B02 perspective notes: how Andrew thinks (2018-08-26 to 2018-11-15, 17 posts)#

What this batch is#

Sixteen of the 17 posts are the weekly run-up to the publication of Films from the Future (FFTF) on 15 November 2018. Most are abridged chapter openings, sometimes with italic framing notes he added when posting. The two exceptions among the sixteen are a self-interview Q&A (2018-10-12) and a stand-alone essay on Transcendence (2018-11-15). The seventeenth post is an original op-ed on tech-company social risk (2018-09-03). Several were first posted on his earlier blog, 2020science.org.

All of the main text is his own prose, written before generative AI. Text that is not his: - the book blurbs (Patrick Lin, Cori Lathan); - quoted declarations and other writers (UN, the 1997 cloning declaration, Stiglitz, Sahakian, Giurgea, Balzac, Ehrlich, Jim Thomas); - the ASU Risk Innovation Lab pull-quote, which is his lab’s institutional line.

There are no Modem Futura posts in this batch. FFTF page numbers are not available from these excerpts, so citations below are by post.

Because these are the opening pages of chapters, the batch shows unusually clearly how he opens a problem. The promotional wrapper (“please buy the book!”) is part of the texture. It is not the substance.

1. How he thinks here#

He opens with a concrete, often odd, real-world story, then moves out to the structure underneath it. Almost every chapter starts with a specific moment rather than a thesis: - the “Baby Eve” cloning claim (2018-09-06 never-let-me-go); - an email he received advertising a “Minority Report-like” trust test (2018-09-10 minority-report); - a journal article whose title caught his attention because it was “unusual” (2018-09-13 limitless); - Occupy’s Zuccotti Park (2018-09-19 social-inequity); - a Musk tweet (2018-09-27 being-human); - a dinner-table question at Asilomar (2018-10-04 superintelligence); - fourteen hangings at York Castle (2018-10-08 enter-the-neo-luddites); - topless nano-pants protesters (2018-10-11 a-nanoscale-perspective); - Ehrlich’s 1969 prediction (2018-10-18 the-honest-broker); - Hurricane Irma emptying Barbuda (2018-10-25 environmental-resilience); - a fourteenth-century friar (2018-11-01 contact-occams-razor); - the view over the Firth of Clyde (2018-11-08 films-from-the-future-the-last-chapter).

The pattern is story, then film, then technology, then the social dynamics underneath. The film is explicitly “a jumping-off point”, not the destination (2018-08-26 in-the-beginning).

He often starts from his own life. He first watched 2001 aged sixteen on a black-and-white TV at his parents’ dinner party (“don’t be such a jerk”, he tells his younger self), and saw Jurassic Park as “a newly minted Ph.D.” (2018-08-26 in-the-beginning; 2018-08-26 jurassic-park). He puts himself in the scene as a curious participant rather than an authority looking down.

He reframes the question someone else has already posed. - Luddite is recast from an insult into a defence of fair and socially just use of technology: “the Luddites were not fighting against technology, but against its socially discriminatory and unjust use” (2018-10-08 enter-the-neo-luddites). - Resilience is recast from bouncing back into protecting value (see §3). - Cosmetic neurology is questioned as a label: “I’m not sure the term “cosmetic neurology” works”, because “cosmetic” trivialises the stakes (2018-09-13 limitless). - The question of whether a film is good becomes whether it is useful: “This doesn’t necessarily make the bad movies good. But it can make them useful.” (2018-11-15 even-bad-sci-fi-movies).

He builds analogies from shared structure, not surface resemblance. - The 1951 Man in the White Suit and the 2005 nano-pants protest are “a preemptive parody” of Transcendence (2018-10-11). - Luddite frame-breaking maps onto the gig economy and AI-driven job loss (2018-10-08). - Caffeine maps onto smart drugs (2018-09-13). - The film’s “lysine contingency” maps onto real biocontainment (2018-08-26 jurassic-park). - A rubber band maps onto ecosystem resilience (2018-10-25). - Of today’s computers he writes: “Our quaint, two-dimensional digital circuits are about as useful to superintelligence as the brain cells of a flatworm are to solving the unified theory of everything” (2018-10-04).

He tests plausibility the way a physicist would. He separates “what is imaginable and what is plausible” (2018-10-04 superintelligence). He takes exponential extrapolation apart: “At some point, all exponential trends stop” (2018-11-15). He points out that a man cured of lifelong blindness could not instantly process vision, because his brain could not “instantly cope” (2018-11-15). Occam’s Razor becomes a rule for prioritising futures (2018-11-01).

He experiments to find out. “Naturally, I took the test.” He scored a Trust Index of nineteen, then asked colleagues to take it too; one scored two. The playful self-experiment exposes a real flaw: a training set of 117 white-collar felons whose traits resemble “curious, independent, and personally-motivated academics” (2018-09-10 minority-report). In the same spirit he asks a student research assistant directly about smart-drug use, and she “sheepishly” produces a pill (2018-09-13).

He holds tensions open rather than resolving them early. - Could against should (2018-08-26 jurassic-park; 2018-09-27; 2018-11-01). - Social responsibility and financial success (2018-09-03). - Nostalgia against access. On Arran he affirms that happiness lies in “food, shelter, warmth, and good company”, then immediately calls it “a sentimental illusion” held from privilege (2018-11-08). - Creativity against critical thinking (2018-11-01, below).

Curiosity and creativity drive the thinking. The Q&A gives the clearest statement of method in the batch. Films are “pretty bad at predicting future technologies”, but “the imagination and creativity they represent is a quite wonderful catalyst for breaking down preconceived ideas and institutionalized thinking” (2018-10-12 everything-you-wanted-to-know). The last chapter adds the other half: films help “not because they are accurate or prescient, but precisely because they are not tethered to scientific accuracy”. Their creativity “and dare I say it, their entertainment value” open our eyes when “seasoned with feet-on-the-ground thinking” (2018-11-08). The method is to use play and imagination to loosen fixed frames, then use grounded science to keep them honest.

He also makes this a principle about science itself. “Critical thinking alone is almost inhuman in its cold impartiality. On the other hand, creativity on its own leads down a path of fantasy and delusion.” Combined, they make science “a disciplined pathway to awe and wonder” (2018-11-01 contact-occams-razor).

Serendipitous juxtaposition gives him evident pleasure. “who’d have thought that Dan Brown and Roger Pielke Jr. would end up sharing the same pages!” (2018-10-18). The book’s original working title was “The Moviegoer’s Guide to Not F***ing Up The Future”, an irreverent “nod to Douglas Adams” (2018-10-12).

2. What matters to him#

Human dignity and worth. Never Let Me Go, which “moves me to tears each time I watch it” (2018-10-12), is about “what it means to have value as a person” and “what separates humanity from inhumanity”. He warns against “the questionable morality of convenient technologies” (2018-09-06).

Fairness: who benefits and who pays. - On Earth in Elysium, the poor are “suppressed by” technology rather than served by it (2018-09-19). - The true Luddite legacy is ensuring technologies benefit “all, not just a privileged few” (2018-10-08). - He deliberately chose films that “weren’t just a bunch of movies about white guys” (2018-10-12).

Agency and a voice for everyone. The book is for “anyone who stands to be impacted … and who wants a say” (2018-10-12). He wants us to “nudge them toward the future we want, rather than one that someone else decides for us” (2018-11-08). “When we’re dealing with science that potentially touches everyone, we all become stakeholders” (2018-11-01).

An obligation to innovate, which is the counterweight to caution. “I would go so far as to say that we have an obligation to explore new ways of using science and technology to improve the world”. Renouncing technology “from a position of privilege” risks denying others “the chance to make their own decisions” (2018-11-08).

Awe and wonder. Dinosaurs are “awe-full in the true sense of the word”, and Jurassic Park is “a celebration of the might and majesty of the natural world” (2018-08-26). The Mandelbrot set’s “stunningly psychedelic images” stayed with him from his undergraduate physics (2018-08-26 jurassic-park). A bioprinted ear was “a quite remarkable series of images” (2018-09-19), and some tech is simply “cool” (2018-09-19; 2018-10-12).

What frustrates him: - myopic science and hubris. Stratton “never thought to ask anyone else what they wanted or needed” (2018-10-11); - “entrepreneurial arrogance” (2018-08-26 in-the-beginning); - the Neuralink line “No neuroscience experience is required” (2018-09-27); - pseudo-scientific prediction, “which I’m still incredulous over” (2018-09-10); - zealotry of any kind; - hyperbole, as in Ehrlich’s and Brown’s “knack for a turn of phrase that transforms hyperbole into an art form” (2018-10-18).

Generosity towards people he does not blame. “I’ve met remarkably few scientists and engineers who would consider themselves to be unethical or irresponsible”. Their problem is absorption in their work, not malice (2018-08-26 jurassic-park). The workers who chase Stratton are “not because they are anti-science, but because they are pro- having jobs that pay the bills” (2018-10-11).

3. Risk as a way of thinking#

This batch shows the risk-innovation mindset forming. It is mostly used to reframe, with a pragmatic, business-facing edge in one post.

Risk landscape, and navigating it (2018-09-03 tech-companies-need-a-social-risk-reboot). - Companies must “navigate an increasingly complex and convoluted risk landscape”. - Drawing on his work across nanotechnology, gene editing, self-driving cars and AI, he argues they need “a total reboot in how they think about risk”. - The key move is to go beyond “safe enough” and regulatory compliance, towards whether products threaten “livelihoods and lifestyles”, discriminate, or “threaten people’s beliefs and aspirations”. The point is “how they potentially threaten what’s important to others within society”. - This is risk as a threat to value in embryonic form. He adds that these social risks “are often poorly understood”, which foreshadows orphan risks, though the term is absent. - The remedy is described as a mindset: “recalibrating their corporate mindset”. It is also, frankly, a pitch for the Risk Innovation Lab and a business case (“competitive edge”). Here the concept works partly as a mental model and partly as a service offer.

Threat to value as a mental model that opens possibilities (2018-10-25 environmental-resilience). This is the clearest example in the batch. He starts from his materials-science foundation, where resilience means a rubber band returning to shape. He then notices the limit of that model: it slips into thinking “that suggests change is bad, and should be resisted”. - He asks “what are we trying to preserve here”. His answer is “what is considered to be “of value””, which may be the environment or health, but “just as equally” “someone’s ability to make a living, or their deeply held beliefs, or even their sense of self-identity and worth”. - The reframed concept is “in itself, adaptable to a changing world”. It “helps illuminate the complex landscape around issues like climate change where different, and sometimes hidden, values may be threatened”. Resilience becomes “getting to where we want to be, rather than simply trying to stay in the same place”. - This is navigating rather than managing, stated through resilience rather than risk. He builds on the physical-science concept rather than throwing it away, and credits Seager et al. (2013) for the systems view.

Threats to value explain backlash. Technologies that “threaten deeply held beliefs and worldviews” provoke resistance “as people protect what’s important to them” (2018-10-08). The White Suit shows “the complex ways in which seemingly beneficial inventions can sometimes threaten the status quo” (2018-10-11). He also calibrates physical hazard separately. The nano pants’ chance of “substantial harm was pretty negligible”, yet the social-value dynamics were real. So low hazard and high social risk can coexist.

A new technology breaks an old safety model (2018-09-27 being-human). The old assumption “that, if everything went pear-shaped, someone could always just pull the plug” is now “quaint and hopelessly outmoded”, because cyber and physical systems are converging. Cyberspace is a human-made “fifth dimension” built while “making up the rules along the way”. This is an early case of a technology that no longer fits the prior risk frame.

Humility against false precision, in both directions. - Chaos theory broke the belief that “if we have enough information, we can predict the outcomes of our actions” (2018-08-26 jurassic-park). - Exponential extrapolation fails because “Minute errors in projected trends can lead to predictions being out by thousands, or even hundreds of thousands, of years” (2018-11-15). - On AI futures: “we’re all looking into a crystal ball … none of us really understand” (2018-10-04). - Occam’s Razor gives “a lower probability … but not a zero probability”. It is “never … more than an aid to decision-making” (2018-11-01).

His quantitative foundation shows in how he handles trade-offs and evidence gaps in Limitless: side effects are not tracked; “the research here is patchy” (2018-09-13). He uses that foundation as judgement, not as an algorithm.

AI risk, told apart by kind. - He doubts superintelligence as implausible, but “I freely admit that I may be wrong” (2018-10-04). - He finds Ex Machina’s scenario “not that implausible”: AIs “learning about our biases and limitations, and using these to manipulate us” (2018-10-12). - He flags training-data bias through his own self-test (2018-09-10) and the hacking of connected minds (2018-09-27). - His scepticism is aimed at speculative extrapolation, not at AI risk as such: “not to say that we shouldn’t take the risks of powerful technologies gone wrong seriously — far from it” (2018-11-15).

4. Scholarship and public writing#

The book is the scholarship, serialised in public. He posted weekly excerpts to publication day, cross-posted from 2020science.org. The Q&A is a self-interview in which he explains his own design choices: the movie selection criteria, the “meta-narrative” of convergence, and what was deliberately left out (2018-10-12).
It is tied to his teaching. “I use these twelve movies to teach an undergrad class about technology and society”. Ghost in the Shell is “the one they struggle with the most” (2018-10-12).
It is tied to his research and policy work. He links his own Nature Nanotechnology “reboot” argument and his lab (2018-09-03), and his Conversation piece “If Elon Musk is a Luddite, count me in!” (2018-10-08). He took part in the 2017 Asilomar beneficial-AI meeting (2018-10-04). He cites peer-reviewed work (Seager et al. in Risk Analysis; Chatterjee in Neurology; Sahakian in Nature) and Pielke’s “honest broker” alongside a Dan Brown thriller.
He works across disciplines by habit. In one batch he draws on physics (chaos, the Mandelbrot set), materials science (resilience), labour history (the Luddites), medieval philosophy (Occam), bioethics (cloning declarations), economics (Stiglitz), neuroscience (Kennedy’s self-implant) and film.
He handles evidence honestly and with a sense of proportion.
He parses statistics rather than repeating them. The “up to 50 percent” figure for student enhancement includes “caffeine-based drinks and tobacco”, while use of off-label drugs is “just a few percent” (2018-09-13).
He gives both cloning declarations a fair hearing and notes the irony of where one is published (2018-09-06).
He admits uncertainty plainly: “it’s surprisingly tough to get a hard fix” (2018-09-13).
Accessibility is built into the method, not added on. He uses spoiler warnings, and tells readers “you definitely don’t need to have seen the films” (2018-10-12). He invites them in with “buckle up, hang on, and enjoy the ride!” (2018-08-26). Everyday analogies do the work: “The dog ate my homework” against “I forgot” (2018-11-01). He wants films that are fun, because “life’s too short for watching deeply earnest films that are boring as old socks!” (2018-10-12).

5. His role as he sees it#

A guide on a shared journey, not an oracle. The book “helps open the eyes of readers to potential consequences, so we can learn together how to navigate around them, and build the future we want” (2018-10-12). It is “a journey of discovery that uncovers the often-hidden links between ourselves and our technologies” (2018-11-08).
Against leaving it to experts, though he is one. “it’s all too easy to leave to “experts” to grapple with. Yet I fear that this is, in itself, an abdication of responsibility” (2018-11-08). He also shows expert speculation causing harm. The group ITS bombed nanotechnologists partly on the basis of Bill Joy’s essay: “fallacious fears spurred on by speculation from experts led to real harm” (2018-11-15).
An insider who stays agnostic. At Asilomar, dinner companions tried “to convert me to becoming a superintelligence believer”. His comment: “I sometimes had to remind myself that I was at a scientific meeting, not a religious convention.” Yet he respects “Bostrom’s brilliance as a philosopher” (2018-10-04), and calls Kurzweil and Drexler “pin-sharp” even while rejecting their extrapolations (2018-11-15).
He actively seeks out critics. In 2009 he invited civil-society critics, including Jim Thomas of ETC, to write for his blog, because “I wanted to get a better understanding of how they saw the emerging relationship between society and innovation”. He was “intrigued to know more about what drove” them, and admits “rather a soft spot” for the Luddites (2018-10-08).
Constructive towards industry. He speaks to tech companies and start-ups as people who can “ride this wave” and be “successful and socially responsible” (2018-09-03). He criticises entrepreneurial arrogance (Musk/Neuralink, Hammond) without demonising business.
What he refuses to do:
He will not stoke anti-technology fear. Asked whether he is worried the book could be seen that way: “Absolutely not.” (2018-10-12).
He will not accept dystopian fatalism. Films without “scientific facts and social realism” can 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).
He will not use “Luddite” as “a term of derision and censorship” (2018-10-08).
He will not join the zealots on either side (2018-10-04).
Changes of mind in the open.
It took “another 30 years” to see how powerful film is for thinking about the future (2018-08-26 in-the-beginning).
In 2008, “AI wasn’t even on my radar” (2018-10-04).
He corrects his own Arran nostalgia within a single paragraph (2018-11-08).
He was “in two minds” about including The Day After Tomorrow (2018-10-25).

6. What is distinctive#

He uses creativity deliberately to break institutional frames, then disciplines it with science. Most technology-risk writing either predicts or moralises. He uses films, including bad ones, because they are unconstrained (“it’s the very absurdity of the movie that makes it useful”, 2018-11-15), then applies physical-science plausibility tests. This is the 2018 form of his later claim that play and creativity are needed to escape stovepiped thinking.
He thinks about value before he thinks about hazard. By 2018 he was already treating social risk as threats to livelihoods, beliefs, identity and aspirations. That explains backlash (the Luddites, the White Suit mob) as people protecting what matters to them, not as irrationality. It also recasts resilience as moving towards valued futures rather than preserving the present. Few AI-risk voices of the time framed things this way; most were split between hazard-and-safety and ethics-principles camps.
He sorts AI risks by plausibility instead of lumping them together. In 2018 he was sceptical of superintelligence and exponential “singularity” logic. At the same time he flagged as plausible AI that learns our biases in order to manipulate us, bias in training data, hackable connected minds, and the loss of the ability to “pull the plug”. And he held his scepticism with explicit humility (“I may be wrong”).
He is neither booster nor critic. He holds an obligation to innovate (grounded in equity and the danger of renouncing technology from privilege) together with a firm insistence on responsibility, participation and humility.
He is accountable for speculation. He recognises that fear-driven expert speculation can itself cause harm (ITS), and that hyperbole does damage. This comes close to an ethics of public scholarship.
His tone is warm, self-deprecating and playful about serious things. Examples: a Trust Index of nineteen; “Goodness knows what our guests were thinking!”; of Occam, “he’d have probably got along well with Palmer Joss”. This tone is part of the method. It lowers defences and invites readers into questions they might otherwise avoid.

7. Posts that best reveal how he thinks#

2018-11-08 films-from-the-future-the-last-chapter. His purpose and role: the obligation to innovate, the critique of privilege, the refusal to leave it to experts, creativity “seasoned with feet-on-the-ground thinking”, self-correction in real time, and place (Arran) as a spur to reflection.
2018-10-25 environmental-resilience-and-the-day-after-tomorrow. The clearest case of reframing a concept around value and future direction. He builds on a physical-science concept rather than discarding it, and the result is a mental model that is “adaptable to a changing world”.
2018-10-04 superintelligence. A plausibility test held with humility. He is an insider at Asilomar, agnostic about superintelligence, respectful of Bostrom, wary of zeal, and ready to admit “I may be wrong”.
2018-11-01 contact-occams-razor. A heuristic used as an aid rather than an algorithm. Critical thinking and creativity have to be combined. Science is a “disciplined pathway to awe and wonder”, and everyone is a stakeholder.
2018-10-08 enter-the-neo-luddites-transcendence. He reframes the Luddites as defenders of what people value, seeks out critics, and refuses derision.
2018-10-12 everything-you-wanted-to-know-about-films-from-the-future. His own account of method: creativity as “a quite wonderful catalyst for breaking down preconceived ideas and institutionalized thinking”, links to teaching, “Absolutely not” to fear-stoking, and irreverence.

Close runners-up: 2018-09-10 minority-report (experimenting on himself in public to expose algorithmic bias) and 2018-10-11 a-nanoscale-perspective-on-the-man-in-the-white-suit (a calibrated view of hazard alongside a vivid picture of social-value threats and scientific hubris).