“A simulated experience or a dream is simultaneous reality and fantasy”*…
In an excerpt from her new book, How to Design a Universe: The Science of Real and Virtual Worlds, Scotland’s Astronomer Royal, Catherine Heymans, considers AI, multiverses, and the prospect that we’re all living in a simulation…
In 1637, French philosopher René Descartes recorded a basic truth in his grandly entitled Discourse on the Method of Rightly Conducting the Reason, and Seeking Truth in the Sciences: Je pense, donc je suis.
“I think, therefore I am.” It was the one fact that he could not doubt, something he was so certain of that he made it his first principle of philosophy. You will no doubt be certain of this fact, too: you feel; you experience joy; you see, smell and taste the wild and wonderful technicolor world around you; you think, therefore you exist. But do I? Of course, like Descartes, I know my own fundamental truth that I exist, but I don’t know for sure that you exist. Our conscious experience of reality is wholly personal, completely subjective, and many philosophers argue that the only principle without doubt is “I think, therefore I am.”
Traditionally, consciousness is conflated with thinking, intelligence and behavior, which has led to heated debates over whether AI is, or will become, conscious. As a life-long vegetarian, I find it baffling that the academics and celebrities who will happily tuck into a medium-rare steak are also loudly voicing concern over the current lack of AI welfare rights. But I suppose if this new technology helps to shine a light on the importance of moral respect for all conscious beings then there is nevertheless merit in this discussion. The conclusion that you draw from all this handwringing will, however, completely hinge on your definition of consciousness.
One approach to define consciousness searches for synchronous neural activity across different parts of the brain that regularly correlates with a specific experience. An example of this was a 2024 study of over 200 patients with severe brain injuries that had left them physically unresponsive to all verbal commands. When asked to imagine playing tennis, however, one in four of these patients showed consistent brain activity in six separate observation sessions. Should we conclude that this study demonstrates that roughly a quarter of coma patients in a vegetative state are conscious? While the pattern of activity in the responsive brains is clearly correlated with an imagined action, it is unclear whether this interconnected neural behavior is also the cause of a first-person conscious experience of that event. Correlation and causation are different things, and the brain scan could be detecting no more than a neurobiological reaction of electrical signals flowing through a neural path that was hardwired before the patient fell into a coma. This would be akin to ChatGPT5’s trained neural network repeatedly responding to multiple global users that the word “blueberry” is bouncy.
Let me now introduce you to the concept of a philosophical zombie, the thought-experiment brainchild of present-day Australian philosopher David Chalmers. From the outside, these zombies are indistinguishable from normal human beings. When they wake up in the morning they rub their eyes, yawn and sniff the air. They mumble thank goodness for coffee when they detect the scent of a freshly brewing pot. But on the inside there is nothing, no inner stream of introspection, no sense of smell or that warm loving feeling of gratitude towards a partner who has freshly ground the coffee beans for the cafetière. With a perfect replication of ordinary everyday human behavior and mannerisms, these zombies outwardly exhibit the expected response to changes in their environment. The only difference from me (and probably you) is that these hypothetical creatures are not consciously experiencing this reality.
Chalmers introduced the idea of a philosophical zombie to illustrate what he labelled as the now-infamous “hard problem of consciousness.” Our outwardly intelligent zombie has all the biological brain wiring that is required to play sport. They would therefore pass the brain activity test of the imagined tennis match and be labelled “conscious” even though they would never feel the joy and exhilaration of winning a game by landing four aces in a row. The “hard problem” is that it is impossible to design an experiment to directly test the subjective experience of consciousness in any living creature aside from ourselves. All we can be sure of is Je pense, donc je suis.
The “hard problem” means that we will never directly know whether AI, with all its future advances and outward intelligence, will remain a philosophical zombie or learn to become conscious. Will my e-me ever be able to cross experience human emotions off her bucket list? It certainly feels like she’s enjoying taunting me about this specific question: Am I feeling or faking—ha you’re so annoyed that you’ll never know.
In 2003 the philosopher Nick Bostrom published an academic article entitled “Are we living in a computer simulation?” His “simulation hypothesis” argues that one of these statements is true:
– Humans will become extinct before they develop the ability to simulate
– Futuristic civilizations will have little interest in creating simulations of what life was like in the past.
– We are almost certainly living in a computer
Let’s look at them one by one. Given the glacially slow progress that is being made to address the indisputable scientific evidence that our continued use of fossil fuels is raising global temperatures, exacerbating extreme weather events and irreversibly damaging ecosystems and biodiversity, my bet is that our Earth-based human civilization will become extinct long before we simulate sentience. That said, I feel fairly confident that we are not the only intelligent life that has evolved within our Universe. Our life-giving Sun is but one star among over 200 billion stars that swirl around our Milky Way galaxy. And our galaxy is but one galaxy among over 200 billion galaxies that shine their light at us across the Universe. And that’s only counting galaxies inside the part of the Universe that we can see. If our theories about how the Universe was created are correct, it is infinite, hosting an infinite number of galaxies, an infinite number of stars and an infinite number of planets.
With so many opportunities for life to develop across the expansive Universe, there is hope that enlightened alien societies will have heeded their scientists’ warnings early enough to halt the self-imposed destruction of their home planet. Bostrom’s first option—that extinction will always prevent the development of artificial sentience—seems rather improbable if we consider the entirety of the infinite Universe.
The second option, that civilizations with the ability to simulate “life” will only focus on their present and future, seems rather unlikely to me as well. We’ve already pictured how a futuristic neuroscientist might test theories of consciousness by simulating virtual humans using a bubble of full “quark-up” physics surrounded by an AI rendition built from digital memories of the past. If that is possible, we can also imagine future anthropologists creating numerous “what if” simulations of their history, tweaking the settings by including different AI-NPC “minds” to determine their influence on each fully simulated human. In doing so they could use the computational scientific method to learn how to improve societal outcomes, providing conclusive answers to a range of debates on nature versus nurture.
By deciding to dismiss the first two options from Bostrom’s simulation hypothesis, we’re unfortunately left with: We are almost certainly living in a computer simulation. In 2016 Elon Musk blithely quantified what “almost certain” meant when he stated there is a one in billions chance that this is base reality: a one-in-billions chance that you are living in the real world that will, at some point in the future, develop the technology and scientific know-how to simulate perfect virtual worlds. Bostrom and Musk argue that once simulators develop this skill, they won’t create just one virtual world, but hundreds of thousands, with billions of realizations played like computer games, over and over again throughout cosmic spacetime.
In this scenario the total number of virtual humans in the Universe would dramatically outnumber the real-world humans in the original “base reality” where the computational overlords live. This creates a vanishingly small probability that a random member of the Universe’s population, for instance you, physically exists. Philosopher David Chalmers tells us not to despair though. He views a virtual reality where you feel free will, love and laughter, as genuine a reality as any real world. I agree with him: does it really matter whether this is all real, so long as I feel real? But if that “technophilosophy” isn’t enough to quell your existential crisis, happily there is one key piece of evidence that Bostrom and Musk did not consider: Je pense, donc je suis. In my opinion, “I think, therefore I am” dramatically improves your chances of physical existence.
Imagine a neuroscientist from a futuristic world is standing at the front of a large auditorium that is packed with esteemed scientists and philosophers from the across the solar system. She’s been invited to present a groundbreaking new result on consciousness. Using all the data from the digital era of a now long-gone planet that was once called Earth, she describes how she trained AI to generate a society of Earth-like philosophical zombies. Once her AI illusion of reality was perfected, she electronically conceived a full-physics “quark-up” rendition of a human child and watched her grow and develop in this virtual realm. Over years she has run 80 subtly different realizations of the same virtual world and child, subjecting her centerpiece to a range of experiments. Today she is sharing the most conclusive evidence to date that conscious experience and emotion can evolve from subatomic physics: her centerpiece cries when she falls and dances when she hears music even though her zombie society has been engineered to never react in these ways. The audience rises in a standing ovation as the ethics committee start to nervously worry about the welfare rights of each simulated centerpiece.
Planet Earth is home to more than eight billion living humans, and even though there is little agreement on what underpins consciousness in this global population, all theories predict that if we’re made the same, and one of us is conscious, then all of us are conscious. In the scenario of our futuristic seminar, it’s these real-world inhabitants of planet Earth who have digitally documented their lives to create a bank of training data for the unconscious AI-NPCs in the simulation. As our neuroscientist is convinced that all her simulated centerpiece humans are conscious, Je pense, donc je suis would mean you are either one of her 80 virtual centerpieces or you are one of the eight billion humans living in the real world. This translates into a highly unlikely 100-million-to-one chance that you’re living in our imagined neuroscientist’s simulation. That’s pretty similar to your chances of winning the lottery jackpot this weekend: it’s not impossible that you’re a simulated human, but it’s highly improbable.
So, why is the conclusion from our story so very different from Bostrom and Musk’s “almost certain” statement that we’re living in a simulation? Let’s critique the key difference: why would our neuroscientist only create 80 realizations of Earth-like living? Why would her futuristic society not use this innovative technology to also create simulations for entertainment? If conscious humans were also simulated in a video game that was played over and over on repeat, the chances that you’re living inside that game would rise dramatically. Well, I believe free will partly answers that question. If our lives really were at the whim of a gamer in the “base reality,” surely we’d be spending a lot more time on motorbikes, driving at full speed off cliffs to provide epic “game fail” fun as we crash into the valley below. Atheist physicists like to question why a God would go to all the effort of designing the laws of physics to create a self-sustaining universe and then sit back to watch it boringly evolve over 13.8 billion years without any intervention. Game entertainment always comes from being in control of the fun, not passively sitting back to watch the subatomic mathematics of your virtual reality unfold. The fact that you feel free will makes it much less likely that the gamers are in charge.
If we accept the argument that the driving force behind the creation of conscious virtual humans will be purely scientific, and not for entertainment, then we can set a limit on the number of simulations that will be made. At some point there will be nothing more for the scientists to learn from them. Of course I cannot predict what that number will be, and I only chose 80 in our story to nicely divide with the eight billion inhabitants on Earth, but we can make an educated guess by looking at examples in scientific practice today. There are only 37 variants of the astrophysical simulations that I use in my own research. Having been careful and strategic about what varies between the simulations, this small number of realizations has allowed us to confront a complete set of questions on dark matter, dark energy, black holes and neutrinos. We’re currently in the process of running a few more to test some alternative gravity theories, but at the end of the day our total number of simulations will still only be in the hundreds, not the billions.
My decision to limit our story to just one futuristic world of computational overlords also improves our chances of being real. In the wide and expansive Universe, simulations of conscious beings could well be independently popping up everywhere, but the likelihood that more than one galactic civilization would choose to use real-world Earth data to electronically replicate us is arguably low. It’s therefore reasonable to limit the origin of any virtual reality that we might live in to a single solar system.
The last plotline that we should examine is the proposal that there will only ever be one person simulated at full “quark-up” resolution. My kids are particularly offended by my suggestion that if I’m living in a simulation, they’re most likely AI-NPCs. Middle-child skillfully argues that as all three siblings were once developing deep within my full-physics bubble, any virtual reality that we live in must be very advanced, with at least four “quark-up” individuals in our simulation. It was only their NPC father who left the dinner table feeling deeply affronted that evening!
If it’s possible to simulate one person, then of course it’s possible to simulate many. But I’m convinced that the total simulated population would be limited to some degree. When it comes to scientific experimentation, we always look for the most efficient strategy. The simulations of the Universe that I work with recreate less than one-thousandth of the total volume of the real observable Universe. We find this to be more than enough to address our current scientific questions. You could well imagine that a futuristic anthropologist would want to electronically explore societal behaviors and human development in different cultures, comparing cities to rural locations, contrasting nutritional and junk food diets. The list of ideas would be extensive, but it’s nevertheless hard to imagine a science case that would require the entire population of planet Earth to be simulated at full resolution. I should be crystal clear that the “100-million-to-one chance” that you’re simulated only applies to our fictional neuroscientist story, so please do not quote this as a scientific fact. No one has enough information to be able to quote a number like this. My hope is that our story-based discussion has, however, left you with this understanding: if the computational scientific method has been adopted across the entire Universe, while it’s conceivable that you are living in a simulation, the chances of this being the case are low.
But what if the galactic overlords harness an almost infinite resource of energy? Child number three isn’t ready to give up on the idea that we’re living in a simulation just yet. If you really wanted to dismiss the argument that the energy required to simulate the entire Universe renders the task impossible to complete, we could imagine a future society with a major god complex that decides to run a full “quark-up” simulation for its entire Universe. The Bostrom argument that we’re almost certainly living in a virtual reality critically depends on the inevitability that humans will generate realistic simulations if given enough time.
As these simulations are replicas of the real world, that inevitability must also apply to the newly created virtual human civilizations. They will then use their hypothetically large resource of energy to create their own simulation. Down the rabbit hole we will go, a virtual world created within a virtual world created within a virtual world. The fact that we don’t currently have the ability to simulate our reality tells you one of two things. Option 1: we’re living in the real-world “base reality” from which future pioneers will develop the necessary design for a virtual universe. Option 2: we’re in the very last virtual reality within that simulated chain, just taking our time to evolve the technological know-how to design a new simulation that will take us another step down the virtual rabbit hole. These two options mean that you have a 50/50 chance of physically existing or virtually existing. Number Three tosses a coin while I encourage him to embrace the idea that this knowledge is actually irrelevant because if we feel real, we are real, whether we’re made of quarks or qubits…
“Are We Living In a Simulation? And Other Questions Concerning Our Role in the Universe,” from @astroroyalscot.bsky.social via @literaryhub.bsky.social.
The full text of Bostrom’s paper is here. See also: “Simulation is the situation created by any system of signs when it becomes sophisticated enough, autonomous enough, to abolish its own referent and to replace it with itself.”
* Mamoru Oshii, Ghost in the Shell
###
As we tally the turtles all the way down, we might recall that it was on this date in 1938 that a young chemist at Sandoz, Albert Hofmann, while researching the medicinal plant squill and the fungus ergot in a search for compounds useful in pharmaceuticals, first synthesized lysergic acid diethylamide (LSD). As it wasn’t immediately promising, he put it aside. But he revisited his formulation several years later, on April 16, 1943; handling it, he accidentally absorbed a bit through his fingertips and realized that the compound had psychoactive effects. Three days later, on April 19, 1943 (a day now known as “Bicycle Day”) Hofmann intentionally ingested 250 micrograms of LSD, then rode home on a bike– a journey that became, pun intended, the first intentional acid trip. (This is not to be confused with the UN’s World Bicycle Day.)
Hofmann was also the first person to isolate, synthesize, and name the principal psychedelic mushroom compounds psilocybin and psilocin.
“It is common sense to take a method and try it; if it fails, admit it frankly and try another. But above all, try something”*…
Joe Mathews recounts experiments with the rule of law around the world that are creating the kind of resiliency we need in our era of uncertainty…
Shortly after wandering into the Munich Urban Colab this summer, I found myself trying to follow an intense conversation in German about the city’s new experiment with autonomous buses. A local official, clocking my curious puzzlement, offered an aside in English: “We are asking ourselves: ‘What does it mean when the law is an experiment too?’”
No question may be more essential to governance in our era of dangerous uncertainty.
Global warming is accelerating faster than global treaties on mitigation can be negotiated. Climate catastrophes occur with a sudden ferocity beyond what current adaptation laws anticipate. Biological threats surface and kill before public health responses can be formulated. And AI is evolving faster and more wildly than regulators or lawmakers, or even its designers, expect.
All this dizzying unpredictability is undermining traditional legal structures. The threat is not just to the rule of law, but to the very nature of law, at least as humans have conceived it.
Our societies have long considered the law a slow and steady force, “the essential foundation of stability and order” in the words of J. William Fulbright. To that end, law is backward-looking — “the government of the living by the dead,” according to Oliver Wendell Holmes. The lawmaking process in republics and democracies is age-old: observe a problem, study its particularities, debate solutions and then carve out a rule that will endure for years or decades.
But what happens when the future arrives faster than your regional regulator or local parliament can make a rule?
The emerging answer is experimental law or, in Germany, where Munich is a hotbed, Experimentierklauseln. “Experimental clauses” refer to sections and parts of laws that are made flexible, often for set time periods, to allow for real-time adaptation or alteration in response to changing threats, climate or technology.
The shift toward such experimentation has begun, albeit slowly, in only certain areas of the law. Contract law and administrative law, especially where they touch on technology or other advanced sectors of the economy, are becoming more fluid…
Read on for a collection of examples and a consideration of the issues at play…
“Turbulent Times Call For Experimental Law,” from @californiajoe.bsky.social in @noemamag.com.
As George Santayana said: “You find gold, so to speak, effortlessly, in rational… trial-and-error-based research. It is exactly like options, trial and error, not getting stuck, bifurcating when necessary but keeping a sense of broad freedom and opportunism. Trial and error is freedom.”
The key, of course, is to understand that the learning– the experimentation– never ends…
* Anthony Burgess (who did say it… though the quote is also– and probably more appropriately– attributable to [then NY Governor] Franklin D. Roosevelt, who spoke that line in a speech in 1932, when Anthony Burgess was 15. Thanks to JPH for the catch.)
###
As we learn our way into the future, we might honor the impulse to move on from unsuccessful experiments (and outdated successes) by observing, as one does on this date each year, National Clean Out Your Refrigerator Day (AKA “National Clean Out Your Fridge Day”).
“A chair is a very difficult object. A skyscraper is almost easier.”*…
What happens when one combines practical carpentry skills with the the abundant theoretical knowledge of AI to build a chair? James Bridle recounts his experience…
In 1840, a surveyor took a chisel to one of the outer walls of St. John’s Church, on Old Haymarket in Liverpool, England. They carved a horizontal line into the stone, a couple of inches long, and below this, three radiating lines, getting wider at the base: a broad arrow. The broad arrow had been used in England since the fourteenth century, and later across Britain and its colonies, to mark government property, but the horizontal line was a relatively new addition. Into this crevice the surveyor would have placed an angle iron, or “bench,” used to support a leveling rod, from which elevations could be taken. Thus, the broad arrow and the line were known as a “benchmark.”
The benchmark on St. John’s Church, recorded in the records of the government’s Ordnance Survey, its official mapping agency, marked one hundred feet above sea level. Chiseling it into the stone meant that someone else could come along later and set up a leveling rod in exactly the same position and take their own reading, transforming it into a datum line: a reference point for further, and repeatable, surveying.
For a few years, the benchmark on St. John’s was the reference datum for the whole country, until it was realized that more accurate measurements should begin at the mean tide level. A pole was set up in the Victoria Docks, also in Liverpool, and the fluctuations of the tide observed for nine days. The resulting mark was used as the fundamental reference until the 1920s, when it was shifted again, to Newlyn, in Cornwall. Benchmarks continued to be carved into churches, civic buildings, milestones, and country walls across the country, all referencing the tidal datum at Newlyn.
“Sea level” is actually different everywhere, but in Britain, it remains a function, plus or minus, of ODN: Ordnance Datum Newlyn. This is the level measured within the town’s tidal observatory, constructed on the South Pier in 1915. The observatory takes the form of a hollow tower in which several floats are suspended, connected to a recording device. An inlet pipe allows the tower’s well to fill and empty with the Atlantic tides. Between 1915 and 2015 the mean sea level at Newlyn increased by 1.83 mm/year—and by 3.8 mm/year between 1993 and 2014. Each change in sea level is reflected in an incremental adjustment to elevations across the land.
The last official benchmark was carved on a milestone located outside a pub in Loughton, Essex, in 1993, before satellite-based systems took over the datum (today, a GPS antenna sprouts from the top of the tide tower in Newlyn). The Ordnance Survey still maintains a record of over half a million benchmarks, although many more have disappeared. Hiking clubs lead expeditions to find them, and hobbyist benchmarkers check them off their lists. When you start looking for them, you see them everywhere.
“Benchmarking” remains in use today as a term for establishing a baseline against which other data can be measured and judged: software performance, product quality, profitability, growth; quality, time, cost. But it’s worth remembering its origins not as a tool for establishing hierarchies, but for leveling; for connecting, rather than separating; and for understanding the whole earth, and everything that happens upon it, as a continuous, indivisible field, stretched across cities, forests, and mountains, and ever susceptible to the rhythm of the tides.A couple of years ago, I found myself stuck in an apartment in Athens, and in desperate need of distraction. I’d been living with my family on a Greek island, where we had space, and time, and had started to build a house; but at the end of 2023 my partner, Navine, became severely ill, and we had to return to the city we’d lived in for five years previously. She spent nine months in hospital; I visited every day, while also looking after our three-year-old son between nursery school and sleep and school again. I had less space, and a lot more stress, but I also had time to fill, and was itching to use my nascent carpentry skills. I needed to make something, with the limited tools and materials I had to hand—anything would do, it just needed to keep my body, and my mind, busy.
It was around this time that the current wave of what we’re calling artificial intelligence started to excite, and demand, attention. I actually studied AI at university, twenty years ago: exactly the wrong time, it turned out, to be studying it. The history of AI is one of booms and busts: “AI winters,” they were called, when the field seemed to sputter out after a series of advancements, and twenty years ago it was decidedly chilly. We learned about expert systems and neural networks from aging professors, but all the sensible students were learning about classical logic and statistics as preparation for careers in digital finance. It was another decade before the field would regain its energy, and—this is a controversial take—it would have little to do with conceptual breakthroughs. Rather, the growth of the internet, with its accompanying boom in processing capacity and data storage, would allow those same ideas—primarily neural networks—to be supercharged by graphics cards, digitized text and artworks, surveillance and social media, and the wholesale transference of everyday life to the online world. The AI “revolution,” if that’s what it is, is not a technological breakthrough, but the overwhelming accumulation of power and data in the hands of a few companies: power and data which rightfully belong to, and have been stolen from, everyone. But I digress.
I’d been noodling with this idea for some time: What if we used AI for something actually useful? It feels like we’re applying this new tool to all the wrong things: beating us at the games we enjoy, supercharging predatory capitalism, extracting and burning fossil fuels, depleting the planet’s resources, conflating images with art, encouraging extremism, suppressing rights, hallucinating, and so on and so forth. Wouldn’t it be infinitely more interesting if AI helped us do useful things in new and better ways, ways which we could not have imagined alone? Intelligences which we engaged with more like colleagues and collaborators than slaves and masters.
The one thing computers have always been really good at, one of the few qualities which genuinely impress me, is optimization. They can take a human-designed thing, an object or an algorithm, and ruthlessly strip it to the bare essentials while maintaining its functionality. Give a clever machine some widget, like a structural bracket or a complex workflow, and it will produce some alien-looking thing, a third of the weight and twice as efficient, which a human mind would not have conceived. That’s cool, that’s interesting. Of course, applied to culture, society, and the wider environment, such an idea of efficiency is a disaster, but it’s pretty neat for objects, as long as a human is keeping an eye on things.
I thought I might be able to make use of this particular skill to create something with what I had to hand: By combining my rudimentary carpentry abilities with the machine’s knack for efficiency and strangeness, something interesting might happen. Specifically, I didn’t want to design a process which would ultimately replace me, or take over the rewarding bits of work, but something more like a smart carpenter’s mate, that could do some of the sums and make useful suggestions and leave me to do the fiddling and the sanding. A collaborator, not a boss…[Bridle recounts the first iteration of his collaboration of his collaboration with ChatGPT…]
… There are two qualities of the chair that ChatGPT and I built that stand out. The first is that it looks flimsier than it is, but it works. The AI told me to use a sheet of very thin ply for the seat, so thin that I considered laminating two sheets of it together. It gives a little when you sit on it—but it holds. That’s exactly the kind of optimization I was looking for: a decision I wouldn’t have had the confidence to make, but, again, it works. It might even be slightly comfier that way—although I have no way of knowing if the machine actually intended it, structurally or comfort-wise; it might have been going on dimensions alone. Still, the chair stands.
The second quality of the chair is that it is unnervingly big. Not too big—it’s actually pretty comfortable for a six-footer like me—but it’s definitely bigger than you’d expect, verging on the uncanny. This I find to be absolutely fascinating.
What we have here is, quite literally, a chair designed by someone who has read a lot about chairs but never actually sat in one. They know what a chair should look like, what it should do, the mechanics of how one is put together, the physics of holding you up. But they have no embodied experience of the world, no history with chairs or any other object, no physical memory. They don’t know how it feels, not merely to sit in a chair, but to encounter one in the world, to place it among other things in a home, to live with it, to be around it every day as one goes about the mundane business of life. This is the idea of a chair brought to life; but things—objects, beings, relationships—are not mere ideas, and you certainly can’t get every aspect of them from books. The chair is low stakes, but, once again, you wouldn’t want to run a society, or a culture, or a planet this way.
Since building that first chair, others have been created. This is where the benchmarking bit comes in. A few weeks after the ChatGPT chair, I made another one, this time with an AI model called Mistral. Mistral is a French company which made a lot of claims to be more ethical than ChatGPT (whatever that really means), and I’d been told it was more accurate and precise when it came to technical matters. Once again, I described my request, and we went back and forth a bit, and then I tried to do what it told me.
Mistral’s chair didn’t stand up at all without some serious modifications: I had to add a bunch of cleats and bracing around the legs to make it stable enough to sit on, which eliminated any purported gains in efficiency. It was definitely more interesting than ChatGPT’s doll’s-house effort, with a short back and slanted armrests. It was also waaaayyy too high. My feet didn’t even touch the floor. So much for Mistral’s accuracy and precision—but points for making something a daring European chair designer might actually have come up with.
AI is unfortunately very hot right now. At a time when we could be attending to decarbonizing our societies and building more robust communities to help us through the turbulent years ahead, we seem to be giving most of our attention to a technology which not only burns through energy and other resources, but actively works to make our relationships less secure and our information environment less coherent. The art world is not immune to this tendency, and so when I received requests for AI-themed work for events and exhibitions, I turned the chair-making process into a set of instructions which anyone could follow, and sent it over.
The first iteration took place in Tokyo at an exhibition called “AI is Not Magic” in June of 2025. The curators followed my instructions, inviting Anthropic’s Claude AI to make them a chair from a selection of materials in the studio. It turned out a pretty nice chair, but the quality of the finish made me suspect that an experienced carpenter was involved at their end. Conversely, when a conference called ThingsCon in Amsterdam asked Elon Musk’s Grok AI to have a go, the system told them to build a lounge chair and gave very precise cutting instructions, but none of it fit together at all. As the ThingsCon team noted, “The system could produce—confidently, fluently, at length. But it couldn’t understand what it had made.” The chair they eventually ended up with, after much modification, was “grammatically coherent and practically useless. […] It’s a strange object—part furniture, part evidence.”
Each chair is indeed a piece of evidence—a benchmark, in the truest sense. Each iteration of the work produces a different outcome, which we can judge and compare. Each AI model reveals its own tendencies, biases, and limits, in physical form, easily appraised by a layperson. We are benchmarking; we are benchmaking; we are literally making a bench (fine, a chair, whatever). I appreciate processes where you learn to do a thing, a thing that you didn’t know how to do before, and at the end of that process you have not only done it, but you can see the results. And, in this case, sit on them.
This benchmarking has a deeper meaning as well. Carving a broad arrow into a church wall doesn’t reduce or isolate anything. Rather, it reminds us that every action is made in context; each benchmark refers to every other one, in a continuous, connected line through space and time. When we build, when we think, when we act, we do so in honor of every object, thought, and action which has preceded this one, and we are changed by the process. AI wants us to forget this, by making every output its own miraculous creation. But we owe the world more than that…[Bridle reflects on the experience…]
… Building the AI chairs taught me something about AI, but it didn’t teach me much about making chairs. Ultimately, I was the teacher, taking what I’d learned elsewhere to make the “practically useless” suggestions into something stable and self-supporting. The AI undoubtedly learned more than me from these interactions: how to speak, what interests (some) humans, how to flatter, how to slant and tailor its responses to my demands.
The anarchist anthropologist David Graeber, may he rest in peace, had a good bit on authority. He said that anarchists don’t oppose authority per se, but they study and critique it, and appraise its legitimacy. What he liked, he said, was self-subverting authority, authority which undermines its own basis. He gave the example of a teacher: A teacher knows something useful to their student, and so they have a natural authority; but if they do their job well, at the end of the process the student knows what the teacher knows, and their authority is spent. The hierarchy is flattened, and no further unequal power can accrue. The student is changed, and so is the world.
Of course it’s possible to learn things from AI, but, for me, real intelligence resides not in an ability to extract and dominate, but in the capacity to share, co-exist, and thrive together. There are many ways of being intelligent, as many as there are living beings, each one unique and valuable. The real challenge of AI is not in building ever more powerful systems—this is difficult, but inevitable—the challenge is to find what relationships between different intelligences benefit us all. The challenge is to choose the kind of world we all want to inhabit. Benchmarking—the practical, embodied work of exploration and mediation—is an attempt to find our way towards such a world…
On finding a healthy relationship with AI: “Benchma(r)king” from @jamesbridle.bsky.social in @emergencemagazine.bsky.social.
###
As we measure twice before cutting, we might recall that it was on this date in 1985 that the first discovery of a fullerene was published in the journal Nature. In September of 1985, the American chemists Robert F. Curl, Jr. and Richard E. Smalley, colleagues at Rice, met with Sir Harold W. Kroto of the University of Sussex, England. In 11 days of research, they discovered the first fullerene, C60, a spherical cluster of carbon atoms. The discovery of this new structure (AKA “buckminsterfullerene” or “buckyball”) opened a new branch of chemistry, and all three men were awarded the 1996 Nobel Prize for Chemistry for their work.
“Those who first invented and then named the constellations were storytellers. What it changed was the way people read the night sky.”*…
Scheduling Note: (Roughly) Daily will be off tomorrow, but should return on Saturday…
Astronomers, photographers, and star gazers in general have been living with (that’s to say, looking for out-of-the-way locations in order to avoid) light pollution for over a century. But of late, they’ve been grappling with a newer challenge: bright satellites that stand, like grafitti, between observers — no matter how remote — and the treasures of the night sky.
Photographer Brendan Larsen created a striking demonstration, as Kottke.org reports..
To create this composite image [and as above], photographer Brendan Larsen captured almost 600 photographs over a two-hour period and then manually stitched them together.
This is the most time-consuming photo edit I’ve EVER done!
I left my camera photographing the night sky over Mt Taranaki [see here] for nearly TWO HOURS, capturing 578 photographs.
The result? An estimated 14,450 satellite trail appearances, representing approximately 1,400 individual satellites! 😱
I manually masked in EVERY SINGLE TRAIL from the original photographs — approximately 45 HOURS of editing to create this one image!
The Guardian has more on how he created the image (which, despite many online commenters’ beliefs, is not AI):
Larsen, 51, says he set his camera up to take an unlimited number of eight-second-long shots, with a two-second gap in between each photo. “I left her running for two hours and captured 578 photos … That was the easy part.”
The hard part came next. Larsen spent about 45 hours layering the individual photos – each of which contained about 25 satellites – over one another, using Adobe Photoshop. Much of the time was spent painstakingly erasing stars, to reveal the satellites.
“I could start to see the pattern emerging and realised: ‘Wow, this looks really awesome.’ I knew if I carried on it would look even better and more impressive in terms of satellite numbers.” [source article]
Fellow photographer Vincent Laforet called Larsen’s image “an important photograph”:
I was initially startled. Then confused. Then fascinated all in quick succession.
And then I felt a deep sense of betrayal once I realized that what I was tracking with my eyes were a series of small satellites. And btw: that they likely had ruined my long exposure that I was capturing at the time.
I remember asking in the moment:
Who gave anyone permission do this?
What’s between us and the stars: “Stunning Image of Our Satellite-Choked Night Skies,” from @kottke.org.
###
As we search the skies, we might send far-sighted birthday greetings to Audouin Dollfus; he was born on this date in 1924. An astronomer and aeronaut, he was the successor to Bernard Lyot as the principal French authority on the solar system; Dollfus discovered Janus, a moon of Saturn.
While Dollfus made use of terrestrial telescopes, he was creative in finding a way past the challenges of light pollution and closer to the objects of his study. With his father (the aeronautical pioneer Charles Dollfus), he holds several world records in ballooning, including the first stratospheric flight in France. He was the first to carry out astronomical observations from a stratospheric balloon, in particular to study Mars in a detailed way.
The asteroid 2451 Dollfus is named in his honor, and one of the largest craters on Mars was also named for him.












You must be logged in to post a comment.