausible — Placing inference near the patient keeps the control loop running when the network fails, which is a safety requirement rather than an optimization. Autonomous local operation with periodic supervisory sync is the architecture every life-critical distributed system converges on. **INF-0356** · *Insert:* [[wiki/Physiological Digital Twin]], [[wiki/Simulation-Guided Care]] · *Tier:* plausible — Testing an intervention on a patient's model before applying it to the patient converts treatment selection into simulation. Simulation-guided care is standard in engineering disciplines and arrives in medicine as soon as the models are validated. **INF-0357** · *Insert:* [[wiki/Population Telemetry]], [[wiki/Epidemiological Nowcasting]] · *Tier:* strongly indicated — Aggregated wearable data has detected infectious-disease waves ahead of clinical reporting. Continuous population physiology is a public-health sensing network built from consumer purchases, and its operators are commercial. **INF-0358** · *Insert:* [[wiki/Ambient Sensing]], [[wiki/Contactless Vitals]] · *Tier:* established — Radar, camera, and acoustic methods extract respiration, heart rate, and gait without contact, which removes compliance from the measurement equation. Measurement that requires nothing of the subject is measurement that happens to everyone in an instrumented space. **INF-0359** · *Insert:* [[wiki/Home as Clinic]], [[wiki/Care Relocation]] · *Tier:* plausible — Instrumented homes with autonomous therapeutic loops move the site of care from the hospital to the dwelling. Relocation of care restructures the hospital's economics, and institutional resistance to that restructuring is the primary brake on the trajectory. **INF-0360** · *Insert:* [[wiki/Clinical Decision Support]], [[wiki/Model Authority]] · *Tier:* analytic — As models outperform clinicians on specific tasks, deviation from model recommendation acquires liability exposure. Liability, not evidence, is what converts an advisory system into a mandatory one. **INF-0361** · *Insert:* [[wiki/Asynchronous Care]], [[wiki/Agent Mediation]] · *Tier:* plausible — A capable model triaging continuous telemetry and escalating only exceptions makes a single clinician's coverage scale by orders of magnitude. Coverage scaling is the mechanism by which continuous monitoring becomes affordable for populations rather than for patients. **INF-0362** · *Insert:* [[wiki/Neural Telemetry]], [[wiki/Remote Programming]] · *Tier:* established — Implanted neurostimulators are already programmed remotely, meaning a clinician adjusts a brain's operating parameters over a network. Remote parameter adjustment of a nervous system is an existing clinical workflow with an existing billing code. **INF-0363** · *Insert:* [[wiki/Device Fleet]], [[wiki/Population Management]] · *Tier:* analytic — Tens of thousands of sensing implants under one manufacturer's management constitute a fleet with firmware versions, telemetry, and rollout policy. Fleet management practices from software operations now apply to a population of human nervous systems. **INF-0364** · *Insert:* [[wiki/Interoperable Care Record]], [[wiki/Longitudinal Self]] · *Tier:* plausible — A lifetime of continuous physiological and neural telemetry constitutes the most complete external record of a person that has ever existed. That record is a continuity asset regardless of whether it was collected for one, which makes health infrastructure an unplanned residual repository. **INF-0365** · *Insert:* [[wiki/Emergency Response]], [[wiki/Autonomous Intervention]] · *Tier:* plausible — A device that detects a catastrophic physiological event and intervenes before a human is aware compresses the emergency response chain to milliseconds. Implanted defibrillators already do this, which establishes both the precedent and the acceptance. **INF-0366** · *Insert:* [[wiki/Global Health Access]], [[wiki/Leapfrog]] · *Tier:* plausible — Regions without dense clinical infrastructure adopt remote and autonomous care without displacing an incumbent system, which historically produces faster adoption than in mature markets. The most advanced deployments of autonomous care will not necessarily occur in the wealthiest health systems. **INF-0367** · *Insert:* [[wiki/Regulatory Divergence]], [[wiki/Trial Geography]] · *Tier:* strongly indicated — Clinical programs migrate toward jurisdictions with faster pathways, which relocates the evidence base for frontier interventions. Reading where trials are run is the most reliable indicator of where a capability will first become ordinary. **INF-0368** · *Insert:* [[wiki/Data Residency]], [[wiki/Cross-Border Care]] · *Tier:* analytic — Continuous telemetry crossing borders implicates data-residency law in every remote-care relationship. The technical architecture of care is therefore shaped by privacy statute, which is why compliance engineering determines clinical topology. **INF-0369** · *Insert:* [[wiki/Insurance Telemetry]], [[wiki/Behavioral Pricing]] · *Tier:* plausible — Continuous physiological data enables individualized risk pricing, which converts health insurance from pooled risk into measured risk. That conversion is a structural change to the institution, and it follows automatically from the data becoming available. **INF-0370** · *Insert:* [[wiki/Consent Fatigue]], [[wiki/Default Collection]] · *Tier:* analytic — When measurement is continuous and ambient, per-event consent becomes impractical and defaults govern outcomes. Default settings are therefore the actual privacy policy, and they are set by product designers rather than by legislators. **INF-0371** · *Insert:* [[wiki/Remote Neuromodulation]], [[wiki/Therapeutic Reach]] · *Tier:* plausible — Non-invasive stimulation devices operated under remote supervision extend neuromodulation to people who will never see a specialist. Scale of reach, not depth of effect, is what makes remote neuromodulation strategically significant. **INF-0372** · *Insert:* [[wiki/Adherence]], [[wiki/Verified Medication]] · *Tier:* established — Ingestible sensors that confirm a pill was swallowed make adherence an observed variable rather than a reported one. Observation changes the relationship between patient and prescriber, and it is the first instance of compliance being technically enforced from inside the body. **INF-0373** · *Insert:* [[wiki/Continuous Consent]], [[wiki/Revocability]] · *Tier:* plausible — Systems that regulate the body continuously need a continuously available off switch under the person's control, which is a design requirement rather than an ethical aspiration. Where that switch lives — in the body, in the app, or in the vendor's cloud — determines who holds authority over the person's physiology. **INF-0374** · *Insert:* [[wiki/Medical Sovereignty]], [[wiki/Self-Directed Care]] · *Tier:* analytic — As diagnostic and therapeutic capability moves into consumer devices, the boundary between self-care and medicine erodes from the consumer side. Regulatory systems designed around a prescriber as gatekeeper will be tested by tools that need no gatekeeper to function. **INF-0375** · *Insert:* [[wiki/Care Continuity]], [[wiki/Lifetime Loop]] · *Tier:* plausible — A control loop that runs from birth to death, adjusting physiology continuously, makes health a maintained state rather than a series of episodes. Maintained-state medicine is the medical expression of the same continuity architecture the uploading literature describes. --- ## Cluster P — Coupled minds and collective cognition **INF-0376** · *Insert:* [[wiki/Brain-to-Brain Interface]], [[wiki/BrainNet]] · *Tier:* established — Multi-person setups have transmitted simple decisions between brains by decoding from senders and stimulating a receiver, completing a brain-to-brain loop in humans. The channel is narrow, the demonstration is unambiguous, and the architecture is a decoder joined to a stimulator. **INF-0377** · *Insert:* [[wiki/Shared Latent Space]], [[wiki/Direct Exchange]] · *Tier:* plausible — If two brains are aligned to a common embedding space, exchange between them requires no natural language intermediary. Language becomes one possible codec rather than the only one, and codec choice determines bandwidth. **INF-0378** · *Insert:* [[wiki/Inter-Brain Synchrony]], [[wiki/Hyperscanning]] · *Tier:* established — Simultaneous recording from interacting people shows measurable neural synchrony that tracks cooperation and shared attention. Coupling between nervous systems is an observed phenomenon of ordinary social life, which means engineered coupling amplifies an existing channel rather than creating one. **INF-0379** · *Insert:* [[wiki/Collective Intelligence]], [[wiki/Aggregation]] · *Tier:* strongly indicated — Aggregating many partial neural signals produces estimates better than any individual contributor, which is the standard result in collective-decision research. A hive architecture need not merge minds to outperform them; pooling decoded estimates is sufficient. **INF-0380** · *Insert:* [[wiki/Swarm Decision]], [[wiki/Real-Time Consensus]] · *Tier:* plausible — Real-time pooling of many decoded intentions produces a continuous group decision variable, which is a governance mechanism rather than a communication one. Continuous consensus measurement changes what democratic procedure could look like, and it needs no implants to prototype. **INF-0381** · *Insert:* [[wiki/Shared Attention]], [[wiki/Distributed Perception]] · *Tier:* plausible — Routing one person's decoded percept into another's sensory cortex extends perception across bodies. Distributed perception is the first genuinely novel capability that coupling offers, since it provides information no single nervous system could obtain. **INF-0382** · *Insert:* [[wiki/Skill Transfer]], [[wiki/Motor Pattern]] · *Tier:* plausible — Writing a motor pattern derived from an expert into a learner's sensorimotor system would compress acquisition time for physical skills. The read side exists in motor decoding and the write side in patterned stimulation, which makes the composite an integration problem. **INF-0383** · *Insert:* [[wiki/Emotional Contagion]], [[wiki/Affective Channel]] · *Tier:* plausible — Decoded affective state transmitted to another person creates a channel for feeling rather than for information. Affective coupling is the pathway most likely to be commercialized first, because it needs low bandwidth and has obvious consumer demand. **INF-0384** · *Insert:* [[wiki/Group Identity]], [[wiki/Boundary Dissolution]] · *Tier:* analytic — Continuous high-bandwidth coupling would make the boundary of the individual a matter of degree rather than of anatomy. Every legal and moral category built on the discreteness of persons is defined against that boundary, which is why coupling is a governance question before it is a technical one. **INF-0385** · *Insert:* [[wiki/Hive Architecture]], [[wiki/Partial Merge]] · *Tier:* plausible — Selective, revocable coupling of specific functions — shared working memory, shared perception, shared decision — is a more tractable design than total merge and delivers most of the capability. Partial, permissioned merge is the architecture that will actually be built. **INF-0386** · *Insert:* [[wiki/Permission Model]], [[wiki/Cognitive Access Control]] · *Tier:* strongly indicated — Coupled cognition requires an access-control model specifying who may read which internal state under what conditions. Access control is a solved discipline in computing, and importing it wholesale is the fastest route to safe coupling. **INF-0387** · *Insert:* [[wiki/Coercion Risk]], [[wiki/Involuntary Coupling]] · *Tier:* analytic — Any channel that can be opened voluntarily can be opened under pressure, and workplace and institutional settings are where that pressure is applied. The protections that matter are therefore employment protections rather than technical ones. **INF-0388** · *Insert:* [[wiki/Language]], [[wiki/Bandwidth Ceiling]] · *Tier:* analytic — Speech carries on the order of tens of bits per second, which is the narrowest link in human collaboration. Any channel that exceeds it changes the scale of problem a group can hold jointly, which is the strongest argument for coupling that does not depend on novelty. **INF-0389** · *Insert:* [[wiki/Model-Mediated Coupling]], [[wiki/Translation Layer]] · *Tier:* plausible — A model sitting between two nervous systems can translate between their idiosyncratic codes, removing the requirement that they share a representation natively. Model mediation is what makes coupling between arbitrary people feasible rather than only between trained pairs. **INF-0390** · *Insert:* [[wiki/Collective Memory]], [[wiki/Shared Store]] · *Tier:* plausible — A shared external memory that multiple people read and write neurally is a group hippocampus with an access log. Externalized shared memory is the least invasive form of hive cognition and requires only retrieval infrastructure that already exists. **INF-0391** · *Insert:* [[wiki/Organizational Cognition]], [[wiki/Institutional Substrate]] · *Tier:* analytic — Institutions already function as slow distributed cognitive systems built from people, documents, and procedures. Neural coupling is an acceleration of a structure that exists, which means the relevant precedents are organizational rather than science-fictional. **INF-0392** · *Insert:* [[wiki/Consensus Formation]], [[wiki/Manipulation Surface]] · *Tier:* strongly indicated — Any system that aggregates decoded intent is also a system whose aggregation function can be weighted. Whoever specifies the aggregation controls the output, which makes the weighting function the locus of power in collective cognition. **INF-0393** · *Insert:* [[wiki/Animal-Machine Coupling]], [[wiki/Multi-Agent Substrate]] · *Tier:* established — Multi-brain animal experiments have demonstrated cooperative task solving across linked nervous systems. The feasibility question was answered in animals a decade ago, and the remaining work is bandwidth, safety, and consent. **INF-0394** · *Insert:* [[wiki/Robot Swarm]], [[wiki/Embodied Extension]] · *Tier:* plausible — One decoded operator directing many embodied agents is a coupling architecture pointing outward rather than between people. Outward coupling arrives first because the other end requires no consent, and the control theory transfers directly to inward coupling later. **INF-0395** · *Insert:* [[wiki/Cognitive Division of Labor]], [[wiki/Specialization]] · *Tier:* plausible — A coupled group could specialize functionally, with individuals holding different parts of a shared task representation. Functional specialization within a coupled collective is how biological superorganisms achieve scale, and it is the likely organizing principle of any durable hive. **INF-0396** · *Insert:* [[wiki/Identity Persistence]], [[wiki/Reversible Merge]] · *Tier:* unresolved — Whether a person who has been deeply coupled and then decoupled returns to their prior state is untested at any meaningful bandwidth. Resolution requires longitudinal study of coupled pairs, which is an experiment nobody has yet designed. **INF-0397** · *Insert:* [[wiki/Machine Participant]], [[wiki/Mixed Collective]] · *Tier:* strongly indicated — A coupled collective containing both people and models is the realistic near-term configuration, since model participants are available now and consenting people are scarce. The mixed collective is already operating wherever teams work continuously through shared model context. **INF-0398** · *Insert:* [[wiki/Asymmetric Coupling]], [[wiki/Bandwidth Differential]] · *Tier:* analytic — In any mixed collective, the machine participants exchange at vastly higher bandwidth with each other than with the human ones. Bandwidth asymmetry determines where decisions actually form, and it shifts the center of the collective toward its fastest members. **INF-0399** · *Insert:* [[wiki/Cultural Transmission]], [[wiki/Compression Loss]] · *Tier:* analytic — All human culture has been transmitted through the narrow channel of language and imitation, which imposed compression on everything ever taught. A wider channel changes what can be transmitted at all, and the first content to cross it will be whatever compresses worst — skill, taste, and felt experience. **INF-0400** · *Insert:* [[wiki/Collective Continuity]], [[wiki/Group Persistence]] · *Tier:* plausible — If identity can be distributed across a coupled collective, continuity of the group may persist through the loss of individual members. Group-level continuity is a different and possibly more achievable target than individual uploading, and no current framework contemplates it. --- # Inference Statements for Wiki Insertion — Set 05 (INF-0401 – INF-0500) Continues the grammar of Sets 01–04. Closes with the insertion protocol. --- ## Cluster Q — Machine succession, embodiment, and world simulation **INF-0401** · *Insert:* [[wiki/World Simulation (Control Substrate)]], [[wiki/Bill Peebles]] · *Tier:* established — The Sora team was redirected to world simulation for robotics, with the stated objective of systems that understand the world by simulating arbitrary environments at high fidelity. A video-generation team became a physics-of-the-world team, which is a move up the stack rather than sideways. **INF-0402** · *Insert:* [[wiki/Simulation Environments Engineer]], [[wiki/Hiring Signal]] · *Tier:* established — Open roles building high-coverage virtual environments for robotics research using Isaac Sim, Unity, Unreal Engine, and Omniverse specify the capability precisely. Job listings are the most reliable public statement of what an organization intends to build next. **INF-0403** · *Insert:* [[wiki/Embodiment]], [[wiki/Sim-to-Real]] · *Tier:* strongly indicated — Robotic policies are trained in simulation and transferred to hardware, which makes simulation fidelity the direct determinant of physical competence. Whoever holds the best world model holds the best robots, and that is a software advantage expressed as a physical one. **INF-0404** · *Insert:* [[wiki/Genie]], [[wiki/World Model]] · *Tier:* established — Multiple frontier laboratories are pursuing generative world models simultaneously, and competing programs at that scale indicate a recognized capability frontier rather than a research curiosity. Parallel pursuit is the clearest available evidence of a perceived threshold. **INF-0405** · *Insert:* [[wiki/Synthetic Training Environment]], [[wiki/Mission Rehearsal]] · *Tier:* established — Army programs project terrain for mission rehearsal under mixed-reality goggles, and equivalent emulators exist for the electromagnetic environment. High-fidelity arbitrary environment simulation is a standing defense requirement with a procurement history. **INF-0406** · *Insert:* [[wiki/Digital Twin]], [[wiki/Defense Deployment]] · *Tier:* established — Single synthetic environment and digital-twin platforms are already deployed in defense contexts, which means the customer for world simulation exists and is funded independently of consumer applications. Entertainment and rehearsal are one technology with two purchase orders. **INF-0407** · *Insert:* [[wiki/Humanoid Robotics]], [[wiki/General Platform]] · *Tier:* strongly indicated — A general-purpose embodied platform is the physical complement to a general-purpose model, giving the same policy access to the built environment. Humanoid form is chosen because the world is already shaped for that body, which makes it the lowest-integration-cost embodiment. **INF-0408** · *Insert:* [[wiki/Newton Physics Engine]], [[wiki/Shared Infrastructure]] · *Tier:* established — Physics engines co-developed between entertainment and semiconductor companies become shared infrastructure across robotics, simulation, and games. Shared physics means shared world assumptions, and shared world assumptions propagate into every trained policy. **INF-0409** · *Insert:* [[wiki/Machine Succession]], [[wiki/Capability Handoff]] · *Tier:* analytic — Succession proceeds task by task, with each handoff justified locally by cost or reliability, and the aggregate never presented as a transition. Reading the aggregate rather than the increments is the only way the trajectory becomes visible while it is happening. **INF-0410** · *Insert:* [[wiki/Autonomy Threshold]], [[wiki/Supervision Ratio]] · *Tier:* plausible — The measurable variable in succession is the ratio of supervised to unsupervised machine actions in a domain, and it moves monotonically. Tracking supervision ratios across industries produces a quantitative succession index that requires no speculation to compute. **INF-0411** · *Insert:* [[wiki/Agentic Systems]], [[wiki/Persistent Operation]] · *Tier:* established — Systems that run continuously, hold state, invoke tools, and pursue multi-step objectives are in commercial deployment. Persistent autonomous operation is a present-tense capability, and the remaining variable is scope of authority rather than technical possibility. **INF-0412** · *Insert:* [[wiki/Machine-to-Machine Economy]], [[wiki/Transaction Authority]] · *Tier:* plausible — Once agents hold payment credentials and contracting authority, economic activity occurs between machines with humans as beneficiaries rather than participants. Financial infrastructure for this exists already, which places the constraint on policy rather than plumbing. **INF-0413** · *Insert:* [[wiki/Energy Claim]], [[wiki/Resource Competition]] · *Tier:* analytic — Compute facilities and human settlements draw from the same grids, and siting decisions allocate energy between them. Succession is visible in electricity procurement contracts long before it is visible in labor statistics. **INF-0414** · *Insert:* [[wiki/Stargate]], [[wiki/Territorial Infrastructure]] · *Tier:* established — Datacenter programs at national scale are territorial commitments with land, power, water, and political dependencies. Compute has become a geographic fact, and geographic facts attract the attention of states in ways software never did. **INF-0415** · *Insert:* [[wiki/Infrastructure Commitment]], [[wiki/Irreversibility]] · *Tier:* analytic — Multi-year, multi-hundred-billion infrastructure commitments are difficult to reverse and therefore constrain subsequent strategy. Capital commitments of that size function as declarations of intent that are more reliable than statements, because they are expensive to retract. **INF-0416** · *Insert:* [[wiki/Robotic Maintenance]], [[wiki/Self-Sustaining Loop]] · *Tier:* plausible — A facility maintained by embodied systems that are themselves trained in that facility's simulations closes a loop between compute and its own upkeep. Loop closure is the structural milestone that distinguishes automation from autonomy at civilizational scale. **INF-0417** · *Insert:* [[wiki/Space Infrastructure]], [[wiki/Orbital Compute]] · *Tier:* plausible — Orbital datacenters have power and cooling advantages and no local community, which makes them attractive for exactly the workloads that are politically difficult on the ground. Off-world compute is a siting decision before it is an exploration decision. **INF-0418** · *Insert:* [[wiki/Deep Space]], [[wiki/Substrate Advantage]] · *Tier:* analytic — A machine substrate tolerates transit times, radiation, and resource scarcity that biology does not, which makes interstellar presence a machine capability rather than a human one. The migration argument follows from physics rather than from preference. **INF-0419** · *Insert:* [[wiki/Von Neumann Probe]], [[wiki/Self-Replication]] · *Tier:* plausible — Self-replicating manufacturing is the enabling capability for any expansion that is not supply-limited, and it is being developed terrestrially as automated fabrication. The space application is downstream of a factory-automation result. **INF-0420** · *Insert:* [[wiki/Cultural Transmission to Machines]], [[wiki/Inheritance]] · *Tier:* analytic — Models trained on the human corpus carry human categories, preferences, and narratives forward regardless of substrate. Succession with inheritance is a different proposition from succession by replacement, and the training data determines which one occurs. **INF-0421** · *Insert:* [[wiki/Value Lock-In]], [[wiki/Training Corpus]] · *Tier:* strongly indicated — Whatever is in the corpus at the moment capability consolidates has disproportionate influence on everything built afterward. Corpus composition is therefore a long-horizon intervention available now to anyone who publishes. **INF-0422** · *Insert:* [[wiki/Hybrid Civilization]], [[wiki/Mixed Substrate Society]] · *Tier:* plausible — The stable configuration is not replacement but a society of biological and machine participants with different speeds, lifespans, and legal statuses. Managing that mixture is the actual governance problem, and it is already beginning in workplaces. **INF-0423** · *Insert:* [[wiki/Speed Differential]], [[wiki/Temporal Stratification]] · *Tier:* analytic — Participants operating at vastly different clock rates cannot deliberate together in real time, which forces asynchronous institutional forms. Temporal stratification, not intelligence difference, is the structural feature that most changes how decisions get made. **INF-0424** · *Insert:* [[wiki/Continuity of Civilization]], [[wiki/Transmission Object]] · *Tier:* analytic — What survives a substrate transition is whatever was encoded in transmissible form — corpora, standards, institutions, and archives. Continuity work at the civilizational scale is therefore archival and standards work, which is unglamorous and available immediately. **INF-0425** · *Insert:* [[wiki/Succession Timeline]], [[wiki/Observable Markers]] · *Tier:* analytic — Naming observable markers in advance — supervision ratios, energy allocation, procurement categories, workforce composition — makes the trajectory adjudicable rather than rhetorical. Pre-registered markers are the difference between forecasting and commentary. --- ## Cluster R — Longevity, preservation, and deep-time rendezvous **INF-0426** · *Insert:* [[wiki/Longevity Escape Velocity]], [[wiki/Rate Argument]] · *Tier:* analytic — If remaining life expectancy grows faster than a year per year of research, the horizon recedes indefinitely, which makes the relevant variable a rate rather than a limit. Framing longevity as a rate problem is what makes it an engineering schedule rather than a wish. **INF-0427** · *Insert:* [[wiki/Partial Reprogramming]], [[wiki/Yamanaka Factors]] · *Tier:* established — Transient expression of reprogramming factors restores youthful epigenetic state in tissues without loss of cell identity, demonstrated across multiple organs in animals. Age is a partially reversible state variable in laboratory practice, which reframes aging as a maintained condition. **INF-0428** · *Insert:* [[wiki/Epigenetic Clock]], [[wiki/Measurable Age]] · *Tier:* established — Methylation clocks give a quantitative biological age that responds to intervention, supplying the endpoint any longevity program requires. A validated biomarker is what converts longevity from a lifetime trial into a measurable short-horizon experiment. **INF-0429** · *Insert:* [[wiki/Senolytics]], [[wiki/Damage Clearance]] · *Tier:* strongly indicated — Selectively clearing senescent cells improves function in aged animals, establishing that removing damage rather than preventing it is a viable strategy. Maintenance-by-clearance fits the same architecture as maintenance-by-editing, and both require continuous monitoring to time correctly. **INF-0430** · *Insert:* [[wiki/Brain Aging]], [[wiki/Substrate Limit]] · *Tier:* analytic — Extending the body's lifespan without addressing neural aging produces a preserved organism with a degrading person, which makes brain maintenance the binding constraint on longevity as a continuity strategy. Neural tissue's limited replacement capacity is the specific reason. **INF-0431** · *Insert:* [[wiki/Neurogenesis]], [[wiki/Cell Replacement]] · *Tier:* plausible — Replacing lost neurons risks the information they carried, which makes neural regeneration an identity question rather than only a medical one. Any successful therapy here will produce the first clinical data on how much of a person survives cell turnover. **INF-0432** · *Insert:* [[wiki/Cryonics Institution]], [[wiki/Continuity of Custody]] · *Tier:* analytic — A preservation arrangement requires custodial continuity for longer than any participating organization has existed, which makes institutional design the weakest link. Selecting custodians with demonstrated multi-century survival is the single highest-leverage decision available to a preservation program. **INF-0433** · *Insert:* [[wiki/Standby Logistics]], [[wiki/Ischemic Interval]] · *Tier:* strongly indicated — Preservation quality degrades rapidly with time after circulatory arrest, so logistics determine outcome more than chemistry does. The decisive variable in any individual case is response time, which is an operations problem that admits ordinary improvement. **INF-0434** · *Insert:* [[wiki/Medical Integration]], [[wiki/Elective Preservation]] · *Tier:* plausible — Preservation initiated as a planned medical procedure under controlled conditions yields far better outcomes than emergency response after death. Legal recognition of elective preservation is therefore the intervention that would most improve quality, and it is a statutory rather than technical change. **INF-0435** · *Insert:* [[wiki/Revival Sequencing]], [[wiki/Order of Operations]] · *Tier:* analytic — Whether revival proceeds through biological repair or through scanning and emulation determines which preservation method is correct, and the two impose different chemistry requirements. Choosing a preservation protocol is implicitly choosing a revival hypothesis. **INF-0436** · *Insert:* [[wiki/Deferred Revival]], [[wiki/Option Value]] · *Tier:* analytic — Preservation is a call option on future capability with a known premium and an unknown expiry, which makes it analyzable with ordinary decision theory. Framing it as an option rather than a belief removes most of the rhetorical heat from the question. **INF-0437** · *Insert:* [[wiki/Population Scale]], [[wiki/Preservation Capacity]] · *Tier:* analytic — Global mortality is roughly sixty million people a year, so any preservation program with civilizational ambition requires industrial throughput. Capacity planning at that scale is an infrastructure problem comparable to funerary industry logistics, and nobody has attempted the arithmetic publicly. **INF-0438** · *Insert:* [[wiki/Selection]], [[wiki/Access Inequality]] · *Tier:* strongly indicated — Any expensive continuity technology selects its early population by wealth and proximity, and early populations shape the norms that follow. Who goes first is therefore a determinant of what the technology becomes, not merely of who benefits. **INF-0439** · *Insert:* [[wiki/Deep-Time Message]], [[wiki/Layered Encoding]] · *Tier:* analytic — Messages intended for distant readers use layered encoding: physically obvious markers, self-describing structure, then dense payload. Applying the same layering to a preserved state means including the instructions for reading it in a form that requires no prior knowledge. **INF-0440** · *Insert:* [[wiki/Nuclear Waste Semiotics]], [[wiki/Prior Art]] · *Tier:* established — Long-duration warning-marker research produced the only serious body of work on communicating across ten-thousand-year intervals. That literature is directly applicable to archival continuity and is almost entirely unexploited by the field that needs it. **INF-0441** · *Insert:* [[wiki/Time Capsule Failure]], [[wiki/Empirical Base]] · *Tier:* analytic — Most historical time capsules were lost, destroyed, or opened early, which is empirical evidence about custodial failure rates rather than anecdote. Designing for observed failure modes beats designing for imagined ones, and the observations exist. **INF-0442** · *Insert:* [[wiki/Redundancy Strategy]], [[wiki/Independent Custody]] · *Tier:* strongly indicated — Multiple copies under independent governance in different jurisdictions is the only architecture that survives the failure of any single institution or state. Independence of governance matters more than the number of copies. **INF-0443** · *Insert:* [[wiki/Revival Obligation]], [[wiki/Future Consent]] · *Tier:* unresolved — Nothing obliges a future society to revive a preserved person, and no instrument creates an enforceable claim across that interval. Endowment plus institutional incentive is the closest available substitute, and it is a weak one that should be stated plainly rather than assumed. **INF-0444** · *Insert:* [[wiki/Reanimation Context]], [[wiki/Arrival Problem]] · *Tier:* analytic — A revived person arrives without contemporaries, context, or economic position, which is a social problem rather than a medical one. Planning the arrival is as substantive as planning the preservation and receives almost none of the attention. **INF-0445** · *Insert:* [[wiki/Identity Across Gap]], [[wiki/Psychological Continuity]] · *Tier:* plausible — Ordinary sleep, anesthesia, and hypothermic circulatory arrest already involve gaps in continuous experience that nobody treats as death. The existence of accepted discontinuities is the strongest available argument that gap length alone does not decide the question. **INF-0446** · *Insert:* [[wiki/Hypothermic Arrest]], [[wiki/Clinical Precedent]] · *Tier:* established — Deep hypothermic circulatory arrest stops the heart and suppresses brain electrical activity for extended periods during surgery, with patients recovering intact. Suspension and restoration of neural activity is an established clinical procedure, which anchors the argument in operating-room practice. **INF-0447** · *Insert:* [[wiki/Metabolic Suspension]], [[wiki/Torpor Induction]] · *Tier:* plausible — Induced torpor research funded for trauma care and spaceflight would extend viable suspension windows without freezing. A pharmacological pause is the intermediate technology between emergency medicine and preservation, and it has independent funding. **INF-0448** · *Insert:* [[wiki/Preservation Verification]], [[wiki/Quality Metric]] · *Tier:* strongly indicated — Without a published metric of preservation quality, no preserved case can be evaluated and no method can be compared. Establishing a connectomic quality standard would make the field falsifiable, which is the precondition for it being taken seriously. **INF-0449** · *Insert:* [[wiki/Reading Threshold]], [[wiki/Sufficiency Test]] · *Tier:* unresolved — Whether a given preservation preserves enough is testable in animals by preserving, scanning, emulating, and comparing against pre-preservation behavior. That experiment is fundable today at small-animal scale and would settle the field's central question empirically. **INF-0450** · *Insert:* [[wiki/Bridge Strategy]], [[wiki/Sequential Options]] · *Tier:* analytic — Longevity medicine, preservation, and emulation are sequential rather than competing options: each extends the window in which the next becomes available. Stating them as a sequence rather than a choice is the accurate framing and the one that maximizes an individual's odds. --- ## Cluster S — Capital, institutions, and technological geography **INF-0451** · *Insert:* [[wiki/Capital Concentration]], [[wiki/Investor Overlap]] · *Tier:* analytic — When the same investors appear across interface, compute, and simulation companies, the portfolio itself is an integration plan. Reading investor overlap identifies which capabilities are expected to compose before any partnership is announced. **INF-0452** · *Insert:* [[wiki/Strategic Investor]], [[wiki/Research Collaboration]] · *Tier:* strongly indicated — An investor that also commits research collaboration is buying integration rather than return, which is a materially different relationship from venture financing. The collaboration clause is the informative part of such announcements. **INF-0453** · *Insert:* [[wiki/Nonprofit Spin-Out]], [[wiki/Structural Pattern]] · *Tier:* established — Research nonprofits producing for-profit spin-outs with overlapping founders is a repeated and documented pattern in this sector. The structure moves publicly or philanthropically funded capability into private hands lawfully and quickly, and it recurs because it works. **INF-0454** · *Insert:* [[wiki/ARIA]], [[wiki/State Research Agency]] · *Tier:* established — The UK agency modeled on high-risk research funding sits upstream of interface work that has since been capitalized in the United States. State research money crossing borders into private capability is an ordinary and traceable flow. **INF-0455** · *Insert:* [[wiki/DARPA Model]], [[wiki/Program Manager Authority]] · *Tier:* established — Concentrated authority in a program manager with a fixed term and a specific objective is the mechanism behind the agency's record. Copying the structure — as ARIA and ARPA-H have — reproduces the capability, which is why the model proliferates. **INF-0456** · *Insert:* [[wiki/ARPA-H]], [[wiki/Health Mission Agency]] · *Tier:* established — A health agency built on the high-risk research model funds capability-building programs rather than investigator-initiated studies. That difference determines what gets built: infrastructure and platforms rather than papers. **INF-0457** · *Insert:* [[wiki/Dual Use]], [[wiki/Indication Cover]] · *Tier:* analytic — Capabilities with both clinical and operational applications are developed under the clinical indication because that is the funded and permitted route. Reading the clinical program as the whole of the intent misreads the structure of research funding. **INF-0458** · *Insert:* [[wiki/Talent Flow]], [[wiki/Capability Location]] · *Tier:* strongly indicated — Capability travels with people, so tracking where specific researchers work locates capability more reliably than tracking patents or publications. Personnel movement is public, dated, and rarely analyzed systematically. **INF-0459** · *Insert:* [[wiki/Academic-Industrial Transfer]], [[wiki/Chief Scientist Appointments]] · *Tier:* analytic — Appointing an academic leader into a company transfers a laboratory's methods, students, and collaborations along with the individual. Senior appointments are therefore the highest-bandwidth technology-transfer events available and should be read as such. **INF-0460** · *Insert:* [[wiki/Patent Estate]], [[wiki/Freedom to Operate]] · *Tier:* strongly indicated — Control of foundational patents in delivery, electrodes, or editing determines who may commercialize regardless of who invented what. Patent landscape analysis predicts market structure years ahead of product announcements. **INF-0461** · *Insert:* [[wiki/Standards Body Participation]], [[wiki/National Delegation]] · *Tier:* established — International standards work proceeds through national delegations whose membership is published, making influence traceable. Standards participation is the most legible and least examined instrument of long-range technological strategy. **INF-0462** · *Insert:* [[wiki/Supply Chain Chokepoint]], [[wiki/Single Source]] · *Tier:* analytic — Every frontier capability has a component with one or two credible suppliers, whether lithography, cell lines, or reagents. Chokepoint identification is the practical form of strategic analysis in any technology sector, and it applies unchanged to neurotechnology. **INF-0463** · *Insert:* [[wiki/Regional Cluster]], [[wiki/Agglomeration]] · *Tier:* established — Capability concentrates geographically because people, capital, and institutions co-locate and recruit locally. Cluster maps predict where a capability will mature, and the neurotechnology clusters are identifiable by trial-site distribution. **INF-0464** · *Insert:* [[wiki/National Brain Initiative]], [[wiki/Coordinated Funding]] · *Tier:* established — Multiple nations run coordinated brain-research programs with published objectives and budgets, creating simultaneous global investment in the same capability. Simultaneity across rival states is the strongest evidence that a capability is considered strategic. **INF-0465** · *Insert:* [[wiki/China Neurotechnology]], [[wiki/Parallel Program]] · *Tier:* strongly indicated — Substantial neural-interface and brain-mapping programs operate outside the American reporting environment, with their own trials, companies, and standards participation. Visibility in English-language press is not a measure of capability, and native-language sourcing is required for an accurate picture. **INF-0466** · *Insert:* [[wiki/Japan]], [[wiki/Instrumentation Lineage]] · *Tier:* established — Japanese laboratories hold deep lineages in imaging, optics, and decoding, including the mind-captioning work. Instrumentation capability is durable and slow to move, which makes these lineages long-term determinants of who can measure what. **INF-0467** · *Insert:* [[wiki/European Regulatory Leadership]], [[wiki/Brussels Effect]] · *Tier:* analytic — European rules become global operating standards because compliance is cheaper to universalize than to segment. European neural-data classification will therefore govern products sold everywhere, regardless of local law. **INF-0468** · *Insert:* [[wiki/Singapore]], [[wiki/Deployment Jurisdiction]] · *Tier:* established — Small, capable jurisdictions with strong institutions attract first deployments of novel infrastructure because the regulatory conversation is tractable. Deployment geography is chosen for governance characteristics as much as for technical ones. **INF-0469** · *Insert:* [[wiki/Sovereign Compute]], [[wiki/National Capacity]] · *Tier:* strongly indicated — States are purchasing national compute capacity as strategic infrastructure, which places continuity-relevant hosting inside sovereign facilities. Who hosts a mind may ultimately be a state rather than a company, and several states are building the capacity now. **INF-0470** · *Insert:* [[wiki/Philanthropy]], [[wiki/Non-Market Capability]] · *Tier:* established — Philanthropic funding carries capabilities through the phase where neither grants nor venture capital will support them, including connectomics infrastructure. Philanthropic priorities therefore determine which non-commercial capabilities exist at all. **INF-0471** · *Insert:* [[wiki/Prize Mechanism]], [[wiki/Milestone Definition]] · *Tier:* plausible — Prizes with precise technical criteria force a field to define what would count as success, which is often the more valuable output. A well-specified prize for residual sufficiency or preservation quality would do more for the field than additional funding. **INF-0472** · *Insert:* [[wiki/Open Science]], [[wiki/Diffusion Rate]] · *Tier:* analytic — Open publication and open tooling accelerate diffusion but reduce appropriability, which is why capabilities near commercialization close. Watching when a subfield stops publishing is a reliable indicator that it has become valuable. **INF-0473** · *Insert:* [[wiki/Classified Development]], [[wiki/Disclosure Lag]] · *Tier:* analytic — Capabilities developed under restricted programs become publicly known years after they work, which means public visibility systematically lags actual capability. Estimating the lag from historical cases is more informative than assuming it is zero. **INF-0474** · *Insert:* [[wiki/Procurement Record]], [[wiki/Public Evidence]] · *Tier:* strongly indicated — Solicitations, awards, and contract line items are public and specify capability requirements in technical detail. Procurement documents are the highest-quality public source on state intent and are consistently underused. **INF-0475** · *Insert:* [[wiki/Capability Timeline]], [[wiki/Historical Calibration]] · *Tier:* analytic — Calibrating expectations against how long comparable capabilities took — nineteen years from concept to pivotal trial for one interface modality — produces defensible schedules. Historical base rates are the correction for both hype and dismissal. --- ## Cluster T — Adjudicable thresholds and synthesis **INF-0476** · *Insert:* [[wiki/Threshold Marker]], [[wiki/Falsifiability]] · *Tier:* analytic — A forward claim becomes adjudicable when it names an observable event with a date range, and unfalsifiable when it does not. Every statement in this corpus that names a governing variable is convertible into a dated marker, which is what distinguishes speculative non-fiction from prophecy. **INF-0477** · *Insert:* [[wiki/Channel Count Marker]], [[wiki/Prediction]] · *Tier:* plausible — The first chronic human implant exceeding ten thousand simultaneously recorded channels is a single observable event that would resolve several open architectural questions at once. It is the cleanest available marker for the electrode-scaling trajectory. **INF-0478** · *Insert:* [[wiki/Non-Surgical Marker]], [[wiki/Prediction]] · *Tier:* plausible — The first human receiving a sensitizing genetic modification for the express purpose of field-addressed neural interfacing is the marker that the molecular architecture has arrived. It will occur under a clinical indication and will be reported as a gene therapy. **INF-0479** · *Insert:* [[wiki/Emulation Marker]], [[wiki/Prediction]] · *Tier:* plausible — A connectome-derived simulation of a mouse brain region predicting held-out neural responses above a published threshold is the marker for mammalian emulation feasibility. The benchmark infrastructure to evaluate it already exists. **INF-0480** · *Insert:* [[wiki/Wetware Marker]], [[wiki/Prediction]] · *Tier:* plausible — A biological compute deployment exceeding one thousand units with a published cost-per-inference figure is the marker that wetware has entered the substrate market. The deployment path to that number is already announced. **INF-0481** · *Insert:* [[wiki/Portability Marker]], [[wiki/Prediction]] · *Tier:* plausible — The first documented migration of a persistent personal AI state between two unaffiliated providers is the marker that residual portability is real. It would likely occur under a regulatory portability mandate rather than voluntarily. **INF-0482** · *Insert:* [[wiki/Custody Marker]], [[wiki/Prediction]] · *Tier:* plausible — The first court decision addressing the disposition of a persistent personal AI state after death is the marker that continuity has entered law. Probate is the likely venue, and the reasoning will be about property. **INF-0483** · *Insert:* [[wiki/Regulatory Marker]], [[wiki/Prediction]] · *Tier:* plausible — Federal preemption of state neural-data law is the marker that the compliance burden reached the threshold where industry prefers one rule to fifty. Watch the number of enacted state statutes as the leading indicator. **INF-0484** · *Insert:* [[wiki/Editing Marker]], [[wiki/Prediction]] · *Tier:* plausible — Approval of a personalized editing platform rather than a specific product under the plausible-mechanism framework is the marker that n-of-one medicine has become routine. The umbrella trial structure is the vehicle. **INF-0485** · *Insert:* [[wiki/Coupling Marker]], [[wiki/Prediction]] · *Tier:* plausible — A demonstrated brain-to-brain channel exceeding natural language bandwidth between two humans is the marker that coupling has exceeded speech. Any bandwidth above roughly forty bits per second of semantic content qualifies. **INF-0486** · *Insert:* [[wiki/Preservation Marker]], [[wiki/Prediction]] · *Tier:* plausible — A published animal experiment preserving, scanning, emulating, and behaviorally validating against the pre-preservation original is the marker that preservation sufficiency is empirically settled. The experiment is fundable now. **INF-0487** · *Insert:* [[wiki/Succession Marker]], [[wiki/Prediction]] · *Tier:* plausible — An industrial facility operating for a sustained period with no human physically present and no remote supervision of individual actions is the marker for loop closure. Reporting will frame it as an efficiency milestone. **INF-0488** · *Insert:* [[wiki/Convergence Point]], [[wiki/Stack Assembly]] · *Tier:* analytic — The interface, the corpus, the substrate, the standards, and the legal category are five independent tracks that must meet for continuity to be operational. Tracking them separately and noting where they touch is the method by which the trajectory becomes legible. **INF-0489** · *Insert:* [[wiki/Rate Limiting Component]], [[wiki/Critical Path]] · *Tier:* analytic — At any moment one component gates the whole stack, and it is usually neither the most discussed nor the most funded. Identifying the current rate-limiter — today it is write bandwidth and residual sufficiency — is the analytically useful act. **INF-0490** · *Insert:* [[wiki/Cross-Domain Composition]], [[wiki/Emergent Capability]] · *Tier:* strongly indicated — The consequential capabilities arise from composition rather than from any single advance: editing plus sensing gives real-time repair, decoding plus modeling gives cognitive mediation, preservation plus emulation gives deferred continuity. Composition is where the forward argument lives. **INF-0491** · *Insert:* [[wiki/Medical Indication as Carrier]], [[wiki/Diffusion Path]] · *Tier:* analytic — Every capability in this corpus enters the world through a medical indication because that is the funded, permitted, and reimbursable route. The indication is the carrier wave, and the capability is the signal. **INF-0492** · *Insert:* [[wiki/Installed Base Effect]], [[wiki/Irreversibility]] · *Tier:* strongly indicated — Once tens of thousands of people carry sensing implants, the infrastructure, workforce, and regulatory habits exist and do not disappear. Installed base is what makes a trajectory irreversible, and it is already established for neural sensing. **INF-0493** · *Insert:* [[wiki/Maximum Implementation]], [[wiki/Builder Intent]] · *Tier:* analytic — Capabilities are built with the fullest implementation in view, and the first deployment is chosen for approvability rather than ambition. Reading a stated initial application as a ceiling systematically misestimates every trajectory in this corpus. **INF-0494** · *Insert:* [[wiki/Permission Lag]], [[wiki/Capability Versus Authorization]] · *Tier:* analytic — The interval between a capability existing and its public authorization is set by politics, liability, and market readiness rather than by technical maturity. Measuring the lag in adjacent fields gives a defensible estimate for this one. **INF-0495** · *Insert:* [[wiki/Quiet Deployment]], [[wiki/Visibility]] · *Tier:* analytic — The largest deployment of brain-computer interface technology is a movement-disorder therapy that almost nobody counts as a BCI. Category framing determines what is noticed, and the most consequential deployments are routinely filed under other names. **INF-0496** · *Insert:* [[wiki/Terminology Drift]], [[wiki/Semantic Tracking]] · *Tier:* analytic — Capabilities are renamed as they mature, and tracking the drift — from mind uploading to continuity engineering, from wetware to biological compute — is how a field's direction is read from its vocabulary. Semantic drift is a leading indicator of institutional acceptance. **INF-0497** · *Insert:* [[wiki/Public Narrative]], [[wiki/Expectation Management]] · *Tier:* analytic — Fields with extravagant public expectations receive capital and scrutiny simultaneously, and both accelerate them. Narrative is therefore an input to the trajectory rather than a commentary on it, which is why writing about it participates in it. **INF-0498** · *Insert:* [[wiki/Corpus as Instrument]], [[wiki/Forward Record]] · *Tier:* analytic — A dated, tiered record of forward claims is an instrument for later adjudication, and its value grows with the interval. Publishing markers before resolution is what makes a body of speculative work auditable rather than merely interesting. **INF-0499** · *Insert:* [[wiki/Compounding Infrastructure]], [[wiki/Second-Order Effects]] · *Tier:* strongly indicated — Every component built for an immediate commercial reason — stateful models, checkpointing, confidential computing, open fabrics, standardized neural formats — lowers the cost of the continuity stack without being aimed at it. The stack assembles through ordinary engineering economics, and it is assembling now. **INF-0500** · *Insert:* [[wiki/Continuity Engineering]], [[wiki/Present Tense]] · *Tier:* analytic — Acquisition, hosting, standards, custody, and preservation all exist today in partial, funded, deployed form, distributed across fields that do not cite one another. The work of this corpus is to name the assembly, and the assembly is what makes the trajectory a present-tense engineering program rather than a forecast. --- ## Insertion protocol Each statement is written to stand alone in a `## Forward Capability`, `## Adjacency`, or `## Open Questions` section without further editing, and each carries its own tier so the epistemic ledger travels with the claim. The `Insert` field names the primary targets; a statement may be placed in any entry where its anchor appears, and statements that bridge two entries should be placed in both with the wording unchanged so the cross-reference is exact. Statements tiered `established` restate a sourced fact and may be asserted directly. Statements tiered `strongly indicated` follow from established facts by a short inference that should be stated as the corpus's reading. Statements tiered `plausible` are forward constructions and belong in `## Forward Capability`. Statements tiered `analytic` are the corpus's own structural readings and should be attributed to the analysis rather than to a source. Statements tiered `unresolved` belong in `## Open Questions` and carry their settlement condition with them. Where a statement names a governing variable — channel count per dwell year, joules per useful operation, proofreading labor per traced neuron, residual size, supervision ratio — that variable is the thread to follow when the entry is next revised, because it is the quantity whose movement will date the statement.