Late Lessons, Jensen Huang and AI
Markdown version

B11 notes: 2023-09-04 to 2023-10-20 (18 posts)#

Reading notes on Andrew Maynard’s Substack posts in batch B11. Only his own prose is treated as evidence. Quotes are exact, including his original typos and curly punctuation.

Batch context. Autumn 2023, about ten months after ChatGPT launched. Five posts are wrappers for the last episodes (10 to 14) of The Moviegoer’s Guide to the Future podcast, in which he reads chapters of his 2018 book Films from the Future. (There are no Modem Futura posts in this batch.) The rest are short commentaries: on public engagement (the PCAST letter), AI and the UN Sustainable Development Goals (co-written), students and ChatGPT, his NSF comments on advanced technology transitions, the film The Creator, a Nature Nanotechnology commentary on lessons from nanotechnology for AI, and Marc Andreessen’s Techno-Optimist Manifesto. There are also two announcements for the new ASU Future of Being Human … Unplugged live-stream series.

Evidence base. Only the post text is in the corpus. Podcast audio, the Zombified interview, the live-stream videos, and the linked Nature Nanotechnology and The Conversation articles are not. Claims about those are limited to what he says about them in the posts.

The five Moviegoer’s Guide posts (09-08, 09-15, 09-22, 09-29, 10-06) all end with the same boilerplate paragraph, “About Films from the Future”. It is his own prose and it states the book’s purpose: - to avoid the polarised stance of most tech-and-future books (“we’re either all going to die … or technology is going to save the world”); - to weave together many kinds of expertise, “including the arts and humanities”; - to favour dialogue over preaching; - to be about “how all of us can think differently about our roles in ensuring the future we’re building is better than the past we leave behind”.

It is noted once here and not repeated under each post.

Relevance summary#

Date Slug Relevance
2023-09-04 why-public-engagement-is-so-important high
2023-09-08 living-in-a-material-world medium
2023-09-10 can-large-language-models-reason none
2023-09-11 its-time-to-get-serious-about-ai-and-sdgs medium
2023-09-15 weaponizing-the-genome medium
2023-09-18 will-ai-transform-how-we-learn low
2023-09-20 what-do-college-students-think-about-chatgpt high
2023-09-22 riding-the-wave-of-climate-change low
2023-09-25 building-a-better-futures-tough high
2023-09-28 the-creator-and-being-human high
2023-09-29 living-by-more-than-science-alone low
2023-10-02 responsible-ai-lessons-from-nanotechnology high
2023-10-05 2023-nobel-prize-for-chemistry low
2023-10-06 looking-to-the-future low
2023-10-08 a-guide-to-responsible-innovation medium
2023-10-12 the-rise-of-ai low
2023-10-19 marc-andreessen-ditch-sustainability high
2023-10-20 do-ais-dream-of-electric-people low

HIGH#

2023-09-04 — why-public-engagement-is-so-important — “Why public engagement is so important for advanced science and technology”#

Provenance. His own prose. It includes one long block quote from the PCAST letter of 29 August 2023, and the bold emphasis in that quote is his. He cites Brian Wynne’s “May the Sheep Safely Graze?”. The post calls it “Brian Wynn’s seminal 1995 article”; the linked reference is dated 1996.

Argument in his terms. - The gap between talk and walk. Public engagement in science and technology is “oft talked about, but rarely practiced effectively”. He speaks as an insider (“I should know”), and says the gap between talk and walk is often “crushingly large”. - Why the PCAST letter matters. He welcomes it as a possible turning point. It marks a shift from assuming public value arises naturally from science and technology to recognising that it “will only happen through intentional two-way engagement with key communities”. - Resistance persists. Decades of scholarship (Wynne) have not removed “deep resistance to listening to and learning from diverse communities”. He gives a current AI example, linking his earlier post on Eric Schmidt (slug erik-schmidt-ai-regulation): claims that AI is “simply too complicated for non-experts to understand!” - Why the pushback is wrong. It shows a lack of understanding of how science and technology serve people “in a democratic society”. It also carries the risk of hubristically ignoring the very communities technologists say they are helping. - The film case. He uses The Man in the White Suit (chapter 10 of his book): a well-meaning scientist who “forgot to engage with the people and communities he claimed he was working to benefit”. - Practical models. He names three: - Participatory Technology Assessment, from ASU colleagues at CSPO/ECAST, including the NASA asteroid work cited by PCAST; - Public Interest Technology and PIT-UN, which put public interest ahead of “the whims and profits of individuals and companies”; - NISE Net, which grew out of NSF funding for informal nanotechnology engagement. - Hypocrisy of developers. Developers “tout the importance of public engagement in the abstract, but strenuously resist it in practice”. - Firmness. He states the conclusion firmly: a better future needs “broad, sophisticated, and meaningful ways of engaging” the people who will build and live in it.

Concepts. Two-way engagement versus a deficit or “trickle-down” assumption about public value. Participatory Technology Assessment (structured, informed deliberation between experts and publics). Public Interest Technology. Informal STEM engagement (NISE Net). Hubris.

Analogies. Nanotechnology (NISE Net), used literally as a proven model of engagement infrastructure. The Man in the White Suit, used conceptually as a parable. AI is the current case where engagement is being resisted.

On AI and AI companies. The only direct AI point is his rejection of the expert or industry claim that AI is too complex for public input. This is an expertise-and-publics argument, not a hazard argument.

Governance. Democratic, participatory decision-making. Experts and developers are not enough.

Quotes. - “this will only happen through intentional two-way engagement with key communities” - “a very real risks of hubristically ignoring the very communities many scientists and technologists claim they are working to help” - “tout the importance of public engagement in the abstract, but strenuously resist it in practice”

2023-09-20 — what-do-college-students-think-about-chatgpt — “What do college students really think about ChatGPT?”#

Provenance. His own prose. It reports an unplanned class discussion with undergraduates in his emerging tech course, from a range of disciplines. The one student quote is flagged as a paraphrase.

Argument in his terms. He is careful to say this is anecdotal (“an n of very few”, “I’m probably over-interpreting”, “I may be wrong of course”). Three observations: 1. “Not a big deal.” About a third or fewer used ChatGPT regularly. Few used it extensively. He sensed students “normalizing” it as one small app among many rather than a transformative technology. 2. “I came to college to learn.” Many saw using ChatGPT as contrary to why they came to college. He criticises the assumption, common among educators, that students are just there for grades: “we forget that many are here because they want to learn”. 3. “My writing reflects who I am.” Some students saw their writing or art as self-expression and “recoiled” from handing it to an app. He agrees and discloses his own practice.

His conclusions: - Educators urgently need to include students in these conversations. - Student views are sometimes more sophisticated and mature than those of “some of my colleagues”. - Attitudes have changed within two to three months, so decisions risk resting on “out of date understanding”. - He suspects “we’ve been asking the wrong questions” in surveys. - Educators should engage students now rather than wait “until we feel we’ve got our own house in order” (by which time it will be too late).

Concepts. - Socialisation and normalisation of a technology: its “socialization” moving fast. - Writing as identity and self-formation. - Students as stakeholders with knowledge experts lack. This is his public-engagement argument applied inside the university.

On cognition and formation. This is the clearest statement so far of a personal stake in keeping human writing un-outsourced. He treats writing as constitutive of the self: to relinquish it “would be to diminish myself”. This is worth tracking against his later practice (the corpus notes that AI and the Art of Being Human was deliberately written with AI assistance). In 2023 his stated default is not to use ChatGPT for his own writing.

On education. He grants that generative AI “used in the right way” can enhance learning. His main target is institutional and faculty assumptions, not students.

Quotes. - “I typically don’t use ChatGPT myself when I write for exactly the same reason.” - “The way I write is personal. It reflects who I am, and to relinquish that to a machine would be to diminish myself.” - “we’re in danger of making decisions on ChatGPT and generative AI in learning and education that are based on out of date understanding”

2023-09-25 — building-a-better-futures-tough — “Building a better future’s tough when you don’t know where you’re going”#

Provenance. His own prose. He writes an introduction, summarises the CHIPS Act provisions (with statutory citations), and then gives a “lightly edited” version of the comments he submitted to the NSF Request for Information on the roadmap for the Directorate for Technology, Innovation, and Partnerships (TIP). The submission is a formal policy document in his name. The full PDF is linked but not in the corpus.

Argument in his terms. - The Inception metaphor. In the introduction, AI, quantum technology, gene editing and similar technologies “seemingly rewrite the rules of what is possible”, like the reality-bending scenes in Inception. Navigating this needs “new understanding, tools, and mindsets”. - Tipping point. “we are at a scientific and technological tipping point in human history”. Futures are departing “in radical ways from past norms, trends, and expectations”. Advanced technologies transform what is possible “in highly non-linear ways and at unprecedented rates”. - AI as the accelerant. AI foundation models alone have shortened the timescale of social disruption “from years to months”. - Precariousness. Humanity is “more interconnected, more resource-constrained, and more socially and environmentally precarious as a species, than at any previous point in history”. - Failure modes are hidden. The dynamic between huge promise and “potentially catastrophic economic, social, and environmental failures” is driven by conventional thinking, naïve assumptions, siloed understanding and limited approaches to responsible innovation. It is “veering toward potential failure modes” that conventional thinking cannot overcome and that established, often outmoded, approaches to innovation obscure. The claim is that current frameworks do more than fail to fix risks: they hide them. - What is needed. New thinking, framings, knowledge, philosophies, skills, jobs and organisational structures. New collaboration across expertise and communities. A new generation of leaders to steer transitions “toward more vibrant, promise-filled, and equitable futures”. - Advanced Technology Transitions as a crosscut. He proposes ATT as a crosscutting frame for the TIP roadmap. He defines technology transitions and notes that existing work is domain-specific (for example energy transitions) or loosely tied to practice. An “integrated, transdisciplinary, and use-inspired approach to advanced technology transitions is lacking”. - Past transitions as material. Gene sequencing and editing, nanotechnology and now AI “provide rich material for novel and impactful research”.

Concepts and frameworks. - Advanced Technology Transitions (ATT). Technology transitions are defined as “theories, frameworks, and practices, that support and enable economically and societally beneficial development, adoption, and use, of emerging technological capabilities”. “Advanced” marks transformative, converging, non-linear technologies. - His proposed research domains. This list is effectively his research programme: - general ATT theory; - “failure modes, best practices, and emerging principles” from historic transitions; - models of complex sociotechnical systems; - responsible innovation; - Public Interest Technology; - governance “spanning the spectrum of public engagement, soft law, agile governance, hard-law regulation”; - equity and equality; - public engagement and democratic decision-making; - the arts and humanities as “modulators” of transitions; - sustainable development and decarbonisation; - ethics; - “Novel theories, models, and approaches to risk in the context of advanced technology transitions” (his risk-innovation thread); - foresight; - misinformation and disinformation; - job loss, gain and displacement; - skills, learning and education.

Analogies and comparisons. Gene editing, nanotechnology and AI are treated as a class of “advanced technology transitions” whose histories yield generalisable lessons. This is a structural comparison. The emphasis on learning from historic failure modes is explicit.

On AI. AI is one of a family of transformative technologies, and currently the fastest disruptor. It is not singled out as unique in kind.

On AI risk. The framing is systemic: economic, social and environmental failure modes, misinformation, jobs, equity. No existential-risk language.

Governance. A full spectrum from engagement and soft law to agile governance and hard law. Transdisciplinary. Includes entrepreneurs, businesses, policymakers and civil society. Public input is built in (he notes the CHIPS Act mandates it).

Quotes. - “we are at a scientific and technological tipping point in human history” - “Emerging AI foundation models alone have seen the timescale associated with social disruption move from years to months” - “obscured through established – and often outmoded – approaches to technology innovation” - “Novel theories, models, and approaches to risk in the context of advanced technology transitions.”

2023-09-28 — the-creator-and-being-human — “The new AI movie The Creator is a must-see for anyone grappling with what it might mean to be human in an AI future”#

Provenance. His own prose. It includes one quote from Alex Godfrey’s review in Empire (not his).

Argument in his terms. - An anti-Terminator. It seems “inevitable” that an “AI-as-existential-threat blockbuster” should appear while AI risks are “hotly debated”. The Creator inverts the genre: “It’s an anti-Terminator movie”. Humans are the aggressors, and the AIs “are simply looking to be free”. - Personhood beyond biology. The film is about “dignity and autonomy in age where personhood is no longer tied to biology”. - His own view on machine moral status. He notes a “groundswell of pushback” against the idea that machines could ever be more than tools “with no rights, no autonomy, no dreams or aspirations”. He treats that pushback as an assumption to be challenged. If human worth “transcends our biological “wetware,”” then sentient machines with a right to freedom and autonomy are a real possibility. He hedges on timing: “We’re still a long way from such a future.” - Skynet as projection. The Skynet scenario reflects “that all-too-human trait” of treating whatever is unlike us or outside our control as a threat. The film shows propaganda (by the US military) that turns benign AIs into threats. - Love. Reading the film through its working title, “True Love”, he suggests that love, and worth as a person, are central to being “human” “in the broadest sense”, independent of substrate. - Uncertainty. He is modest about relevance: “How relevant this will be … I don’t know”. He is firm that “we already need to be thinking seriously about what it means to be human in a future that transcends our biological origins”.

Concepts. Personhood and worth independent of substrate. Dignity and autonomy. AI rights. Existential-threat narratives as projection of human fear of the other. Love as a marker of personhood. “Being human” extended beyond biology.

Analogies. Elon Musk’s Tesla Bot is mentioned as a real-world parallel to the film’s humanoid AIs: helpful “but not too capable”. Terminator and Skynet are the foil. The comparisons are conceptual.

On AI. Open, and speculatively sympathetic, to the possibility of machine sentience and moral status. This is consistent with his 2018 Ex Machina chapter’s question about AI rights. He is sceptical of the reflexive “tool/slave” framing.

On AI risk. He implicitly downgrades the “AI as existential threat” story. It is a narrative that can carry human aggression and propaganda. He does not claim that AI poses no danger.

Quotes. - “a movie about what it means to have dignity and autonomy in age where personhood is no longer tied to biology” - “if what gives us worth as humans transcends our biological “wetware,” there is every possibility that sentient machines will one day become a reality” - “we already need to be thinking seriously about what it means to be human in a future that transcends our biological origins”

2023-10-02 — responsible-ai-lessons-from-nanotechnology — “Responsible AI: Lessons from Nanotechnology”#

Provenance. His own prose. It is a short pointer to a co-authored commentary with Sean M. Dudley in Nature Nanotechnology (“Navigating Advanced Technology Transitions Using Lessons from Nanotechnology”, 2 October 2023) and a companion article in The Conversation. One phrase is quoted from the commentary. Neither article is in the corpus.

Argument in his terms. - Why nano and AI belong together. They may seem unrelated, but they share “surprising similarities when it comes to avoiding failures in a society where the success of transformative technologies depends on far more than technical knowhow alone”. - His credentials. He co-chaired the interagency Nanotechnology Environmental and Health Implications (NEHI) working group and was science advisor to the Wilson Center’s Project on Emerging Nanotechnologies. - What nano taught. “Plenty of mistakes” were made, but the field learned two things: - the importance of working with the arts, humanities and social sciences, alongside nano scientists; - that “broad stakeholder and public engagement are absolutely critical to success”. - AI is not learning these lessons. “Development is still being driven by a small group of experts and companies who believe that they have all the understanding they need”. There is “a reluctance to engage a diversity of voices”. He calls this “a serious mistake” to be corrected “as soon as possible”. - The GMO lesson. GMOs were “a masterclass in how naivety, hubris, greed, and a lack of broad engagement, can create near-insurmountable roadblocks to progress”. Early responsible-nanotech investment “drew heavily on lessons learned from the GMO debacle”. - Higher stakes. For AI, “the stakes here are far higher than they were with either nanotechnology or GMOs”. - The remedy. Investment in understanding advanced technology transitions that “bridge disciplines and sectors” (quoting the commentary).

Concepts. Advanced technology transitions. Lessons across technology waves. Transdisciplinarity. Broad stakeholder and public engagement as conditions of success, not only of safety. Hubris.

Analogies and comparisons. This is the batch’s central past-technology comparison. Nanotechnology and GMOs are used structurally, as governance and engagement precedents: how a transformative technology succeeds or fails in society. They are not literal comparisons of hazard. He does not compare the physical or toxicological risks of nanomaterials with AI harms here. The framing of GMOs is notable: the failure was a roadblock to progress caused by naivety, hubris, greed and non-engagement. This is a pro-innovation reading of responsible innovation.

On AI companies and leaders. He criticises a narrow, self-confident group of experts and companies for believing they have all the understanding they need.

Governance. Broad, diverse, transdisciplinary engagement, driven by learning from the past.

Quotes. - “Development is still being driven by a small group of experts and companies who believe that they have all the understanding they need.” - “a masterclass in how naivety, hubris, greed, and a lack of broad engagement, can create near-insurmountable roadblocks to progress” - “if anything, the stakes here are far higher than they were with either nanotechnology or GMOs”

2023-10-19 — marc-andreessen-ditch-sustainability — “Marc Andreessen: Ditch sustainability and technology ethics if you want a better future”#

Provenance. His own prose. Short phrases from Andreessen’s Techno-Optimist Manifesto are quoted as its claims.

Argument in his terms. - His own position. He welcomes the debate (“I’m all for this”) and places himself as pro-technology: “I live and breathe advanced technologies in my work. I revel in their potential.” He is equally aware of harms and “the growing need to be innovative in how we navigate emergent risks”. His work is about innovating in ways that are “socially responsive and responsible”. - Overall verdict. The manifesto is “a spaghetti mess of cherry picked ideas that does anything but inspire optimism in me”. “Responsive” and “responsible” “seem to have no place in Andreessen’s market-driven and permissionless techno-future”. - A manufactured enemy. He rejects the “we are being lied to” framing: “this doesn’t sound like the world I live in”. Andreessen “manufactures a stage on which he can do battle with a mythical foe”. - The listed “enemies”. Andreessen names sustainability, the SDGs, social responsibility, trust and safety, tech ethics, risk management and the precautionary principle as enemies. Maynard wryly adds that responsible innovation would have been on the list “given half a chance”. He wonders whether Andreessen does not understand how these ideas “put people and humanity at the center”, or is just being provocative. - Technology’s record cuts both ways. Technology “can and does lead to harm”; “pretty much every problem we face in the world today has its roots in previous technological innovations”. - Partial concession. Andreessen “gets this right” that the long-run trend has been improvement through innovation. But this holds “especially if you’re selective in your metrics … and you brush over some of the details”. - Technological foreshortening (his named concept). Collapsing the complexity of technological history into over-simple upward trends, so that “the pain and suffering in the detail is lost”. Zoomed out and utilitarian, we have clear material and medical gains. Zoomed in, the human cost is stark, especially seen through “different values and expectations around what brings meaning to life and personhood”. Foreshortening becomes a way of saying some will suffer for “someone else’s dream of a better future, and that’s OK”. He rejects this utilitarianism, “especially when I begin to ask who decides who will suffer and who will thrive”. - Why past success does not guarantee the future. Consequences are “increasingly unpredictable” because we are pushing against “a finite planet within finite resources, but with near-infinite ambitions”. His analogy is the elastic band, which cannot be stretched forever. The key point: “past technological successes are no guarantee of future wins”. - Three reasons ethics and sustainability matter. 1. The “granularity” of impacts. 2. “The growing non-linearity of the relationship between technological cause and societal effect”. 3. Uncertainty about whether we can always “innovate our way out of the problems caused by previous good ideas”.

For these reasons sustainability, trust, ethics and responsible innovation “are not the enemies of a vibrant and promise filled future, but critical components of achieving it.”

Concepts. - Technological foreshortening. - Granularity of impacts: aggregate progress versus distributed harm. - Non-linearity of cause and effect. - Limits of a finite planet (the elastic band). - Problems as the legacy of “previous good ideas”. - Permissionless innovation, now applied to Andreessen. This echoes his earlier critique of Thierer. - Responsive as well as responsible innovation. - “Who decides”, as a question of justice.

Analogies. Technology history in general. The elastic band is a physical metaphor for the limits of extrapolation.

On tech leaders. This is his most direct critique of a Silicon Valley figure in the corpus so far. It is sharp but not ad hominem. He grants Andreessen’s historical trend point and welcomes the debate.

On precaution. He does not defend the precautionary principle specifically. He defends the cluster of ideas Andreessen attacks, and his closing list names sustainability, trust, ethics and responsible innovation. This fits his usual stance: he is neither a precautionary maximalist nor a permissionless one.

Quotes. - “pretty much every problem we face in the world today has its roots in previous technological innovations” - “technological “foreshortening” where the deep complexities of our technological history are reduced to over-simplistic trends” - “who decides who will suffer and who will thrive” - “These are not the enemies of a vibrant and promise filled future, but critical components of achieving it.”

MEDIUM#

2023-09-08 — living-in-a-material-world — “Living in a Material World and the movie The Man in the White Suit. The Moviegoer’s Guide to the Future Episode 10”#

Provenance. His own prose. The post introduces a podcast episode in which he reads chapter 10 of Films from the Future (2018). The chapter text and audio are not in the corpus. The chapter’s section titles are listed: “There’s Plenty of Room At The Bottom”, “Mastering the Material World”, “Myopically Benevolent Science”, “never Underestimate the Status Quo”, “It’s Good to Talk”. This is his own book narration, not Modem Futura.

Argument. - He teaches a responsible innovation class on the 1951 film every year, and students enjoy it despite its age. - The film is a tale of a “well meaning but desperately naive and myopic scientist” who assumes his vision is universal. Things go wrong largely because “he didn’t think to engage with the people his work impacted”. - He recommends the film to “anyone who’s serious about being in the business of developing new technologies”. - The episode also covers nanotechnology. He says “there are deep parallels between the story in the film and modern day nanoscale science and engineering”, including the role of public engagement in safe and beneficial technologies.

Concepts. “Myopically benevolent science”. Never underestimating the status quo (incumbent interests, labour, everyday life). “It’s good to talk” (engagement). Parts of the boilerplate are relevant: the book as an anti-polarisation project, and arts and humanities alongside science.

Analogies. A polymer-textile innovation compared to nanotechnology, structurally, through engagement and stakeholders. Nano is also discussed literally (“nanotech insider baseball”).

Quotes. - “he didn’t think to engage with the people his work impacted” - “dialogue and discussion are far more important than preaching”

2023-09-11 — its-time-to-get-serious-about-ai-and-sdgs — “It’s time to get serious about artificial intelligence and the UN Sustainable Development Goals”#

Provenance. Co-written: “Written with guest contributor José Lobo” (ASU School of Sustainability; urban economic development). The post uses a joint “we”, and the authors’ contributions cannot be separated. The sections on invention as combinatorial search (Paul Romer), labour transitions and informal settlements closely match Lobo’s expertise. Treat the post as co-authored evidence of positions Maynard chose to sign and publish, not as his sole voice. It also cites Vinuesa et al. (2020), the Li et al. water-use paper, the WEF Top 10 Emerging Technologies list and a PNAS paper.

Argument. - AI is “a highly complex multi-edged sword” for the SDGs. - Harms include data-centre water and energy use (ChatGPT “drinking” 500 ml per 20–50 exchanges), bias, access and inequity. - Vinuesa et al. found roughly a 2:1 ratio of positive to negative impacts, but that was before LLMs. - Labour: AI will reshape skills and occupations “in both advanced economies and in the Global South”, at a time of growing inequality. - Energy: AI could make grids smarter, but also risks “accelerating energy and resource use” and behavioural changes that threaten gains (a rebound-style concern). - AI is “hype aside, a transformative technology”, and treating it as “a minor perturbation” in sustainability work “would be a serious mistake”. - The positive case. Invention is search through a space of combinatorial possibilities. Generative AI offers “an augmentation of the very process by which humans find solutions to problems” (“augmented intelligence”), not just automation. Suggested uses include LLMs with open geographic data for informal-settlement planning, sanitation and decarbonisation. - Call to action. SDG 17 supports an urgent, global, multi-stakeholder discussion. The UN should “center artificial intelligence”.

Concepts. Multi-edged sword. Automation versus augmentation. Invention (discovery of novelty) versus innovation (adoption and diffusion). Combinatorial possibility space. Technology transition affecting labour.

On AI. A transformative, dual-edged technology with a material resource footprint. This is the first time environmental costs (water, energy) appear as an AI risk in the corpus notes.

Governance. Multi-stakeholder, international (UN), across science, the private sector, civil society and policy.

Quotes (co-authored). - “AI is a highly complex multi-edged sword when it comes to the SDGs” - “to treat it as a minor perturbation to what is considered the important business of building sustainable futures would be a serious mistake”

2023-09-15 — weaponizing-the-genome — “Weaponizing the Genome and the movie Inferno. The Moviegoer’s Guide to the Future Episode 11”#

Provenance. His own prose, plus a block quote from his own 2018 chapter 11 (on Spanish flu, pandemic likelihood and virus traits). Both are his writing, at different dates.

Argument. - He reflects that the chapter, written in 2018, was “eerily prescient” about a bioengineered or natural pandemic. The 2018 excerpt says a 1918-scale pandemic is “highly likely”, and recommends identifying which viruses might mutate into threats “so we can get our defenses in order”. His 2023 comment: “Sadly, we weren’t prepared”. - The chapter focused on gain-of-function research. That was niche in 2018 but is now familiar “because of concerns that COVID may have had its origins in such research”. He reports the concern without endorsing any origin claim. - The chapter now feels “a little dated” because “we’ve literally lived the future it warns against”. - The enduring issue is how scientists navigate “the fine line between research, social responsibility, and activism”. He recommends Roger Pielke Jr.’s “honest broker” concept and his Substack.

Concepts. Gain-of-function research as dual-use risk. Preparedness. The honest broker (Pielke Jr.). The science–activism tension.

Analogies. Biotech and pandemics, discussed literally.

Change of view. None. This is a retrospective validation of 2018 foresight, with the lesson being preparedness failure.

Quotes. - “Sadly, we weren’t prepared, and we suffered global (and continuing) consequences as result.” - “how scientists struggle to navigate the fine line between research, social responsibility, and activism”

2023-10-08 — a-guide-to-responsible-innovation — “A guide to responsible innovation like no other …”#

Provenance. His own prose: a reflective post on Films from the Future as the podcast series ends.

Argument. - Why it matters now. Responsible innovation is “complicated”, but it has “never been more important”. The AI wave reveals “just how vulnerable we are to powerful technologies that are wielded without foresight or understanding”. We live where “small missteps can have outsized and irreversible consequences”. - Everyone needs a systems view. Everyone who influences or is affected by emerging tech needs “a nuanced and interconnected understanding of how these technologies intersect with society”. That includes the public, developers, investors and policymakers. The understanding must cross disciplines, educational attainment and socioeconomic status. - Stories as engagement. After “over twenty years” of experimenting with communication and engagement, he turned to stories, and to sci-fi films in particular. They let people from very different backgrounds explore challenging ideas together, “neither preachy or judgmental”, with movies as “the catalyst for the journey rather than its destination”. - A frank self-assessment. The book was “a failure” commercially: it fit no category, the title misled, and it was “nuanced … in an age where certainty rather than subtlety sells”. Yet it contains “some of my most thoughtful work”. - Academic snobbery. Colleagues told him to stop talking about a popular book because it was “professionally embarrassing and undermined my credibility”. This is a critique of academic norms about public-facing scholarship. - Prescience. He claims the book foreshadowed AI, brain-machine interfaces, predictive justice, COVID and gain-of-function research, and that most topics are “at least as relevant now”.

Concepts. Responsible innovation as needing broadly shared “interconnected understanding”. Storytelling and science fiction as engagement tools. Irreversibility. Nuance against polarisation.

Criticises. Academic gatekeeping of popular writing. A market that rewards certainty. Implicitly, AI developers wielding power “without foresight or understanding”. He links to his nanotech-lessons post and his earlier posts on responsible innovation for entrepreneurs and on education.

Quotes. - “we’re increasingly living in a world where small missteps can have outsized and irreversible consequences” - “in an age where certainty rather than subtlety sells” - “stop talking about the book because it was professionally embarrassing and undermined my credibility”

LOW / NONE#