F5. The public scholar: thinking with people, not at them#

Facet synthesis, 27 September 2026. Built from the perspective notes (B01–B32, FFTF-A, SUPP-self), then checked against the original posts, the text of Films from the Future (FFTF) and his papers and testimony. Only his own prose counts as evidence. AI-written passages, guest posts and co-written books and articles are excluded (the Maynard & Garbee 2019 chapter counts as his). Three sources have mixed provenance and are flagged where used: the 2026 King’s College lecture (his spoken words, smoothed by Claude from his transcript and then line-edited by him), the April 2026 andrewmaynard.net essays (his approved self-account, parts of which recycle AI-origin material), and the 2026 orphan-risks paper (his line-by-line rewrite of an AI draft). Posts are cited by date and slug stem. The key at the end gives full file names.

1. The role in his own words#

The best place to start is a sentence he wrote in May 2026, on the eve of his first sabbatical. For nearly twenty years, he says, public writing has been central to his conviction that academic work should be as accessible as possible, “Not as an add-on to my research and scholarship, but as something that’s integral to how I explore, test, and share new ideas and insights” (2026-05-17 the-nonsense-i-write). For him, public writing is not where finished scholarship goes to be explained. It is where the thinking happens. The rest of this facet follows from that.

He has described the role in several ways over twenty years, and the descriptions fit together:

• An obligation that comes with privilege. “I believe rather strongly that 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, no matter who they are or where they are” (2024-09-04 succeeding-at-science-on-youtube).
• A stance: the honest broker. Taking Roger Pielke’s term, “This is the role I try to carve out for myself in my public-facing work, 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” (FFTF p.246).
• A relationship with readers. When he described his own writing so that ChatGPT could imitate it, he wrote that he aims “to invite thinking and discussion while offering new insights – I do not write pieces that are polarizing, that are preachy, that push an ideology or an agenda, but rather these are articles that encourage people to join me in thinking deeply” (2025-01-30 ai-at-a-crossroads, his own prompt text).
• A purpose. Of the four reasons experts talk to publics (instruction, ego, impact and empowerment), he places himself with empowerment: “providing others with access to information that they are able to utilize on their own terms” (2025-05-25 why-parasocial-communication-is-important).
• An institutional form. His vision for the Future of Being Human initiative “was not one of a research center, or even an educational endeavor (although we do both), but of an initiative that would catalyze thinking at scale” (2024-12-22 why-modem-futura).
• A known price. “the cost of the decision I made to put public good before academic prestige”, which is “only OK if there really is public good that comes from my writing” (2026-05-17).

Taken together, these say that his job is to widen the circle of people who can think well about technology and the future, and to give them what they need to do so on their own terms. It is not to bring them round to his conclusions. The same shape runs through his thinking about risk. He values risk concepts that “open up new ideas and possibilities” rather than prescribe (2016-01-11 thinking-innovatively-about-the-risks-of-tech-innovation). He asks for an AI risk framing “that opens up new possibilities rather than closing down conversations” (2023-05-31 existential-risks-of-ai). The last word of that phrase matters. For him, opening up a risk concept and opening up a public conversation are the same commitment.

2. Thinking with people rather than persuading them#

Why he does not preach#

His refusal to preach is sometimes read as a matter of temperament. In fact he has reasons for it, and he has stated them repeatedly.

• What risk communication shows. “One is the realization that most people are reasonably smart.” If “you preach to people—you tell them that they’re wrong, you try to blind them with science… you’re not going to get anywhere” (TechTrends interview 2023, his quoted words, p.5).
• How stories work. “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 to exploring possibilities that they would otherwise reject out of hand” (2024-01-21 how-can-stories-unlock-pathways-to).
• The failure of the deficit model. It is still “very common to find experts who believe that ‘bad’ decisions are simply due to a deficit in understanding — which they are happy to fix!”, despite a model “repeatedly shown not to be effective” (2025-05-25).
• Respect for other people’s values. The honest-broker stance “avoids mistaking personal values for the ‘right’ values, and respects deeply held beliefs and values in others”, and it is “a path toward empowering others while trying not to let your ego get in the way” (FFTF p.246).

Behind all four sits a view of who holds expertise. “Most people have a pretty high level of expertise in what’s important to them and their communities” (FFTF p.222), and “one thing we’re all qualified to do is think about what the possible consequences of technology innovation might mean to us and the people we care for” (FFTF p.26). In 2023 he carried this straight into AI: “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). Risk understood as a threat to value becomes a reason for democratic participation. People are qualified to take part because they know what matters to them.

What invitation looks like in practice#

• Questions in place of conclusions. He closes a lecture “in the spirit of taking on the role of quantum tunneling catalyst” by leaving three questions (2021-04-09 bounded-infinities). He publishes fifteen questions for educators and deliberately does not answer them, because the answers “should be the responsibility of instructors and their schools and colleges” (2023-08-02 fifteen-questions-about-generativeai). He offers ten questions about AI and higher education “that I don’t have good answers to” (2026-04-11 ten-questions-about-ai-and-higher).
• Readers as co-investigators. He publishes prompts, drafts, data and code, and asks readers to judge. He asks them to engage on informed terms: “do read the paper and the process first before providing opinions” (2025-02-04 openai-deep-research-ai-scholarship). He puts two versions of a paper side by side and asks readers which is better, “and why” (2026-07-19 publish-or-perish-ai-vs-human-vs-human).
• Books designed to leave readers their own conclusions. Future Rising’s sixty reflections were “purposely designed not to be preachy or dogmatic, or driven by ideology… to take readers on a journey that helped them develop their own ideas” (2024-05-19 future-rising-short-history-of-tomorrow). Films from the Future aimed to take readers “on a compelling journey into the future where their thoughts and ideas were just as important as mine” (2023-07-07 moviegoer-guide-chapter-1-beginning).
• A guide who shares the journey. Asked whether FFTF would stoke anti-technology fear, he answered “Absolutely not.” The book helps readers see consequences “so we can learn together how to navigate around them” (2018-10-12 everything-you-wanted-to-know-about-films-from-the-future).
• Tools handed over, not rules handed down. His “do not do this with AI” rules are “personal”, and readers should “copy them, share them, even modify them… even if you don’t use this particular list, at least develop one that is useful to you” (2026-05-10 do-not-do-this-with-ai). He asks readers to think about a papal encyclical “and not just paste it into an LLM” (2026-05-21 magnifica-humanitas-and-being-human).
• Self-mockery about the guide’s pose. He notes that academics love Plato’s cave because it is “a pretty powerful way to explain why people should be paying attention to you if you are one” (FFTF p.155). His own guide, like Douglas Adams’s, “fails miserably on the ‘where to get the best drinks’ front” (FFTF p.290).

The relational turn: a theory of his own practice#

In 2025 he gave this practice a theory. Relational, and especially parasocial, communication earns the audience’s “permission… to talk about things that matter”. It can combine “the reach of transactional approaches with the impact of more intimate public engagement”. It aims “not just to bring about pre-determined outcomes, but to enable participants to benefit from this flow in ways that benefit them on their own terms”. His verdict on his own career: “In many ways it’s what I’ve aspired to for much of my career: communication and engagement that focuses on empowering others through relationship building” (2025-05-25). When the word “parasocial” soured six months later, he conceded he had “read the tea leaves wrong” about the word but kept the practice: “a concept that I practice in my own work, and still stand by” (2025-11-19 parasocial-relationships-problematic). The distinction he draws there, between what he got wrong and what still holds, is typical of how he changes his mind (§6).

3. What he refuses to do, and why#

His refusals have been stable since his earliest posts, and each has a reason.

He refuses to polarise. In 2018 he rejected “Luddite” as “a term of derision and censorship” and turned it into an argument for asking questions (2018-10-08 enter-the-neo-luddites-transcendence). He faults most books about technology and the future because they “take a polarized stance… These sell — people love reading about extremes. But they’re not that helpful” (2023-07-07). Asked whether he is a techno-optimist or techno-pessimist, he says the question is like asking whether he is “an oxygen pessimist or optimist” (2024-03-31 we-have-a-technology-problem-and). In 2026 he calls himself “neither an AI optimist nor an AI pessimist” (2026-09-24 being-an-academic-in-an-age-of-ai; AI-smoothed transcript). He applies the same rule to his own side. Dismissing embryo-screening advocates as a “dangerous Silicon Valley tech-bro fantasy” would be “lazy and narrow minded” (2024-04-14 welcome-to-the-age-of-swipe-and-select-embryos). Writers and academics who spin the burning of a Waymo car into a story of backlash “are legitimizing such behavior” (2024-02-18 setting-fire-to-self-driving-cars-is-bad). Refusing to polarise means turning down easy allies as well as easy enemies.

He refuses to preach. The reasons are set out in §2. Even when his moral source is religious, as when he takes design principles from the Beatitudes, he offers them as “One intriguing starting point” (2019-03-31 design-principles-for-de-marginalizing-the-future).

He refuses to fear-monger, and he has two reasons from experience. - Expert speculation can do harm. A group that tried to murder a nanotechnologist had acted partly on expert-voiced fears: “fallacious fears spurred on by speculation from experts led to real harm” (2018-11-15 even-bad-sci-fi-movies-can-teach-us-something). - Fatalism takes away people’s agency. Stories without “scientific facts and social realism” 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” (2018-11-08 films-from-the-future-the-last-chapter).

Refusing to fear-monger is not the same as refusing to talk about risk, and he keeps the two apart. “It never ceases to amaze me how many people equate talking about risk with fear mongering. And yet, it’s pretty much impossible to manage risks if you don’t talk about them” (2026-09-15 will-ai-really-kill-us-all, n.1). His 2018 AI-risk video was made “not to stoke fears (not my style), but to help prepare the ground for informed approaches to navigating them” (same post). In 2026 he put the safety message first, knowing it was “probably not a smart move for my reputation and readership”, because “everything we know about human behavior and risk management tells us that putting the safety message first is necessary” (2026-05-10). He is just as unwilling to dismiss. The extinction statement deserved “pausing to consider… rather than dismissing it as merely fear mongering or naive speculation” (2023-05-31). It “would be embarrassing if we were all wiped out by something because we didn’t have the imagination to foresee it”, and yet there are ways to deal with such risks “without running around like headless chickens” (2026-09-15, n.5).

Balance has a cost, and he pays it. Films from the Future flopped partly because “it is nuanced… in an age where certainty rather than subtlety sells” (2023-10-08 a-guide-to-responsible-innovation). In 2026 he apologises “to the attention economy gods for having the temerity to be balanced!” (2026-05-10).

Refusing polemic does not mean refusing judgement. Where dignity or consent is at stake he says so plainly. On using language models to predict criminal behaviour: “Here I should lay my cards on the table” (2023-05-22 can-large-language-models-be-used). He calls OpenAI’s handling of the Johansson voice “childish irresponsibility” (2024-05-21 openais-problem-with-the-movie-her). He admits to a “poorly disguised tone of unease” over the US AI Action Plan (2025-07-23 americas-ai-action-plan). What he rejects is the performance of certainty, not the making of judgements.

4. Convening: building rooms rather than delivering conclusions#

For him, convening is as central to the role as writing. From early on he used his platforms to bring in other people, and especially people who disagreed with him or were not being heard.

• Bringing critics in (2009). He invited civil-society critics to write on his blog, including Jim Thomas of the ETC Group, whose “highly critical position” he knew, because he “was intrigued to know more about what drove him” (2018-10-08; FFTF p.191).
• Building bridges as a job. At the Project on Emerging Nanotechnologies, “I had to be an expert in everything, and I had to be able to build bridges fast” (TechTrends 2023, p.2).
• Bridge-builders as a value. Among his secular Beatitudes: “Listen to those who are quiet, who doubt themselves, and whose voice is easily drowned out” and “Elevate people who mend rifts and build bridges” (2019-03-31).
• Quiet voices in AI governance. He warns that loud voices will fill an “insights vacuum”. He lists the quiet ones: academics with little visibility, developers “buried within layers of institutional structure”, “experts in fields that no-one has realized yet have something important to bring to the table”, and “ordinary people who sometimes have extraordinary insights” (2023-04-10 as-ai-goes-to-washington-whats-being). The world congress on AI he proposed in 2023 would include “smart and collaborative thinkers from the humanities, arts and social sciences — diversity of thought and perspective will be essential here” (2023-04-04 what-are-the-alternatives-to-calling).
• Designing the room. His live conversations set aside “the usual long-in-the-tooth experts”: “two of my three guests are undergraduate students”, with “absolutely no guarantee as to where we’ll end up going” (2023-09-18 will-ai-transform-how-we-learn). He paired strangers from different fields on purpose, “I intentionally set things up this way”, and then held back from steering: “I’m glad I didn’t” (2024-03-15 liz-lerman-and-jonathon-keats-on).
• Handing over the platform. On AI and Native American communities he asked colleagues and a former student to reflect in their own voices, and stepped back (2024-12-15 advanced-ai-challenges-opportunities-native-american-communities). A week after praising a planetary-health report, he hosted a colleague’s “alternative and compelling perspective” on it (2025-11-16 can-ai-help-redesign-the-technosphere). He uses his reach to “carve out a bit of space” for lesser-known writers (2025-01-05 five-voices-five-pieces).
• Invisible groundwork. Looking back on his World Economic Forum council work, he argues that “The ‘conditions for possibility’ are underrated as a form of contribution”, and that “Institutional experiments that fail are still formative… The ideas keep working after you stop pushing them” (2026-04-08 Before-the-Fourth-Industrial-Revolution, web essay).

What ties these together is that he convenes in order to make unexpected insight possible, not to steer a room towards a conclusion he already holds. This is where the public-scholar facet meets his principles of curiosity, play and serendipity. The rooms are designed, but what comes out of them is left open.

5. Engaging industry and policymakers without capture#

He has spent his career inside the rooms where technology is promoted and governed: congressional hearings, PCAST, National Academies committees, the Wilson Center, WEF councils, and now as a daily user of frontier AI models. The question is how he stays close without being captured. The record shows seven consistent habits.

1. He credits the other side first. His rebuttal of Eric Schmidt is titled “Respectfully Erik Schmidt…”, and in it he writes “In fairness to Eric Schmidt, I suspect that he has a more nuanced perspective… I get this” (2023-05-15). Of the debate Andreessen’s manifesto stirred up he writes “I’m all for this”, and he grants that “Andreessen gets this right” about the historical trend, before calling the document a “spaghetti mess of cherry picked ideas” (2023-10-19 marc-andreessen-ditch-sustainability). Of Musk he says he would “be remiss” to ignore him, “I get where he’s coming from”, and he admits enjoying the conversation: “I suspect this will divide my readership” (2024-08-04 7-key-takeaways-from-elon-musk-and-lex-fridman).
2. He explains by structure, not villainy. “I’ve met remarkably few scientists and engineers who would consider themselves to be unethical or irresponsible” (FFTF p.36). The problem is benevolent myopia, which he sees “in many of my peers, and even myself at times” (FFTF p.220). In 2026 the same diagnosis is applied to AI companies: “sincerity almost always operates inside an incentive field”, so “remedies aimed at sincerity, such as exhortation or public shaming, are unlikely to have the desired impact” (2026-07-16 orphan-risks-frontier-ai-maynard; the paper is his rewrite of an AI draft, but the no-villains stance matches his FFTF-era view).
3. He speaks their language without pretending the translation is enough. Responsible innovation maps onto customer discovery, pivoting and failing fast, but “In their native form, they do not necessarily lead to responsible innovation, merely to successful innovation” (2019-08-13 responsible-innovation, the Maynard & Garbee chapter). Risk innovation works “not necessarily by impressing on researchers and developers the need to ‘do the right thing’, but by making it clear that their success is intimately intertwined with how they impact… others” (2024-12-17 navigating-the-challenges-and-opportunities-of-advanced-biopreservation).
4. He keeps his red lines. “Industry can’t get AI governance right on its own” (2023-05-15, title). In 2008 he told Congress that safe nanotechnologies would not emerge “if the promoters of the technology are calling all the shots” (2008 House testimony, PDF p.8). In 2026: “Good (as in technically capable) as some of these companies are, they simply do not have the perspective… to be able to decide for humanity what this future looks like” (2026-09-24).
5. He declares his interests, often with a joke. “Waymo once sent me a pair of socks… they are nice socks… Cruise have never sent me socks”, followed by the serious part: “I have never worked for or been paid by Waymo” (2023-11-09 waymo-safety-study-shows-benefits). On AI governance history: “full disclosure, I was there at the 2017 Asilomar meeting” (2023-04-04).
6. He builds in independence. He told Congress in 2006 that “my comments here are my own personal opinions” (2006 House testimony, rec. p.50). His 2008 WEF proposal specified a body that was “Non-advocacy” and “Non-partisan”, operating “independently of the funders” (via S3). He moved from Slate and The Conversation to his own Substack because those outlets “come at an editorial price, and you end up compromising content for clicks” (2026-05-17, n.3).
7. He criticises close to home. He celebrated students getting ChatGPT through ASU’s partnership with OpenAI (2024-04-16), then condemned OpenAI’s conduct five weeks later (2024-05-21). He reviewed AI principles written by a committee he sat on and concluded they “may not have been as useful as we’d hoped” (2026-05-03 are-design-principles-for-responsible). He named ASU among universities whose ChatGPT pages say nothing about mental health (2025-11-09 universities-chatgpt-mental-health, n.2). He wrote that “Even in my own institution, it’s near-impossible to have an open and honest conversation around potential risks” (2026-05-10).

With policymakers the pattern holds over twenty years. His 2006–08 congressional testimony was forceful and quantitative. He showed the NNI’s claimed $68 million of risk research to be about $13 million of highly relevant work. Yet it opened with human questions: “Who will decide how it is used, and who will pay the cost?” (2008 House testimony, PDF p.2). It also refused the trade-off framing: innovation and harm-avoidance “need not be, nor should be, mutually exclusive” (PDF p.5). At PCAST he argued for public engagement and was told “That sounds like a very bad idea” (FFTF p.222). At another meeting of senior science advisers, suggesting that scientists listen to ordinary people was “a really bad idea” (2016-01-12 can-citizen-science-empower-disenfranchised-communities). He makes advice public: his remarks to a National Academies committee were posted “in case you’re interested”, and in them he admitted his own team’s analysis had been “somewhat naïve” (2020-10-15 the-ethics-of-advanced-brain-machine-interfaces). He offers frames rather than complaints, as in a formal NSF submission that opens with Inception (2023-09-25 building-a-better-futures-tough). He stays non-partisan under partisan pressure. The morning after the 2024 election, the ideas mattered “whoever’s at the political helm” (2024-11-06 ai-in-a-world-of-trump). The AI Action Plan was “probably not as bad as some will be making out, or as great as others will argue” (2025-07-23). On the Genesis Mission he declined to “fall back on protectionist critique”: universities claiming a place “from a position of entitlement is not going to fly” (2025-12-07 universities-genesis-mission).

A caveat. Closeness sometimes shows. His post on ASU’s OpenAI partnership (2024-01-18) and his advocacy of enterprise ChatGPT for universities (2023-08-29) are among his least hedged writing, and his early enthusiasm for Deep Research (February 2025) ran ahead of his usual caution. He usually corrects course in a coda, an update or a later post, but not always. His main safeguard against capture is transparency, not distance: he discloses, shows his working and lets readers judge. This matters more now that the tools he studies are also the tools he writes with.

6. Candour and changing his mind#

Changing his mind in public is part of his method, not an occasional event. He described his attitude directly: “When I discover I’m wrong… I find it amazing, and that becomes fuel to my creativity” (TechTrends 2023, pp.2–3). The record runs through the whole period:

• 2011: “Five years ago, I was a proponent of a regulatory definition of engineered nanomaterials. I have changed my mind” (Nature 475:31).
• 2018: He criticises his own field, which sold nanotechnology with fudged science and “went along with it because of the promise of funding”, “for all my cynicism” (2018-02-21 the-bs-and-the-science-of-nanotechnology). On superintelligence: “I freely admit that I may be wrong” (FFTF p.170).
• 2020: He tells a National Academies committee that his own analysis was “somewhat naïve” (2020-10-15).
• 2024: “I tend not to sign open letters like this as they’re often deeply naive… But in this case I did” (2024-02-25 ai-rollercoaster-of-a-week). He publicly corrects “suffering” after a disability scholar’s challenge (2024-01-30 first-in-human-trial-of-neuralink-bci).
• 2025: “Clearly I read the tea leaves wrong back in May!” (2025-11-19).
• 2026: On the early OpenAI APIs his students showed him: “It’s a toy. It’ll never catch on… I was wrong” (2026-09-24). After AI models preferred a machine’s version of his paper to his own: “Ouch!”, and the experiment left him “more uncertain than ever” (2026-07-19).

His candour takes several forms beyond reversals. - He publishes his failures. Films from the Future was “By most standards… a failure”, and colleagues told him it was “professionally embarrassing” (2023-10-08). Future Rising sold “a mere 596 copies” (2024-05-19). His posts “rarely make much of a dent” (2026-05-17). - He names his own motives. He worries his public writing may be “an ego trip… (and maybe it is — although I hope it isn’t)”, and reports being called “a self-centered egotist” in a faculty meeting (2026-05-17, n.4). “Being a consummate academic, my first question was far more opportunistic” (2025-12-07). - He tells the audience how to calibrate him. “You should know a little about where I’m coming from, just so you can calibrate… What I see scares the life out of me sometimes. Some of this is rational, some of it is irrational” (2026-09-24; AI-smoothed transcript, line-edited by him). He treats his fear as information and says how much weight to give it. - He shows where AI was involved. He publishes prompts “typos and all”, separates what a model wrote from what he wrote, and gives an AI sole authorship when that is the truth (2026-09-04 anthropics-fable-5-1-as-an-original-scholar).

Two self-examinations go deeper than any single correction, because they question the role itself. In 2024 he admitted to “a form of technology apologetics that’s been part of my professional life for decades — and one that intentionally seeks to support others where they are at in their thinking. But I’m beginning to question its wisdom” (2024-03-31). His habit of meeting colleagues where they are may have kept him from telling them that technology is part of what we are, not something outside us. In 2026 he asked whether his public writing reaches anyone at all: “my worry is that what I write has no real value to those who might read it… because it makes no sense to them”, and “to write without care for your readers is a very academic trap to fall into” (2026-05-17). Few public scholars examine their own practice this openly.

The candour does real work. In 2020 he argued that institutions must earn trust by being trustworthy, which “takes awareness, empathy and humility, and a willingness to embrace the messiness of being human” (2020-12-15 why-trustworthiness-matters). His openness is how he tries to meet that standard himself. It is also the humility he asks of AI developers, practised in public by the person asking.

7. Responsibilities to students, readers and future generations#

Students#

• Making room for those easily lost. His pizza seminar was for students who might become “buried and lost”, “especially if you’re a first generation student or a bit of a loner”, in spaces where they are “heard and validated, no matter what their backgrounds or perspective” (2023-05-02 pizza-and-a-slice-of-future).
• Treating students as sources of knowledge. Students’ views can be “better informed, more sophisticated, and more mature, than those from some of my colleagues!” (2023-09-20 what-do-college-students-think-about-chatgpt). Their aspirations are “easy to dismiss as the dreams of the young and naive”, and he declines to dismiss them (2019-08-13). A student’s question about how to navigate a future you are not exposed to “does put the onus back on us as educators” (2025-05-06 what-students-really-think-about-ai).
• Standing up for them. He was “invariably disappointed when my students make the effort to do stuff like this and it lands on deaf ears”, after a class letter to the Tesla Bot team went unanswered (2023-08-28 dear-mr-musk-can-i-have-my-book-back). He asks advisors whether their AI use is “robbing them of their dignity” (2025-10-26 ai-misuse-in-student-advisor-collaborations). He argues that universities providing ChatGPT owe students a duty of care over mental health (2025-11-09).
• Discretion. A student’s story that revealed a new risk is “not mine to share” (2025-10-05 when-chatgpt-turns-snitch).
• Obligation over tradition. “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). Among his ten questions: “What do we owe our students in an age of AI?” (2026-04-11).
• Access and the long view. He made his book available as a free podcast “to give students an alternative to purchasing the book” (2023-07-03 a-new-tech-podcast). He notes that a teacher’s recommendation took him “40-odd years” to follow, a reminder that “sometimes these things take decades to percolate through” (2026-09-24).

Readers#

He treats accessibility as an ethical duty, not a marketing concern. Bridging “the gap between my academic work and people who might potentially benefit from it” is “a public responsibility that I take very seriously” (2024-05-19). In 2016 he proposed that excellence in online resources for casual learners should count “toward academic tenure and promotion” (Nature Nanotechnology 2016, nnano.2016.167, p.735). The same year he argued that public universities should add “a fourth leg of community service” to how faculty are judged (2016-01-31 public-universities-must-do-more). He treats readers’ time and understanding with care: transcripts for readers “simply hoping for a quiet read” (2025-05-11 designing-responsible-technological-futures), tl;drs, and apologies for length. He also treats readers as adults who can form their own judgements, which is the point of the honest broker.

Future generations#

The obligation to the future is written into why he writes at all. “Responsibility here also means that we collectively need to give a damn about the future we’re creating, and whether it’s the future we want for ourselves and for generations to come. This is important—it’s partly why I wrote this book” (FFTF p.286). The initiative’s founding statement, which he wrote, imagines “what it might mean to be human a hundred years from now” (2024-12-22). In a post about the digital archive, he asks “who determines how we appear to future generations, or even whether we appear at all?” (2019-09-04 how-to-ensure-our-digital-legacy). His sense of obligation reaches across place as well as time: renouncing technology “from a position of privilege” risks “denying too many people without the same privileges the chance to make their own decisions” (FFTF p.288). So the duty he feels to future generations is to keep their choices open, not to decide for them. That mirrors what he owes readers now: the means to decide “on their own terms”.

8. Nuances and tensions#

(a) Honest broker or advocate? The record is U-shaped. Before Congress in 2006–08 he was a forceful advocate for strategy, money and leadership. In 2018 he chose the honest broker as his self-description. By 2026 he was naming the strain himself: “With AI, that’s become harder. The questions are more urgent, the stakes are higher, and the temptation to advocate for particular positions is stronger” (2026-04-12 Stick-Figures, web essay). Two things reconcile these positions in his own terms. First, his honest broker always had a stated limit: where silence would become “tacit support for not taking action”, advocacy is justified, preferably through institutions rather than lone voices (FFTF pp.246–247). Second, what he mostly advocates is process and capacity: who is at the table (“‘it’s complicated’ simply isn’t an excuse for excluding people”, 2023-05-15), and people’s ability to navigate. He also advocates, as a risk professional, putting the safety message first. His “mixed feelings” that a WEF proposal was softened from “policy” (to prescribe) to “intelligence” (to inform) (2026-04-08) show that he never thought informing was always enough.

(b) Relational persuasion and manipulation. He champions stories and parasocial relationships because they get past the defences that preaching sets off. The same mechanism is at the centre of his fear about AI: machines that learn “our cognitive vulnerabilities” (FFTF p.177) and fluent language that slips past epistemic vigilance. He acknowledges that history is “replete with examples” of parasocial manipulation (2025-05-25). When he asked GPT-4o how it would nudge him towards his own values (2024-10-20), the plan it produced used many of the tools of his own public scholarship: curiosity, open questions and invitations to join a community (my observation, drawn from B19; he does not make the link). His implicit answer is honesty about what each party is (“generative human-machine interactions where each is clear about what it is”, 2024-09-29 chatgpt-as-author-part-2-the-conversation) and communication aimed at the listener’s own ends. He has not yet set out a full test.

(c) Everyone has a stake, but who acts? “Everyone has the right to play some role” (2023-05-15) sits beside “you cannot hand a problem of this magnitude over to everyday people and say, ‘Solve it for us’” (2026-09-24). He reconciles them by dividing the roles. Publics are experts on what matters to them and must have standing. Designing responses falls to institutions. In 2026 he casts universities as “an accelerator and a catalyst for what other organizations can do” (2026-09-24). That is his public-scholar model made into an institution, though he argues it more than he demonstrates it.

(d) Accommodation and challenge. His generosity towards people has not changed. His patience with institutions, including his own, has thinned since 2024: the “technology apologetics” doubt, the academic “crab bucket” he saw while chairing promotion and tenure, and universities as “guardians of the past more than leaders toward the future” (2026-08-30 do-universities-have-a-place-in-bill). The tone darkens, “really, really bad days” (2026-09-24), but he still refuses contempt.

(e) Presence and ego. Relational scholarship requires him to be visibly present, which exposes him to the charge that it is self-promotion. He neither denies the charge nor lets it stop him. He names it, argues that academia confuses empowerment with ego (2025-05-25), and asks himself the question again on sabbatical (2026-05-17).

(f) Closeness to the technology. He uses frontier models daily, and some of his writing about them is AI-assisted and openly labelled as such. His own warning that AI is “beginning to train us to think like them” (2026-07-19) applies to his workflow too. His answer so far is to keep the final judgement for himself, line-edit against his own sources and show the process. That is honest, but it is a partial safeguard.

(g) Reach and nuance. He accepts small audiences as the price of nuance, and keeps trying new routes to reach people: stick-figure videos, podcasts, book-as-website, AI-legible text. Whether this adds up to public good is the question he has taken into his sabbatical, and he has said openly that he does not know the answer.

9. How this facet connects to the others#

• Risk as a change of mindset, and as a threat to value. His readiness to treat publics as experts in what they value comes from the same source as his reading of resistance as the protection of value, not ignorance. His examples include the mob in The Man in the White Suit, the Luddites, and COVID-vaccine hesitancy, which he recasts as a question of whether institutions are trustworthy (2020-12-15). The reciprocal idea in his risk thinking, that a company’s success depends on what it threatens in others, is also what lets him speak to industry and to its critics at once.
• Navigating rather than managing. He describes his role in navigational terms: a guide on a journey, a “pilot” through “perilous waters” (2024-03-31), a convener of routes. He does not describe himself as an authority who manages an outcome.
• Play, curiosity and serendipity. His convening formats (pizza, unscripted conversation, pairing strangers) are built for serendipity. His humour is one of the main ways he avoids preaching: it lowers defences without closing questions.
• Stories and film. Films are “democratic entry points” (2026-04-12 Stick-Figures) and spaces “neither preachy or judgmental” (2023-10-08). His theory that stories open minds where sermons close them is the story-facet form of the same commitment.
• Quantitative foundations and humility. His candour and his habit of telling audiences how to calibrate him are the public face of his humility against false precision. His refusal to amplify implausible fears is a plausibility check applied to his own public speech. The congressional testimony shows the quantitative risk scientist already acting as a public scholar.
• What it means to be human. What he protects in his readers, their ability to think for themselves and on their own terms, is what he fears AI may wear down (cognitive surrender, bypassed epistemic vigilance). His public role is therefore partly a response to the risk he studies. It tries to equip people to keep their own judgement.

10. What is distinctive, and what it offers the AI discussion#

1. Public writing as the scholarship, chosen at a cost, with a theory behind it. Few people in the AI debate treat public writing as their main research medium, “not as an add-on”, and also give a worked account of why: relational communication aimed at empowerment “on their own terms”. Fewer publish the costs, including the colleagues who do not read them, the 596 copies, and the suspicion of ego.
2. An honest broker who admits the strain. The AI debate is dominated by advocates for safety, acceleration and ethics. He defaults to equipping people rather than recruiting them, says openly that the pull to advocate has grown, and when he does advocate, it is mostly for process (who is at the table) and capacity (people’s ability to navigate).
3. Refusing both hype and dread while insisting on talking about risk. He holds that “talking about risk” is not fear-mongering but the route to benefit: “you can only begin to realize the benefits of a technology if you understand what can possibly go wrong” (2026-09-24). This lets him take catastrophic concerns seriously without joining a camp. He did not sign the pause letter or the extinction statement, and he published his reasons both times.
4. A long memory of engagement. He watched “it’s complicated, leave it to us” backfire with GM crops, argued for public engagement at PCAST and was told it was “a very bad idea”, and was shaped by nanotechnology’s relative success in bringing people in. He brings that experience to AI’s more exclusionary governance.
5. Convening as design. He brings in quiet voices, unlikely disciplines, students and critics, sets up the conditions, and lets conclusions emerge. He values “conditions for possibility” over planting flags.
6. Engagement without capture. He manages this through fairness, disclosure, structural rather than villain-based explanation, meeting innovators in their own language without letting that language define responsibility, and criticising his own institution and even his own committee’s work.
7. Candour as a source of trust. He dates his changes of mind, publishes his failures and shows how AI was involved in his work. In a debate full of confident prediction, he practises in public the humility he asks the AI field to show, and he applies it to the role of public scholar itself.
8. Responsibility framed as agency. Students, readers and future generations are owed the means to decide for themselves, not conclusions and not only protection. That is also how he would have society approach AI: keep the conversation, and the futures it can reach, open while there is still time.

Source key#

Posts (corpus file = slug + .md)

Book, papers and other sources

• FFTF = Films from the Future (Mango, 2018), printed pages, from working/maynard/book/FFTF-*.txt.
• 2006 House testimony = Resources/maynard-papers/web/2006_House-Science-Testimony_Nanotech-EHS-Research_2006-09-21_Maynard-statement.pdf (hearing-record page).
• 2008 House testimony = Resources/maynard-papers/web/2008_House-Science-Testimony_NNI-Amendments-Act_2008-04-16.pdf (PDF page).
• TechTrends 2023 = Resources/maynard-papers/papers/2023_TechTrends-Interview-AI-Responsible-Innovation_author-copy.pdf (his quoted words only).
• Nature 475:31 (2011) = Resources/maynard-papers/Maynard supplied/475031a.pdf; Nature Nanotechnology 2016 = …/Maynard supplied/nnano.2016.167.pdf.
• 2026-04-08 Before-the-Fourth-Industrial-Revolution = Resources/maynard-papers/web/2026_Before-the-Fourth-Industrial-Revolution-Institutional-Prehistory.md.
• 2026-04-12 Stick-Figures = Resources/maynard-papers/web/2026_Stick-Figures-Sci-Fi-Movies-Obligation-to-Make-AI-Accessible.md (approved self-account; weight as self-presentation).
• The 2008 WEF proposal wording is taken from working/maynard/supplement/S3-testimony-institutions-2020science.md, §B1.