Late Lessons, Jensen Huang and AI
Markdown version

F2. What matters to him: the values and commitments behind Andrew Maynard’s work#

The facet. What is important to Maynard: flourishing and “being human”; agency, dignity and identity; responsibility and the obligation to improve the world; who benefits and who pays; openness and accessibility; joy, wonder and play; and the future as something people can shape. The aim is to set these out in his own terms, as they work in his writing, not as a list of principles.

Evidence and conventions. - Reading. All perspective notes (B01–B32, FFTF-A, SUPP-self) first. Every quotation was then checked against the original: working/maynard/corpus/, working/maynard/book/, and the Future Rising, Rethinking Risk, TechTrends, Nature Nanotechnology, 2006 testimony, 2008 Bulletin and 2009 2020science files in Resources/maynard-papers/. The 05 map and themes were used only for leads. - Citations. Posts by date and slug (Medium-era hash suffixes dropped). FFTF = Films from the Future, FR = Future Rising, RR = Rethinking Risk (2017_Rethinking-Risk_VVEV-CSI-chapter.pdf), all by printed page. TT = 2023_TechTrends-Interview-AI-Responsible-Innovation_author-copy.pdf, PDF page, his quoted words only. T2006 = 2006_House-Science-Testimony_..._Maynard-statement.pdf, record page. - Weight flags. [AI-edited] = the 2026-09-24 lecture body (his words, smoothed by Claude). [self-account] = April 2026 andrewmaynard.net essays. [prompt] = his 2025-01-30 self-description written for ChatGPT. AI-written text, guest posts and co-written text are excluded.

In brief#

Maynard’s values are not an ethics module attached to a risk scientist’s toolkit. They come first, and much of the rest of his work follows from them. They start with a public-health conviction he stated in 2009, that “people matter”: risk work exists for “the people it ultimately impacts”. Seven commitments grow from that root. 1. What makes us “us” is most worth protecting. This means worth, dignity, identity, idiosyncrasy, the inner life and the capacity for wonder. Yet “being human” is for him an open question, not a fortress. 2. People should author their own lives and futures “on their own terms”, toward the future they want “rather than one that someone else decides for us”. 3. Power over the future carries an obligation to use it to improve lives. That includes an obligation to innovate, not only a duty of caution. 4. The test of a technology is who benefits and who pays. 5. The “we” that shapes the future should be “as big and inclusive as possible”, and knowledge should reach “anyone who can benefit from it”. 6. Joy, wonder and curiosity are part of what is at stake, not decoration. 7. The future is something people can shape. Hope is warranted, but never guaranteed.

What makes these distinctive is how he holds them. He holds them pluralistically: his risk framework uses value, “not to be confused with ‘values’”, and lets others define it. He holds them from inside, applying them to himself, his field and his institution, and he holds them without preaching. Above all, they are built into his definition of risk, so that what people value decides what counts as harm. There his values stop being sentiments and start doing analytical work, and that is what they most distinctively bring to the AI discussion.

1. “People matter”: where the values come from#

His values come less from moral philosophy than from occupational and public health, and the root shows in his earliest policy writing. - 2006. In House testimony, as a scientist counting research dollars, he framed the moral stakes around the people who carry risk: “it is ultimately the public—as workers or consumers, for instance—that may bear many of the potential risks” (T2006 rec. p.53). It was “irresponsible to spend millions of dollars on building a better microscope in the name of risk research when we cannot tell workers how effective their respirators are” (rec. p.57). - 2009. His list of ten things everyone should know about nanotechnology safety puts one item first: “people matter”. Risk research “isn’t about the buzz of new discovery. It isn’t about getting rich and famous.” Its “primary focus… should be the people it ultimately impacts”. He locates this in a professional culture, “that ingrained culture of putting others first”, and warns that getting nanotechnology “right” will be “a hollow achievement if we end up neglecting the very people who will make its success possible” (2009-08-29, 2009_2020science_Ten-Things-Everyone-Should-Know-About-Nanotechnology-Safety.md).

The “buzz of new discovery” line matters because his own work runs on that delight (§7). Even in 2009 he ranked people above it.

The root stays visible. In 2023 he describes workplace risk as a theme he knows well, “having worked in occupational health for over a decade in my early career” (2023-08-11 social-inequity-elysium). In 2025 his one declared commitment is “I do place human wellbeing and flourishing at the heart of my work” (2025-01-30 ai-at-a-crossroads [prompt]). In 2026 the people his own framework might still miss are “Data workers in annotation supply chains… communities carrying the environmental costs of compute, or people affected by systems they never chose to use” (2026-07-16 orphan-risks-frontier-ai-maynard). Twenty years on from the 2006 respirator-wearers, it is the same kind of person: someone carrying the cost of a technology they did not choose.

This root explains why his best-known conceptual move is a value move: “Risk starts with something that is worth protecting” (nnano.2016.28 p.211). The public-health ethos becomes a definition of risk.

2. “What makes us ‘us’”: being human, protected but not fixed#

His clearest statement of purpose is: “what drives my work more than anything is the possibility that… our technologies begin to fundamentally change who we are — or even what we are” (2024-01-01 the-future-of-being-human-in-2024). Reading a papal message on AI, he names the task: “understanding what’s worth protecting lest we discard it without realizing what we’re losing” (2023-12-15 pope-francis-artificial-intelligence).

What he means by “us”. Taken across his writing, it is not intellect, and never “humanity” in the abstract. - The individual person. He chose “the future of being human” as a frame because it focuses on “each of us personally, rather than the rather generally handwaving around ‘humanity’” (2023-04-04 welcome-to-the-future-of-being-human). - Worth. The line from Never Let Me Go that “stays with me” is that technology can “rob us of our souls, even as it sustains our bodies… because it makes us forget the worth of others” (FFTF p.62). - Feeling, not intellect. “Being smart doesn’t make you good” (FFTF p.108). Part II of Future Rising, “Uniquely Human”, runs through feelings, faith, imagination, curiosity, creativity, art, fear, loss, despair, possibility and hope. - Messiness, spoken of fondly. We are “a hot mess of social instincts and individual desires” (FR p.197). The “threads of hope” include “our eccentricities and irrationalities” (2020-12-10 we-need-to-rethink-our-relationship-with-the-future). By 2026: “our irrationality, our attachments, our stubbornness are not bugs to be fixed” (2026-04-12 The Future We’re Building [self-account]). - Idiosyncrasy. His sharpest recent worry is AI that “robs us of the eccentricities, weirdness, and glorious diversity of personalities, perspectives and ideas”, with most at stake for “anyone who lives on the edge of convention” (2026-03-08 ai-linkedinification). - Curiosity and wonder. These are “core to what makes us us” (2024-11-10 is-ai-poised-to-suck-the-soul-out-of-science). - One’s own voice. “The way I write is personal. It reflects who I am, and to relinquish that to a machine would be to diminish myself” (2023-09-20 what-do-college-students-think-about-chatgpt). - The embodied and the transcendent. He cares about “the transcendent core of what it means to be human that defies reduction to transactional interpretation” (2026-05-21 magnifica-humanitas-and-being-human, postscript).

This is why the “fix” worldview troubles him more than almost anything else. If AI can fix anything, “the only logical conclusion you get to is that this includes ‘fixing’ people — including who we are, what we do, and what makes us ‘us’” (2024-10-06 the-double-or-nothing-bet-on-ai-fixing-the-climate, note 1). He objects to arguments for embryo screening because they “boil human flourishing down to a balance sheet” (2024-04-14 welcome-to-the-age-of-swipe-and-select-embryos). AI sharpens the concern because it acts on the self, possibly bringing “alterations in this base code without our agreement or permission” (2024-01-01).

Being human is an open question. Yet he will not treat “being human” as a fixed thing to be defended. This is the nuance most easily lost. - Technology is part of what we are. Asking whether he is a techno-optimist is “a bit like asking if I’m an oxygen pessimist or optimist”. Assuming “technology is something we do and not something we are” is part of the problem (2024-03-31 we-have-a-technology-problem-and). - Growing humanity, not only defending it. His 2025 keynote turns the question round: “rather than grappling with the challenges of being human in an age of AI, how do we learn how to be human in an age of AI?” (2025-03-30 reimagining-education-in-an-age-of-ai). - No immutable bedrock. It is “getting harder to hold on to what we’ve thought of as the bedrock of being human as if it’s an immutable truth” (2024-12-29 fantasy-top-ten-lists-2025, note 12). - Respect for tradition, with a limit. He “genuinely felt seen” reading Magnifica Humanitas, yet judged that it “stops short of what I suspect is needed”. It sets out to preserve being human “as we have understood this for millennia”, while AI “is beginning to force a reckoning with long-held assumptions” (2026-05-21). - A widening moral circle. Future Rising imagines “personhood” long after “we’ve outgrown the constraints of our biological heritage” (FR p.193). The extinction framing of AI risk is “too human-centric” (2023-05-31 existential-risks-of-ai). We must avoid “dehumanizing” possible machine minds “to justify how we control and use them” (2023-08-18 being-human-in-an-augmented-future), and must not “devise ways of enslaving AIs” (2023-08-23 could-we-build-conscious-ais-in-the-future).

What connects these is not human exceptionalism. It is a refusal to let anyone be counted as less. In Future Rising he names “a particularly ugly and insidious part of being human”: the tendency to exclude from the future “those we consider to be in some way less valid, less entitled, even less ‘human’ than us” (FR p.191). He had made the same point about augmented bodies in 2018, warning against the “dark, deep tendency” to label the different “non-human” or “sub-human” (FFTF p.139). Dignity and worth are fixed points for him. The boundary of “human” is not.

3. “Rather than one that someone else decides for us”: agency, dignity and identity#

Authorship of one’s own life. The phrase that recurs most is from the close of Films from the Future. The task is to “nudge them toward the future we want, rather than one that someone else decides for us” (FFTF p.288). Agency is what he protects when he writes about: - Implants. “As you incorporate more tech into your body, who owns you?” (2020-08-28 navigating-the-complex-world-of-advanced-brain-machine-interfaces; FFTF p.146). - Behavioural prediction. Pre-emptive “justice” assumes “other people do not have agency over their destiny”. “There will always be an element of chance and choice that determines our actions” (2020-01-01 the-science-of-predicting-bad-behavior; FFTF pp.72, 76). - AI choice engines. The danger is people becoming “engines of value creation rather than the primary recipients of created value”, and he asks “who decides what is good for individuals” (2024-07-13 ai-choice-engines-sunstein).

Consent. His objection to the bioterrorist of Dan Brown’s Inferno is not to the method but to its lack of consent: “what gave him the right to take this gamble in the first place? Not the people whose futures he was playing with” (FFTF p.249). The same question comes back in his AI writing, where it targets benevolent control, not only malicious control. - Value alignment. History shows people who “propagated evil on the back of an absolute certainty that the solution to value alignment is to rob others of their right to hold values that don’t align with theirs” (2023-05-25 leading-ai-expert-says-we-should). - Persuasion for good ends. If voice AI could nudge people towards climate action: “But who decides what is good for society? Who decides what you should believe, and what you should do?” (2024-09-01 is-chatgpts-new-voice-mode-dangerously-persuasive). - Making lives “better”. On AI remaking mental health, he affirms the need to help people “lead lives that they consider to be better” and adds: “great care needs to be taken in who decides what ‘better’ means” (2024-10-13 amodei-machines-of-loving-grace).

The standard that follows is informed choice, not prohibition. Of ChatGPT’s memory feature: “that’s fine. As long, that is, they are made fully aware of the potential privacy concerns” (2025-10-05 when-chatgpt-turns-snitch).

Dignity: where he draws lines. Dignity is where the non-preacher takes firm stands. - Predictive policing. “Here I should lay my cards on the table”. Using LLMs this way would violate human rights “especially around dignity, equality, and self-determination” (2023-05-22 can-large-language-models-be-used). - OpenAI and Scarlett Johansson. His generosity to industry ends at disregard for consent: “a reality that sometimes seems childish irresponsibility” (2024-05-21 openais-problem-with-the-movie-her). - Advisers and students. He asks advisers whether their AI use is “robbing them of their dignity” (2025-10-26 ai-misuse-in-student-advisor-collaborations-1). - Why numbers are not enough. Risk numbers “will be meaningless to people who may stand to lose their health, livelihood, or dignity” (RR p.194). Reading a short story, he frames a dying astronaut’s choice as “her dignity in being able to have control over when and how she dies” (RR p.198).

Identity in the definition of risk. In 2018 he wrote his value commitments into how risk is defined. Films reveal “threats to dignity, belonging, identity, belief, even what it means to be human” (FFTF p.23). Value includes aspiration: “something we have and can’t face losing, or something we aspire to and cannot bear to lose sight of” (FFTF p.24). That aspiration clause makes his risk thinking forward-looking in a way hazard-based frameworks are not.

4. “An obligation to explore new ways of using science and technology to improve the world”: responsibility#

A two-sided obligation. The key passage is the scene on Arran at the end of Films from the Future. - He admits his nostalgia for a slower life and calls it “a sentimental illusion” (FFTF p.287). - He gives the reason: “if we’re tempted to start renouncing technologies from a position of privilege, we risk denying too many people without the same privileges the chance to make their own decisions”. - He states the conclusion: “we have an obligation to explore new ways of using science and technology to improve the world”. It is an obligation “that comes with some tremendous responsibilities” (FFTF p.288).

Responsibility for him therefore points two ways. There is a duty not to harm, and a duty not to withhold. The second is argued from fairness, not from enthusiasm.

The positive half is old and steady. - 2006. Fear-driven rejection of nanotechnology could “deal a severe blow to the quality of life” (T2006 rec. p.52). - 2015–16. “Technology innovation alone will not solve these and other challenges. But without it, many will not be solved at all” (2015-01-30 responsible-development-of-new-technologies-critical-in-complex-connected-world). The “greatest risk” is losing sight of technologies “that make our world a better place, not just a different one” (2016-03-02 how-risky-are-the-world-economic-forums-top-10-emerging-technologies-for-2016). - 2023. Answering Andreessen: “I live and breathe advanced technologies in my work. I revel in their potential.” Ethics and sustainability are “not the enemies of a vibrant and promise filled future, but critical components of achieving it” (2023-10-19 marc-andreessen-ditch-sustainability).

Power is to be used well, and responsibility is shared. “The challenge we face is not to abdicate power, but to develop ways of understanding and using it in ways that are socially responsible” (FFTF p.44). Technologies are “developed by large teams of people, and their uses and abuses are tacitly sanctioned by millions more” (2019-04-26 avengers-endgame-a-cautionary-tale-of-technological-innovation). So responsibility “also means that we collectively need to give a damn about the future we’re creating”, which is “partly why I wrote this book” (FFTF p.286). In 2025: “each one of us has a role to play” (2025-08-31 holding-on-to-our-humanity-age-of-ai).

Stewardship is the emotional centre. The chapter “Stewardship” in Future Rising opens with him in tears at a 2011 Davos panel with three young people, which left him with “a deeper appreciation of what it truly means to be a steward of the future”. The picture was of a future “where each and every one of us has the responsibility to inspire others and add value to their lives” (FR p.209). He does not claim we are good stewards. “What remains unclear… is whether we have what it takes to be good stewards” (FR p.210). The book ends: “We’ve been given the gift of being the architects of our own future. And because of this, we have a responsibility to ourselves, and to future generations, to learn our trade” (FR p.216).

Privilege creates obligation, including his own. - Knowledge. “the privilege of scientific insight should come with the responsibility to use this insight with care and consideration” (nnano.2014.196 p.659); “the privilege of academic scholarship and research comes with an obligation to ensure that the knowledge we unearth is accessible to anyone who can benefit from it” (2024-09-04 succeeding-at-science-on-youtube). - Himself. “how privileged I am to be paid to think, to write, and to teach… And, of course, the responsibilities and obligations that come with this” (2026-08-02 what-we-can-learn-with-ai-by-not-trying-to-learn). He chose “to put public good before academic prestige”, but “it’s only OK if there really is public good that comes from my writing” (2026-05-17 the-nonsense-i-write). - Students and universities. “we owe it to them to put their success before our own traditions and egos” (2026-03-29 can-ai-create-an-undergraduate-degree-plan). Universities need “to be of service to society, rather than service to our organization or ourselves” (2026-09-24 being-an-academic-in-an-age-of-ai [AI-edited]).

Responsibility as care. Implanted devices “come with a lifetime responsibility to patients” (2024-09-18 neuralink-blindsight-brain-computer-interface). He wants care to be a “hard” concept, a basis for “policy, governance, and decision-making” (2025-03-09 the-hard-concept-of-care-in-technology-innovation, note 1). He applies it to his own craft: “to write without care for your readers is a very academic trap to fall into” (2026-05-17), and good writing signals “a level of care and effort in its formation that indicates it’s worthy of someone’s time to read” (2026-07-04 just-how-good-is-anthropics-fable-as-a-research-assistant).

He implicates himself. As a young aerosol scientist he “was quite happy to coopt a narrative of social good so that I could continue to satisfy my scientific curiosity” (FFTF p.219). He sees “benevolent myopia… in many of my peers, and even myself at times” (FFTF p.220). And: “All of us, it has to be said, have a bit of Sidney Stratton in us” (FFTF p.227). His sense of responsibility is not something he preaches at others. It is something he expects to fail at himself.

5. “Who is reaping the benefits… and who is paying the price?”: fairness#

He put the question in those words in 2008: “Who is reaping the benefits of new nanotech applications, and who is paying the price?” (2008-01-14, 2008_Bulletin_Setting-the-Nanotech-Research-Agenda.md). It runs through everything since. - The Luddites. Their legacy is ensuring technologies are “developed for the benefit of all, not just a privileged few” (FFTF p.193). - The digital divide. Robotics, AI and the Internet of Things could “widen the gap between the privileged and the disadvantaged” (2016-01-11 thinking-innovatively-about-the-risks-of-tech-innovation). - Designing the future. Without inclusion, “the default will a future of the powerful, by the powerful, for the powerful” [sic], yet “the future belongs as much to them as it does the powerful” (2019-03-31 design-principles-for-de-marginalizing-the-future). His response is characteristically unexpected: a secular rewriting of the Beatitudes, including “Listen to those who are quiet”. - Who carries the risk. “It’s easy to make risk decisions when you’re not the one who has to suffer the consequences” (2020-07-30 life-on-mars-astrobiology-and-thinking-differently-about-risk). - Techno-optimism. His sharpest statement is against the long-run trend argument. Technological “foreshortening” becomes “a neat way of saying that some people will inevitably suffer as we strive to build someone else’s dream of a better future, and that’s OK”. This “doesn’t sit well with me, especially when I begin to ask who decides who will suffer and who will thrive” (2023-10-19). - Aggregates. “it’s hard to imagine a truly humane future where we use technology to benefit the privileged at the expense of those who aren’t so privileged”. This is aimed at those who think “on aggregate” but “have little time for the individuals who make up that aggregate” (2025-03-09).

In education the same value turns concrete. - College essays. They became “less about who you are and more about where you come from”. With AI, applicants “just need to be themselves” (2023-07-27 chatgpt-and-college-applications). - Paid AI tools. A $20 subscription is “too much to expect from every student — especially those who are already on the margins” (2023-08-29 chatgpt-enterprise-game-changer). - “Free” AI. In 2026: “what does it mean when we say it’s free — because it isn’t? Somebody is paying somewhere” (2026-09-24 [AI-edited]).

Two features make his fairness distinctive. - It argues for innovation as well as against harm. The renounce-from-privilege argument (§4) is a fairness argument. His hope for AI is a future “substantially better than might might otherwise be–for everyone, not just a privileged few” [sic] (2020-11-12 is-artificial-intelligence-going-to-kill-us-all). He wants pathways “where we can reduce those iniquities while still ensuring the broad benefits of the technology” (TT p.5). - He applies it to his own framework. The pragmatic heart of risk innovation is “your risk is my risk”: firms attend to what stakeholders value because harm comes back to them. In 2026 he names its limit. Its benefits run through channels that “are not equally open to everyone”, so its limitations “fall hardest on the people with the least leverage” (2026-07-16).

6. “As big and inclusive as possible”: openness, inclusion and accessibility#

The most revealing passage is a self-correction in Future Rising. “Up to now in this exploration of the future, I’ve been rather loose with the term ‘we’” (FR p.191). A homogeneous “we”, he goes on, hides the habit of excluding others. “Instead, we should be striving for this ‘we’ to be as big and inclusive as possible”, building “a future where everyone has the right to thrive, as long as in doing so, they don’t deprive others of this selfsame right” (FR p.192).

Who gets to take part. In practice this means insisting on who is at the table. - “everyone has the right to play some role”; “‘it’s complicated’ simply isn’t an excuse for excluding people”; and “people don’t need to understand the inner workings of AI to discuss and explore how it might potentially impact their lives and threaten what’s important to them” (2023-05-15 erik-schmidt-ai-regulation). - In 2023 he told interviewers that AI conversations “are driven by white guys who head up tech companies”. Otherwise “the few people that have that power and that understanding craft the future that they want” (TT pp.4–5). By 2026, “‘together’ means considerably more people than are currently in the room” (2026-04-12 What Nanotechnology Taught Me [self-account]).

Respect for ordinary people’s intelligence. This rests on a view of people. - “most people have a pretty high level of expertise in what’s important to them and their communities” (FFTF p.222); “most people are reasonably smart” (TT p.5). - The people “all too easily overlooked by scientists and engineers” belong in the conversation, “in fact, especially these individuals” (FFTF p.18). - He rejects the view that “‘nonscientists’ should revere, but not interfere with, science”, and wants experts “willing to serve the needs of citizens — and not just co-opt them for their own ends” (2016-01-12 can-citizen-science-empower-disenfranchised-communities).

Public fear is information, not ignorance. This is one of his most consistent and least conventional commitments. - In The Man in the White Suit, “everyone is shrewd enough to see how change supports or threatens what they value” (FFTF p.225). - He built a timeline of technology moral panics explicitly not to treat them as “something to be mocked” (2025-06-01 vibe-coding-moral-panic). - He refuses to confine AI’s harms to “the vulnerable”: “we all have some degree of vulnerability here” (2025-08-31). - He turns the deficit model back on its users. After the 2024 election, “the most glaring ‘deficit’… is the lack of understanding that some people have of how individuals and society work” (2025-05-25 why-parasocial-communication-is-important).

Accessibility as an obligation. It follows from the privilege argument (§4). Future Rising’s poor sales stung because bridging the gap to “people who might potentially benefit” is “a public responsibility that I take very seriously” (2024-05-19 future-rising-short-history-of-tomorrow). He gives away course materials, tools and a book; the audiobook podcast was partly to “give students an alternative to purchasing the book” (2023-07-03 a-new-tech-podcast). He wants knowledge that is not “a commodity to be bought and sold” (2025-03-27 ai-agent-creates-online-course-in-minutes), and communication that lets people benefit “on their own terms” (2025-05-25).

The honest broker is itself a value choice. His chosen role, from Films from the Future, is “trying not to judge others or advocate for a specific course of action, but to help people make the best-informed decisions for themselves and their communities”. He justifies it on moral grounds. It “avoids mistaking personal values for the ‘right’ values, and respects deeply held beliefs and values in others” (FFTF p.246). The role, he says, is “rooted deeply in values that I consider important”. His restraint is a form of respect, not neutrality.

7. “Reveling in the awe and wonder”: joy, wonder and play as things of value#

For Maynard, wonder is not ornament. It is something whose loss is a real loss. - Complacency as a hazard. Near the end of Films from the Future he names the risk of “complacency”, of “not reveling in the awe and wonder of the world we’re building” (FFTF p.286). The slide from “wow” to “meh” (p.285) hands the future to “those that do care” (p.286). - Science as delight. Physics was “awe-inspiring and humbling” (FFTF p.219). “Science is a love language between us and the universe, where we can be awed and delighted and amazed” (TT p.2). - The soul of science. This is the clearest case of joy treated as a value technology can threaten. “The soul of science lies in the delight and wonder of exploring the unknown”. Reduced to production, science becomes a tool “where curiosity, joy, and wonder, have no place”, and “a little bit of all of us dies in the process”. The harm is reciprocal: “who would become a scientist who sets out to make the future a better place if all the soul had been sucked out of what you love?” (2024-11-10). - Joy as a neglected category. “The concept of joy is much under-appreciated in how we think about the roles and impacts of technology in our lives”, and advances “have a nasty habit of sucking the joy out of what we do without us realizing it!” (2024-12-29, note 7). By 2026 it is “a deeply under-appreciated metric of intellectual and academic achievement!” (2026-09-20 reasoning-llms-just-want-to-have-fun, note 4). - Value, broadly. His list of what “value” includes runs from health and wealth through “dignity, equity, justice, kindness” to “our capacity to experience awe and wonder” (2025-03-30).

Joy is relational. In Future Rising it is joy “from building a future that values, nurtures, and celebrates others” (FR p.210). At the WEF meeting in Tianjin what moved him to tears was not the world leaders but 61 paintings by local children, “personal, authentic, and mesmerizingly beautiful”, and, in his capitals, “NOT GENERATED BY AI” (2025-07-20 still-human-61-inspiring-paintings). He worried that “there’s simply not enough curiosity left in the world” for people to wonder about the slow mechanical message board he installed on campus: “I hope that isn’t the case” (2024-01-07 the-future-of-being-human-is-analog). In his King’s lecture, joy is what academics bring to the AI transition: “the joy of playing around and serendipitously discovering something” (2026-09-24 [AI-edited]). He remarks that “Joy is a word I don’t use often”, and his footnote adds: “Actually, I suspect I use it more than I realize.”

Play protects agency. Asked how people will learn to live with socially agentic AI, he answers “observation, play, and experience… albeit with intent”, not formal training (2024-10-20 learning-to-live-with-agental-social-ai). For students he asks for “permission to play” (2025-03-15). Play as method belongs to another facet. The value is that people keep the freedom to explore and find out for themselves.

Joy never outranks people. The 2009 line about “the buzz of new discovery” still holds. “I’m not sure there is a strong causal link between curiosity and benevolence” (2023-07-19 elon-musk-maximally-curious-agi). He traces the lure of permissionless innovation to the same curiosity he prizes, and to his own PhD all-nighter (FFTF p.161). Joy matters, and it needs other people to keep it honest.

8. “Architects of our own future”: a future people can shape#

The seventh commitment holds the others together: the future is something people can and should author. “As a species, we are profoundly talented architects of our own future”, yet there is “a gaping chasm between what we are capable of and what we achieve” (FR pp.17–18). Among the harms on the first page of Future Rising are people “being denied the futures they aspire to because of the narrow-mindedness of others” (FR p.17). In a line he later chose to republish, we are “disturbingly good at stealing the futures of others when it suits us. Or worse, imprisoning them within the walls of futures we impose on them” (2024-09-08 a-journey-from-the-past-to-the-edge-of-tomorrow, quoting FR ch.26). Risk thinking serves this aim: it is “not just about protecting what we have, but protecting pathways to futures we desire” (2023-11-26 everything-youve-heard-about-ai-risk-is-wrong).

Against determinism, in both directions. “Technology is not deterministic — it doesn’t just happen to us”. Humans are “incapable as a species of not innovating”, but “we do have the agency to determine what futures we aspire to and how we get there” (2025-03-30). He rejects stories that leave “a misplaced impression that we’re careering toward a hopelessly dystopian technological future, and there’s not a lot we can do about it” (FFTF p.289). He rejects the fatalism of inevitability too: “a flood can’t be halted, but it can be directed” (2025-08-31). “What happens next is up to us” (2024-03-31).

The future is fragile. His chosen object for it is a soap bubble, “full of wonder and promise, but at the same time, in need of care if it’s to survive and thrive” (2020-10-22 what-if-the-future-was-an-object). He is humble about design: “I’m not yet convinced homo sapiens have the tools and abilities to truly design the future” (2019-03-31). And we are “committed to a future where someone, somewhere, is not going to be happy” (FR p.197).

Hope that is grounded, not guaranteed. Future Rising ends: “I hope with all my heart that we do” (FR p.216). Films from the Future is “optimistic enough to believe that we have the collective ability to develop new technologies in ways that work for us, not against us” (FFTF p.290). The 2026 lecture is starker. He is “neither an AI optimist nor an AI pessimist”, has “really, really bad days”, and places his hope, conditionally, in organisations stepping up (2026-09-24 [AI-edited]).

9. How he holds his values#

Value, not values. Since the concept’s public launch he has kept apart “tangible value (not to be confused with ‘values’)” (2016-01-11). By 2023 the frame is “agnostic to particular worldviews, ideologies, or ethics” and lets organisations “define ‘value’ in their own terms” (2023-11-21 ai-and-risk-innovation). This is deliberate. It lets him work with entrepreneurs, regulators, AI labs and the Vatican without asking them to adopt his ethics. His own firm values sit mostly at two levels: - Process. Who decides, how big the “we” is, whether people act on their own terms. - Floors. Dignity, consent, and not counting anyone as less.

He rarely pushes a picture of the good life. Even flourishing comes with a qualifier, “assuming that this is something people aspire to”, and with attention to “what it does not look like” (2024-03-10 harari-teaching-future).

Without preaching. His theory: “Preach to someone about the future, and most people will shut down… But tell them a story that resonates with them, and you open their mind” (2024-01-21 how-can-stories-unlock-pathways-to). In his own self-description the refusal to preach and the commitment come in the same breath: “I do not write pieces that are polarizing, that are preachy, that push an ideology or an agenda… That said, I do place human wellbeing and flourishing at the heart of my work” (2025-01-30 [prompt]).

Enacted, not declared. His values show most in what he does. He gives tools and books away, credits students and colleagues by name, discloses AI use and publishes his failures. He keeps a student’s story private as “not mine to share”, and will not even probe his own ChatGPT account “as if I was someone else fishing for dirt” (2025-10-05).

Where his generosity ends. He is fair to Musk, Andreessen, Altman, Bostrom and others. The exceptions are specific: disregard for consent and dignity (2023-05-22, 2024-05-21), and a national AI plan that “prioritizes power before people and… values US exceptionalism over global wellbeing, And this worries me” (2025-07-23 americas-ai-action-plan).

10. Tensions and nuances#

  1. The obligation to innovate and the duty of care. He holds both and resolves the tension by context and by asking who decides, not by rule. On Arran he states his own nostalgic preference, then overrules it on equity grounds (FFTF pp.287–288). He should not be flattened into either a booster or a precautionist.

  2. Honest broker or advocate? He chose the honest-broker role but admits it “has its problems” when neutrality becomes “tacit support for not taking action” (FFTF p.246). In practice he advocates firmly for process (inclusion, disclosure, informed choice) and for dignity floors, rather than for outcomes. The balance is moving: in 2024 he questioned his decades of “technology apologetics” (2024-03-31), and in 2026 he led with risk though it was “probably not a smart move for my reputation and readership” (2026-05-10 do-not-do-this-with-ai).

  3. Defending the human while refusing human exceptionalism. He protects “what makes us ‘us’” while saying it is no “immutable truth” and extending the moral circle towards machines. The fixed point is worth and dignity, not a human essence. The position is easy to misread in either direction.

  4. Flourishing without paternalism. He is committed to flourishing but will not define it for others (“who decides what ‘better’ means”). The cost is that his positive vision is often procedural: it says who should shape the future more than what the future should contain.

  5. Pragmatism and justice. “Your risk is my risk”, and the business-facing language of 2018–20, meet innovators where they are. He names the limit himself: value-based self-interest works least well for “the people with the least leverage” (2026-07-16). His fairness commitment outruns his most adoptable tool, and he says so.

  6. Curiosity and people. Curiosity and joy are central, yet since 2009 people come first, and he distrusts curiosity without moral grounding, his own included (§7).

  7. Access and voice. His equity case for AI in college essays (2023-07-27) sits beside his conviction that his own writing is part of who he is (2023-09-20); there he did not ask whether a translated voice stays the student’s own. His 2023–24 enthusiasm for ASU’s OpenAI deal is also less hedged than his later candour about his university (2026-05-10). The values are consistent; the calibration varies with the moment.

  8. Privilege, hope and dread. He writes from tenure and security and says so (2026-08-02). The renounce-from-privilege argument is his own check on the comfortable positions privilege makes easy. He keeps hope and dread both on the page (“really, really bad days”; “I hope with all my heart”) and lets neither settle the analysis.

11. How this facet connects to the others#

12. What is distinctive, and what it brings to the AI discussion#

  1. Values are built into the definition of risk. The AI debate usually separates technical “safety” from ethical “principles”. In his framework, what people value is what counts as harm. Dignity, identity, belonging, aspiration and joy become things that can be “named, mapped and watched — even where [they] cannot be measured” (2026-07-16). For a technology whose deepest effects fall on the self, that is the right place for values to sit.
  2. The unit of concern is the person and the relationship. He treats as serious risks an AI-drafted email, a memory setting, an adviser’s AI edit of a student’s draft, and a quirky identity flattened into a stereotype. “Each of us personally” corrects a discourse that swings between civilisational catastrophe and aggregate productivity.
  3. Equity argues for innovation as well as caution. Few voices in AI say both that renouncing technology from privilege harms others and that “who decides who will suffer and who will thrive” must be asked of every optimistic trend line.
  4. Flourishing without paternalism. He challenges benevolent control: value alignment by a narrow group, persuasion towards good ends, “fixing” people. The challenge cuts against accelerationists and safety-minded aligners alike.
  5. Public intelligence and public fear are taken seriously. Where the field often treats concern as a literacy gap, he reads resistance as people protecting what they value, and he turns the deficit model back on experts.
  6. Joy and wonder can be quietly taken away. No AI risk register lists the draining of joy from science, work or learning. He treats it as a real and reciprocal loss.
  7. Being human is kept open, with dignity fixed. He is neither a humanist defending a fortress nor a transhumanist. He protects worth and dignity while accepting “a reckoning with long-held assumptions around what it is to be human” (2026-05-21).
  8. A twenty-year record. The same commitment runs from workers without tested respirators (2006), through “people matter” (2009) and the big “we” (2020), to data workers and compute-burdened communities (2026). Because he holds it with humility, humour and self-implication, people across the AI debate can hear it without feeling preached at.