Applied AI reviews, build logs and essays on systems, validation, agent workflows, decision-making and how ideas survive contact with reality.
Why I write Working Stories Kitchen Notes
I think 2026 may be remembered as the year we finally started questioning one of the strangest assumptions in the software industry: That you can put your intellectual property on somebody else’s computer and continue co
Read essay ->
Well, this is going to upset some people. Linus Torvalds has published a series of Linux kernel patches written with Claude Code. Yes, that Linus Torvalds. The man who created Linux, maintains its kernel, and has spent d
Read ->
Every few days, someone on the internet announces that another profession has officially become obsolete. Programmers? Finished. Designers? Pointless. Writers? Extinct. If you’re not running 20 AI agents simultaneously,
Read ->
I’ve been following the AI employment debate for a while, partly because I’ve invested in several AI-native companies and spend a lot of time thinking about how they operate. And I find both popular narratives increasing
Read ->
British and Australian researchers have finally investigated one of the great mysteries of the modern internet: How do AI slop factories produce so much garbage? The answer is surprisingly sophisticated. A new study publ
Read ->
Another interesting entry in the growing collection of scientific problems where AI has started doing something more useful than summarizing papers. Physicist Matt Landreman from the University of Maryland has published
Read ->
Daron Acemoglu has been returning to a question that I think deserves much more attention: what happens to democracy if AI massively concentrates economic power? Acemoglu is an interesting person to make this argument. H
Read ->There is a fairly comfortable story about AI. Models get better. Productivity rises. Intelligence becomes cheaper. Everyone gets richer. I think the first three parts are probably true. The fourth is much less obvious. A
This is almost too perfect a Silicon Valley story. A 23-year-old student drops out of college. Builds an AI assistant. Seven months later, investors are throwing billions at him. The startup is still invite-only. Most pe
OpenAI just expanded OpenAI Academy into a fairly serious training platform. The catalog now contains 14 free, self-paced courses , organized around four broad roles: people using AI at work, developers building with AI,
A few days ago I wrote about AI shopping already being here . My argument was that everyone is waiting for agents to press Buy , while the important shift is already happening earlier. AI searches, compares prices, build
I already have AI agents constantly monitoring hotel and flight prices. And I regularly give Codex a very sophisticated financial instruction: “Find it cheaper.” It works surprisingly often. Not in the futuristic sense w
Over the weekend, a few friends and I had a small Zoom debate about AI, civilization, work, power and the usual cheerful subjects one discusses instead of simply opening a bottle of wine and talking about football. We di
a16z recently published a column by NYU professor Ruby Thelot arguing something mildly offensive to anyone over thirty: the internet is probably fine. You’re just getting old. His target is the now-famous idea of “enshit
I’ve written quite a lot about the skills adults should develop for a world with increasingly capable AI. The reason is simple: I think we’re heading into one of the largest macroeconomic shocks of our lifetimes. Not nec
Sixty-five years ago, America was already asking whether machines would produce more with fewer workers, destroy entry-level jobs, make retraining pointless, and eventually break the link between work and income. They go
There is a particularly good AI experiment from METR that I keep thinking about because it measured something almost nobody measures: whether AI coding tools actually make experienced developers faster. METR: Early-2025
I’ve been working in IT and around technology and information for more than a decade. I can clearly see the accumulation. Every year I know more. I recognize patterns faster. I make fewer stupid mistakes, or at least inv
More than half of new articles published online may already be AI-generated. And yet humans still seem to be winning the attention game. A recent Graphite analysis found that AI-generated articles had overtaken human-wri
I increasingly think one of the biggest risks in AI right now is surprisingly mundane: spending the next three years building something that isn’t ambitious enough. Everything that used to qualify as a decent software st
People happily put “AI propaganda,” “cyberattacks” and “human extinction” into the same bucket called AI risk , despite these being events with wildly different probabilities and consequences. So here is my completely su
For most of its existence, Pictiq had a slightly embarrassing problem. It was supposed to be a visual language, but actually writing anything in it required knowing rather too much about GitHub, JSON, internal IDs and th
There is a new internet habit that is beginning to drive me insane. You disagree with someone in the comments. You write a few sentences explaining why. And five minutes later you receive 1,400 words copied directly from
