Demis Hassabis - Bryant McGill

# Demis Hassabis **Domain:** Artificial Intelligence / Scientific Discovery / Cognitive Science **Doc Type:** Person Node **Maturity:** Developing **Related:** [[wiki/Google DeepMind|Google DeepMind]], [[wiki/Mechanistic Intelligence|Mechanistic Intelligence]], [[wiki/Artificial General Intelligence|Artificial General Intelligence]] --- ## Definition **Demis Hassabis** is an artificial-intelligence researcher and entrepreneur who co-founded DeepMind. His work connects general-purpose learning systems, planning, game-playing, and AI-assisted scientific discovery. ## Corpus Context [[articles/Mechanistic Intelligence Is Humanity's Greatest Liberation|Mechanistic Intelligence Is Humanity's Greatest Liberation]] places Hassabis and [[wiki/Google DeepMind|Google DeepMind]] in the article's laboratory lineage because their systems demonstrate that capabilities formerly treated as uniquely human can be decomposed into trainable architectures and objective-driven computation. The article uses those systems as evidence for tractability, not as proof that any present system is conscious. ## Simple Reminders, Quotations, and Thoughts  > "[The risk of catastrophic AI outcomes is] worth very seriously considering and mitigating against." > **— Demis Hassabis**, *2025, discussion of non-zero catastrophic AI risk* *Verification Status: Editorially quarantined — terminal fragment without the risk being evaluated. Recover a complete, self-contained passage before promotion.* [[reminders/Existential Risk/Catastrophic AI Risk Deserves Serious Mitigation by Demis Hassabis|Catastrophic AI Risk Deserves Serious Mitigation by Demis Hassabis]] > "Can we make sure that– we can keep control of [increasingly autonomous and powerful AI] systems? That they're aligned with our values, they– they're doing what we want that benefits society. And they stay on guardrails." > **— Demis Hassabis**, *2025, 60 Minutes interview* *Verification Status: Editorially quarantined — `the systems` are not identified. Recover a complete, self-contained passage before promotion.* [[reminders/AI Control/Keeping Control of AI by Demis Hassabis|Keeping Control of AI by Demis Hassabis]] > "Proceed with exceptional care." > **— Demis Hassabis**, *2023, interview with UK Prime Minister Rishi Sunak* *Verification Status: Editorially quarantined — the thing that must proceed carefully is absent. Recover a complete, self-contained passage before promotion.* [[reminders/unverified/Proceed with Exceptional Care by Demis Hassabis|Proceed with Exceptional Care by Demis Hassabis]] > "In the situation where there's a high degree of uncertainty, but huge potential impact either way, I think [the only right way] to proceed [with increasingly powerful AI systems] is with the precautionary principle." > **— Demis Hassabis**, *2023, discussion of AI development* *Verification Status: Editorially quarantined — `the only right way to proceed` omits what action or program is at issue. Recover a complete, self-contained passage before promotion.* [[reminders/AI Control/The Precautionary Principle for Advanced AI by Demis Hassabis|The Precautionary Principle for Advanced AI by Demis Hassabis]] > "We must take the risks of AI as seriously as other major global challenges, like climate change." > **— Demis Hassabis**, *2023, media interview* [[reminders/Existential Risk/Treat AI Risk as Seriously as Climate Change by Demis Hassabis|Treat AI Risk as Seriously as Climate Change by Demis Hassabis]]  > "Step one: Solve intelligence. Step two: Use it to solve everything else." > **— Demis Hassabis**, *2010, DeepMind founding-era mission formulation* *Verification Status: Unverified — exact wording/source has not yet been independently confirmed against a primary source; source clue: DeepMind founding-era mission formulation, 2010.* [[reminders/unverified/Step One Solve Intelligence Step Two Use It to Solve Everything Else by Demis Hassabis|Step One Solve Intelligence Step Two Use It to Solve Everything Else by Demis Hassabis]] > "We believe AGI will be the most profound technology ever invented." > **— Demis Hassabis**, *2026, DeepMind, '10 years of AlphaGo'* [[reminders/Machine Succession/We Believe AGI Will Be the Most Profound Technology Ever Invented by Demis Hassabis|We Believe AGI Will Be the Most Profound Technology Ever Invented by Demis Hassabis]] > "Many people join artificial-intelligence labs because the work is prestigious, intellectually interesting, technically deep, well paid, and good for their careers. But others join because they are true believers in the idea that human intelligence can be replicated in silicon. Many of DeepMind's early researchers were among the first true believers in that mission. Shane Legg and Demis Hassabis believed early on that the human mind was an existence proof of general intelligence and that studying it while advancing machine learning would eventually produce human-level AI." > **— Vishal Maini**, *Palisade Research interview, September 29, 2026* [[reminders/Machine Succession/Human Intelligence Is an Existence Proof for Machine Intelligence by Vishal Maini|Human Intelligence Is an Existence Proof for Machine Intelligence by Vishal Maini]] ## Sources / Provenance - [Google DeepMind — about](https://deepmind.google/about/) - [Nobel Prize facts — Demis Hassabis](https://www.nobelprize.org/prizes/chemistry/2024/hassabis/facts/) ## See Also [[wiki/Google DeepMind|Google DeepMind]] · [[wiki/Mechanistic Intelligence|Mechanistic Intelligence]] · [[wiki/Artificial General Intelligence|Artificial General Intelligence]] · [[wiki/AlphaFold|AlphaFold]] · [[wiki/Reinforcement Learning|Reinforcement Learning]]