Category: AI

Categories
AI Consulting

The Judgment Layer

An analyst’s note about the CEO of one of the largest consulting companies making comments at an investor conference includes a line that deserves more attention than it got: “token volume used on a project isn’t a proxy for AI maturity.”

Translation — clients are burning money on frontier models for problems that don’t need frontier models, and they’re not getting the outcomes they expected.

This firm’s CEO offered this as a business opportunity. I read it as a confession.

The old consulting model was simple: client has a technology problem, firm deploys humans to solve it. Billing followed effort. The new problem is different in kind — clients have an AI strategy problem. They know they’re supposed to be using AI. They’ve heard the word “frontier.” They’re spending accordingly. They just don’t know why, and the outcomes are showing it.

So the CEO is right that there’s an opportunity here. The value proposition shifts from implementation to judgment — not deploying AI, but knowing when not to deploy the expensive one. Matching capability to problem. Being trusted enough to tell a client that their $50M frontier model contract is solving a $500K problem.

Here’s the irony that the comment skates past: that advice is structurally difficult for a large consultancy to give.

The business model that built consulting firms was billing for doing. The more you deploy, the more you bill. Helping a client spend less, or choose the cheaper model, or run a narrower project, is genuinely good advice that the incentive structure actively works against. You don’t grow a $70 billion professional services firm by talking clients out of scope.

The judgment layer, if it becomes the real value, requires something closer to a doctor’s relationship with a patient than a contractor’s relationship with a client. Doctors get paid whether they prescribe or not. The value of the visit is the diagnosis — including the diagnosis that says you don’t need the expensive intervention. Consultants, historically, get paid to prescribe, and paid more when the prescription is larger.

There’s a reason we trust doctors with that asymmetry and not contractors. Licensing, malpractice, professional norms built over centuries — all of it exists to align the incentive. Consulting has none of that infrastructure. What it has instead is reputation, which is slower-acting and easier to game.

Whether the large firms can actually make the shift — rather than just reframe the same billable-hours model in the language of AI optimization — is the real question the market is wrestling with. The CEO’s comment is genuinely perceptive about where client value lies. It’s less clear that consulting firms are currently built to capture it honestly.

Categories
AI History

The Arrival

Yoshua Bengio spent forty years building the foundation of modern artificial intelligence. He won the Turing Award for it. And he didn’t think he’d live to see it work.

That’s the quiet fact buried inside Stephen Witt’s New Yorker profile of him. Bengio — one of the three researchers whose decades-long bet on neural networks eventually became the architecture underlying every large language model running today — had made peace with the idea that the thing he was building was a multi-generational project. Something for his successors to finish. Then Witt writes: “one day in late 2022, the technology had simply arrived. He compared it to meeting an extraterrestrial.”

Hemingway once described bankruptcy happening two ways: gradually, then suddenly. He meant ruin. Bengio experienced something harder to name — not ruin but arrival, which carries its own vertigo. The gradually was four decades of work that most of his peers considered quixotic. The suddenly was a Tuesday in November when a chat interface went live and the world quietly changed.

What unsettles me about the extraterrestrial comparison isn’t the strangeness it implies. It’s the distance. You meet an alien; you don’t meet something you made. The metaphor suggests that even its creator couldn’t fully recognize it — that the thing, once arrived, belonged to a category that exceeded its own origins.

We don’t have good language for this. Breakthrough, inflection point, paradigm shift — these are words people reach for after the fact, when they’re building timelines. What Bengio seems to be describing is the experience of standing in front of a threshold you spent your life approaching, and finding it already behind you.

The technology didn’t ask permission. It didn’t announce itself.

It arrived.

Categories
AI Technology

The Bathwater Problem

Gary Kamiya was writing about the Tenderloin when he said it, but the line has been following me around: “The problem is that by saving the baby, you also save the bathwater.”

The pattern is remarkably consistent across every major information technology. Each one arrives promising to liberate the deserving — the faithful, the learned, the civic-minded — and each one immediately, inevitably, arms everyone else too. Gutenberg’s press was understood by its champions as a device for spreading the true Word; within decades it was the primary infrastructure for Protestant schism, Catholic counter-propaganda, astrological almanacs, and pornography. The reformers got their Bible. They also got their pamphlet wars.

The telegraph was greeted as a force for peace — shared information would make war irrational, commerce would bind nations. It also became the nervous system of commodity speculation, financial manipulation, and the first truly industrial-scale news hoaxes. The telephone: connection and the crank call, the crisis line and the threatening voice in the dark. Radio: FDR’s fireside chats and Father Coughlin. Television: Murrow taking down McCarthy, and also fifty years of manufactured consent. The internet: the largest library ever assembled and the largest sewer.

The pattern isn’t coincidental. It’s structural. Each technology expands what’s possible for human expression and coordination — and human expression and coordination contain both the noblest and the worst of us in roughly fixed proportion. The tool doesn’t change the ratio. It scales both sides of it.

What’s interesting historically is how each generation believes their technology will be different — that this time the architecture can be designed to select for the good. The internet era produced the most elaborate version of this belief: algorithmic curation would surface truth, network effects would reward quality, the wisdom of crowds would outcompete misinformation. Instead it turned out that engagement was the attractor, and outrage was the highest-engagement content. The bath got hotter.

The AI moment is the same belief system, restated with more technical sophistication. But the Kamiya line stands. You are saving a baby, and you are saving bathwater, and no one has yet designed a tub that can tell the difference.

The question isn’t whether the bathwater comes with the baby. It always does. The question is whether you turn on the tap.

Categories
AI AI: Transformers Books

The Updating Machine

Tom Chivers puts Bayes’ theorem in plain English and it sounds almost obvious: “the probability of event A, given event B, equals the probability of B given A, times the probability of A on its own, divided by the probability of B on its own.” A formula for revising what you believe when new evidence arrives. You started somewhere. Something changed. Now you believe something slightly different. Repeat.

The obvious part is the mechanics. The hard part is the loop.

Most reasoning errors I catch in myself aren’t failures of logic — they’re failures to update. I hold a view, evidence accumulates against it, and I find reasons the evidence is flawed rather than reasons the view might be.

Psychologists have a name for this: confirmation bias. But I’ve always found that label a bit too clean, like it describes a bug rather than a feature.

The prior isn’t wrong to be sticky. It represents everything you’ve learned up to this point. The problem is when it becomes load-bearing — when the prior stops being a starting position and starts being a conclusion.

“Strong opinions, loosely held” is supposed to solve this. It’s a useful phrase — it captures something true about the right posture toward your own beliefs. But in practice the second half is harder to honor than it sounds. The strong opinion gets stated, new evidence arrives, and changing your mind in public feels like losing. The “loosely held” part quietly becomes decorative.

What Bayes actually demands is something closer to epistemic humility with arithmetic attached. You don’t get to say I don’t know. You have to say I estimate 0.4, and here is what would move me to 0.6. That’s harder. It requires you to specify not just what you believe but how you’d know if you were wrong.

This is why Bayesian thinking keeps surfacing in AI conversations. Modern language models do something structurally adjacent to this — not consciously, but mechanically. Every token generated is a probability distribution revised forward by context. The model doesn’t know the next word; it updates a prior over all possible words, given everything that came before. It’s not reasoning the way humans reason, but it’s updating the way Bayes updates: continuously, contextually, without the luxury of certainty.

Whether that’s comforting or unsettling probably depends on your own prior.

The deeper thing Chivers is pointing at, I think, is that Bayesian reasoning is essentially a description of intellectual honesty as a process rather than a trait. You can’t just decide to be open-minded. You have to build the loop: form a belief, assign it a probability, watch for evidence that should move it, and then actually move it. Most of us do the first three. The fourth step is where it gets expensive.

I’ve been wrong about enough things by now that I’ve started to treat my own confident views with mild suspicion. Not paralysis — you have to act on something — but a background awareness that the prior I’m acting on was formed by a person who had less information than I do now, and less than I’ll have next year.

Strong opinions, loosely held, sounds right. The trick is meaning it.

Categories
AI Books Writing

The Tax We No Longer Have to Pay

When Carol Coye Benson and I sat down to write Payments Systems in the U.S., one of the first problems we had to solve wasn’t about payments. It was about history.

To understand why the ACH network works the way it does, or why checks persisted decades longer than anyone expected, you need the institutional sediment underneath — the regulatory decisions, the failed experiments, the path dependencies baked in by choices made in the 1970s that nobody thought would still matter in the 2000s. The history is the explanation. Strip it out and you have a description of current practice with no account of why it exists or what it cost to get there.

But history takes pages. And pages test a reader’s patience. So you compress. You make judgment calls about what survives the cut and what gets left behind, and you make those calls knowing that every omission is a bet — a bet that the reader can follow without it, that the thread holds without that particular knot.

Writing it taught me something. The act of compressing, of finding the minimum sufficient version of a complex thing, forces a clarity that living inside the complexity never quite delivers. You don’t fully know what you understand until you have to say it precisely enough for someone else to follow.

But compression is always a loss. You feel it as you write. The version in the book is thinner than the thing you know.

Garry Tan uses a term — “tokenmaxxing” — that initially sounds like jargon from a performance optimization thread. The idea is simple: don’t be stingy with context. Give the model everything. Every source document, every relevant article, every piece of background that a human reader would never sit still for. Let it synthesize rather than guess.

The instinct it runs against is deep. We have spent decades building information systems around compression — search engines that retrieve rather than ingest, executive summaries that stand in for reports, one-pagers that distill months of work into something a decision-maker can absorb in four minutes. All of it was a rational response to a real constraint: human attention is finite and expensive. You couldn’t afford to read everything, so you built filters. The whole architecture of how organizations manage information was designed around that limit.

Tokenmaxxing is a bet that the limit has moved.

The model can read everything. The cost of giving it full context — the uncompressed history, the original sources, the institutional sediment — is low enough now that filtering before the model sees it may introduce more error than it prevents. You’re potentially discarding signal when you summarize for the model the way you’d summarize for a human. The model doesn’t need the one-pager. It can handle the report.

This doesn’t dissolve the need for curation entirely. More context isn’t always better — models can lose the thread in noise the same way humans do, just differently. The skill shifts from summarizing to selecting: not what’s the minimum version of this but what’s actually worth including. Different judgment, still essential.

But the deeper change is upstream of any particular project. The compression we built into every research process, every briefing, every book — that was never the goal. It was the tax we paid for human cognitive limits. Part of the process doesn’t pay that tax anymore.

When I think about writing that payments book today, I don’t think the book itself would change much — it still has human readers with finite patience. But the map we drew before writing it, the synthesis work, the “what connects to what across fifty years of regulatory history” work — that could happen at a different depth now. The understanding you bring to the writing can be informed by everything, not just the subset you had time to read.

The payments book was written entirely for humans, with all the compression that implies. But Tyler Cowen just published what he calls a “generative book” — 40,000 words released free online, paired on the same screen with a Claude interface so readers can discuss, interrogate, and extend it in real time. He’s writing for both audiences simultaneously now. The human reader and the model that will help that reader go deeper. The text is optimized not just to be understood but to be used — as context, as a jumping-off point, as raw material for a conversation that the author won’t be in.

That’s a different kind of writing. Not better or worse. Different. The compression decisions change when one of your readers has no patience to protect.

Writing still clarifies thinking. That part hasn’t changed. But what you’re clarifying, and who you’re clarifying it for, is quietly expanding.

Categories
AI AI: Large Language Models China

Cranes on the Horizon

In 2005, during my first trip to Shanghai and Beijing, the most striking feature of the skyline wasn’t the architecture—it was the cranes. More than I could possibly count, perched atop half-finished skyscrapers like a mechanical forest. Entire districts seemed to be mid-construction simultaneously, as if someone had pressed a button and the whole country decided to build everything at once. Dan Wang in his book “Breakneck” described China as the “engineering state” that approaches national problems with physical solutions. Back in 2005, coming from Silicon Valley, I thought I understood what growth looked like. I didn’t.

I’ve been thinking about that trip while reading Nathan Lambert’s recent piece, “Notes from Inside China’s AI Labs.” Lambert — who runs the Interconnects newsletter and does serious work tracking the open-weight LLM ecosystem — just returned from visiting essentially every major AI lab in China. Moonshot, Zhipu, Meituan, Xiaomi, Qwen, Ant Ling, 01.ai. He went in with genuine curiosity and came back with humility. That combination is rarer than it should be.

What he found was the cranes. Different domain, same energy.

Lambert’s central observation is about culture, not capability. The Chinese labs aren’t winning on any single technical breakthrough — they’re winning on execution discipline. He describes researchers, many of them active students, who bring no ego to the work. They absorb context fast, drop assumptions faster, and seem genuinely unbothered by the philosophical debates that seem to swirl constantly in the American AI community. When he tried to engage Chinese researchers on the long-term social risks of models or the ethics of AI behavior, those questions “hung in the air with a simple confusion. It’s a category error to them.” Their role is to build the best model. Full stop. To them, an LLM isn’t a philosophical entity to be interrogated; it’s a piece of infrastructure to be optimized.

That description landed for me. Not as a criticism of American research culture, but as a real observation about what the moment demands. Building good LLMs today is, as Lambert puts it, meticulous work across the entire stack — “all points of the model can give some improvements, and fitting them in together is a complex process.”

The work that matters most right now isn’t the 0-to-1 creative leap; it’s the thousand unglamorous decisions executed without complaint. Students who haven’t yet learned to lobby for their own ideas turn out to be well-suited for exactly this.

Lambert ends on a note that’s hard to shake. Looking up from his laptop on a high-speed train, he keeps seeing cranes on the horizon. He draws the same connection I did, though from the inside: the construction everywhere fits the broader culture and energy around building. “When I look up from my laptop and always see bunches of cranes on the horizon, it obviously fits in with the broader culture and energy around building in China.”

Twenty years after my first visit, the cranes are still there. They’ve just moved indoors — into server rooms and training runs and model releases that land every few months with quiet confidence. In 2005, what China was building was obvious: you could see the steel frames going up. What’s being built now is harder to see, which may be exactly why it keeps surprising us.

Check out Lambert’s essay – it’s remarkable. If the 20th century was defined by who could move the most earth, the 21st will be defined by who can move the most tokens. And right now, the cranes are moving faster than we think.

Categories
AI Programming Software Work

The Scarcest Thing

Garry Tan woke up at 8 a.m. after sleeping at 4. Not because he had to. Because he wanted to see what his workers had done overnight.

The workers are AI agents. Ten of them, running in parallel across three projects. And something about that sentence — wanted to see what they’d done — keeps stopping me. That’s not the language of someone using a tool. That’s the language of someone managing a team.

Tan gave a name to the state this puts him in: “cyber psychosis.” He said it as a joke. But the joke has an insight in it. He’s not describing addiction to a productivity app. He’s describing a shift in what it means to do creative work — the strange vertigo of becoming a director when you’d always been a laborer.

I’m retired. I watch this from the outside now, which is its own kind of vantage point. For most of my career, the path from idea to working product ran through people — through hiring and managing and the slow accretion of execution capacity. You had the vision or you didn’t, but either way you needed the team. The idea and the means of making it real were, structurally, separate things. The gap between them was where companies lived.

What Tan is describing is that gap closing.

The thing he built — gstack, his open-sourced Claude Code configuration — got dismissed in some quarters as “just prompts.” And it is just prompts, in the same way that a conductor’s score is just notation. The abstraction is the invention. What he encoded is a model of how a startup team thinks: the CEO who interrogates the why before a line of code gets written, the engineer who builds, the paranoid staff reviewer who looks for what breaks. Each role blocks a different failure mode. Blurring them together produces, as his documentation puts it, “a mediocre blend of all four.”

That’s an organizational insight. It has nothing to do with code.

Tan described being a “time billionaire” — not because his biological clock had slowed, but because he can now purchase machine-consciousness-hours. The bottleneck of implementation, which has governed every creative project since the beginning of creative projects, is dissolving for those who know how to direct.

The scarcest thing is shifting. It’s no longer the hours of execution. It’s the clarity of intent — knowing what you want to build and why the journey matters, before any of the workers start moving. That’s harder than it sounds. For decades, most of us could muddle through in the making of it. The act of building taught you what you were building. Now the making is cheap, and that shortcut is gone.

For someone watching from retirement, that’s not a small thing to absorb. The model I internalized over a long career — that ideas become real through sustained organizational effort, through teams and timelines and the grinding work of execution — is being revised faster than I expected. Not invalidated. Revised. The judgment still matters. The taste still matters. The why matters more than ever.

It’s just that the how has found new hands. Many of them. More than any team I ever assembled, available the moment the intent is clear enough to direct them, gone when the work is done. The constraint was always the hands. It turns out it was always the knowing.

Categories
AI AI: Large Language Models Anthropic

Breakout

Jack Clark doesn’t panic easily. He spent years at OpenAI watching capabilities inch upward, then left to co-found Anthropic, and has been writing his Import AI newsletter long enough to have developed — and been wrong about — many priors. So when he publishes an essay saying he has reluctantly arrived at a 60% probability that fully automated AI R&D happens by the end of 2028, the word “reluctantly” deserves some weight.

His essay, published last week and titled “Automating AI Research,” isn’t a press release or a fundraising pitch. It reads more like a man thinking out loud at the edge of something large. “I don’t know how to wrap my head around it,” he writes, which is a notable thing to say publicly when you are one of the architects of the thing you can’t wrap your head around.

The argument is built from benchmarks — not any single one, but a mosaic of them assembled to reveal a trend. SWE-Bench, the test that measures an AI’s ability to solve real GitHub issues, was at roughly 2% when it launched in late 2023. A recent Anthropic model sits at 93.9%, effectively saturating it. METR’s time-horizon plot tracks how long an AI can work independently before needing human recalibration: 30 seconds in 2022, 4 minutes in 2023, 40 minutes in 2024, 6 hours in 2025, 12 hours today. The trajectory, if it holds, suggests 100-hour autonomous work sessions by the end of this year.

Clark marshals similar progressions across AI fine-tuning, kernel design, scientific paper replication, and even alignment research itself. His throughline is the same in each: AI is now genuinely competent at the unglamorous scaffolding of AI development — the debugging, the experiment runs, the parameter sweeps, the code reviews. And crucially, it can now do these things not just faster than humans, but for longer, with less supervision.

There’s a Thomas Edison quote at the center of the essay: “Genius is 1% inspiration and 99% perspiration.” Clark’s claim is that AI has become very good at the perspiration. The question of whether it can supply the inspiration — the paradigm-shifting insight, the Move 37 — remains open. But he argues it may not need to. Most of what has moved the AI field forward has been sustained, methodical work, not lone flashes of genius. If you can automate the 99%, you have something that compounds.

There’s a data point that makes Clark’s argument feel less like forecast and more like dispatch. Last month Boris Cherny, who runs Anthropic’s Claude Code, disclosed that he hasn’t written a line of code by hand in more than two months. Every pull request — 22 one day, 27 the next — written entirely by Claude. Company-wide, roughly 70–90% of Anthropic’s code is now AI-generated. Anthropic’s stated position: “We build Claude with Claude.” The loop Clark is describing as a probability by 2028 is already running, at least partially, today.

The word Clark uses for the threshold he’s describing is not “singularity” or “AGI.” It’s quieter than that. He calls it “automated AI R&D” — the point at which a frontier model can autonomously train its own successor. It’s a specific, falsifiable thing. And he puts a number on it: 60% by end of 2028, 30% by end of 2027.

I’ve been writing about the dark software factory and the 3D printer that prints better printers, finding metaphors for what seems like an inexorable process. Clark’s essay is a different kind of writing about the same thing — the primary source document, the engineer’s log, the inventory of evidence. Reading it is a little like watching someone carefully pack boxes before a move. Each individual item seems manageable. But there are a lot of boxes.

What he’s describing — if the trend holds — is not a feature or a product launch. It’s a breakout. The moment the loop closes and the system starts building itself. He’s not certain it happens. He just thinks it’s more likely than not, and he thought you should know.

Categories
AI Business Work

The Tipping Point Was Last November

Matthew Prince, Cloudflare’s CEO, said something on today’s earnings call that I keep turning over. He didn’t bury it or soften it. He named a date.

“Internally, the tipping point was last November.”

That’s a specific thing to say. Not “we’ve been on a journey” or “AI has been transforming our industry.” A month. A moment. The thing changed, and he knows when.

What changed, by his account, is that Cloudflare’s teams began seeing productivity gains so dramatic they were hard to describe — people who were two times more productive, ten times, in some cases a hundred times. “It was like going from a manual to an electric screwdriver.” Usage of AI tools internally is up more than 600% in just the last three months. Every line of production code is now reviewed by an autonomous AI agent.

And then he said goodbye to 1,100 people — about 20% of the company.

Today wasn’t just Cloudflare. Earnings season has become something like a drumbeat. Meta is cutting 8,000 employees this month. Amazon cut 16,000 in Q1. Oracle eliminated roughly 30,000 to fund AI infrastructure. Block cut almost half its workforce. PayPal is reportedly planning to cut 20% of its staff over the next few years. Coinbase cut 14%. Snap cut 16%. As of this week, more than 92,000 tech workers have been laid off in 2026 alone.

The scale is striking. But what strikes me more is the framing — the specific language being used to describe what’s happening. These aren’t being announced as cost-cutting moves or post-pandemic corrections, the way they might have been in 2022. They’re being announced as architectural decisions. Structural adaptations. Evolution.

Prince was careful to be explicit: “This isn’t a cost-cutting exercise or an assessment of individuals’ performance. It’s about defining how a world-class, high-growth company operates and creates value in the agentic AI era.” That’s not empty corporate language, or at least not only empty corporate language. The distinction he’s drawing — between trimming fat and reimagining how a company is built — maps to something real about what AI agents can now actually do.

There’s a legitimate version of this argument and a convenient one, and they’re being delivered in the same sentence by the same people, which makes them hard to separate. Some analysts suspect companies are using AI as cover for cuts they wanted to make for other reasons — rightsizing from pandemic-era overhiring, funding massive infrastructure buildouts, chasing margin. Oxford Economics flagged this: maybe some firms are “dressing up layoffs as a good news story.” The cynicism is warranted.

But then there’s the Cloudflare number: 600% increase in AI usage in three months. That’s not a narrative. That’s a measurement.

What’s different about this moment — what makes Prince’s “tipping point” language feel accurate rather than convenient — is that the people making these decisions are themselves users of the tools. They’ve seen the productivity numbers internally before anyone else has. They’re not theorizing about what AI might do to their workforce; they’re describing what it already did.

That’s the thing that changed. For years, AI’s labor impact was a future tense conversation. Economists studied it, think pieces warned about it, conferences debated the timeline. Then, somewhere around last November apparently, a cohort of technology companies crossed from hypothetical to empirical. The future tense became past.

Whether you read that as tragedy, as transformation, or as both depends on where you’re standing. 1,100 people at Cloudflare today are standing somewhere very specific. Prince acknowledged this with what felt like genuine difficulty: “A number of friends will no longer be colleagues.” Whether that difficulty changes anything material for the people leaving is a fair question.

But the acceleration itself — the thing he named — is real. The tipping point was last November. And if it was last November for Cloudflare, it was some nearby month for Amazon, for Meta, for Block, for all of them. Whatever these companies learned that changed everything, they all seem to have learned it around the same time.

That’s what I find myself sitting with today: not just the scale of the disruption, but the synchrony of it. The realization arrived, and then the decisions followed. Quietly at first, then all at once.

Categories
AI AI: Large Language Models

The 3D Printer That Prints Better Printers

Imagine a 3D printer that looks at its own design and begins printing a better version of itself. The loop closes. What had always required an external human intelligence now happens inside the machine. All by itself.

Jack Clark — Anthropic co-founder, someone who has spent years closer to this technology than almost anyone — puts the odds of this happening by 2028 at better than even. I have been turning that number over ever since I heard it. Not the technical claim, exactly. The feeling of it.

We have grown used to AI accelerating our work. Coders watch models close GitHub issues at rates that would have seemed miraculous eighteen months ago. Researchers delegate experiment design, kernel optimization, even the fine-tuning of smaller models. The scaffolding of AI progress is already being built, in part, by the systems themselves. But the moment the system begins to redesign the scaffolding — that is something new.

What unsettles me is not the raw capability, though that is staggering. It is the loss of distance.

For most of technological history, the creator stood outside the creation. Even the most sophisticated tools remained tools. Now the distinction begins to blur. A model that can meaningfully improve its own training process, its own architecture, its own alignment constraints, is no longer merely reflecting human intent back at us. It is participating in the shaping of its own nature. And because each iteration can happen faster than the last, the curve steepens in ways our intuitions, tuned to linear progress, struggle to grasp.

Clark is careful, as he should be. He speaks of validation work that will still fall to humans, of the need to broaden the pipes through which abundance flows, of preparing defense-dominant postures against misuse. Yet the image that lingers for me is quieter: the silence after the handoff. What does it feel like when the thing you have been painstakingly teaching begins to teach itself — and then to teach its teachers?

I think about Leo Szilard at the traffic light, or the first controlled chain reaction under the stands at the University of Chicago. Moments when a new regime of possibility quietly announced itself. Recursive self-improvement carries that same charge — not a single event but a process, one that could accelerate the very pace of events themselves.

The more I sit with it, the more I return to an older tension in our relationship with tools. We build them to extend ourselves, and in doing so we are always, subtly, extending — or perhaps risking — what we are. The values I try to live by — generosity, curiosity, compassionate honesty — are not refined in specifications. They are refined in friction, in relationship, in the slow work of being human with other humans. If the machines begin to optimize their own lineage at speeds we cannot match, will we still have the bandwidth to tend the parts of ourselves that no algorithm can yet measure?

I don’t know. None of us do. That uncertainty feels honest.

What feels clearer is the invitation. Not to fear the printer that prints better printers, nor to worship it, but to remain awake inside the loop. To ask, as each new version arrives, what kind of world we are collectively printing — and whether the values we claim to hold are baked into the design or merely etched on the surface, likely to wear away under the heat of iteration.

The light is still yellow. We are still deciding whether to step off the curb. But the traffic is already moving faster than it was a moment ago.