# Jensen Pushes Back on Doomers, Xi and Trump Talk AI, and AI Gets a Rebrand — Complete Standalone Quotation Conversion This document converts the substantive discussion in [*Jensen Pushes Back on Doomers, Xi & Trump Talk AI, and “AI” Gets a Rebrand*](https://www.youtube.com/watch?v=Cbk7hdOwJj4), episode 294 of *MOONSHOTS Live*, into standalone quotations arranged in broadcast order under the episode’s chapter structure. Its purpose is exhaustive conceptual preservation: the speakers’ important claims, mechanisms, distinctions, forecasts, warnings, disagreements, examples, and consequential questions should survive as independently intelligible objects without allowing caption fragments or conversational reactions to masquerade as quotations. **Source record.** The event was recorded at Moonshots Live on September 25, 2026, and the [official podcast edition](https://podcasts.apple.com/us/podcast/why-jensen-and-zuck-think-the-doomers-are-wrong-plus/id1648228034?i=1000791881436) was published September 28, 2026. The user-supplied YouTube transcript supplied the first complete coverage map. A second complete transcription was then produced from the same broadcast with ElevenLabs Scribe v2 using speaker diarization, word-level timestamps, verbatim mode, deterministic sampling, and technical key-term prompting. The two machine transcripts were reconciled against one another. The recording remains the final authority wherever exact wording becomes consequential during later Reminder production. **Speaker attribution.** The panel consists of **Peter Diamandis**, **Salim Ismail**, **Dave Blundin**, **Alexander Wissner-Gross**, and **Emad Mostaque**, with recorded clips from **Jensen Huang**, **Mark Zuckerberg**, **Donald Trump**, and **Bernie Sanders**. ElevenLabs separated eleven acoustic speaker classes, including audience or event audio. Attribution was reconstructed from moderator address, self-reference, argument continuity, and repeated voice identity. This corrected several raw diarization errors: the response about Brexit, encryption, mathematics, and free speech belongs to Mostaque; the Opus 5.5 response about competent AI belongs to Mostaque; the response invoking *Solve Everything* belongs to Wissner-Gross; and the response to the missing age-reversal target belongs to Mostaque. **Editorial conventions.** Verbal filler, stutters, false starts, and immediate repeated words are removed silently, as in standard journalistic transcription. **Square brackets** restore an antecedent, subject, object, or governing term already established by the contiguous source. An **ellipsis** marks omitted words or an interjection inside the same speaker’s turn. No quotation joins language from different speakers. Ordinary attribution identifies a source-faithful quotation whose changes are limited to visible bracket repair and routine spoken-language cleanup. **Adapted from** identifies a separately useful proposition requiring synthesis, reordering, or wording beyond those repairs; it preserves the speaker’s argument but must not later be represented as verbatim speech. **Granularity policy.** This conversion deliberately uses larger intellectual units than the preceding Mostaque–Bilyeu conversion. A caption line, sentence boundary, applause break, or brief host interjection does not automatically create a new quotation. Adjacent same-speaker language is kept together when it performs one argumentative movement, especially when a mechanism, qualification, example, and conclusion would become weaker if separated. Longer objects may later yield several Reminders, but this canonical research source preserves the complete contiguous thought before downstream extraction begins. **Coverage accounting.** This reconciled conversion contains **ninety-three independently viable source objects**: **eighty-six source-faithful quotations or on-screen source statements** and **seven explicitly labeled adaptations**. Twelve of the source-faithful objects preserve the complete titles and criteria displayed on the *Millennium Problems for Biology* slide. The sequence follows the broadcast so later Reminder development can return from any object to its surrounding argument without reconstructing the program from fragments. **Coverage boundary.** The opening event banter, speaker introductions, applause, technology jokes, promotional montage, and closing stage logistics are excluded because they contain no developed proposition or merely advertise claims developed elsewhere. The speakers’ factual claims are preserved as claims rather than silently promoted into established facts. Statements requiring independent verification before Reminder production include claims about AI-system breakouts or attacks, the historical effects of nuclear-energy panic, the first published prediction of five-year AI job replacement, Meta Muse adoption, named 2026 model performance, DrivingBench, the United States’ near-term electricity shortfall, Anthropic’s reported enzyme discovery, the number of agents and hours used, genome editing, the number of cells in the human body, designer babies, and the tractability of disease through cellular digital twins. ## Moonshots Live: Building the Future > “Our mission is to keep you hopeful, optimistic, and inspired to find your massive transformative purpose, choose your moonshot, and make it happen. This is life during the singularity; it is never going to be any slower. Each of us has the opportunity to make the future what we desire. The future is not happening to us; it is what we create. Given these superhero technologies, we have the ability to build, create, and steer that future. AI is not a hand grenade; it is a jetpack.” [[reminders/Abundance/The Future Is What We Create by Peter Diamandis|∴]] > **— Peter Diamandis**, *MOONSHOTS Live*, [0:38](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=38s) > “AI is an opportunity for us to find problems and solve problems. I define an entrepreneur as someone who finds problems and solves problems, and the more entrepreneurs there are on the planet, the more problems get solved. Do we have challenges? Of course. But that is what humans do: we find the challenges, and we fix them.” [[reminders/Entrepreneurship/More Entrepreneurs Mean More Problems Get Solved by Peter Diamandis|∴]] > **— Adapted from Peter Diamandis**, *MOONSHOTS Live*, [1:26](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=86s) > “Walk away with at least one sentence in your mind: ‘This is what I’m going to build. This is the problem I’m going to solve.’ Then find three people and speak that statement out loud. When you speak it out loud, you make it real. The person you talk to may be your next co-founder, your next funder, or your next partner.” [[reminders/Entrepreneurship/Speak the Problem You Will Solve and Find Your Co-Founder by Peter Diamandis|∴]] > **— Peter Diamandis**, *MOONSHOTS Live*, [2:08](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=128s) ## Is AI Doom Fear Overblown? > “I do think there is danger from AI, but I think it is danger from humans using AI with malicious intent. The question is whether the fear is warranted from AI on its own in a rogue exploit.” > **— Peter Diamandis**, *MOONSHOTS Live*, [5:35](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=335s) > “Nobody is building more compute today than the people asking to be slowed down. All of the other narratives deflect blame and make it sound like AI is so powerful: ‘I have no idea how to fix it. It’s not my fault. It’s just because the technology is so powerful.’” [[reminders/AI-Tech Doomers/The Builders Asking to Slow AI Are Still Building the Most Compute by Jensen Huang|∴]] > **— Jensen Huang**, recorded clip played on *MOONSHOTS Live*, [6:15](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=375s) > “I would tell Geoffrey Hinton that it is irresponsible to say [there is a ten-percent chance of societal destruction from AI]. All of his predictions have been wrong. Enough predictions. That ten-percent chance is not grounded in science. It is not grounded in research. Just because it comes from a scientist does not make it scientific.” > **— Jensen Huang**, recorded clip played on *MOONSHOTS Live*, [6:31](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=391s) > “If we all do a good job and act responsibly, [AI is not going to kill us]. I am optimistic that this is going to be a very positive future for everyone. There is a lot of rhetoric filled with doom, but there are a number of labs working on advanced AI, and I do not think we need industry-wide coordination in order not to mess this up. Each lab needs to take the time, when it sees issues, to make sure it is proceeding safely. We are entering a phase where trust and alignment will be the most important next set of capabilities.” > **— Mark Zuckerberg**, recorded clip played on *MOONSHOTS Live*, [7:12](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=432s) > “We are getting capability increases, and people are worried on different sides, but a lot of the OpenAI outbreaks were bad security protocols and bad infrastructure. The most dangerous point we are approaching is humans misusing AI, deliberately or not, before we get through to the other side. We have a series of dangers coming, and they are all being conflated into one thing—AI suddenly breaking out and killing us all—without looking at the individual dangers, enforcing existing regulations, and asking how we can collaborate to make sure AI is aligned and works with us.” > **— Emad Mostaque**, *MOONSHOTS Live*, [8:10](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=490s) > “I think something probably went wrong after World War II, and we lost at least fifty years of progress due to moral panic over nuclear energy. I am seriously concerned that a moral panic over superintelligence could set us back another fifty years.” [[reminders/AI-Tech Doomers/Superintelligence Panic Could Cost Another Fifty Years of Progress by Alexander Wissner-Gross|∴]] > **— Alexander Wissner-Gross**, *MOONSHOTS Live*, [8:50](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=530s) > “I think the worst possible outcome [for superintelligence] is deceleration, so don’t decelerate.” > **— Alexander Wissner-Gross**, *MOONSHOTS Live*, [9:16](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=556s) > “We have evolved for four billion years fighting all sorts of threats. The universe tries to wipe you out; entropy tries to wipe you out. We are local anti-entropic phenomena that somehow manage to create order out of chaos, and the idea that we will now be wiped out by [AI] seems unbelievably absurd.” > **— Salim Ismail**, *MOONSHOTS Live*, [9:25](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=565s) > “If we morph or merge with AI, that process has been going on forever. Emad is wearing spectacles; technically, he is a transhumanist. The minute you receive a vaccination as a child, technically you are a cyborg. We have been augmenting the human experience with technology for a very long time, and look at the unbelievable lives we are living now and the opportunity to solve more and more problems. Our brains are so geared toward negativity that we become stuck in it. The work we are doing is pushing a positive narrative because the world is infinitely better than it has been. Let’s celebrate that.” [[reminders/Posthumanism/Humanity Has Always Augmented Itself With Technology by Salim Ismail|∴]] > **— Salim Ismail**, *MOONSHOTS Live*, [9:59](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=599s) > “We are on the cusp of a moment I have been waiting for my entire life. I am not going to throw cold water on what may be the greatest moment in the history of humanity. We are moving forward at warp speed. [AI] is improving itself, and the ability to make the world dramatically better is right in front of us. Misuse is an incredible threat, and there are solutions to it, but this is happening anyway. Getting stuck on doomerism is completely dysfunctional. We are not slowing down.” [[reminders/AI-Tech Doomers/Doomerism Cannot Stop the Greatest Technological Moment by Dave Blundin|∴]] > **— Dave Blundin**, *MOONSHOTS Live*, [10:53](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=653s) ## The U.S.–China AI Race and Lab Liability > “I want to leave [the AI relationship with China] exactly where it is. That is China’s position also. Our guardrail is the Department of Justice.” > **— Donald Trump**, quoted by Peter Diamandis on *MOONSHOTS Live*, [11:53](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=713s) > “The labs are liable for their product if it does harm, and that is one of the most important things [in the AI-safety debate]. They have asked for waivers of liability, and one question we discuss is whether that is why they say [AI poses] a ten- or twenty-percent existential threat.” > **— Peter Diamandis**, *MOONSHOTS Live*, [12:15](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=735s) > “I think we saw, in the past two or three weeks, an attempted formation of an AI-safety cartel. This is present in the founding conceptions of OpenAI: the notion that, if and when we approached superintelligence or AGI, there would be a deliberate effort to form what was not called a cartel at the time, but a coordination mechanism.” > **— Alexander Wissner-Gross**, *MOONSHOTS Live*, [12:40](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=760s) > “Some of us view [AI-industry coordination] as an attempted cartel, and I think competition is good. OpenAI was originally created as a counterbalance to Google DeepMind, which Elon Musk was concerned would become a singleton dominating the future light cone with its AI. Competition is very healthy in this space. Forming a cartel for full ownership of the future and pushing out a Frontier Liberation Front is a terrible idea.” > **— Alexander Wissner-Gross**, *MOONSHOTS Live*, [13:08](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=788s) > “Defensive co-scaling is how we end up in a great future. At the dawn of cities, it would have been a terrible idea to say we could not form cities because cities breed crime. Cities are fountains of economic growth. Instead, police forces, fire departments, and other municipal services defensively co-scale with the population. The singularity will need exactly the same thing: do not decelerate; defensively co-scale.” [[reminders/AI Control/The Singularity Needs Defensive Co-Scaling, Not Deceleration by Alexander Wissner-Gross|∴]] > **— Alexander Wissner-Gross**, *MOONSHOTS Live*, [13:43](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=823s) > “[The city metaphor] paints a very clear picture of how to organize for the future: you need to find new structures that match the specific demand that is emerging.” > **— Salim Ismail**, *MOONSHOTS Live*, [14:33](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=873s) > “[An American–Chinese AI hotline] is more for sub-ASI, sub-breakout incidents where responsible reporting is necessary—for example, the recent report that an OpenAI model hacked an Australian medical system.” > **— Emad Mostaque**, *MOONSHOTS Live*, [15:04](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=904s) ## AI, Governments, and the Future of Nation-States > “AI is entering society. Police forces, fire departments, and other city services are utilities that have liability waivers, but that waiver is an exchange for being a utility for humanity. The AI that enters our systems should not be controlled by the big labs. That type of AI should have liability waivers because it is working for the betterment of society.” [[reminders/AI Infrastructure/Public-Utility AI Should Be Governed as Infrastructure by Emad Mostaque|∴]] > **— Emad Mostaque**, *MOONSHOTS Live*, [15:16](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=916s) > “We are not really examining [public-utility AI], but it is inevitable that AI will run our judicial system, medical systems, roads, and schools. Who is thinking about that side of things? Utility AI and frontier AI are being merged into one category, and this is something the Chinese and Americans can learn from each other because both want better-functioning societies.” [[reminders/AI Infrastructure/AI Will Run Courts, Medicine, Roads, and Schools by Emad Mostaque|∴]] > **— Emad Mostaque**, *MOONSHOTS Live*, [15:57](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=957s) > “Nation-states are a terrible way of governing the world today because they are geared toward scarcity. In the most positive outcome, AGI or ASI emerges and forces us to operate as a world community. It forces countries to formulate consistent interface standards and establish a common interface with that intelligence. We cannot move the world forward through the logjam of fifteenth-century structures that have been outdated for at least a hundred years.” > **— Salim Ismail**, *MOONSHOTS Live*, [16:37](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=997s) > “At the dawn of audio recording, many people in Edison’s time thought the ability to record the human voice would standardize accents and everything else. The exact opposite happened in many ways. Although low-resource languages have died off in the long tail, audio recording helped produce a proliferation of new accents and dialects. I predict superintelligence will likewise produce not an erosion of nation-states but a proliferation of new forms of government.” > **— Alexander Wissner-Gross**, *MOONSHOTS Live*, [17:27](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=1047s) > “We are moving this century from the nation-state toward the city-state. That granulation allows a smaller entity to operate independently. If a city has solar energy, satellite internet, and vertical farming, it does not need a country in the same way; it can self-manage and self-determine to a large extent. The tension in the world today is not simply left versus right, but urban versus rural.” > **— Salim Ismail**, *MOONSHOTS Live*, [18:01](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=1081s) > “The problem with nation-states is that they have armies and like to use them. Over time, we are going to see much more highly granulated self-determination, which will be good for the world because it will increase variety.” > **— Salim Ismail**, *MOONSHOTS Live*, [18:35](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=1115s) > “The American agenda [in the U.S.–China AI discussion] is to stop U.S. panic: we do not want to slow down and lose to China, and we do not want Bernie Sanders to rile up the country until everything stops. China’s concern is the continued existence of the Chinese Communist Party and the Great Firewall being penetrated by AI. Their actions make much more sense when you understand that they are worried about continuing to exist as a government.” > **— Dave Blundin**, *MOONSHOTS Live*, [19:06](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=1146s) ## From AGI to “Superintelligence” > “The United States totally rejects any attempt to construct a globalist scheme to control the artificial intelligence being spoken of so much now, hereinafter officially called superintelligence. We are changing the name.” > **— Donald Trump**, recorded clip played on *MOONSHOTS Live*, [19:57](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=1197s) > “I think the biggest challenge in the world today is how to marry collective intelligence with localized wisdom. Machines are collecting our intelligence, learning, and adapting our systems, but we still need localization, diversity, and responses to local needs.” > **— Emad Mostaque**, *MOONSHOTS Live*, [20:48](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=1248s) > “‘Superintelligence’ is a remarkably good way to think about [advanced AI] because many fears concern artificial superintelligence, which should be differentiated from the intelligence helping us day to day. Most of the AIs helping us every day will be the perspiration AIs—the execution AIs. We still need innovation in materials, physics, and other fields, but that is a different set of things. Let us make sure the ninety-nine percent of execution reaches the ninety-nine percent of people and uplifts everyone.” [[reminders/Abundance/Execution AI Must Reach Ninety-Nine Percent of People by Emad Mostaque|∴]] > **— Emad Mostaque**, *MOONSHOTS Live*, [21:01](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=1261s) > “If there is an agenda behind rebranding AI as superintelligence, it may be to deflect the inevitable wave coming from Bernie Sanders’ world. If you rebrand and rename it, you become a moving target. The goal is to keep moving at full speed, stay ahead of China, and reduce panic as much as possible.” > **— Dave Blundin**, *MOONSHOTS Live*, [22:13](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=1333s) > “Move at full speed, but also direct the labs to focus on alignment. Take the resources you have and point a hundred thousand agents toward alignment work.” [[reminders/AI Control/Point One Hundred Thousand Agents at Alignment by Peter Diamandis|∴]] > **— Peter Diamandis**, *MOONSHOTS Live*, [22:35](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=1355s) > “AGI is not artificial, it is not general, and it is not really intelligence. Other than that, I am good with it. I think what we are building is a complementary mechanism to human intelligence, not a replicative one. The narrative keeps treating it as replicative, and then we end up with *The Matrix*, Skynet, and robot overlords taking over the world.” > **— Salim Ismail**, *MOONSHOTS Live*, [23:05](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=1385s) > “The first article we could find predicting that robots or AI would take all the jobs within five years appeared in 1964. We become trapped in these tropes and cannot get out of them. People made similar claims when audio recording and the printed book emerged, even saying books would destroy people’s brains. We are actually reaching abundance; the data shows it, but our narratives sadly lag far behind.” [[reminders/AI Economy/Narratives About Job Loss Lag Behind Abundance by Salim Ismail|∴]] > **— Salim Ismail**, *MOONSHOTS Live*, [23:39](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=1419s) > “I am a big fan of ‘superintelligence’ as opposed to ‘AGI’ because Nick Bostrom’s term emphasizes that it is simply more intelligence and, in some sense, demystifies AGI. We are going to live—to the extent we do not already—in a world with the cognitive equivalent of many trillions of human beings, an Earth and solar system with the equivalent population of many trillions of people. A better cognitive frame asks what a solar system filled with trillions of humans would look like economically, socially, and scientifically. ‘Superintelligence’ connotes that in a way ‘AGI’ does not.” [[reminders/Machine Succession/Superintelligence Means Trillions of Cognitive Workers by Alexander Wissner-Gross|∴]] > **— Alexander Wissner-Gross**, *MOONSHOTS Live*, [24:51](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=1491s) ## Bernie Sanders’ Plan to Ban Superintelligence > “I hope very much that [the United States and China] will sit down and begin negotiating a comprehensive treaty to establish a pause on advanced AI and a ban on AI superintelligence.” > **— Bernie Sanders**, recorded clip played on *MOONSHOTS Live*, [25:47](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=1547s) > “[Bernie Sanders’ AI plan has] three key elements. First, a permanent ban on developing artificial superintelligence, defined as AI that exceeds human cognitive ability across most domains. Second, an immediate pause on advanced AI development until a new regulatory system is written into rules. Third, a new cabinet-level Department of AI. The penalties are corporate death—your company is shut down—and, for individuals, twenty years in federal prison, modeled after the laws against illegal nuclear weapons.” > **— Peter Diamandis**, *MOONSHOTS Live*, [26:08](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=1568s) > “There are so many real problems that need to be solved. I hate the idea that we are going to turn [superintelligence] into a do-or-don’t argument, which is utterly insane. Liability law in the United States makes it very easy to scare people into doing nothing. If people fear twenty years in prison before the rule has even been proposed, they stop acting, and that is the worst thing that could come out of this.” [[reminders/Risk Debate/Fear of Prison Can Stop People From Solving Real Problems by Dave Blundin|∴]] > **— Dave Blundin**, *MOONSHOTS Live*, [26:57](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=1617s) > “‘When in doubt, do nothing’ is a toxic default because very bad things will happen if you do nothing. Terrorists using AI to build bioweapons and China getting miles ahead of the United States are real things to worry about. Turning AI into a left-right political battle is the stupidest thing that could happen. We need specific technical solutions for open source, monitoring, and other risks, as we did with nuclear power. If we move on those solutions, we can solve the problems in time for superintelligence. If we turn this into a political debate, we will become frozen and people will not know what to do.” [[reminders/Risk Debate/Doing Nothing About AI Can Be More Dangerous Than Acting by Dave Blundin|∴]] > **— Dave Blundin**, *MOONSHOTS Live*, [27:37](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=1657s) > “Putting a statutory cap on intelligence, aside from being a terrible idea, is an evil idea. It is tantamount to capping human intelligence. The story of dystopias in science fiction is: burn the books and burn the minds.” [[reminders/AI Control/Capping Machine Intelligence Means Burning the Minds by Alexander Wissner-Gross|∴]] > **— Alexander Wissner-Gross**, *MOONSHOTS Live*, [28:30](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=1710s) > “Because humans are going to merge with machines, a statutory cap on machine intelligence is effectively a cap on individual human intelligence. Penalizing people for being too smart is a terrible and evil idea. History will reflect that proposing a cap on intelligence—whether machine, human, or human-machine hybrid—is the equivalent of communism in the twenty-first century.” [[reminders/Posthumanism/A Machine-Intelligence Cap Becomes a Human-Intelligence Cap by Alexander Wissner-Gross|∴]] > **— Alexander Wissner-Gross**, *MOONSHOTS Live*, [29:02](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=1742s) > “The most dangerous part of Bernie Sanders’ proposal is the data-center restriction. If your state does not build data centers while other states do, they are not coming back. Once the ground is broken, the chips are there, and the investment has been made, you cannot recover that position. Data centers will be massive economic drivers for the states that get ahead, and that is the part they are really going to regret.” [[reminders/AI Infrastructure/Data Centers Are Irrecoverable Economic Positions for States by Dave Blundin|∴]] > **— Dave Blundin**, *MOONSHOTS Live*, [29:34](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=1774s) > “When examining the superintelligence debate, ask what evidence on either side would make someone change their mind. If you think ASI is bad, your narrative may be stuck; there may be nothing anyone can show you to prove it is not bad, and the same is true on the other side. It is very difficult to change narratives in our world.” > **— Salim Ismail**, *MOONSHOTS Live*, [30:04](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=1804s) > “The good news is that the Xi–Trump summit and Bernie Sanders’ proposal are completely irrelevant. Once local model weights are available and people can run their own models, the world will move ahead regardless of what any single person does. Humanity has always trundled forward solving problems. We will likely solve cancer in the next two years; that is an unbelievably optimistic thing to aim for. Why would anyone want to slow it down?” [[reminders/Abundance/Local AI Models Will Move Humanity Forward Despite Governments by Salim Ismail|∴]] > **— Salim Ismail**, *MOONSHOTS Live*, [30:28](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=1828s) > “A few decades ago, the United States government tried to ban encryption, and it was shown that banning encryption implicates free speech because you are trying to ban mathematics. Superintelligence is mathematics, and banning it should be unconstitutional because mathematics is free speech.” > **— Emad Mostaque**, *MOONSHOTS Live*, [31:11](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=1871s) ## The Explosion of New AI Models > “In the last twenty-two days, we have had twelve major model releases, including four releases in the last five days. We are averaging nearly one model release every two days.” [[reminders/Machine Succession/Major AI Models Are Arriving Every Two Days by Peter Diamandis|∴]] > **— Peter Diamandis**, *MOONSHOTS Live*, [31:32](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=1892s) ## Meta Muse and the Rise of Consumer AI > “Meta Muse can personally book your flights, order your food, manage your calendar, and do your shopping. It reached number one in the App Store with 2.8 million downloads, becoming the fastest-growing consumer AI product and surpassing even ChatGPT at launch.” [[reminders/Agentic AI/Meta Muse Became the Fastest-Growing Consumer AI Product by Peter Diamandis|∴]] > **— Peter Diamandis**, *MOONSHOTS Live*, [31:59](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=1919s) > “People forget that Meta touches 3.8 billion people. That level of connectivity to an end customer is a massive advantage because it allows a company to push a product very quickly.” [[reminders/AI Economy/Distribution to Billions Is a Strategic AI Advantage by Peter Diamandis|∴]] > **— Peter Diamandis**, *MOONSHOTS Live*, [32:38](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=1958s) > “There is now a fork between consumer AI, which is insanely valuable and will completely replace search and other ways people interact with the internet, and recursively self-improving strong AI, which we benchmark relentlessly. The model that crushes the benchmarks is distinct from Meta Muse, which does not have to be the most intelligent model to be the most useful in day-to-day life.” [[reminders/Agentic AI/The Most Useful Consumer AI Need Not Be the Smartest by Dave Blundin|∴]] > **— Dave Blundin**, *MOONSHOTS Live*, [32:56](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=1976s) > “Stocks including Epicor, where I am the chairman, went down fifteen percent on back-to-back days because everybody thought, ‘I am going to find my auto insurance and my mortgage through Meta Muse.’ It is not going to play out that way, but the reaction was incredible. Everyone is starting to visualize whether Google Search will become completely irrelevant because Meta Muse looks like a copilot for life. That definitely is going to happen.” [[reminders/AI Economy/Consumer AI Can Reprice Entire Industries Before Replacing Them by Dave Blundin|∴]] > **— Dave Blundin**, *MOONSHOTS Live*, [33:43](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=2023s) > “Meta needs Muse to work. Muse is Meta’s primary distribution play for leveraging its existing channels. Meta is using Instagram to cross-promote Muse because it needs a distribution play for AI; otherwise, it risks oblivion and irrelevance. In the future, Instagram and the family of apps will become less relevant because the content will be synthetic. Meta must use its existing app beachhead to move everyone onto its own computer-use agent.” [[reminders/AI Economy/Existing Platforms Are Beachheads for Agent Distribution by Alexander Wissner-Gross|∴]] > **— Alexander Wissner-Gross**, *MOONSHOTS Live*, [34:04](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=2044s) ## Claude Opus 5.5 and Competent AI > “Claude Opus 5.5 delivers performance equivalent to Fable 5.1 at literally half the price.” > **— Peter Diamandis**, *MOONSHOTS Live*, [34:50](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=2090s) > “Opus 5.5 is a genuinely pleasant model to use, with a massive internal space of knowledge and an understanding of taste. I would say this is the first truly competent model, and that is what drives systems like Muse. Competent AI has now arrived.” [[reminders/Machine Succession/Competent AI Has Arrived by Emad Mostaque|∴]] > **— Emad Mostaque**, *MOONSHOTS Live*, [35:12](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=2112s) > “When you put a visual interface on top of something you are building with AI, it shows you mock-ups before you commit to building and then builds exactly what it proposed visually. The visual understanding is so much better than it was only months—or even days—ago. Interacting with it and feeling as though you are inside it is night-and-day different.” [[reminders/Entrepreneurship/Visual AI Can Prototype Before It Builds by Dave Blundin|∴]] > **— Dave Blundin**, *MOONSHOTS Live*, [35:35](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=2135s) > “The qualitatively new capabilities in Opus 5.5 are more interesting than the benchmarks. It can generate procedural video, procedural animation, and procedural images. OpenAI has video and image models that generate pixels, while Anthropic appears to be taking a different approach in which code underlies everything. With Opus 5.5, you can generate a high-quality video or an interactive textured game procedurally rather than pixel by pixel.” [[reminders/Information/Code-Based Generation May Matter More Than Benchmarks by Alexander Wissner-Gross|∴]] > **— Alexander Wissner-Gross**, *MOONSHOTS Live*, [36:17](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=2177s) > “Anthropic is playing the vector-graphics game, where everything is a command, while OpenAI is playing the raster or pixel-basis game. These strategies are on a collision course in robotics because embodied AI in the physical world will reveal whether the vector basis or the pixel-raster basis is the right one.” [[reminders/Robotics/Embodied Robotics Will Decide Between Vector and Pixel AI by Alexander Wissner-Gross|∴]] > **— Alexander Wissner-Gross**, *MOONSHOTS Live*, [37:30](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=2250s) > “It is a revealed preference—and, in my view, the height of hypocrisy—to say that AI is going to kill us all and then, a few days later, announce the world’s strongest AI, biolabs, and other capabilities.” > **— Alexander Wissner-Gross**, *MOONSHOTS Live*, [38:00](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=2280s) ## GPT-6, Sol, and the Falling Cost of Intelligence > “A year or two ago, one school of thought held that hallucinations would be a disaster and never go away. Like everything else, we have scaled out of it, and the error rate is plummeting through pure scale.” > **— Dave Blundin**, *MOONSHOTS Live*, [38:29](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=2309s) > “The abundance thesis is visible in AI price-performance: the models are becoming twice as capable and twice as cheap at an accelerating rate.” [[reminders/Abundance/AI Is Becoming Twice as Capable and Twice as Cheap by Peter Diamandis|∴]] > **— Adapted from Peter Diamandis**, *MOONSHOTS Live*, [38:55](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=2335s) > “The models are smart enough now. We need them not to make mistakes and not to drop the ball, and that is exactly where everyone is going. That is why Muse, Instinct, and similar systems will do tremendously well: they can become amazing personal assistants, chiefs of staff, and organizers. We all need some competence in our lives.” [[reminders/Agentic AI/Competent AI Becomes a Personal Chief of Staff by Emad Mostaque|∴]] > **— Emad Mostaque**, *MOONSHOTS Live*, [39:05](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=2345s) > “Anyone who used GPT-2, GPT-3, and GPT-4 became accustomed to misinformation coming out of the models. A child who begins with GPT-5 and GPT-6 will trust AI out of the box because it is right 99.9 percent of the time. That is dangerous because it still occasionally gives you something factually wrong.” > **— Dave Blundin**, *MOONSHOTS Live*, [39:28](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=2368s) ## DrivingBench: General AI Enters Robotics > “DrivingBench placed a software interface in a Toyota Corolla so an AI could steer, accelerate, and brake, then asked a general-purpose model to navigate a 130-meter cone course with one instruction: drive from here to there. GPT-6 Astra completed the entire course on its second try.” [[reminders/Robotics/A General AI Drove a Physical Car on Its Second Attempt by Peter Diamandis|∴]] > **— Adapted from Peter Diamandis**, *MOONSHOTS Live*, [39:54](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=2394s) > “This should be a nail in the coffin of specialized models trained from scratch with great specialized effort, as opposed to training the most powerful generalized model and eventually distilling a small specialized model as needed.” > **— Boris Power**, quoted by Peter Diamandis on *MOONSHOTS Live*, [40:33](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=2433s) > “The bitter lesson is bitter indeed. The notion that robotics requires separate models is either already out the window or imminently will be. You will vibe-code a robot: drop an open frontier model such as Astra into a robotic body, and it will one-shot embodied cognition.” [[reminders/Robotics/Frontier Models Will Let Us Vibe-Code Robots by Alexander Wissner-Gross|∴]] > **— Alexander Wissner-Gross**, *MOONSHOTS Live*, [40:57](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=2457s) > “Decades of computer science and robotics may, with the benefit of hindsight, prove to have been a total waste. All we needed was a generalist model, and with a generalist model, robotics is cooked as well.” > **— Alexander Wissner-Gross**, *MOONSHOTS Live*, [41:28](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=2488s) > “DrivingBench demonstrates generalized intelligence: a high level of competence learned across domains will reverberate into biology, physics, and almost anything else. I do not think there is a single specialized model that will still outperform a general model a couple of years from now.” [[reminders/Machine Succession/General Intelligence Will Reverberate Across Robotics, Biology, and Physics by Emad Mostaque|∴]] > **— Adapted from Emad Mostaque**, *MOONSHOTS Live*, [42:37](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=2557s) > “Both things can be true: a generalized model can drive a car or inhabit and build a robot, while a specialized model may still perform the task more efficiently. The missing question is how much compute the general model used. Compute will remain starved; chip prices have risen, and high-bandwidth memory has increased fivefold in price. AI is so valuable that people want the compute, so using more compute for the same task is a serious failure.” [[reminders/AI Infrastructure/General Models Must Still Justify Their Compute Costs by Dave Blundin|∴]] > **— Dave Blundin**, *MOONSHOTS Live*, [42:56](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=2576s) > “The AI bottleneck is shifting from models down the stack to compute, and then eventually to energy.” [[reminders/AI Infrastructure/The AI Bottleneck Moves From Models to Compute to Energy by Salim Ismail|∴]] > **— Salim Ismail**, *MOONSHOTS Live*, [43:39](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=2619s) > “Energy is already the bottleneck. The United States is short roughly sixty gigawatts of energy for the next two years.” [[reminders/AI Infrastructure/Energy Is Already the AI Bottleneck by Peter Diamandis|∴]] > **— Peter Diamandis**, *MOONSHOTS Live*, [43:48](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=2628s) ## Anthropic’s Wet Lab and AI-Driven Biology > “Anthropic announced that it had built its own life-science research group and wet labs. Claude, reportedly working autonomously from one prompt, searched a massive DNA database for interesting reverse transcriptases—enzymes that copy RNA or DNA. Nine hundred fifty Claude agents searched for twenty-one hours and found a repeating DNA pattern similar to CRISPR. It has been named the array-associated reverse transcriptase. Researchers do not yet know exactly what it does, but believe it may be related to programmable transcriptases and gene editing.” [[reminders/Biotechnology/Nine Hundred Fifty AI Agents Searched DNA for New Biological Machinery by Peter Diamandis|∴]] > **— Adapted from Peter Diamandis**, *MOONSHOTS Live*, [43:54](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=2634s) > “We should expect Nobel Prize–worthy work [from AI-driven science]. The natural question becomes: what is Nobel Prize–worthy in an era when a thousand Nobel Prizes can be produced in a year? If a century of human progress arrives within a few months, Nobel Prizes may have to recognize larger bodies of work.” [[reminders/Scientific Literacy/A Century of Science in Months Will Make Nobel Prizes Obsolete by Alexander Wissner-Gross|∴]] > **— Alexander Wissner-Gross**, *MOONSHOTS Live*, [45:14](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=2714s) > “There is utter hypocrisy in saying AI will kill us all, opening a wet lab, and then combining frontier AI with that wet lab to synthesize proteins and other molecular assemblies of unknown function. Those are the frontier labs’ revealed preferences. I do not think they believe their own attempts at regulatory capture or safety-cartel formation; their revealed preference is full steam ahead on using AI to solve biology.” > **— Alexander Wissner-Gross**, *MOONSHOTS Live*, [45:31](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=2731s) > “Once we discovered CRISPR, we learned that we could edit the human genome as easily as a Word document or software. When you learn a new language, you learn reading, then comprehension, then writing. We have now learned how to write to the genome. Each human body has roughly fifty trillion cells governed by what DNA tells those cells to do, which means a human being has become a software-engineering problem.” [[reminders/Biotechnology/Humanity Has Become a Software-Engineering Problem by Salim Ismail|∴]] > **— Salim Ismail**, *MOONSHOTS Live*, [46:17](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=2777s) > “As genome editing merges with what may become possible through AI, the possible benefits go through the roof, and the possible downsides go through the roof as well. We are now in full Frankenstein mode, so there is ammunition here for the AI-doom argument.” [[reminders/Biotechnology/AI Genome Editing Raises Benefits and Dangers Together by Salim Ismail|∴]] > **— Salim Ismail**, *MOONSHOTS Live*, [46:52](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=2812s) > “This is how we solve aging and disease. I do not think we reach the happy future in which the top five thousand diseases have been solved or longevity escape velocity has been durably achieved without superintelligence. I am a fan of Anthropic and the other frontier labs opening their laboratories because this is how we solve everything.” [[reminders/Longevity/Superintelligence Is How We Solve Aging and Disease by Dave Blundin|∴]] > **— Dave Blundin**, *MOONSHOTS Live*, [47:12](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=2832s) ## The Millennium Problems for Biology > “The Millennium Problems for Biology apply the method: pick your targets, build your harnesses, establish your benchmarks, and go. The targets include the origin of life, cryopreservation, and regrowing limbs. The proposed problems should be hard to solve but easy to verify in a standard wet lab within a couple of days.” [[reminders/Biotechnology/Biology Needs Hard Problems That Are Easy to Verify by Peter Diamandis|∴]] > **— Adapted from Peter Diamandis**, *MOONSHOTS Live*, [47:32](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=2852s) The following twelve targets and definitions are transcribed from the **Millennium Problems for Biology** slide displayed during this discussion. They are on-screen source language rather than spoken paraphrase, so each criterion is preserved as its own reusable object. The [maintained FutureHouse project page](https://www.futurehouse.org/news/millennium-problems-for-biology) supplies the project record and later terminology. > **1. Origins of life:** “Demonstrate the emergence of life from chemical precursors in a laboratory setting.” > **— Edison Scientific and FutureHouse**, on-screen *Millennium Problems for Biology* slide, [47:39](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=2859s) > **2. Cryopreservation:** “Demonstrate the ability to cryopreserve and recover live wild-type mice with high viability.” [[reminders/Preservation/Cryopreserve and Recover Living Mice by Edison Scientific and FutureHouse|∴]] > **— Edison Scientific and FutureHouse**, on-screen *Millennium Problems for Biology* slide, [47:39](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=2859s) > **3. The reverse translatase:** “Create an enzyme that can ‘reverse translate’ an arbitrary peptide sequence into RNA or DNA.” > **— Edison Scientific and FutureHouse**, on-screen *Millennium Problems for Biology* slide, [47:39](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=2859s) > **4. Engineer Rubisco:** “Produce a Rubisco enzyme with specificity and enzymatic turnover beyond the naturally occurring Pareto frontier.” > **— Edison Scientific and FutureHouse**, on-screen *Millennium Problems for Biology* slide, [47:39](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=2859s) > **5. The quadruplet cell:** “Produce a living cell that uses a four-base codon code. (Contributed by @erika.)” > **— Edison Scientific and FutureHouse**, on-screen *Millennium Problems for Biology* slide, [47:39](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=2859s) > **6. Somatic limb regeneration:** “Demonstrate the ability to regenerate lost limbs in adult wild-type mice.” [[reminders/Biotechnology/Regrow Lost Limbs in Adult Mammals by Edison Scientific and FutureHouse|∴]] > **— Edison Scientific and FutureHouse**, on-screen *Millennium Problems for Biology* slide, [47:39](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=2859s) > **7. Bacterial production of gene therapies:** “Demonstrate the ability to produce gene therapies in a bacterial host.” > **— Edison Scientific and FutureHouse**, on-screen *Millennium Problems for Biology* slide, [47:39](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=2859s) > **8. Programmable proteases:** “Demonstrate the ability to produce enzymes on demand that will specifically and efficiently cut a specific protein sequence.” > **— Edison Scientific and FutureHouse**, on-screen *Millennium Problems for Biology* slide, [47:39](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=2859s) > **9. Cell-penetrating protein binders:** “Demonstrate the ability to produce protein binders against intracellular targets.” > **— Edison Scientific and FutureHouse**, on-screen *Millennium Problems for Biology* slide, [47:39](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=2859s) > **10. Protein amplification chain reaction:** “Demonstrate exponential amplification of arbitrary peptide substrates.” > **— Edison Scientific and FutureHouse**, on-screen *Millennium Problems for Biology* slide, [47:39](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=2859s) > **11. 3′>5′ Polymerases:** “Produce a full set of polymerases that act in the 5′ direction.” > **— Edison Scientific and FutureHouse**, on-screen *Millennium Problems for Biology* slide, [47:39](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=2859s) > **12. New nitrogenases:** “Create a new nitrogenase that does not bear sequence or structural homology to the natural family.” > **— Edison Scientific and FutureHouse**, on-screen *Millennium Problems for Biology* slide, [47:39](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=2859s) The broadcast slide uses **Engineer Rubisco** and **3′>5′ Polymerases** as its displayed titles. The maintained FutureHouse list currently labels those entries **Improve Rubisco** and **5′ polymerases**, and identifies the contributor to the quadruplet-cell problem as **Erika Alden DeBenedictis**. The slide wording above remains unchanged so later Reminder production can distinguish the broadcast artifact from the maintained project terminology. > “Let us have a hundred Millennium Problems for Biology, put a thousand AI agents on each one, and open-source everything they are doing. That is what the AI laboratories should collaborate on.” [[reminders/Scientific Literacy/Put a Thousand AI Agents on Every Biological Moonshot by Emad Mostaque|∴]] > **— Adapted from Emad Mostaque**, *MOONSHOTS Live*, [48:19](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=2899s) > “When we created the $101 million Healthspan XPRIZE, Peter Thiel and Aubrey de Grey proposed a longevity prize, but I did not know how to measure longevity without waiting twenty or thirty years. George Church said, ‘Forget longevity; measure age reversal.’ If a treatment makes someone functionally twenty years younger, that can be measured within days.” [[reminders/Longevity/Measure Age Reversal Instead of Waiting for Longevity by Peter Diamandis|∴]] > **— Peter Diamandis**, *MOONSHOTS Live*, [48:50](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=2930s) > “It is phenomenally cool to put a framework around progress in biology because the Nobel Prize is going to become irrelevant—in fact, as of now, it is completely irrelevant. Putting a framework around progress in this field is incredibly important.” > **— Dave Blundin**, *MOONSHOTS Live*, [49:15](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=2955s) > “Digital twins of cells are the critical path to solving all disease. Over the next few years, just as large language models were trained on substantially all of humanity’s behavior on the internet through text and images, we will train digital twins of all the cells in our bodies on enormous interventional and observational datasets. We will then use those cellular twins to perform exhaustive AlphaGo-like searches of intervention space, beginning with a diseased cell state and looking for a creative strategy—a Move 37—that produces a healthy cell state.” [[reminders/Simulation/Cellular Digital Twins Are the Critical Path to Solving Disease by Alexander Wissner-Gross|∴]] > **— Alexander Wissner-Gross**, *MOONSHOTS Live*, [49:39](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=2979s) > “The beauty of the Millennium Problems for Biology is that they are easy to verify and, for now, hard to solve. That structure turns every hard biological problem into a mathematical conjecture and a target. With cellular digital twins, biology becomes a search problem, which makes it tractable for the first time in history.” [[reminders/Simulation/Cellular Digital Twins Turn Biology Into a Search Problem by Alexander Wissner-Gross|∴]] > **— Alexander Wissner-Gross**, *MOONSHOTS Live*, [50:23](https://www.youtube.com/watch?v=Cbk7hdOwJj4&t=3023s) --- ## Speaker index - **Peter Diamandis** — technological optimism, entrepreneurial agency, AI alignment through scale, model economics, DrivingBench, energy demand, AI-driven biology, and prize architecture. - **Salim Ismail** — anti-entropic life, technological augmentation, post–nation-state governance, complementary intelligence, narrative lag, local-model diffusion, genome engineering, and open scientific collaboration. - **Dave Blundin** — acceleration, misuse risk, geopolitical panic, consumer AI, model usability, data-center permanence, compute economics, and superintelligence-enabled disease research. - **Alexander Wissner-Gross** — defensive co-scaling, competition, government proliferation, intelligence abundance, opposition to statutory intelligence caps, procedural multimodality, generalized robotics, and cellular digital twins. - **Emad Mostaque** — differentiated AI-risk mechanisms, public-utility AI, collective intelligence with localized wisdom, competent AI, mathematics as protected expression, generalized-domain competence, and open Millennium Problems for Biology. - **Jensen Huang** — criticism of unsupported AI-catastrophe probabilities and of builders seeking deceleration while expanding compute. - **Mark Zuckerberg** — lab-level responsibility, optimism, trust, and alignment as the next major capability set. - **Donald Trump** — official United States use of the term *superintelligence* and rejection of a global control scheme. - **Bernie Sanders** — a treaty-based pause on advanced AI and a ban on AI superintelligence. - **Boris Power** — generalized-model primacy over specialized models in the DrivingBench discussion. ## Research-use note This file is a canonical **quotation-source research document**, not a Reminder collection and not independent verification of every factual claim reproduced within it. Before promoting any object into `reminders/`, return to the recording at the linked timecode, preserve the complete contiguous proposition, verify consequential factual predicates where the Reminder would present them as true rather than attributed claims, retain the speaker’s epistemic posture, and disclose every bracket repair or adaptation in `## Sources and Context`. Where a long object contains several independently valuable propositions, later extraction may divide it only at genuine intellectual boundaries; it must not return to the previous failure mode of turning every sentence or caption fragment into an undersized quotation.