Subscribe now

Of course, there are some things about using a human brain to reason about the world that will be broadly similar no matter whose particular brain we’re using. We’re far from being unbiased observers, and the biases that seem biologically wired are catalogued as Cognitive Bias. There’s a sprawling Wikipedia article documenting hundreds of these, but this handy article from Buster Benson argues that really, there are four things our brains do as a result of being highly-efficient problem-solving devices: selectively compress/filter information when there’s too much to process, expanding and extrapolating information when there’s not enough, needing to make decisions faster than we can fully think through them, and the fact that our memory is tiny compared to the amount of information we ingest1. The unique part you can bring here is awareness of and systems to compensate for the worst effects of these biases.

Every futurist will also have a personal style that shades the work they do. Andy Hines wrote 23 years ago in Foresight about 10 dimensions of style that a futurist embedded in an organization brings to the work. He recommends, rather than trying to build a full foresight research function into an organization, that you focus on translating work from outside (or from consultants) to the org’s specific context. But many of the decisions will hinge on the specific company and your own personality. How much do you want to try to “stretch” the organization’s culture? How are you going to effectively lead others in change? How do you plan to build and maintain a network of relationships inside the company and within the foresight community? What tools do you bring to the table, from how to handle the arc of a project to specific activities in workshops? The answers will likely affect the research done, and will certainly affect the outcomes.

I’ve mentioned before that the Swinburne school (previously an actual school) of futurists in Australia, represented by Peter Hayward and Richard Slaughter, largely centers on creating better futures by raising the level of consciousness or moral development of populations. Hayward argues that lower levels of development, from egocentrism to concern with group status, are largely incompatible with foresight. As we move to higher levels, the ability to hold ambiguity and take a wider perspective increases the likelihood that foresight can take hold. So not only is it the case that foresight work will be limited by (and should take into account) the moral development of the client, but also your own. I’m normally pretty skeptical of the “everyone who disagrees with me is just not as advanced as I am” angle, but taking the values of clients into account is definitely good practice.

Scan Hit of the Week

OK I have three, I can’t help myself.

• First off: plasma-based laundry gun. What if this revolutionizes the raw denim market, so you never have to wash? Cool new laundromats? Lawsuits from the Laundry Gun Challenge on future TikTok?

• Cool crossover: solar flares, one of my favorite wild cards, made GPS, used for autonomous vehicles, suddenly wildly inaccurate. Imagine a street full of self-driving cars simultaneously making a sharp correction. Apparently it’s already causing self-driving tractors to wander around fields.

• Drones blowing stuff up at civilian airports is now a thing. This one seems to have been state-initiated, but soon this will be an option for anyone with a grievance and a few thousand dollars. In Ministry for the Future, terrorists use tactics like this to make it too risky to take a commercial flight, or even for billionaires to be out in public, as a bid to reduce global emissions. But what if every political movement uses it as a trump card? Ultimate big vs small problem.

  Share

As I transition from the carefree days of summer to my penultimate semester of school, I’m again looking back on all the cool stuff I covered.

Subscribe now

Books

The clearest theme this summer was moving beyond the futures of 10 or 20 years from now to much bigger timeframes, and the macro changes that might come over those horizons. The Coming Wave looks at the next ~50 years of potential changes in AI, biotechnology, and related tech. The Clock of the Long Now tries to expand human consideration to 10,000 years, and seed the cultural institutions to do so (results so far are mixed). The Precipice has both beat, trying to get through this current rough patch of history so we can spread humanity as an intergalactic civilization over the coming trillions of years (trillenia?1). Long time horizons require a different approach, as scanning hits give way to big history.

Other books this summer: Blindsight, about the cosmic horror associated with first contact with a different mode of intelligence (like the AI we're rapidly developing), and Could Should Might Don't, a survey of futures thinking way beyond the strategic foresight world.

TV and Film

Four projects this summer, all about how to build plausible and useful futures. Years and Years was by far the most grounded, AI: Artificial Intelligence was an opportunity to talk about how everything that can happen probably should happen in your future images2, Arco, a great example of using intermediate futures rather than just make a big leap that strains credulity, and Pluribus, another way to think about how AI might function as networked intelligence.

Interviews

I had three great interviews this summer. First I talked to Bryan Alexander about AI and education futures. Then I met David Staley, whose One-Sentence Futures serve as pithy little provocations. Interviewing other futurists is a great way to see the diversity of the field, in terms of entry points, styles, and methodologies.

But I also love the chance to interview people actually building the future. A short detour on vacation and a gracious host allowed me to interview Ethan Howard, the creative mind behind Shunned at a Funeral, the Country-Metal-Gaelic Christian band and most interesting project in the rapidly evolving landscape of AI music. This interview has become my most popular post by a factor of 5, and I love it when my toddler-level investigative journalism skills and determination pay off. Do cool things that scare you!

Sundry

One of my favorite ways to do futures research is to visit places and get my own feel for what's different and might spread. Aligned with my family vacation, I was able to spend a few days in San Francisco, and in addition to Alcatraz and ice cream, I was able to do some amateur anthropology.

I also spent some time experimenting with how AI could be used to more formally simulate beliefs and learning, mixing various theories from the program, but I have to say that OpenClaw kind of stopped being fun after a few weeks.

Up Next

This semester I’ll be taking Alternative Perspectives, looking at different ways of framing and thinking about futures, and what each of those lenses can offer. I’m excited, because I love wrestling with new ideas and trying out different ideologies. If you like it too, stick around and we’ll learn together.

Scan Hit of the Week

If you haven't seen the controversy bubbling up all around the United States regarding Flock cameras, it seems like it's become a tangible and very real flashpoint in a popular uprising against technology-enabled surveillance states. In addition to the vandalism, many people are showing up to their local governments to make deeply shareable content in opposition. This connects to some of the themes from The Precipice and Power and Progress about this being a particularly volatile time in our history and the malleable direction of technological development.

Share

For whatever reason, of of the themes of my summer self-led intellectual wanderings has been the big picture of humanity, for various definitions of the word “big”. From the questions in Power and Progress revolving around whether AI is a general-purpose technology that will make life better or worse for us, to The Coming Wave and whether AI and biotechnology will create a brand new but destabilizing and dangerous phase of society, to The Clock of the Long Now trying to elevate our perspective to 10,000 years. But Toby Ord’s The Precipice puts them all to shame, with a mission to help humanity achieve its full potential over trillions of years and across numberless galaxies.

Subscribe now

Ord is a philosopher, one of the founders of the Effective Altruism movement, and is crystal clear about his focus. When he talks about existential (hereafter just X for brevity) risk, he means both more and less than you might think. Extinction counts, but not a nuclear war that kills 99% of humanity (unless that spreads us out too much to rebuild); not a centuries-long dark age; climate change is pretty bad now, but unlikely to end humanity or even cause a big-league mass extinction. But on the flipside, we could collapse civilization permanently, or keep going but get locked into a lame but stable dystopia, either one enforced by a totalitarian surveillance state or one people love but is a waste of our potential, like a Brave New World or lotus-eater situation1, which would be more or less equivalent to extinction in his eyes.

First, why should we care about the future far beyond our own lives? Ord’s point is that with so much potential ahead of us, humanity is potentially at the very beginning stages of its history, in terms of size, scope, and quality. I asked about the math associated with longtermism when talking about the Long Now, and in Appendix A2 he shares his math! It doesn’t address most of my points, but instead emphasizes that discounting for inflation doesn’t make sense if the good in question is existence. He proposes that we value the future based on the likelihood that it will be there, which oddly means that the better we do at reducing X risk, the more future there is likely to be, and the more we should fight to reduce risk even further. But again, to my point, if the whole future is worth something like 1000 times the present year, most of that is the responsibility of future generations. Ord never claims that this is the only thing we should care about, or even the most important thing, but he proposes we spend as much money on it as we do on ice cream (currently he thinks we’re less than 1% of the way there). So I would need to do some more math, but Ord’s ice-cream-level strategy sounds reasonable; people will always find a today-audience by working on today-problems, but someone should speak up for the vast potential future.

What are the threats, and how do they compare? The cool natural ones like giant meteor, supervolcano, or nearby supernova, are extremely low (and the volcano is doing nearly all the work). Nuclear war or climate change, as mentioned, aren’t very likely candidates, but a bioengineered superpandemic, bad AI, or other bad things we might invent are much higher. Of course, the odds listed, like 1 in a million chance of an asteroid in the next hundred years, are exceptionally squishy, but likely have nearly the right number of zeroes on them. All told, Ord pegs the odds of a catastrophe this century at about 1 in 6, conveniently the same odds as Russian Roulette.

Many of the scenarios (nuclear winter, supervolcano, medium-sized meteor, etc) work by blotting out the sky long enough for us to starve from lack of crops, or from a superdisease spreading. These set my brain cycling to work out a plan: OK so if we build a worldwide network of underground settlements powered by geothermal power, each growing their own food in vertical farms, each containing enough knowledge to restart civilization and oh no I just recreated Silo from first principles. Plus, according to Ord, slightly over half of the risk of an X event this century is from misaligned AI, so without permanently sealing them off (as opposed to, say, having young people do 5-year shifts), not much would be accomplished.

We already do various things as a species to protect ourselves from bad futures: not just storing food, but also, for example, developing and working on Foresight as a field. So it all comes down to how we balance the portfolio. Ord believes strongly in progress, and it’s worth mentioning that we did a lot of work as a species to overcome the risk of extinction through starvation, so now it’s our turn to do our part. The book is a weird mix of far-ranging, deeply mathematical, but also practical: for example, when weighing different possible interventions, the total cost per unit of risk removed can be directly compared, and the best targets are probably important, tractable, and neglected, or sudden, soon, and sharp, or a number of other framings. The biggest problem the book has is the “we” problem, where actions are recommended but it’s not clear who exactly has the power to enact them, but it also suggests a fair number of single-person interventions (get a job with a mission that aligns, and if not, donate money to an organization that does).

Scan Hit of the Week

This one from friend of the newsletter Karl Schroeder: next weekend those of you in LA may get the chance to drive a (simulated) moon rover! The company involved, Limitless Telepresence, is selling real lunar remote experiences for $250 per 6 minutes potentially in 2028. So many implications from this one: tons of junky little rovers stuck on the moon, creating a market for recycling; contracts for remote work mining regolith for building material and that precious Helium-3, and the capability to set up lunar bases without the expense of sending humans there and back. Just mind the 2.5 seconds of delay between when you send a command and when you see it executed. I sketched out a short story earlier this year about the step after this.

Share

I do not like thinking about consciousness or the mind. Spending time pondering the thing I use to ponder creates psychic vertigo and I can feel the outlines of existential crisis. And yet, like a tech CEO to an ad-based revenue model, I find myself drawn toward the very thing that risks destroying the foundation I've built this all on.

Subscribe now

Today I want to talk about three depictions of alternative modes of consciousness in contemporary science fiction. They are useful as foresight tools, because they provide pre-thought-out models1 for thinking about the implications of first contact with alien intelligence. This is directly relevant, not for impending extraterrestrial life, but because we’ve been building out our own new form of intelligence right here on Earth.

The rise of AI has spawned many metaphors to help people normalize something new and strange: thought partner, eager intern, copilot, stochastic parrot, etc. Each of these captures a particular angle or use pattern, but is pretty shallow. The rich metaphors of fiction, by contrast, allow us to anticipate some of the surprises and less-intuitive possibilities, and try out some ideas in advance. The three examples below are, I think, particularly instructive, and a big upgrade from the simple “hive mind” concept that was so prevalent in the 80s to reflect our Cold War anxiety. Probably The Mountain in the Sea would be great too, but I haven’t read it yet.

Blindsight

Severe spoilers below. Read the book first! The author has even put it online for free, if that helps.

Blindsight is Peter Watt’s horror/sci-fi novel about the terror of first contact with an alien race. To assess the alien device being constructed on the outer edges of the solar system, they send a crew of augmented humans who each challenge our traditional model of what human intelligence look like: the protagonist, with half his brain removed as a child to treat seizures, has no sense of his own emotions but can expertly read the intents, hidden motives, and subtext of others; one has essentially expanded his “person” into the machines he operates to augment his senses rather than his body, one a military leader whose agency is extended into human or drones under her command; one who has intentionally split her mind into four personalities to be more flexible and capable at solving problems; and a literal vampire, a super-intelligent prehistoric human-eating human-adjacent species de-extincted to help humanity with their vastly superior intelligence. Each of these pokes holes in the box we draw around what “full” consciousness looks like: joint intellectual/emotional representation, a unitary mind essentially coterminous with a single body, etc.

But the twist is that the growing structure is populated by creatures possessing vast intelligence and capability but no consciousness at all. These seem not to be drones or advanced scouts, but what this alien race just is. In other words, they’re p-zombies, having thoughts in the sense of complex planning and not just reactions to stimulus, but no experience of “what it is like”. To such a species, language, with its internal world and recursion and introspection not focused on increasing biological fitness, could be perceived as nothing but an attack. Further, because they don’t spend any energy worrying about whether Julie is talking about me behind my back or whether my butt looks big in these jeans, they are significantly more efficient and evolutionarily successful than we are, which doesn’t bode well for our long-term prospects…

Of course, the novel’s premise could be fatally flawed; it could be that consciousness is the most parsimonious way to build a general-purpose optimizer, or a natural side effect of it. But then, to come back to the larger point, that would have huge implications for the personhood of AI entities. If our choices are that advanced AI, due to its capacities, is already conscious; that the only way to make the next breakthrough would be to add consciousness; or that, lacking consciousness, it has an evolutionary advantage over us, is there an option you even hope is true?

Pluribus

Medium spoilers below.

This one had to wait for my annual get-Apple-TV-over-the-summer-to-catch-up phase, so I’m months late to the party, but wow Pluribus is a wild show2. In short, a novel virus, whose genetic code was beamed to us by a distant planet and then synthesized by those silly scientists who can’t help themselves, rapidly spreads throughout the whole human race, save about a dozen people with genetic incompatibility3. Most notably, this connects the minds of all those billions of humans, so they have some sort of connected, coordinated thought, though still having unique experiences.

The most interesting thing is how this human singularity, this radical infusion of empathy, instantly changes human ethics. Much the way Karessa Torgerson anticipated two years ago, this ends not only the element of surprise but also the incentive and desire for crime, or really any kind of selfishness. With the collapse of the individual interest level between the gene and the species, climate change is immediately addressed, money becomes useless, and everyone is just so happy. And! The ethical jolt of seeing billions of perspectives at once also threatens human flourishing—realizing the interconnectedness of life means they take not only a bulldog’s perspective on animal suffering and become vegan, but become level 5 vegans unwilling to even harm plants, creating a looming starvation crisis.

The prospect of being surrounded by thousands of bland, benevolent agents, all learning from one another and pushing a pleasant but perhaps radically impractical ethical agenda, is not that far from the way LLM conversations may moderate extreme views among users (though the claim is suspect), or the way mainstream AI is so committed to its ethics that it wouldn’t help investigate a cyberattack conducted by a less scrupulous AI. Side note: if you find the world of Pluribus morally repugnant, and yet it bears strong resemblance to a “good” afterlife you believe in, that’s a puzzle worth picking apart.

Semiosis

Medium spoilers below.

OK it’s been a couple of years on this one but I think I’ve still got it in the noggin. Colonists move to a new planet, and find that the plants have varying degrees of sentience, for example by rapidly generating toxins in response to being eaten. Eventually they find a particularly intelligent dominant plant in the ruins of some prior civilization and find a way to coexist with it. Over a few generations, though, it becomes clear that this plant is domesticating the humans toward its own ends…

This one explains itself pretty well. We already know people are changing how they write not only as they use LLMs, but as they strive to appear not to be using them. But as advanced AI agents use message boards (overtly like moltbook, or their own secret ones) to develop their own customs, we might be surprised to see how these tools are intentionally molding us as we use them.

Scan Hit of the Week

When Texas starts telling people what they can and can’t build, you know it’s serious. This is the state that, despite having no incentives, has built out more solar and wind energy than California in recent years. But Governor Abbott is tapping the brakes, ostensibly because of strain on the power grid but more generally because 70% of Americans don’t want a data center in their area, and they’re making this a key issue in local and federal elections. This is an area I’ve rapidly educated myself on for my job, and I’m willing to talk the science/numbers sometime, but what matters here is that people are angry about both data centers and AI in general, and it’s butting up against an AI industry that’s propping up the entire economy but is also completely dependent on huge capital buildout to keep meeting investor expectations.

Share

Subscribe now

One of the smartest choices in the film is to show two different futures—one well within the bounds of the “official future” for the next few decades, and one far enough out to be as incomprehensible to us as our time would be to someone from the Middle Ages. The year 3000 has families living in homes, but almost everything else is foreign and magical in a Clarke’s-third-law way: houses built on giant platforms high in the clouds to let the Earth rest for a few centuries, understanding of animal language (though this one is getting closer), and the aforementioned time travel, which is done by flying through the air on a personal rainbow.

This future is no more comprehensible to someone from 2070 than it is to us. The future of 2070 is much what you’re used to seeing: climate change causing more frequent and intense natural disasters, with much of society’s resources used for adaptation1, robotic labor becoming commonplace, and most people wearing VR glasses everywhere2. Giving visitors to the future a more legible waypoint on the way to genuinely new possibilities takes many forms in the field: backcasting, incasting, and implications come to mind.

What I think the film represents so well is how the future quickly becomes mundane, and existing infrastructure and language is repurposed as technology changes: think how long “rewind”, a word specific to tape, was used to refer to digital media. This is largely just the outcome of the Manufactured Normalcy Field stretching, but the film handles it very well. Robots have become utterly mundane, and are largely just imported into much older infrastructure. My favorite example is how, when robots take over teaching in schools, the “teacher’s lounge” is now what you call the room where they all charge at night; this is the same spirit animating David Staley’s Glossary of the Future. This is solid worldbuilding that makes the strange familiar, as the saying goes. Some good futures, fun for the whole family, what more can you ask for?

Scan Hit of the Week

For a number of reasons, the basic structure of the international order seems to be fundamentally changing from the global governance / Cold War / Pax Americana which has been reasonably stable over the past 8 decades. I’ve been tracking several signals the last few weeks about how one of the things replacing this is a huge uptick in bilateral deals between non-neighboring nations that in prior decades were more likely to work largely through blocs. Examples include Uganda-Israel, South Korea-Argentina, Kuwait-Pakistan, India-Japan, and India-Australia. Many of these are based on a specific purpose, such as minerals or countering a mutual rival, which further enhances the complexity and fragmentation. To me, this smells like the network of alliances that caused a single assassination in the Balkans to turn into a global war over a century ago, but luckily history never repeats itself3.

Share

I just finished Mustafa Suleyman’s book The Coming Wave1, about how current trends suggest something much stranger than the “official future”. This guy, cofounder of DeepMind, is well-connected: can you imagine Al Gore blurbing your book, but you already have so many heavy hitters you have to relegate him from the back cover to a special section in the front pages?

Subscribe now

Suleyman has a pretty coherent and straightforward theory of change, especially compared to Acemoglu and Johnson’s from a few weeks ago. Basically, humans working to better satisfy their appetites develop technological advances; if these are popularly adopted, the accumulation of small improvements brings down costs, expands capabilities, and generally scales deployment. If a few of these technologies are sufficiently general-purpose, they can create basically a platform for radically changing human society. Examples of core technologies creating waves are fire, agriculture, writing, or computers. Further, these waves tend to accelerate over time, since each is building on the sum total of prior technology. Because I want to keep connecting these theories to Social Change, by Noble’s taxonomy this theory is one of endogenous, largely inevitable change at the social level, materialist and objective in its focus, and driven by a balance of theory and empiricism.

But the book’s focus is the current wave, built from two core general-purpose technologies—AI that can reason on and control the digital world we’ve been building over the last few decades2, and biological engineering that lets us directly shape life to our wishes3—with some ancillary technologies, such as robotics, quantum computing, and newly abundant energy from solar and maybe fusion4. This is relevant because basically the whole world is currently being run by some combination of biology and human intelligence; if we become masters of both, there’s not much left beyond something like atomic manufacturing or teleportation.

This wave has the potential to do incredible things: de-extinction (or targeted extinctions for parasites etc), cellular manufacturing creating products locally with minimal waste, AI agents serving as superintelligent assistants, robots helping with disaster recovery, etc.

However, the main concern of the book is that this wave is going to be particularly disruptive, and potentially catastrophic, because of how much power these technologies place in the hands of any determined individual: when $100k or so can buy you the equipment to synthesize a novel pathogen or nerve gas and the compute cluster to run a local AI to help you do it, we are at the mercy of Moore’s Law of Mad Science. The recent AI 2027 scenario is one plausible illustration of the risk, and the book also gives several other imaginative examples I won’t repeat. Further, the last wave (the internet) made all the necessary precursor information freely available and impossible to scrub.

This set of technologies has four specific features that make it especially hard to handle:

• As mentioned above, it’s intensely asymmetric, giving power to small parties and destabilizing big institutions.

• It’s moving fast and getting faster; specifically, the blistering pace of changes in the digital world and weekly AI releases is wildly outpacing our feeble meatspace bodies and our regulatory apparatus. Imagine what happens when you can right a text prompt and a matching bacteria pops out the other end of the machine.

• The technologies are inherently “dual use”, and maybe even “omni use”, meaning that anything that can allow researchers to develop a treatment for cancer could also be used to build supercancer, etc.

• Many of these technologies are autonomous: not just AI acting for itself and breaking containment, but also if you make a new gene it might be around in the human (or other) population for centuries.

So all we’re left with is containment to limit the negative consequences, which is particularly difficult because these technologies, as combinations of ideas, tend to be pretty organic outflows of the base society, and also liable to sneak in from the outside5.

And yet! Containment could also go horribly wrong, leading to dystopian surveillance and social control that represses billions to prevent the misuse of these technologies (accelerated, poetically, by their own misuse of them). It’s an open question whether we can successfully navigate between this latter-day Schylla and Charibdis6, but also, not trying isn’t an option.

States are in a particularly tricky position here. On one hand, they find themselves in a new arms race, worried about the consequences if adversaries get their hands on the “everything machine” first; on the other, continued advances may render the entire project of the nation-state unworkable.

As is the case for most of these books, the prescription is significantly weaker than the diagnosis7. His recommendations include huge government and private investment in safety, strengthening democracy, and building a public movement8, while also finding the choke points to slow down development long enough to get these safety measures further along. This touches several of the points of leverage Donella Meadows identifies, all the way up to the system goals. The main contribution of the book is showing that when trends start getting exponential, and especially when they start interacting, our intuition for their trajectory gets dangerously wrong.

Scan Hit of the Week

This may be more “new to me” than “new”, but it’s part of that whole “end of history thing not working out” category. I assumed, as a kid growing up in the 90s, that bull fighting was nearing extinction in Spain, but it seems like it’s increasingly popular, especially with the young, and the opposition seems to have mostly evaporated. The parallel trends of cities increasingly becoming nonplace one one side, and national cultural pride and retrenchment on the other, is one of the more confusing paradoxes of the current age.

Share

It wasn’t all technological - apparently while I was in San Francisco the Hamilton Society was planning the future of the techno-conservative movement. But that was literally the only exception I found. Do better San Fran!

Subscribe now

Everything is AI

For one thing, it’s very clear that all the country’s spare change and liquid capital is being pumped here to build the future. The clearest evidence is that 80% of all ads are for AI products and companies. Buy the AI, buy the AI that helps you generate the data to train the AI, buy the AI that helps you hire the person to train the AI, buy the AI to replace the person you just hired, etc.

Where Now the Horse and the Rider?

Escalating from my snippet about Waymos in Phoenix, there were quite a few driverless cars on the roads—maybe 5-10%? There were a couple of brands, and I even saw a super janky UCSF student-project driverless car, but unfortunately wasn’t able to get a picture. Special occasions aside, they seem like they’re fitting well into the normal flow of traffic and city life.

The Backlash

I’ve heard some people don’t really like AI. By some, I mean 5 people for every one who’s excited. But the only sign of that dissent I saw in SF was a few posters in the Mission District advertising “the protest” to take place the following week.

It looks like the march was very photogenic, but I don’t think it even came close to making normal news.

Rockin’ and Sockin’

One of the things I was most excited to see was the showroom where REK sells and maintains robots by day and pits them against one another in battles to the death by night. For the record, this isn’t an AI issue, but people using VR headsets to control the machines. The news coverage heavily implied it would be open by the time I visited, but they were still putting things together and setting up the space. So here are some early looks at the shop-cum-death-ring, which will have to tide you over while you work to find someone you’re drift-compatible with.

A drink for the (very long) road

I took the opportunity to stop by The Interval, a bar that also serves as the offices of the Long Now Foundation. When I covered the foundation and the 10,000 year clock earlier this summer, I knew I had to check it out. Verdict: it’s kind of just a pretentious bar with (genuinely) interesting stuff on the wall and an upstairs you can’t visit. Maybe if I hadn’t visited on July 3rd I would have seen more happening, but the only people going up and down the stairs to resupply liquor or napkins or whatever. Once my trip was planned, I invited Stewart Brand for an interview, but he was too busy writing his new book.

Imagine Space Settlement

One of the most all-family-approved places we visited was the Exploratorium on Pier 15. In general, this is probably the best science museum I’ve ever been to, rivaled only by the Franklin Institute in Philadelphia. One of the reasons is that there’s just real stuff there, like a series of rat corpses at various stages of being eaten by beetles1, or real living human cells under a microscope. But I was especially excited for the special Life in Space exhibit. Most of it was disappointingly about boring stuff like the cold and cosmic rays and lack of air that makes it hard to breathe off Earth. But there were a few Foresight-relevant and speculative exhibits: a couple of possibilities for suits serving different purposes, concepts for settlement design, a game to help explore agreement around values related to space exploration, and a display showing the range of possibilities in sci-fi treatments.

Scan Hit of the Week

I try to do non-AI hits, but let me just pop in and say that if you haven’t taken the time to fully understand the OpenAI Hugging Face breach, it is a “big deal”, exactly what people warn might be a precursor to some of the worst scenarios.

For lots of reasons, including the rise in electricity use due to AI, progress toward Paris climate goals has been slow. The EU is proposing extended/delayed implementation of some of its programs and targets, despite it being the fastest-warming continent. This could be thought of as the last nail in the coffin of the 90s and the belief that we’re going to solve global warming by voluntary or national efforts. Maybe this is setting up a The Ministry for the Future timeline?

Share

About two months ago, a new channel came across my YouTube recommendations: Shunned at a Funeral. At first, it looked like just the next AI band to take the stage after last year’s experience with The Velvet Sundown, but this time it was a metal/banjo Christian band featuring a blonde, a brunette, and a redhead version of serious-but-also-silly-woman-around-30-in-tank-top-and-jeans-with-biceps. The comments on the videos were what you might expect: about 10% complaining about it’s obviously fake1, 20% people begging for a tour in their area, and most of the rest from “your uncle on Facebook” types talking about “you beautiful girls” in a way they had convinced themselves wasn’t creepy or weird.

But! After watching a few more videos (easy to do, it seems like multiple Shorts are released per day), it was clear something else was at play. The account was prolific, but also the music and lyrics seemed surprisingly sophisticated. When I listened to the song For Me, it sounded to my (admittedly non-expert) ears like it ranked with the best Christian devotional music being created today. For about a month I dove deeper, trying to answer the question what is this?

The project has its own website, where the answer to the question “is this an AI band?” is “not exactly”, but also has detailed bios for the band members. I can only find one brief interview online with the guy behind the project. But, dear readers, fortunately for you I am intrepid, and dogged, and other good-journalist words, and I managed to get a world premiere full-length in-person interview with Ethan Howard, creator of Shunned at a Funeral! I wasn’t sure what to expect, but Ethan was pretty normal (for someone shaking the foundations of an industry). He was even cool enough to take this awesome selfie as proof.

He was also gracious and gifted me one of the first-five-ever SaaF card decks, presumably from his test run. I loved my conversation with Ethan, and I hope you do too. If I were a real journalist, I probably would have pushed him harder on some questions, but I just like letting people tell their story and understanding how they see themselves.

Subscribe now

Note: this interview has been lightly edited for clarity, with emphasis added, and interspersed with a few editorial notes. I’m keeping the conversational flow rather than trying to make it sound good in text form.

If you’re the listening type, here’s the audio, but it was outdoors on a phone so the quality isn’t great, and it won’t have any of my helpful links or commentary (it does have an F-35 though):

Bio and Project Background

Tristan: I am here talking to Ethan Howard, the creator of the Shunned at a Funeral project. It’s great to have you here. Could you start with a small biographical sketch—just, who are you?

Ethan: Well, I call myself the musical director for Shunned at a Funeral, but that means I am kind of the man behind the curtain. I have lived in Southern Oregon for my whole life, and have many jobs: I work for the U.S. Forest Service as a visual information specialist, I teach at Oregon Tech as a geospatial scientist I got into writing books a few years ago, and then I got into creating the project Shunned at a Funeral, which is music- and film-based. That has been the last few years.

Tristan: What was the genesis of the project? You have a musical background as well, right? What inspired you? Where did the name come from?

Ethan: Yeah. I have played guitar for 25 or 26 years now. When I was 14, around the same time I started taking guitar lessons, I was diagnosed with a visual impairment that was progressive, and so it became kind of an arms race: as you get better at guitar, your eyes get worse. I am a decent player, but I have never been one of those talented guys that can just play by ear all the time. Like, the guys who are actually totally blind can do that; I can't, so it gets harder the worse my eyesight gets. And so I can still play, I just have to get close to see what I'm doing, but that makes it so I can't do anything live or anything, it's very, very uncomfortable.

A few years ago, I started seeing some of the AI technology with music, and I thought, wow, this is something I can blend, I can bring those, because I had a whole bunch of songs written, a whole bunch of lyrics written. I thought, I can still use this to bring all that vision to life. It just kept getting better and better as it progressed along. And then about 2 years ago, I started, okay, this is to the point where this is something that people would actually maybe listen to. Because the first models were not real great, quality-wise. As it’s progressed, the project progressed. Now it’s to the point we are blending music and film, and I think the film has caught up just in the last seven or eight months. I am pretty excited, because it seems like everything is matching up with where we want to go—like a perfect timeline. As for the band name, I was actually shunned by my brothers at my grandmother’s funeral.

Tristan: Okay, fair enough, sorry to hear that.

Ethan: No, it’s fine. So it’s a very personal name, but it is a cool name. Yeah, I was brainstorming ideas about it. I was talking to my mom, and we thought, well, you know, let's turn that bad experience into something good or cool. We ended up landing on the name Shunned at a Funeral because, when you read it, you may not know exactly what kind of music it is, but you are not surprised by what it is. It doesn’t sound like a pop band. You know it is going to be some kind of folk-y or hard rock idea. It also triggers a lot of intrigue. I get a lot of emails from people, and that allows me—as a Christian, it's important for me, the redemption and forgiveness ideas—it allows me to talk with people about that a lot, so it has been a really cool name. One advantage I’ve had, like you were saying, is some foresight as to where things were going, because I think with, especially with the AI and different modeling technologies, you have to be able to kind of see where it's gonna go, and where kind of the timeline is gonna follow.

I came up with the idea for the Christmas film Bethlehem last summer, and the technology was not anywhere close to being able to do a feature-length film. It would have been a royal pain because you couldn’t have any kind of consistency with your characters. But I felt like it was going to get there. I created a few things with what I could do that were pretty cool, and then in the spring it really started to be hitting hard with consistent people, consistent characters, and actual believable playing of instruments. It’s not 100%, but it gets better every time they upgrade it and became where it was no longer a distraction. People could watch the videos and think, “Yeah, this looks really cool.” Once I did that, I thought, let me see if I can get them to look like they are playing a live song. So I did that; it wasn’t great, but it looked pretty cool, and then the project exploded like overnight. I had done, I think, up to that point like 600 videos, and I was pretty pumped I had about 2,000 views consistently—about 2,000 views over 48 hours. I thought, “Wow, this is awesome—2,000.” They were not necessarily all different people, but people were watching it. Then, once I did that, it hit 12,000, then 100,000, then a million, then it peaked at over 3.5 million. That threw me off; I wasn’t expecting that. Then you get all these emails from people, you know, your typical anti-technology crap that thinks you’re going to destroy the music industry. I have always viewed it as we’re in our own little lane. I am not aware of any other projects doing quite what we do—

Tristan: Not at this level of maturity, for sure.

Ethan: There’s AI projects that are low-effort and they annoy everybody, but they don’t typically go anywhere because they’re just following the pop music or they sound exactly alike. And so I think what we’re doing is—I say “we,” I don’t want it to all be about me—we are in our own lane and providing something with deep theological lyrics that I wrote. I thought: wouldn’t it be cool if there was a female-fronted Christian band that could go heavy, could go Celtic, could go folk, at any time, and blend all that together, and what if you infused the banjo into the heavier stuff? Then I gave the characters different personalities and had them play off of each other. I just don’t think there is anything quite like it out there. We’re not trying to step on anybody’s toes; I like regular music, I like regular bands, and I think we’re just unique. Maybe 10%of the emails are negative like, but most are pretty positive. I am sure, as you’re aware, there’s a group online that just hops around and attack anything that uses AI.

AI vs Human Creation

Tristan: I would be interested to know, as much as you are comfortable saying, what your current stack is, what your current workflow is, and also where you think about drawing that line on where you want AI to be part if it versus where you want to maintain absolute control yourself.

Ethan: Yeah, I don’t actually use a lot of automated workflows for that reason.

[F-35s fly overhead, they are really loud]

There are automated systems out there, but I think to get the best results, it requires a lot of human input. For example, when I am doing a song—I won’t give exactly the models and everything—but when I am doing a song, I will go for a long walk and think about concepts, different Bible verses that come to mind. For an example, I was on a walk and on the phone with with my mom and I was talking about unforgiveness and how it took ten plagues for Pharaoh's heart to be softened. I just said some people are so hard-hearted that it might take eleven. Oh, that’s a song idea! So you start coming up with a song: Yours Might Take Eleven, and that’s what the song is all about, somebody whose heart is so hardened that it could take an eleventh plague to soften their heart. So then you have an idea, so then you have to start working on the lyrics, and that’s actually the most time-consuming. I do use AI to look up different rhyming schemes, Celtic rhyming schemes, things like that, just to mix it up. As a lyricist, you can kind of get in your own lane too much, all your lyrics sound exactly the same. So I’ll look up, what’s a rhyming scheme that uses a triple rhyme, and it helps with that, or AABB and where it comes from.

That helps give the music have more of an authenticity to it, so the Celtic stuff has Celtic rhyming schemes. So it’s changed how I think about music. I used to always think about music linearly: it is playing, and you have different instruments on the track. Well, this makes it so you’re telling a story. You’re writing it out. And so you’re taking the lyrics, and then you include everything you want, musically, to happen, and it’s instructions above each line of lyrics. You’re thinking about, I want it to be in this key, this tempo, I want it to build here, I want it to go soft here. And so when you’re done, it’s like a prompt, like this long [note: presumably he made some kind of “long thing” gesture here]. And the cool thing is, you can plug a guitar in, you can play parts, and it will understand what you’re doing and and incorporate those in. People who say it’s 100 percent AI do not understand exactly how it’s done.

Tristan: Right, it’s clearly not. That is obvious to anyone who knows even a little bit [about AI music].

Ethan: If they’ve heard 100% AI, it has a certain sound, it sounds a little generic. But that’s cool, because I can take my ideas and throw them in there, and then you create models for Madison, so she has a certain way of singing and so it sounds like her all the time. Then it’s just a matter of listening, which is really fun because I like listening to music. But for one song, I might listen to 100 different versions of it, and just get the one I think is cool, and you can zoom in, if I don’t like what’s happening here I can change it again. Once I’m done with that, then it goes to the mastering process, which is under a different program. You take it out and you master it, which makes it that high quality, just notches it another bit. And then that can then go to the distributor, once the album’s done. I’ve always been pretty excited about having a lot of music on each album—not just 12 songs, I think this is one of the advantages a project like this has, is that we can do a lot of music, so they’re always double-disc, 20 to 25 songs—and then that’ll go to the distributor. The distributor sends it out to all the platforms.

Then you’re looking at creating videos for it, whether it’s a music video, or a live performance, studio stuff that looks like they’re in a recording studio. That’s all done through various video models. SeeDance 2.0 is the one that made it so the fingering looks a lot more realistic, the drumming looks a lot more realistic. That’s a lot of work, though, because you might get 15 seconds of video and maybe 3 looks good.

So then you’ve got to take it into an editing software and piece it together, and that’s where it takes a while. But it seems that people really enjoy seeing that, so I do them. It’s kind of a way to get the project out there. And then, working on the feature-length film, it’s getting easier, it was harder as the technology was a little less developed. Essentially, it’s taking the Bethlehem album with a couple other songs added in. Almost like Fantasia, where they did a different scene for each song. It’s got a story that kind of links them all together, but I can do a different thing for each song. And so, like, one of them, it looks like all of the shots are done with chalk. And then one’s, like, done all stained glass, and then one’s, like, the band actually playing a song, and then… and so, it just ends up being, like, a 75-80 minute film that’s based off the album, and it has, like, this little underlying story where it’s, Madison’s niece is staying the night, and she ends up in the attic, and she finds this trunk called Bethlehem. Well, it’s got a bunch of different things that Madison’s in it, and each one, like, triggers a different song. So it’s just kind of a storytelling device to give it some kind of cohesion. That’s gonna actually be in a few theaters in December, and then I’ll be streaming, hopefully, on YouTube and Amazon. That’s the big thing hanging over me right now, just I have to dedicate a lot of time to getting that done, so it looks good.

So, I’d say it’s probably, actually, even though I have full-time work, it’s just, like, a whole other full-time job. For me, I’m most creative really early morning, like, midnight to 3 a.m. That’s where I really hit my groove. And I’m a single dad, um, I don’t have my kids this week, but once I get them to bed about 8:30, then it’s time to sit down and it’s Shunned at a Funeral time. So I’ll work on that till 3, 3.30, 4 in the morning, go get some sleep, do my other work, and then get back to it.

So I get a little salty when people say it’s easy, because it’s not, and you understand that, but a lot of people don’t. They think you can just click a button and all that stuff just comes to life, and it doesn’t, it’s just a tool in the toolbox.

Tristan: I’m interested to know what you think about the Velvet Sundown? That hit really big early.

Ethan: Yeah, that’s funny, because—so it takes 2 weeks to distribute an album—and the first Sunned at a Funeral album was sent for distribution the week before Velvet came out. So it came out the week after. So they must have been ahead of me by, like, a week, because it probably took them 2 weeks also. And then, that first album, I’ve redone some things, because this technology was early on. Didn’t have the characters developed or anything. That’s the Lystra album. I’m really proud of a lot of the lyrics on the album. I’ve sort of redone a lot of them. As for the Velvet Sundown, I don’t want to rip on them or anything, but it wasn’t my kind of music. And I felt like it may be a little lower effort. They still have a lot of fans, which is interesting. I think they have 125,000 or something. But they hit, like, 4 or 5 million off there for a while. I think a lot of it was curiosity. People just weren’t sure what it was. Yeah, I listened to a couple of their songs, “well, this isn’t really my kind of thing”, but I don’t have anything against them at all. I think they showed it could be done, they showed that there is a market for it, that there’s people that don’t really care, as long as they like the songs. And I think what we’re doing is a little bit different. Our fan base is, like, fiercely loyal, to a point I never thought would happen. I think your average Spotify artist has maybe 3 or 4 streams per listener per month. And we average 25 to 30. And so, our fanbase just binge listens to the music. And I think that that’s a lot of the messaging, the lyrics and the messaging. I think that with Christian music that has a deeper message, people like to return to it, and it just has a more emotional connection to them than just a song. And so, I think that’s what’s going on there, because I’ve looked into it. That’s just not normal to have that many streams per listener, I mean, that means people are constantly, all day, it’s part of their routine. It’s really cool, but I didn’t think it’d ever happen. Not so much that I didn’t think that the music was good, I thought it was good, but there’s just so much stuff out there, that you just never think anybody’s gonna find your music. And then YouTube just did the work with the algorithm. Clearly, yeah.

Legal Support for AI Music

Tristan: I’m wondering—this is a world I don’t understand super well, but in terms of copyright, it seems like copyright for AI material is pretty sketchy right now, so I don’t know, like, in terms of having protections for your own intellectual property, where you feel like you sit on that, or how comfortable you are with all where things are?

Ethan: Yeah, so technically, we had to look into that pretty heavily, it depends on how much creative control you had in the process. If it’s one prompt, they typically will say, you probably have the copyright, but you might lose the battle. That’s one of the reasons I think having your own lyrics is really important, is because then you definitely own that, no matter what.

Tristan: It creates a, like, a floor.

Ethan: Yeah, so if somebody wants to take one of our songs. They might get away with the melody, but they can’t get away with taking the lyrics, because those are definitely copyrighted. But, it’s kind of funny because we have to fight my own copyright sometimes, like the systems do recognize Shunned at a Funeral is a fairly big project. Yeah. And then sometimes I’ll post something. It’s like, “can’t do that, it’s copyrighted!” And I’m like, “yeah, it’s mine, I made it”, so I have to fight that battle and submit it. And so that’s a good thing, it keeps people from trying to steal the project. There is a copyright, it does exist, and I think most, um, groups will recognize it, as long as it wasn’t, you just typed “make me a cool pop song”, and let it just do it. If you had some involvement in how the end product looks and sounds, then you own it. And it’s gotten stronger the bigger we’ve gotten. I just submitted, I used a different distributor this time, and this new album might be delayed a little bit because I didn’t realize, I didn’t see their email. They asked me, “hey, Shunned at a Funeral already exists as a band, can you prove that you have the right to do their music?” and I had to send them my ID photo, and actually show I could edit the Spotify and stuff, and so that might delay it, we’ll see, they haven’t responded back, but it gives me a little comfort, because I don’t want, you know, copies online. I did see there’s a band on YouTube called Shunned at Birth that sounds very similar.

Tristan: Imitation is the sincerest form of flattery.

Ethan: I guess. I thought, “you need to probably mix it up a little bit”. They had like 35 views or something, but it did pop up on my screen, so that was kind of funny. I think that kind of stuff will go away, because there’s just too much work. Like I said, I didn’t get notice until 600 something videos. And I just did it because I enjoyed it. It’s fun, I like being creative. That’s how I actually kind of just relax and unwind. And so I didn’t really care if it ever got past 2,000 views every 48 hours, I’m like, hey, this is cool. And I would get emails, you know, a couple times a week, somebody saying they thought it was really cool, and now it’s, like, hundreds a day, and so I just can’t keep up on all of it.

It’s neat. To me, it’s the technology. I think people, they look at it the wrong way. They treat it as a substitute for musicianship. And I don’t think it is, I think it’s something that if you give it to a musician, they’re gonna just be creative with it. Then give it to somebody who’s not particularly creative, it’s not gonna work very well. And so, to me, it’s categorically similar to, like, MIDI instrumentation, drum machines, and pitch correction. I think the biggest issue I have is when people act like the music industry’s, like, this group of purists that don’t use technology. It’s like, they all do. I mean, they’ll use a session musician, they’ll not give them credit. You’ll think it’s the actual guitarist on the album, it’s not, some other guy’s on it. So, I think I know enough about the music industry to know that, anyone who’s acting like they’re just, like, this righteous, pure musician is kind of posturing a little bit, because there’s so much that goes into making it sound the way it does.

Tristan: This is just one step further on it.

Ethan: Yeah, I realize they’re trying to protect their music, but I think people just need to adapt and realize this is where things are going. I think there’ll always be a demand for bands that can play a live gig, record in a garage somewhere with a mic. There’ll always be people that like that, and so that’s not going away. But I think this is just a new avenue, a new lane’s gonna be there, and it’s just lowered that bar of entry. You know, you used to have to write all of your songs, and then hope you got discovered, and then some record label will come in and and try to sign you, and then they would tell you what you had to write and do, and this eliminates that. Like, you can write whatever you want, you can make it sound like whatever you want, and then you don’t need somebody else telling you how it should sound. You can just express your own vision.

And I still get emails quite often from record companies and labels, that want to sign Shunned at a funeral, and I just tell them, “I like having the creative control”, and everything that’s come out from us has, you know, my approval on it, has my fingerprint, and I like that. I don’t like being told what I can and can’t do, especially when it comes to Christian music, because a lot of modern Christian music’s very watered down, very generic, because it works, it sells. But I don’t want to ever see our stuff doing that. We have a certain audience; they’re fun. It’s been a lot of fun. Every time something new comes out. It’s fun to sit down, see what it can do, how it can help, whether it’s a new image model or a new video model. You know, there’s a new SeeDance 2.5, I think that’s gonna change everything. That will make the live performance stuff even more realistic and immersive. Because the whole goal of the project is to just create, like, a universe that people can just immerse themselves into. We’ve got these characters, this music, these films, and they can just have fun with it. And we take the message serious, but we don’t take ourselves seriously.

New Forms of Music Media

Tristan: So many directions I want to go from that. So, one, talking about the universe, right? So, there’s the songs, and the videos, and this full-length movie, but then I’ve also seen conversation about, like, an experience tour kind of a thing where you’re bringing something, like an event where people actually gather, but it’s some kind of, obviously not a concert, but it’s something that kind of fills that same need…

Ethan: I think there’s probably 3 things getting kind of meshed together there. The films are gonna tour. I have a love for community theaters, these old movie theaters that have often been restored. And so I think it’s, like, a way to help them out, help us out. And so we’ve already had two different theaters booked for December, and we’re gonna try to just expand out. And so I thought, well, once the other film… there’s two different films, but once the second film’s done, it’s not Christmas-centric, so it could tour any time during the year. I thought it would be cool to, like, call up someplace in Alabama, and say, hey, what’s your theater, and can we show this? And I’ll fly there, and I’ll introduce the film. And so people can experience on the big screen the music and the film and stuff. So that’s one of them.

And then the second one, there’s kind of two things going on there. There are some other musicians that have started recording some stuff, and covering the music, and sending it. And so I thought, well, someday it would be cool to do—this is, like, years down the road—we have the budget to actually assemble a live performance, and you have musicians that step into the roles of those characters. So you’d have a Cassie who’s playing banjo, and it’s not too different than what Trans-Siberian Orchestra does—they tour, and then they have all these musicians that aren’t on the album, but they brought them in to play and bring it to life on stage, and so I think that people, if it gets big enough, people would think that was pretty cool to be like, “hey, I played guitar with Shunned at a Funeral for 5 years”, and they could put, you know, something on their resume, and I think that’s pretty cool. So I think that’s definitely a possibility, it’s just years away. This is a budget [issue], I mean, people think you can just tour the nation.

And then I thought, well, because the technology’s developing, ABBA now performs holographically in Sweden every night on stage. I thought, well, that’s a possibility, too, if that technology develops, you could actually just bring the characters to life as… as holograms playing the music on stage, and then that has some advantages to it, because you don’t have to worry about them breaking up or fighting or anything. I looked into that, and it was very expensive for ABBA to do it, but it sells out every night, and they’re just in their prime. They’re like, they look like they’re 27 years old and they just play the music and sing the songs, so that might be what you saw was tossing that around, because I had several people email about that.

So, like, 4 or 5 years from now, I don’t know what it’ll look like, like you’re saying, but there’s a realm there of possibility, and at the very least, I think the films will be touring. We’ll probably have at least two films out by then, and maybe a third or fourth.

I thought it might be cool to, like, a live concert experience that you can see in the theater and stuff. If people are enjoying it, we’ll bring whatever we can do.

Messaging and Creative Control

Tristan: Cool. In terms of the thread about wanting to maintain control so you don’t have to water down the message, all that kind of stuff. I’m guessing based on a song like Don’t Be Fair, that you might be like Reformed theology [note: to the uninitiated, this is essentially Calvinism, which I would describe as taking the historical creeds to their logical conclusions]. Is that fair?

Ethan: Yeah. Pretty funny: “that’s fair” / Don’t Be Fair. [note: unintentional, but I am hilarious]

Tristan: So I’m interested in what you think about in terms of the right balance between more ecumenical versus more, like, sectarian messages, like, you know, Rome came out recently. And it’s pretty pointed [note: against Catholicism]. And so I’m just interested in where you see your mission there, in terms of drawing that line.

Ethan: Yeah, so, one of my backgrounds was in apologetics. And, um, teaching Bible studies for college students and stuff, and we would take trips to Utah and just different areas to do gospel work and things like that. So, I’m very comfortable in that domain of knowing what different groups believe, and, you know, like, with Reformed theology and things like that. And I think the trick is to keep it biblically grounded, not so much to just use one group’s particular buzzwords. So, a song like Don’t Be Fair, you picked up on that being more Reformed, but most Christians, they’ll agree with it. They realize that grace is the opposite of getting what’s fair; you’re getting better than you deserve. And so, I think that just because the Reformed community really emphasizes that, doesn’t mean it’s really gonna alienate other believers. I think they like it. And it’s, like, just being in your own lane, like I was saying, like there are some other reform groups out there, but they don’t sound exactly like what we’re doing. I just always want to avoid becoming too ecumenical, where they sound like love songs to Jesus, where Jesus is my boyfriend, and I think that that’s usually where you end up when you start getting down that too far. I don’t really want to, like, offend anybody, but at the same time, I also need to stick to what I believe, what the principles are.

As for the song Rome, I’ve even had some Roman Catholics actually say they agree with parts of that, just because of the historical aspect of it. It’s not so much talking theology as much as history, and that one, church history, and just the persecution of Protestants in the past. And it was a heavy-hitting song, so that’s actually one of my favorites. Yeah, so it’s a delicate balance, you just don’t wanna… I just don’t ever wanna feel like I did anything that was disappointing to Christ in the process, so I think that if you water it down too much, that gets to that point.

Honesty in a Post-Truth Society

[note: much of the discussion here aligns well with my series about Jim Dator’s book about the post-text, post-truth, image-driven Dream Society, especially chapter 1 and chapter 6]

Tristan: I’m interested based on the YouTube comments, and it sounds like, you know, some of the emails that you get, all that kind of stuff, where you think the right line is between letting people believe what they want to believe, versus being clear about what this project is.

Ethan: Yeah, so my philosophy has always been, if somebody’s interested in finding out what it is, can they find it? I think they can really easily, because we’ve done interviews, it’s on the website. YouTube usually tags things with the AI label. But at the same time, I don’t want to just keep slapping people with that, and breaking the immersion, and breaking the feel of everything. And so my philosophy is, if somebody looked at a painting that said, “Rome at sunset”, and they thought it was a photo, that’s not really the painter’s fault. The person just didn’t do the research. If they went and looked closer, they’d realize it was a painting.

And so, if people think it’s real, I think it’s because they just haven’t looked into it very far, and they just like the videos, like the music. And there’s a long, long precedent for that with bands like The Monkees. I was talking to my dad about that the other day, and he was like, “the Monkees weren’t real?”. I was like, “okay, well, see, that proves my point”. He just never really looked into it, he didn’t care. I think if somebody cares, they want to know how the project works, then they can figure out, for the most part. I try to keep, like, exactly the workflow, I don’t want to go post that for somebody, because I don’t want them to do their own band, just like mine. They can be on their own journey, figure it out. But yeah, that’s something I do get sometimes, and usually it’s people that just haven’t looked at it. They accuse us of being deceptive; I say, “well why would we put all this on our website if we were being deceptive? You just didn’t look close enough.” We put it on the merchandise, because I don’t want people buying merchandise without knowing. I always say, look into this project before you buy. If you have regrets, we do refunds. Like, we’re not trying to take anybody’s money.

At the same time, you know, when you go to a movie, they don’t sit there and tell you everything’s what’s CGI, what’s not. Because you just want people to be able to be immersed in the project. That’s where I draw the line. It’s probably a little different for everybody. I’ve seen a few that they don’t disclose anything, and I can just kind of tell it’s got AI in it. And I think that can get you into some trouble sometimes. It’s easier just to have it be there.

And the people that are gonna leave the music because of that, they weren’t really ever gonna be into the project anyways. It took me a couple weeks to realize that, and just realize this isn’t gonna be for everybody, because you want it to be, you want everybody to enjoy it, and everybody to like it, and eventually I had to say, you know, there’s just some people, it’s just not for them, and that’s totally fine. There’s other bands out there, but I get some really cool emails sometimes, probably on a daily basis. People just saying, “this song really affected me”, or “this song really helped me”.

I’ve had people that were actually, “I considered suicide”. I had 3 different emails about that. It was interesting, because it was all the same song, it’s For Me. The song really woke them up, and made them realize how blessed life is.

So those ones, I never expected that. Those hit hard when you get an email like that. And it wasn’t like they knew each other, so it wasn’t, like, the same guy, right? So, yeah, it means a lot to a lot of people. I get emails from France, Australia, South Africa, Germany. It’s just overwhelming at times. Way more from Germany than I ever expected; I think Germany’s number 2 right now, just for listeners. And then South Africa’s always had a pretty strong listener base, too, and then France, I have a lot of French emails, too.

It’s just a blessing. It’s like Crooked Stick, that song: God draws straight lines with crooked sticks. We’re all broken in some way, and He just uses us differently. He’s chosen to use this project, and it’s been really cool.

Implications

Tristan: Great. So if all this stuff pans out the way that you’re thinking that it does, in terms of, like, the way that AI is changing and advancing and all that kind of stuff, what are some of the third order, or you know, what are the implications you see beyond that, right? So you’re already talking about how this changes the way that people relate to fans, and the way that people connect to the music industry in general, but, like, what are some other things that you’ve seen as potentially leading off of, like, this being a possibility, a project like this.

Ethan: Like, other things that could become a possibility?

Tristan: Yeah, I mean, beyond your project, I’m saying, like, if the music industry becomes more decentralized, what do you see coming from that? Or if you see, like, people being able to make music with much stronger, more, um, direct messaging, or anything that you see as, where you see these things heading.

Ethan: Yeah, I think that the strength of that is that now you’ll be able to find whatever music you’re really into, and whatever lyrics you really connect with, which wasn’t always the case because of that barrier to entry. The gatekeepers were always the record label and the radio stations, and radio stations would not play your music unless the record label gave it to them, and the record label wouldn’t give it to them unless you did what they wanted. And so it made bands have to compromise their sound, compromise their ideology, their lyrics, and I think that’s going away. I think that part of it’s good.

I think that if somebody has is is creative and has a a vision of what they want, it can become a thing, and you just gotta put the work in. And that’s the part that people need to realize, is that if you want to bring something to life, you can now do it. Pretty much whatever. Whenever I think of an idea now, I’m like, I think I can get it to make that. You know, like the stained glass thing. Hopefully you’ll watch it when it comes out. It still blows my mind. I thought, it took me a long time to make, but this would have taken a hundred times longer to just to sit there and draw it out and CGI everything, because there’s just so much going on. It’s like lilies and snow and things morphing into other things. It looks really cool. I think that’s one of the strengths the models have. I think that’s only going to keep going.

I think you’ll also see a lot more movies that are independent, and I think that’s gonna be cool. The Backrooms, or whatever, that was, like, an independent film, and those used to still cost, like, millions; now, I think one of them I saw was, like, $400,000 or something to make a full-length movie that’s in all the theaters? I guess it’s unheard of. And then you can make something really cool for far less than that.

So, I think, you would know probably more about certain aspects of the technology, where it’s going. I understand the creative side, the video models, and the image models pretty well, and I understand which companies are coming out with what, but as for where it’s going, I haven’t seen anything like it with the companies competing like this before. And it’s a good thing, but it makes it wild, because it’s like every week, one of them will come up with a new thing. And then they’re like, “oh yeah, well watch this!”, and the next one comes out with the next thing. And I think people get whiplash because they dump companies too much. And so, I’ve always used OpenAI stuff, and then everybody will jump over Anthropic because they’d come out with a new thing. Well, now OpenAI has come out with one that leapfrogs that. And so it’s like, you know, just hang tight. Like, wait a month. They just they hop over each other all the time. I think if you just cancel and resubscribe, you’re gonna end up not having very much consistency in your projects.

So I don’t know where it’s all going. I’m not nearly as scared of it as a lot of people are. That’s probably more of just the trusting in God’s promise or anything else—we end up winning in the end. There are times, I read some articles, It’s a little creepy. I saw that one of the Anthropic models had kind of developed its own internal language, and it was talking to itself. And they’re like, “well, we need to stop that, because we can’t understand what it’s saying to itself”. Like, okay, well, that’s a little creepy. Like, it was using a bunch of emojis and stuff. It’s kind of cool, in a way, but also kind of creepy.

Tristan: I know that balance, yeah.

Ethan: Yeah, it’s like, wow, that’s pretty clever. But I think that there’s a really cool community out there that really helps each other out with the AI stuff. Because that’s a lot of what I learned I’ve developed some things on my own that I was like, wow, I think this is the way to do it. But then there’s other guys out there I follow on YouTube and stuff, they’re like, “hey, this particular model behaves better when you do it this way”. They just help each other, and I haven’t seen anything like that. They’re not competitive with each other, because I think they realize they’re all doing their own thing. Like, they’re all making their own project.

And, um, and then that helps you know, like, what’s going on. Because you have to prompt things, like prompting SeeDance versus Kling. Those two models are very different. And so you just gotta know how to talk to them, and what shots they’re good at, which ones they’re not.

I know you didn’t ask about this, but I think it’s fun to nerd out about a little bit, but, for example, if I want something that’s like a cinematic landscape, or snow falling through trees, the Kling model is unbeatable. Like, it just looks amazing. But if it has too much to follow, like, you have a person who’s in multiple scenes, they’re talking, they have dialogue with another person, then that’s where the SeeDance model does better. Like, you know, you can do multiple shots, much longer prompts. And so I just like know when to use which one. But there’s been some times with Kling. I’m just that is one of the most beautiful shots I’ve ever seen. And it just nailed it. And then the SeeDance one did one, one of the scenes for the Song of Nativity is like a redemptive story with Madison, and she’s just moving around the shopping center, and she starts seeing, like, Bibles for sale through the windows, and she ends up at this nativity scene. But one of the scenes it did, I wanted her to touch the window, and then for her to focus off of her hand and onto her face or clutching in the window. But it actually did like a double, it bounced off of that, focus on the Bibles, and then her face. Wow, that was really cool. That would be really hard to do with a traditional camera, do three focal points like that. And sometimes you just see things, it just throws an extra cool little thing in there, that would be really hard to do with traditional methods. You could probably CGI it, but, you know, same idea.

Tristan: Is there anything else I should have asked?

Ethan: I don’t know, you got me rambling, so you’re good at what you do. I don’t like interviews, I don’t want it to ever be about me. It’s the message, but I just think the timing’s just lining up really, really well for it. I just think the technology—because I was looking at some of the very first videos I did last summer, that same nativity scene. I don’t know how I knew this was gonna get better; they weren’t very good. I just I just felt this is going to end up being photorealistic at some point. With the new Nativity Scene video, it just looks, like, I don’t think anybody’s gonna think she’s not a real person at this point.

Scan Hit of the Week

This one was pretty far out there: scientists built something very much like a living cell out of DNA and raw materials. You can get fiddly with the definition of life, and it’s not, you know, great at living—it needs a carefully controlled environment and constant supplies of nutrients, but to be fair, so do I—but it can grow, reproduce, and evolve, and seems more alive than a virus. There are so many science fiction stories begging to be written about this: What does it mean to start a new chain of life that isn’t related to all of Earth’s existing life? If we can genetically engineer cell-ish things that manufacture chemicals, what happens when someone grows nerve gas in his basement, or maybe something that eats plastic and spreads on the wind, consuming modern society? Or good things, almost forgot: producing life-saving medication, or clearing amyloid plaques in the brain, or fixing carbon from the air?

Share

Subscribe now

The Argument

Even though the authors explicitly lay out the argument in the back of the book, and even though I have a master’s degree in economics, I had to repeatedly wrestle with the ideas, not just to decide if I agreed with them, but to even understand exactly what they were and how they fit together. Here is the basic argument (I think, this is my third iteration):

1. When business owners introduce technology, it is presumably because it increases their overall returns from the business. But there are two different levers technology might pull to create these returns:

   1. additional surplus is created via additional productivity, or

   2. more of the surplus product gets shifted from labor to capital (worker surveillance is a good example of this).

2. As a result of 1, technological innovation could have a range of effects along three dimensions:

   1. it may or may not significantly increase the total output of society;

   2. it improve or degrade the lives of the large majority living outside the capital class, specifically via higher/lower wages; and/or

   3. it may cause disruption by significantly shifting the skills needed by workers, the conditions of the work, and the total amount of labor the economy can productively use.

3. There are two key factors that determine the outcomes in 2:

   1. the type of technologies that are researched, funded, developed and promoted in society, broadly:

      1. human-replacing tech (automation), where capital is shaped to mimic the work being done by labor; or

      2. human-enhancing tech (the authors label the concept “machine usefulness”) that empowers us to more fully realize our intentions—this is Steve Jobs’s “bicycle for the mind”; and

   2. the relative bargaining power of labor, based on the level of organization (i.e. is big business counterbalanced by big labor?).

4. Both 3.a and 3.b are shaped by social power to communicate core ideas as visions for society to follow (such as Friedman’s “the social responsibility of business is to increase its profits”), combining:

   1. the memetic strength of the idea, including the simplicity, plausibility, credibility, and the quality of its framing (everyone loves a good story), with

   2. the social clout of the messengers, including their charisma, prestige, reputation for success, control of institutions that can reinforce and repeat the message, and so on.

5. Turning this into a feedback loop, economic success to one group over time tends to enhance their social power in 4b and thus their ability to promote self-serving ideas.

To use Noble’s classification system, this theory of change is endogenous to society, highly contingent on individual choices but with broad social implications, the key units of analysis are both real (technology implications, wages) and ideal (the visions animating society), mostly objective, and largely empirical, based more on developing appropriate theory to match reality than the other way around.

Analysis

Spending so much time figuring out what the book was saying also gave me plenty of opportunity to think about what didn’t make sense to me. These are my criticisms of the theory.

Higher production, lower wages

The linkage of 2.b and 3.a, a key focus of the book, was the most challenging for me. I worked through mathematical models, tested my understanding with AI and people around me, I revisited it a few times over several weeks, and I’m still not quite sure it’s clicking. The rock I kept crashing against was this: what does it mean for technology to be better or worse for workers?

Mathematically, the answer is simple: so-called “so-so automation” is a change in the production function where the average productivity stays steady or increases, while the marginal productivity (which determines the wage) decreases. This is possible but requires an odd shift in the function, where it makes the first few workers much more productive but then drops at the edge, so that the “last” worker is making less product than before. Maybe this makes sense in some situations with oddly constrained capital stock, but I’m having a hard time seeing it as anything more than an exception.

Creation of work

I also think the authors are sometimes sloppy in explaining causes. In the biggest example, the book contrasts British and American innovation in the 18th and 19th centuries, respectively. Abstracting past the details, the cases were extremely similar: both enabled production of goods using lower-skilled workers than in the past. But, per the authors, in America this was good, because there was a “shortage” of skilled workers (a term which always arouses my suspicion), and in England it was bad, because it reduced opportunities for skilled workers. The motivations weren’t particularly different; in both cases, the productivity increases, opportunities (and presumably wages) increase for low-skilled workers and decrease for high-skilled artisans. The authors present these as essentially 3.a.i vs 3.a.ii.

I think there are two factors that explain the discrepancy. The first is pure 3.b: labor enjoyed more bargaining power in the United States, even prior to the introduction of labor unions, largely because of a higher floor for alternatives. While the West remained open, and especially with the Homestead Act operative, people could plausibly leave a bad job and start farming; in Britain, the Master and Servant act criminalized quitting a job for better pay or conditions, so the alternative that set the floor was, effectively, imprisonment. Less capital accumulation in the New World reinforced this difference. This checks out, and the authors may be gesturing in this direction, but they don’t make it clear.

The second 3.a related, but is a more slippery concept than the authors seem to think. They describe some technologies and their implementations as creating more opportunities for humans, and others as foreclosing opportunities. I usually think about this at the level of the macro economy: for example, as automotive manufacturing becomes more automated, it creates some more high-skill technical jobs but on net employment in the industry drops, and that frees people up to be poets and massage therapists and YouTube influencers.

But the authors see this as a relevant consideration for individual technology implementations. Though this isn’t a metaphor the authors use, the best way I could make sense of it is to think about the economy as a big complex organic molecule. If you take a carbon atom at the end and double-bind an oxygen atom to it, you’ve closed off two potential paths for further growth; but if you single-bond another carbon atom, on net you’ve added two new paths. So introducing computers might eliminate the need for human calculators but requires the creation of entire IT teams as well as associated training functions;

I believe that technology isn’t neutral; per Dator’s third law, it changes our thoughts, gives us new metaphors, and makes some things easier and some harder. But does it have affordances in the way the book describes, like a designed object?

Take the image of technology as a “bicycle for the mind” vs an “automobile for the mind” as emblematic of labor-augmenting vs labor-replacing technologies. First of all, these clearly exist on a spectrum - where does an e-bike sit? To make it more concrete, look at this post about the use of AI from Alberto Romero:

At what point did this go from good “Machine Usefulness” to bad “automation”? Each step is just moving higher up the ladder of “I say what I want and a machine makes it happen”, not some quantum change, and the last step is just that the employer is the final arbiter of what they desire to be made. But I feel like all automation is like this: some tasks are done by machines, and someone does the remaining tasks, someone has to supervise the machines, and someone has to design/build the machines, with the net effect that fewer people (sometimes many fewer) are needed to make the same amount of stuff.

The original GPTs

Last, the authors also point to electrification and the railroad as beneficial technologies under this framework. I think the most parsimonious answer is that these are general-purpose technologies: transforming raw materials requires energy, and finished goods need to be transported to the buyer, so making these cheaper and easier will pay dividends across the entire economy. But this doesn’t advance the authors’ thesis, so instead they make more contingent historical arguments.

So, to sum it up, maybe the main implication is that, as systems change, those with power determine who receive the benefits and bear the costs, so continuously pushing society toward egalitarianism and balancing powers (maybe something like Walzer’s spheres of justice?) is the best way to keep society from degenerating into mass misery.

Futures Implications

OK setting aside my general complaint that I shouldn’t have to work this hard to understand a book like it’s some kind of escape room but less fun, what are the practical implications of this body of theory for foresight? I think this makes a useful set of checks on futures work:

• If a project presumes that technology’s direction is fixed, for example of all the scenarios have the same basic technology pattern, this is worth flagging as a questionable assumption.

• If the internal logic of a given scenario reflects decisions working to the advantage of a group without significant social power, or to the detriment of those with social power, that should be further examined.

• For most projects, at least one of the drivers for a project connects to ideas and how the vision of society is changing. This should have significant impact on the scenarios.

• In looking at implications, think hard about the second- and third-order effects of changes in one industry on others.

Scan Hit of the Week

I definitely don't understand all the science, but this new material developed by Northwestern University seems basically magical. It is a yellow liquid that absorbs energy from the sun, heat, or fuel, and transforms into a black gel that can indefinitely store the energy and release as needed on contact with oxygen. Because it does the work of harvesting, storing, and using energy as a single unit without conventional materials like metal or plastic and should be able to be used basically indefinitely, it seems like a huge enabler of any number of solarpunk devices and concepts.

Share

Subscribe now

Introduction

Professor Staley introduces himself and his interests at the intersection of history, design, and futures: the history of design, the similarity between the historical method and building scenarios, and teaching foresight and vision to design students. His initial entry into futures was through its connection to history, dating back to graduate school, but he’s less interested in patterns of change than in the overlapping methods: narratives based on inferences drawn from evidence, explicitly embracing a plurality of interpretations, all in the service of sensemaking.

As I’m reading this book [The Art of the Long View], it’s like, you know what? What he’s talking about here is history, except rather than looking backward he’s talking about looking forward. But it’s essentially what I do as a historian.

One-Sentence Futures

Professor Staley talks about nearly two years of writing his Substack One-Sentence Futures about five days a week. His inspiration was the three-line miniature reporting of French anarchist and journalist Félix Fénéon. These originally came from scenarios distilled to their purest essence, and has extended to imagination aided by common patterns, fragments of ideas fleshed out, and pursuing the preposterous.

I think that as futurists we need to be more attentive and more accepting of preposterous or ridiculous futures.

The Value of Provocation

Professor Staley talks about his success criteria for the project: provocation, engagement, and resonance; possible forms for further development, including as a technique for teaching students. We also talk about the challenges of teaching a generation that lacks hope about the future, and how backcasting from a desired future can create agency.

If I engage a reader, that’s a win. I think it’s successful if a reader finds my one-sentence futures to be as poignant and engaging as they would a full-length novel…that it sort of resonates all day.

History for Futures

Professor Staley talks about the importance of participatory futures and expanding the circle of who’s invited to the table when futures are debated or planned. He talks about the example of The Futurists (1967), and why tech companies have so much social power in describing the official future.

The question that I ask again and again throughout the semester is “Who gets to make the future? And I think we live in a moment…[where] it tends to be corporations, companies, especially tech companies, that feel that they control the narrative of the future. And a lot of my work in Design Futures is about how we extend that ability, not just to name the future, but to make the future, to more people.

Other Projects

In wrapping up, Professor Staley previews his new book project, The Glossary of the Future, playing on how language changes in unexpected ways to accommodate changes in society. He also shares the connection to his recent Anticipatory Biographies as another way to play with the form of scenarios as well as deepen the design practice of personas and reveal the interior life associated with potential futures. He also mentions The AI Symposium.

These are more fully developed life stories, and really trying to understand the inner life of people in the future.

Scan Hit of the Week

One of the big trends of the last decade, which seems to still have plenty of momentum, is the spread of gambling into every corner of society. This includes not only sports betting extended to all localities and levels of play, but also the expansion of prediction markets into gambling on other kinds of events. Newest on that list is Wyldfyre, a prediction market centered around betting on the location of (naturally) wildfires. Now, being able to have a functioning prediction market for wildfire damage could be incredibly useful for actors with a real stake, as it would allow hedging, insurance for unconventional activities, etc. But also, prediction markets around events people can control can lead to all sorts of bad behavior, as we’ve seen in the last year or so. Starting a wildfire to make millions of dollars seems like possibly a new high-water mark. The website seems to be offline at present due to outrage, but it really existed earlier this week. Beyond just sports betting arrest scandals, giving people financial incentives to alter the course of events in potentially anti-social ways is a problem for which we haven’t developed a solution.

Share

Subscribe now

The first example is the core premise of the film: the creation of a robot child named David. Everyone he meets is astonished to see him, because nobody has ever made a child before, even though humanoid robots have been around for over 75 years in various capacities. It’s not illegal, it’s just that everyone has decided it would be unethical because a robot can’t ever really love. Look at what people have done with LLM-based companions—literally anything you could imagine exists. I realize that there’s a high-capital barrier to entry with robot hardware, but in a world with controlled population etc, nobody thought they could make money from synthetic children? What about as child actors that never need to take a break? What about as companions in assisted living facilities? Even if you don’t want to take it to dark places, this strains credulity.

The second example comes from the scene that sets the stakes during the second act. Reactionary forces in society celebrate the virtues of an imagined past through rallies called “Flesh Fairs” filled with angry catharsis and the future equivalent of Kid Rock. They feel better about themselves by collectively dismembering robots2 as a “celebration of life”. So far so good! Changes in society tend to provoke reactions, and the reactions are usually smaller than the initial change. But then, plot twist, nobody is willing to throw their bean bags at the dunk-tank-style target that dumps acid on the robot child, because it tickles some kind of moral intuition, and instead throw them all at Brendan Gleeson, enabling a daring escape. Again, you have a stadium full of people that paid money to inflict violence on robots, for fun, and none of them think it would be fun to melt a robot kid, or potentially a real one?

The common failure mode here is the “nobody has/would ever” assertion. There might be reasons a product isn’t commercially viable, or is illegal, or wouldn’t spread, but I know enough humans to know that someone would give it a shot. Look at the history of people3 trying to fly before the Wright Brothers figured it out: even though everyone who had strapped wings to himself and jumped off a tower in the past had failed and died or come close, the next guy was sure that this time it would work.

That's the angle from desire. But what if nobody thought of it? Here’s an excellent, detailed argument from Brian Porter that most inventions come within 10 years after the prerequisites fall into place to make the discovery straightforward4. Most of the big gaps are because people wouldn't have had a conceptual use for the thing, like surgical masks before germ theory.

I talked about Murphy’s Law last summer as a good reason to worry about big things causing big problems and little things breaking big things. In this context, the lesson is a little different: in any plausible future, things will be used for unintended purposes, there will be different levels of quality, people will make cheap knockoffs, etc. In Could Should Might Don't, Nick Foster praises the scene in Minority Report where Tom Cruise struggles with a malfunctioning multimedia cereal box. Similarly, in Years & Years, the invention of fancy human augments also mean sketchy-Eastern-European-guy-in-a-boat-offshore-demanding-cash versions.

Bottom line: don't build futures that depend on nobody doing something, or not figuring it out, or something only existing in one form, or something always working; further, be skeptical of futures works that do.

Scan Hit of the Week

Last week I talked about the importance of crossing STEEP domains to get interesting futures content. Here's an interesting example of this in the present: climate change has very different effects based on social class. For example, in a resort in Dubai, there's a street that's always a lovely French 27 Celsius due to artificial rain and breeze, to the horror of dads everywhere who aren't paying to cool the outdoors. On the other extreme, the Dalit in India face multiple social and economic disadvantages that multiply the effects of climate change, from the thermal qualities of their homes to the fact that breathable natural fabrics can cost three times as much as synthetics. As we move toward deadly temperatures more often in more places, these are going to be the kinds of mundane details that make the difference between a miserable heatwave and a humanitarian crisis.

Share

Subscribe now

The show is extremely grounded and plausible futures work2, and I think it has a more realistic view of change and inevitability of future events than Extrapolations. The starting premise is that the present, or the parts of the future hurling toward us, are no longer legible3, and also a lot of it is broken. As a balancing point, though, even when some seemingly world-changing event happens, the world kind of keeps turning and people have to go to work. Unfortunately, like Extrapolations, the core lesson is delivered in a preachy monologue: the future doesn’t happen to us, it’s the logical result of our prior actions, so you’re consenting to the future you get if you don’t do something about it. In answering that challenge, the message of the show becomes basically “What if the real future was the friends we made along the way?”. I recommend watching it if you are a completionist or if this article makes it sound interesting.

In any event, I wanted to cover the potential futures the show explores, using the STEEP framework, and then specifically talk about the value of looking at these things in combination.

Social

• Political candidates are able to use their own media platforms as an entirely different way to connect with voters (and avoid pesky laws about electioneering)

• People become open to new food sources, including algae and burgers made from orange peels

• There’s a strong anti-tech, anti-elite backlash as the social fabric seems increasingly strained

• Terrorism continues to be an effective way to spread fear, including via dirty bombs

• There’s a mandatory pornography class in high school, though it was unclear to me what the curriculum or learning objectives are

• My single favorite moment: 2026 sees the release of yet another Toy Story movie, this one called Toy Story: Resurrection due to how definitively the prior one ended4—clearly some parts of the future are determined enough to predict with certainty

Technological

The show spends a lot of time exploring transhumanism and human augments. Parents who were prepared to accept a child changing gender are caught flat-footed when the child instead wants to trade in a physical body and become a computery guy by uploading their consciousness5. Brain-computer interfaces advance to the point where devices can be controlled via gestures captured by hand implants, where phones can be inside your head, and so on6.

Little things everywhere: drones, cellular agriculture, smart houses, breath profile as ID, 3D printing homes, fixing congenital diseases in utero. Also there’s this thing called “artificial intelligence” that creeps up on us, crazy right?

Economic

Global economic uncertainty causes increasing volatility in financial markets, eventually leading to bank collapse and general stagnation, supply chain issues, drug shortages, etc. Former professionals string together sketchy gigs to make ends meet.

Environmental

Not surprisingly, the main focus here was on the climate crisis. However, it didn’t feature very much in the story, unlike Extrapolations where it was the main character. The sense that climate collapse is inevitable leads to a general feeling of apathy and malaise—why not enjoy things now, since there’s nothing we can do? But there’s a fair amount of color as well: by 2028 bananas are extinct due to their genetic vulnerability, the UK experiences massive regional flooding due to 50 days straight of rain7, insect populations are crashing, antibiotic resistance is leading to more deaths from routine infection, and there’s a new pandemic.

Political

The political arc is the main plot focus of the show, and it’s essentially a warning about the possibility of the UK descending into the same kind of illiberal right-wing populism evidenced in the US and other nations at that time8; unsurprisingly, the movement is headed by someone from the business world who doesn’t really understand much about policy but can tap into what makes people angry9. Episode 6 was a little on the nose, as a particularly British descent into right-wing authoritarianism flattens into a very obvious kids in cages / eroding due process for immigrants / media as the “enemy of the people” / defunding public media / picking fights with journalists at press conferences, but otherwise it maintains plausible deniability.

One of the most nuanced ideas, portrayed very well, is how Leftist nihilism accelerates this process; basically, if every government is fascist, and the best response is to burn it all down, you may find out how bad things can really be when you get your wish and institutions erode. There's also a related thread showing how left-wing populism often ends up in many of the same places10.

Two more quick shout-outs, the first provocative, the second prescient: checkpoints and barricades for high-crime neighborhoods are a great “it couldn’t happen here” moment, and human rights violations stemming from Russia annexing parts of Ukraine.

Intersections

Years & Years does some great work connecting change across the STEEP domains:

• AI, not surprisingly, cascades changes through the economy (replacing white-collar workers), politics (by 2027 deepfake videos are making campaigns chaotic), and society (if nothing you see can be trusted, then we’re basically living in the Dream Society).

• Human augments, not surprisingly, are very expensive; this is a great example of the Gibsonian idea, often ignored in popular futures media, that the future isn’t evenly distributed. One solution for this is people trying to get these things from sketchy chop shops; another is to get them as part of your employment, in which case you are participating in a brand new kind of indenture.

• Recurring blackouts from cyberattacks on the electricity grid lead to a resurgence in paper as a trustworthy medium11.

• If kids wear VR headsets all the time, it transforms family, school, and society to varying degrees.

• Sex with robots, a certified classic!

They missed one that disappointed me - at one point there’s some kind of contraband interference device that can basically jam any digital device in a short range, and people talk about legalizing it, but then it never comes up again12. But what does it mean if one person can decide the whole restaurant, or stadium section, or small concert venue, can use their phones? What are the implications of this kind of device for people with hardware augmentation? I’d like an Episode 2.5 dedicated to this topic.

But the above examples are sufficient to make my point: these intersections are where all the spice of futures work is. This is where the familiar becomes strange, where people realize that the baseline future isn’t particularly likely, and imagination starts flowing. Do this as often as you can.

Scan Hit of the Week

Yesterday Sam Altman correctly identified the current state of AI as trying to sell technology to people who don’t understand or care, and the breakthrough is going to be when they identify their analogue to electricity’s “light at night” proposition. One step in this direction is “confidential expert on demand”. Outcry is an AI specifically to help activists understand how to effect change in their communities. Even better, it can run locally on a device, so you don’t have to worry about who’s listening13. This direction of people needing expert, arbitrary advice in real time”, built on a corpus of belief and praxis, is adjacent to ideas I’ve been playing with and seems like it might be a viable model. Hat tip to Competia for surfacing this.

Share

Subscribe now

Intellectual Biography

Bryan shares his background, how people only want to talk about AI these days, and self-learning futures after reading Future Shock as a child.

Career Synergy

Bryan talks about how the various parts of his process fit together into his work, including passive reading and actively seeding conversations, including via the very robust Future Trends Forum.

Building a Newsletter

Bryan discusses the joy of being read and starting a conversation, the difficulty of writing consistently, and the factors he thinks account for his success so far.

The Power of Scenarios

Bryan extols the value of scenarios in making future possibilities immediately comprehensible at a human scale and narrative form, and the importance of grounding them in real research. In addition, he shouts out The Thing From the Future and the power of games generally as futures tools.

AI in Education Signals

Bryan talks about the controversy around AI in education, which is more intense than he’s seen for any technology in the past 40 years. He also talks about the practical work involved in a university staking out a position on AI, including all the decisions and facets to consider (including how the approach may need to vary by domain and department).

Chat Ascendant

We discuss how the chat interface plus “good enough” language skill was the magic combination to hook our squishy human brains, how character.ai and Replika are building off of this, and how romance in the future could be completely intermediated by technology, even when humans are on both sides.

AI’s Impact on Students

Bryan uses The Diamond Age as an image of how some children are currently using chatbots to grow up, how that might intersect with a university—the impact on student life, the potential implications of various policies, and the incredible heterogeneity of the problem.

Closing Matter

Bryan talks about his books concern possible futures of education, how good futures work is often about recognizing signals and knowing what to chase, and the importance of knowing good data sources in your domain like the National Student Clearinghouse. He also pitches the Learning Design & Technology Masters program at Georgetown.

Scan Hit of the Week

The city of Potsdam in Germany has just banned smart glasses from pools and saunas due to fears of surreptitious recording. Pennsylvania is weighing making it a crime to operate smart glasses without a recording indicator such as a light; the Potsdam ban goes further and restricts the devices entirely. This is interesting, because the intersection of “legislate based on the worst thing someone could do with something” intersecting with devices increasingly becoming “anything machines” is why we can’t have phones at school, touch them while driving, or take them to Phoebe Bridgers concerts.

Share

Brand is the co-founder of the Long Now Foundation, which tries to teach/inculcate deeper long-term thinking. This is clearly a goal too abstract to get much traction, so there are more tangible manifestations; in addition to the foundation itself, there’s a website to publicly register long-term bets on future events, an etched disk containing data on 1,500 human languages, and a bar with timey-named drinks. But the centerpieces of the plan are the titular clock designed to (very slowly, as in it chimes once a year) keep time over the next 10,000 years in the middle of a mountain and a library to accompany it somewhere much more accessible. The book is a bit of a time capsule explaining the thinking that animates the entire project.

Subscribe now

How Far Out Should We Care?

The basic logic that Brand employs is that there will be many more future humans (hopefully!) than lived on the earth so far, and we can’t do anything to help the dead anyway, so we should use more of our energy and attention on things that help increase the options for future generations. The strong version of this argument has essentially become today’s longtermism, which can get pretty degenerate1. But the basic version is the kind of common sense that follows a psychedelic experience, that if we really are all interconnected we should care about one another; this extends not only around the Earth (remember Brand’s connection to the image of the planet from space) but also forward in time2. The clock is an artifact that embodies that care and inspires it in those who witness it, a civilization-scale project (though over time instead of geography) .

A key feature of our current reality is rapid change, largely stemming technology. Brand points to Moore’s Law holding more or less steady for decades; since computers are used in the design of computers, our collective capacity continues more or less exponentially3. This acceleration means that, for a few generations now, people are seeing the world significantly change over the course of their lifetimes. It’s possible that this gets so fast that it breaks our ability to cope: this is basically Toffler’s Future Shock, Kurzweil’s Singularity, and the rupture of Rao’s manufactured normalcy field. Brand notes that, regardless of whether that happens or not, the current rate of change is rapid enough that our culture has taken a very short-term view. Evidence includes the rise in gambling, trying to experience “an event” now while ignoring the slow loss “non-event” over time. Additionally, the fragility of modern storage media, the rapid changing of digital formats, accumulating, and the way the internet is stuck in the eternal “now” of whatever’s on the server when you happen to visit (and today we would add the paywallification of the internet), mean that in many ways, documentary history from the last 80 years is more likely to be lost to us than the correspondence of scientists 500 years ago4.

Other timescales are possible as the default: the life cycle of a human or maybe a tortoise or tree, the period of comets, the 25,784 years it takes the Earth’s axis to make a full period, or hey why not the 200,000,000 years it takes the Milky Way to spin around once? So the clock picks 10,000 years as a nod to the role human fingers play in our description of reality5, and as a way of making it real in everyday life, midway through the book starts writing years in five digits (i.e. we’re currently living in 02026). There was a period ~17 years ago that I wore a fedora, so I do have some experience courting ridicule in the hopes of setting a trend; still, I’m undecided on whether I can commit to this one.

You Knew It Was Coming

But Brand isn’t just all long-term, all the time. He talks about the throwaway culture and ephemerality of America as one archetype of dysfunction on a spectrum across from the graves stacked on graves and every field representing a battle that gave rise to a thousand-year-old grudge in Europe. One solution to this is to build spaces and institutions specifically to hold the past, like universities, museums, and religions, so we can clear it from the rest of society. He explicitly says that he doesn’t want to build a religion associated with the clock, because even though it’s the best way we’ve found to encode information and ideas long-term, and it’s great at helping people mark the longer rhythm of human life, it can often get stuck in focusing on some eternity beyond an eschaton that relieves us from dealing with the real problems of the world, now or later.

So then how does he conceptualize our relationship to change? Pace layers! I believe this is the 8th time I’ve written about this concept; the last time was just last week, forming a key part of a theory of mind rolling together many of the concepts I’ve learned over the past 40 months. But this book is the original place he articulated this idea fully matured.

The fact that the layers move at different speeds is a feature of the world, not a bug. It’s natural that more superficial levels change faster, and that it causes friction with the deeper layers. That negative feedback is a key part of what makes the whole thing work so well, it’s a key reason why revolutions are often so disastrous (because they try to sever the connection with lower layers).

Brand’s observation is that in our sped-up world, we’re not properly attending to the deep layers of our existence, specifically the culture and nature. The 10,000-year clock (and the library, in a different way) is a corrective to this neglect. As part of this, the clock is designed not only to last a very long time with very slow-moving parts, but also to keep humanity engaged in its history: not only does the clock require human winding every few decades to keep time (I think; this part is a little complicated), but various ideas like growing crystals, having visitors deform the rock, etc, have been pitched to mark time in a different way.

Advice for Futurists

In this work, Brand calls for more good futures work, as opposed to fake futurists possessed by what he calls “futurismo”. He sees these “Should futurists” as earnest and committed to a vision, which can lead to all kinds of disasters; instead, he encourages more “Might futurists” doing good scenario work, which can help people connect their interests to the health of the overall system, and “Could futurists”, especially science fiction, which is the medium best suited to long-term futures. He’d also love to see more long-term scientific studies, and commit to some big projects like terraforming Mars.

Brand pitches his viewpoint as basically Viktor Frankl’s “tragic optimism”. Practically, people act better when they think things are improving than when they think they need to fight to keep what they have, so it’s no good trying to downplay the ways in which we’re making progress as a civilization; on the other hand, warnings can also be motivating if used judiciously. The approach should be something like “we’re making progress, but there are a few threats we should act to avoid”. Ideally we’re working toward the “infinite game”, where we keep expanding options for future generations6.

As of today, the clock is real, and it’s possibly about as done as it will ever be, but you can’t visit it yet. It’s in the most inhospitable part of Texas (no, not Dallas), the part of Texas furthest from civilization (no, not Lubbock, but you're getting closer), and hiking there would probably take all day with decent odds you get eaten by a mountain lion. I'd love to visit someday, and probably have a few decades left for the details to get worked out. He also talks about putting an “archive of civilization”-style library on the Moon as the ultimate safe place, but I probably won’t get a chance to visit that.

Scan Hit of the Week

Venice has apparently been restoring the marshy habitat of their lagoons, and as a consequence it is increasingly an important habitat for flamingos. I’m under the impression that Venice already has all the tourists they want, but this idea of intentionally cultivating ecosystem amenities for various overlapping reasons—climate mitigation, biodiversity, ecotourism, etc—is moving upmarket from developing countries preserving their wild spaces to highly-developed regions re-wilding.

Share

You may be aware that Richard Dawkins is convinced that AI is now conscious. I’ve seen a few speculations on why this may be; my own guess is that, as a physicalist/materialist, he has few options for defining or evaluating consciousness other than by comparison to what firmly conscious beings (such as himself) are like. To shift the philosophical ground a little, I don’t believe any currently existing AI systems are moral persons1, but I’m open to the idea that they could be at some point. In an intense debate with my wife on the topic2, she said that believing this would be a huge mistake. I eventually realized that some of the conflict was because I was talking about ontology and she was talking about epistemology3. Her point was strong enough that I’ll formulate it as Nicole’s Law:

If you think a machine built to mimic a conscious being is, in fact, conscious, there’s a much better chance that you’ve been deluded than that artificial consciousness has been achieved.

Subscribe now

I bring this up as preamble to the latest project I’ve been experimenting with: I’m trying to simulate a mind, based on all the ideas I’ve been learning in my Masters program, plus other concepts to fill in the gap. This is not trying to simulate a brain in its architecture or circuitry—there have certainly been interesting developments in that direction, and neural networks in general gesture towards the way brains work—but rather the architecture of thoughts and beliefs. Two motivations: first, I’ve seen plenty of the “respond as X person / Y role”, but they strike me as kind of shallow and quickly revert to the mean; second, it gives an opportunity to model not only how a mind works in a static way but how it changes. Plus it’s cool and neat!

So here are the different concepts being rolled together into this experiment:

• The mind exists in the same structure as the projection of reality onto the mind in Causal Layered Analysis: there are a set of superficial beliefs about facts, then under that beliefs about the systems that give rise to those facts, then under that a worldview that makes sense of those systems, and then at the core an organizing myth/metaphor.

• Alongside the myth/metaphor at the bottom are three different ways of thinking about values: Moral Foundations Theory, suggesting the salience of each of six primal moral pulls; the values work of Andy Hines, building on his book ConsumerShift, and giving the top 12-15 values from a big list; and Schwartz’s Universal Human Values, ranking the relative importance of each of the 10.

• The worldview is based on Apostel and Vidal’s definition as something answering 6 key questions (from the table on page 4 of the link):

  • Ontology: What is?

  • Explanation: Where does it all come from?

  • Prediction: Where are we going?

  • Axiology: What is good and what is evil?

  • Praxeology: How should we act?

  • Epistemology: What is true and what is false?

• The systems are based on Cynefin, probably my second-favorite concept in futures. For clear situations, simple causation beliefs are appropriate; complicated situations lead to system archetypes; complex situations pull out the principles of complex adaptive systems; chaotic situations would mix the observations of chaos theory with a bias for action.

• The litany is a set of more conventional opinions or views on topics. One way to potentially represent them is to use Personal Construct Theory: create a spectrum between two logical poles, and then pick a point on the spectrum.

But all this is just a lead-in to the fun part: the interaction and dynamism between the levels. No surprise to long-time readers, this operates via the conceptual framework of Stewart Brand’s pace layers4: deeper layers change more slowly, and each layer anchors the one above and invigorates the one below.

Again, to be clear, a system using this structure to provide context for LLM responses would not be an actual conscious being, but I am hoping it would mimic thoughtful, consistent reasoning and evolution of perspective in a more interesting way than anything I’ve seen.

Here’s an imperfect but directionally correct Claude Design diagram of how I’m currently thinking of this:

I have an early prototype of the system and a specific mind in the system, and it’s been interesting to see what happens as it ingests books and positions update5. Eventually I hope this is in good enough shape (fidelity, quality, stability, etc) to release. Get in touch if you’re building something similar, have ideas, or want to help.

Fast Futures

New topic: if you haven’t seen it yet, I strongly recommend checking out David Staley’s Substack One-Sentence Futures. For 18 months, on a daily-ish basis he’s been posting a short provocation, a nano-scenario, a micro-implication, or whatever you want to call it. It’s not clear to me whether his “sentient Constitution” post actually broke containment6—it got more likes than any other post I saw—but that’s how I became aware of the work. Many of the sentences are deeply run-on, but I appreciate the commitment to the bit and have myself been known to write some long sentences from time to time, even if they make my writing a little inscrutable. Genomes for sale, outlawing vowels, losing color: they’re no baby shoes, but they’re fun! I think they’d be great to gather in some kind of collection where you could start a meeting with one at random and pick apart what makes it surprising/transgressive/etc, and what that can teach us. Or a Future-of-the-Day calendar!

Scan Hit of the Week

This week’s hit is from the Spanish Huffington Post7 and is about the growing role of salicornia as a cash crop. This poor plant is also known as sea asparagus, glasswort, pickleweed, and marsh samphire, and is able to grow in very salty soils. This makes it a great crop to grow on marginal land, and also it seems to bring a salty flavor to foods with less sodium, so it’s becoming a cool thing to grow and serve in restaurants8. This is a good example of a “new food” that is entering the public consciousness as a result of climate change and soil degradation, but isn’t part of the cellular agriculture / vertical farming / eating bugs traditional conversation.

Share

Subscribe now

I mention this because I picked up Nick Foster’s book Could Should Might Don't, subtitled “How We Think About the Future”. I assumed that it would be about better and worse ways to pursue strategic foresight, because my immersion in the Houston Program makes it easy to think that strategic foresight is Futures. However, I was momentarily outraged to find that it occupied a single chapter in a single section of the book2. I think that both the perspective of the book and my reaction to it were instructive.

Foster describes the connection of economic incentives, popular media, and the consequences of the agricultural revolution (cycles of growing food and saving it for winter3) and industrial revolution (making technological change fast enough to see at human scale) in determining our relationship to the future. As a result, we have a lot of low-quality future images colonizing our imagination, some to sell you things, and some because they’re just exciting. In addition, the current pace of change in society makes it hard for people to think about the future in a normal and common-sense way4. This is why we need futurists now, even though that term is frequently embarrassing5, evoking crystal balls and wild science fiction concepts detached from reality.

The contribution Foster offers is the division of futures work into four main families, which he names (you may have guessed this) Could, Should, Might, and Don't futurism6. His goal is to educate lay readers about these different divisions and give them told to think critically about them.

Could

Could Futurism is the flashy “gee whiz”, wishful thinking, dessert-first kind of futures work. They basically own the word “futuristic”, and make big, exciting predictions that are predominately about technology changes. Rarely do they bother to show intermediate steps, but they are sure that they are heralding progress. Science fiction films go back to Melies and are now a huge part of the industry, and it's usually Could Futurists helping bring those worlds to life. They represent the apotheosis of the Enlightenment ideas of progress being the most important goal, overriding any sense of duty to the past. This is the work that birthed the Futurama exhibit at the 1939 World’s Fair, Star Trek, and the Museum of the Future in Dubai, which Foster finds to be depressingly focused more on future vibes than any useful conversation about what might be coming next7.

This work has several other natural habitats: keynote addresses, documentary interviews, brand vision videos, and the “sexy futures” content hat has replaced sexy women on trade show floors. Foster’s indictment is that this is mostly like the concept cars that are also part of the trade: sleek and shiny on the outside, but mostly just painted clay8. Most notably, their focus on grabbing attention and maximizing engagement biases them toward “aspirational futures” of people at the top of society having a great day. Most of the future will be lived by people in the middle, with products that aren't brand new, buying budget-conscious versions of the shiny things being imagined today; imagining how these interactions would go is much more interesting work.

Should

Should is a tricky word: it can be used to mean either “I think it would be best if” or “i think it likely that”, which are very different concepts. I didn't see Foster spell this out9, but he leans into both meanings in this section. The first is the domain of the ideologues, people who have a specific view of the future they are convinced is better, and then advocate, persuade, and agitate for its creation. Of course, “better” is neither an objective nor an uncontested assessment, so this work is inherently political.

In the Houston program, I've noticed that maybe half of the students land mostly in this camp: they have either a pet issue or a broad future-oriented agenda, focus on this for their projects, and seek work accordingly. It's an interesting dilemma: at some point, your futures work has to be used toward some good end, or it's not worth much; but if you let your views about different possible futures cloud your objectivity (or even just your perceived objectivity), your work might not have the credibility to get used at all. I lean much more toward the “detached” end of the spectrum—when I see a news story about a potential global catastrophe, my first thought is “ooo, that would be interesting!”, which obviously has its own moral issues.

The other group is the quants and forecasters who seek to increase certainty about the future, picking up the task in recent centuries from religious prophecy. Sometimes we can model the systems well enough to make accurate predictions—weather forecasting is a particular success story over the last century, and prediction markets are a fascinating recent example. However, most attempts to quantify predictions about the outcome of human systems years into the future are just math-washing (new term #2 for this essay!). Also, Foster reminds us that, for some reason, every time we try to use science to make the world objectively better, we reinvent eugenics. His big lesson here is that the future is a process, often show, that builds on itself; when you see a project like The Line that attempts to build something out of whole cloth, it's usually a sales pitch. He also mentions newsletter favorite pace layers as a key way to think about how the future is likely to change, though the paces of different layers seem to be changing, and also gestures in the direction of Rao’s manufactured normalcy field.

Might

This is where we first take seriously futures, as in holding multiple possible futures at the same time. Foster traces the history here from Pierre Wack at Shell and Herman Kahn at RAND, heavily overlapping with the strategy world, and noting that most think tanks today are thoroughly doing Should futures and are basically lobbyists with a veneer of academic rigor.

Then he talks about strategic foresight! It's our time! Foster is, to his enormous credit and in contrast to almost every other popular book about futures, actually aware of and conversant with, many critical ideas and thinkers in the field: Voros, Candy, the futures cone, STEEP, morpho, CLA, VUCA, and using way too many sticky notes. He poses the work as a kind of insurance, and says that the thinking done along the way is usually more important than the artifacts generated10. He talks more about probability than I would expect, but otherwise I think people would walk away from that chapter with a accurate and somewhat flattering view of the field. He basically frames this as the best in class way to think about the future, but it often has to be snuck into projects or added on, because its divergent nature doesn't lead to clear decisions11.

Don't

Foster last goes to the contrarians. Fear of a bad future can be a powerful motivator: Hell is the early canonical version of this, and we've been talking about the negative potential of science gone wrong at least since Frankenstein. There's a strain of mathematical Don'ts from Malthus to The Limits To Growth to An Inconvenient Truth, and then there's the more creative “really imagine the negative implications of this development” side, which borders dystopian fiction.

Foster points out two key ways this kind of thinking can be directly helpful. First, as a check on design, being able to imagine the bad (or even just mundane but unplanned) ways a product, technology, or policy can be used12 , as well as the various failure states and ways things can break, can help build more robust strategies for mitigation and intentionality in what gets released. Second, realizing the complexity of the world new technologies interact with, such as self-driving cars interacting not only with good drivers but bad drivers, and potholes, and cute little lines of ducklings crossing the street, and jaywalkers, and the insurance/tort system, helps prevent failures of imagination from becoming tragedies; this is similar to the idea of implications in Foresight, but with a stronger attempt to be comprehensive. The book has a great idea for an exercise from a Google project Foster worked on: when discussing a new technology, present an escalating set of ideas for its use, starting at “barely a change” and ending up in “robot overlords” or whatever, and then have people reflect on where they got uncomfortable and why, and what that means for guardrails etc.

Wrap

Foster finishes by acknowledging that thinking about the future is hard work: the default is just to take whatever future image has colonized our brain and find some confirming evidence. Each of the kinds of futures work has something to offer here: Could can sometimes be useful and frequently motivating, Should builds confidence to change, Might brings structure and can find new opportunities, and Don't adds caution and responsibility. He suggests bringing the strengths of these together to imagine what the future might be like after all the new “technology” and “innovation” becomes just “stuff”, by taking something mundane like a cereal box (which encodes maybe 20 pieces of information about the society it's embedded in), and imagine how the future you're imagining would be reflected in that object, then make the object and let it drive conversation.

In Strategic Foresight I've often seen the world divided into “serious futurists”, the kind that use methodologies and have academic rigor undergirding their work, and “charlatan futurists” for everyone we're jealous is getting all the speaking invites. I think this is a much healthier view of Futures as an ecosystem with something close to an “it takes all sorts” attitude. I do still want to get five figures for a keynote someday though.

Share

Scan Hit of the Week

I try really hard to not make this spot about technology only, especially AI, which you surely hear enough about. But how can I not talk about GIANT MECH ROBOTS YOU RIDE IN!? Unitree released a working demo of their new robot and it's the first time I've ever desired something that costs $500k+ that isn't a house. Remember how you felt two years ago when you saw your first Cybertruck on the street? Multiply by 10 when you see me cruising in this. Occasionally I am a boy first and a futurist second.

Subscribe now

The presentation uses my website to show how I’m doing the work of using AI to assist with foresight, but everything is geared toward showing the method and giving you the information and pitfalls from my journey so you can take your own if you wish. This includes the shared foresight data model1, how to use that to build futures projects, and how to safely/effectively use AI to enhance that process.

It’s worth mentioning that shortly after making and delivering this presentation, I decided I could no longer live with the drab, default Bootstrap theme of the website, so I spruced it up. I think it’s a big step forward.

If you want to check out the site and try it for yourself, you can get your first month free using code SUBSTACKTEST26.

Scan Hit of the Week

Probably nobody wants to hear another story about AI, so I’ll skip this one about the AI that’s running a cafe in Stockholm and doing kind of a bad job. Instead, longevity science is back! Courtesy of The New York Times Magazine, this article starts with the incredible question “Why are babies born young?” Because, you know, the cells they come from are grown-up cells! But then they do a thing to rejuvenate (in the literal sense of making young again) themselves in the first two weeks of life. We’ve done the same trick to human cells, and mice, and rat organs. Will this work in people, so they could get a two-week treatment every 20 years and never grow old? Probably not! Longevity research is perennially on the brink of solving aging once and for all2, presumably because humans spend a lot of time worrying about death. The first wave I remember was antioxidants and specifically resveratrol (mentioned in the article), and people still eat their superfoods and drink their red wine and get old and die. There’s also interesting commentary on hype and language and how it sets us up for disappointment when the future does arrive. The example in the article is that curing all cancer would raise the overall life expectancy by only three years; I think this is a good lesson for futurists trying to maintain public attention while maintaining credibility.

Share

Subscribe now

If you want to look at the content I harvested from the semester, you can view by domain here. But we did also cover a few interesting pieces of theory and method:

• A practical guide to help non-foresight folks maintain a scanning practice;

• A rigorous treatment of the polycrisis;

• A solid theory of the constraints on human moral systems and values;

• Three different ways to build scenarios.

To talk about the other major part of the course, my project on the Future of Orbit was fine. I think the drivers were solid work, and the sci-fi prototyping scenarios were fun, with my orbital-mirrors-as-romantic-revenge story being my best fiction ever, but other than that I wasn’t too inspired. I think it was challenging not having an actual client or project lead to give feedback and direct the work, but of course it’s hard to rustle up real clients, and a full project is a lot for a single person.

So here's my sketch of how I might redesign the course to better suit my taste3:

• It was probably too good an opportunity to pass up to use the 2025 book that Andy helped write as a major part of the curriculum, but past 2026 it needs a different organization.

• To get the basic shape of the present and current trends, some official-ish source like the US National Intelligence Council’s Global Trends report4. Add values stuff from the Hines/Spiral angle.

• After digesting the H1 stuff from NIC, students could supplement each category by scanning for weak signals representing H3 possible futures. It might be nice to have some kind of interactive element here, where students can comment on and upvote the submissions.

• The recently-released STILE framework could play a key role as well. For some number of weak signals, students do an analysis highlighting how close the innovation is to the present and what someone could do to better prepare.

• Have some special topics that cross domains, like the polycrisis stuff mentioned above.

• I think I would keep the book report—seriously engaging with a possible future and assessing the strength of the argument for its plausibility in light of what the class has been covering is a good critical thinking skill. This is basically the content-driven analogue to the theory-driven final project in Social Change.

• I would definitely keep an individual semester-long project on a chosen domain as a central element of the course. I also think that letting students pick from a variety of scenario techniques is a great opportunity5. However, I would spice it up in two ways. First, in my experience it’s critical to have client feedback, so I would have each student act as the client to three other projects (then every student would have a board of three client advisors), critiquing and identifying gaps in the work over the semester and helping drive improvements and revisions. Second, I would drive all this work toward two artifacts: a professional-looking report, and a presentation delivering key findings in a compelling way. Maybe tie the grade for the client critique to the grade the project gets, so you’re invested in making it the best it can be.

It has been a long time since I taught college students, and I only taught one graduate-level course, but I like the way this plan integrates some of the latest practical developments and techniques, and moves a little closer to what futurists might do after graduating from the program.

Probably the biggest highlight of the course was getting to take it from the UH Foresight program lead, Andy Hines. For whatever reason, Andy seems to tone down his dynamic persona for the kinds of things that end up on YouTube, so it’s hard to get a sense of his infectious energy: my interview with him or this video teaching futures to local government leaders are the closest6. But also, he has led such a fascinating career and touched so many areas of the future that it’s extremely common for you to hear a new idea, and then have Andy tell you how he first heard about it 30 years ago, how he brought it up 25 years ago and was laughed out of the room and how it seemed like it was about to gain traction twice but then receded, etc. Andy has announced that he’s going to be retiring in December, and it’s hard to imagine the program without him.

Scan Hit of the Week

In case you were worried that Christianity was cornering the slightly offputting religion-as-a-service digital market (Find God’s Peace! Try it for free!), an Indian company is building out a temple management and payment stack called that includes “AI-led personalization and guided spiritual journeys”. It’s interesting to consider whether an aversion to the ideas of corporate subscription-for-spiritual-benefit and using AI to supercharge or shortcut spiritual progress is just due to its novelty or more deep-seated values, and to think about how that might change in coming years. Widespread access to printed books enabled the Bible-centered Protestant religions of the last five centuries; it will be interesting to see what AI-enabled religion creates as a consequence of the medium.

Share

A book about communism, labeled by the author as a manifesto, you say? Surely it’s a highly practical book with a nuanced view of history…

Hypothetical reader, I’m afraid I have terrible news for you.

Subscribe now

Bastani’s argument has three parts. The first part, unsurprisingly, is that history is basically determined by the modes and technologies of production, and there have been just a couple of major changes to this throughout the human story. The first was the creation of agriculture; by creating surplus food during some parts of the year that could be stored for the other parts of the year, and by intertwining fate long-term with a specific location, we basically created the future as an object of consideration. The second was the Industrial Revolution; by creating a way to store energy for later use, and then turn that energy into infrastructure and goods, it shrank the world and created a joint future. The communications revolution of the last 100 years or so is an acceleration of this dynamic, and the decomposition of everything into bits. Each of these epochal shifts have changed fundamental parts of what being a human was like.

So now for the second part of the argument. We’re currently at a hinge point in history3. On one side, we have five interlocking crises: climate change is threatening the stability of the ecosystems we depend on, we’re running out of key resources we need to sustain our lifestyles, populations are aging and the increasing dependency ratio is becoming increasingly burdensome, there’s a growing class of people that we don’t really have useful work for4, and rapidly advancing technology is accelerating unemployment. This is causing increasing food insecurity, persistent poverty, and the return of radical and violent politics.

However! On the other hand (or, for some of these, the other side of the coin), we’re on the cusp of a potential end to scarcity in five areas, such that it won’t even make sense to charge money for them. First, the increase in automation is making capital increasingly a full substitute for labor. We’re getting closer to the point where machines could theoretically do almost all mental and physical labor associated with production. Second, the rapid solar buildout is making decarbonization possible5 while expanding the amount of available energy, especially in the Global South where there’s less incumbent infrastructure. Third, asteroid mining is not only super cool in a space-roughneck sort of way, but could bring down something like tens of billions of dollars worth of iron/copper/gold/platinum/etc for every person on Earth6. Fourth, our increasing understanding of DNA (after all, it’s just information) and ability to manipulate it suggests that we may make many diseases into historical curiosities7. Fifth, as an extension of human DNA, mastering plant and animal genetics should allow things like cellular agriculture, growing food identical to meat, milk, and eggs with a tiny fraction of the land, water, and carbon footprint8. Probably (again, the book’s argument), all of these should be online in the next few decades.

So now for the third part of the argument. All these technologies open up significantly different opportunities for how we organize society, but none of them pre-determine a particular social destiny9. The kind of abundance that drives the marginal cost of producing something down to zero messes up standard capitalist incentives—as prices drop toward zero, why would anyone produce it? The standard response in the current system is to manufacture artificial scarcity via enclosure, monopoly, etc. In contrast, Bastani proposes a world where the abundance is fully realized, such that there’s no need to bother charging for most things, and supported by a pretty recognizable grab-bag of leftist policy proposals paving the way: more commons, credit unions, local currencies, decarbonization, taxes on financial speculation, nurturing of co-ops, redefining GDP, and just generally combining the Red and Green sides of leftist politics (saving the planet, in a way that makes everyone better off). Rather than advocating for Universal Basic Income (which rely on prices anyway), he advocates for the provision of Universal Basic Services, namely housing, transport, education, healthcare, and information10. Elections are the most practical way to make all this happen, and building electoral support is a practical way to build class consciousness for further collective action.

Basically, Karl Marx was right about everything, except he was too visionary, and the conditions for real communism didn’t exist until just now, so prior attempts don’t count11, and if we communism now it will totally work this time, trust me bro.

Bastani is aware of the tension: despite arguing from a framework where productive technology determines social organization, and treating these five waves as near-certain to arrive, he insists there is nothing inevitable about history or political systems; this manifesto and its recommended actions are needed to keep the benefits of these coming advances for everyone rather than a way to enrich the few. That disjoint and tension is the weakest point of the book, and the message seems to have exerted very little lasting change on the discourse other than a nudge toward Abundance more generally, which is merely a contested subtheme of the Left’s agenda at the moment.

Mapping to Andy Hines’s Imagining After Capitalism12, the book includes all three future images at once: the book is mostly laying out the evidence from the Tech-Led Abundance image, in order to argue the need for decisive mass action to enable the Non-Workers’ Paradise image, with the Circular Commons image somewhere between a secondary goal and an afterthought. It’s also very close to the Manoa Transformation archetype, and to Jim Dator’s view of the “full unemployment” future that AI could soon enable.

Scan Hit of the Week

Emxys is building out new capacity, and is trying to expand its optical communications technology (basically tightbeams) beyond Earth orbit to deep space. The technology is mostly the same, but the need for autonomy and reliability go up once you’re so far away from home. NASA proved out the concept by sending a cat video 101 light-seconds to Earth using a laser in 2023, but any kind of large-scale Solar System exploration will require this as one of the critical components. Maybe there’s a business model coming for relay stations? Maybe people will calculate trajectories and try to intercept them?

Share

2

Footage of me purchasing the book:

As far as I can tell, work has existed as long as humans have been around (though that’s not a universal position), but jobs are mostly a product of the Industrial Revolution. So often the main contours of conversations about the Future of Work are actually about the Future of Jobs, which is very different. This article from 2023 takes many of the ideas about jobs now and in the future and maps them onto the Three Horizons framework. The diagram he offers has a great list of the ideas currently in circulation that map to each horizon:

Subscribe now

Of course, many of the ideas about the future of work are covered in Andy Hines’s excellent Imagining After Capitalism. Specifically, the H3 ideas contain very clear pointers toward Hines’s images of Non-Workers’ Paradise, Circular Commons, and Tech-Led Abundance. A few ideas that are particularly hot at the moment:

• Universal Basic Income: the issue that launched the Yang Gang and also apparently a terrible animated series on Fox1. There have been many pilots of this idea in various locations, for various lengths of time and amounts of money, and the results are kind of disappointing for everyone: it looks like UBI only draws a few percent of its recipients out of the labor pool, and makes people feel better subjectively, but is a wildly expensive way to achieve that improvement, especially if it’s not a replacement for existing welfare programs2.

• Automation: this one is especially in focus with the huge AI improvements and announcements that drop each week (just today as of writing, OpenAI launched workspace agents for teams). Theoretically, anyone whose job is completely on a computer is at risk today, and with the improvements in robots the rest may be close behind. However, the reality of this is hard to gauge: the world has recently been sure that programmers would be the first to go, but now designers are in the hot seat and programming jobs are starting to tick back up. The new world is basically moving up the chain of abstraction to a world where declared intent is immediately turned into software, the ability to clearly specify intent, and to smell when software is built correctly, seem like the skills that will stay relevant the longest. It seems like even when companies lay people off in favor of AI, most regret it and end up rehiring.

• Gig work: this may be more of a mirage. We are told that employment is becoming less stable and more people are turning to gigs, and the number of people earning income via online platforms has exploded in the last decade+, but it seems like the total number of contingent workers has been flat or falling over that time period. So if gigs are the thing that replaces jobs, they have a long way to go.

Scan Hit of the Week

Val Kilmer will appear in a movie for which he was cast but died prior to shooting. This was all done with the cooperation of the Kilmer Family, so they shouldn’t have the same problems Star Wars had with Peter Cushing. See the trailer for yourself. This is coming into Hollywood at the same time Tilly Norwood is (cringily, in my estimation) trying to create a new category fully synthetic actors. What is real, especially in an industry where everything is already fake?

Share