Several years ago I read Jeff Hawkins’s A Thousand Brains, which dug into his theories on how our brain works. In it he argued intelligence comes from thousands of cortical columns that each learn…

🔗 Opinion | Is There an A.I. Bubble? A Debate. - The New York Times: “So Paul, make the case to me that this is a bubble. I mean, it’s...

Everyone seems to be asking how to build trust in AI. But maybe we should be focusing on being trustworthy? Thoughts on AI shit storms, where and when AI is used, and focusing on showing rather than…

We release Video-Index 1.1 with agentic evaluation setup.

Leveraging REVIEW.md to calm down Claude's code reviews on GitHub.

I've built with 1 agent, 8 agents in parallel worktrees and a software factory that ran for days on its own. My sweet spot is level 2 or 3. Here's why.

Nathan Wilmers, Regina Catipon, Maxim Massenkoff, and Peter McCrory, in an Anthropic research article: How has AI restructured work and employment at Anthropic? As a company at the frontier of AI…

Yes, LLMs are all the craze right now, for a variety of reasons. One of our great community members made the TinyBrain little LLM that can run natively on the MEGA65. It's only 325K weights, but…

by now, most people should be aware that we're in a cost of compute crisis. The cost of CPU and memory is through the roof. There just isn't enough. Meanwhile, there definitely aren't…

This Labor Day I was finally able to stop and take a breath about how AI is going to affect the game industry. Like many of you, I feel like AI news, content, and hype have permeated every aspect of…

All 35 major model prices as of 17 October 2026, from Meta's $0.10/$0.20 Contributor tier to GPT-6 Astra Fast at $20/$100, with what each model is actually for.

Models are changing the loop through which organizations learn, build, and refine what they express through software

Rhyme, and reason In the early days of the World Wide Web (WWW) cartoonist Peter Steiner created a cartoon depicting a dog seated at a computer console talking to a dog nearby with a caption that…

Futurology is weird. Our prefrontal lobes compel us to make up stories about things to come. Narratives that are sometimes enhanced by scientific projections and models and plausible scifi…

Hello Coders! 👾 Something about the way many people talk about AI and creating software has started to bother me a lot lately. I keep seeing people use the term vibecoding for all AI assisted coding…

Just a normal set of lyrics to a normal blogpost

My first ever “blog post” (though not in this blog) was about how to write a research paper. I wrote it after reviewing a batch of badly-written papers for a conference, probably ITP or CADE. Later —…

A few months have passed since my last "How I use AI" post anad a good chunk has changed in said months. Lucky for me, still only $20 a month is going into a codex sub, but how I interact…

Over the last few months, AI has been making undeniable leaps. I hadn’t digested news of the Jacobian conjecture being conquered when notifications popped up alerting me that Navier-Stokes had…

I built KillSwitch, an esoteric language designed to confound LLMs and coding agents, and benchmarked the major models against it.

Recently, both their SDK and all their apps have been completely rewritten. Unfortunately, instead of doing it themselves, they had everything vibecoded. You can see this by checking the repositories…

A model can now rebuild anything that runs on your machine. The companies will answer by moving everything onto their servers, and the small developers pay first.

I am a huge fan of controlling agents by voice, and talking to AI in general, as if talking to a real person. OpenAI's Advanced Voice Mode and the more recent GPT-Live were both mind-blowing the…

A log check flagged 16 of 31 Claude Code turns that ended on a failing test. None ended on a false claim; it missed the one that did. Why, and 3 rules to try.

I strive to add accessibility descriptions to all images I post online.1 Yet, I do not have any skill whatsoever at writing them, nor do I want to invest time in learning, so I call upon the IMatch…

It is Friday. The French used to say “Le roi est mort, vive le roi!”. For those of us raised with less melodious languages, this means “The king is dead, long live the king!” And since we don’t use…

An evidence-based comparison of Mistral with American and Chinese foundation models, and an assessment of what model capability contributes to European digital sovereignty.

A sealed 240-call exhibit grounds the boundary between decision spaces an agent should enumerate exactly and those it should delegate to a typed judge.

No, really it is. But not for reasons you might think. This isn’t some tedious anti-AI rant. It’s a disgruntled moan at Microsoft being Microsoft and half-arsing something just to keep up.

It was recently reported that Palantir CTO Shyam Sankar claimed that "many AI researchers have started smoking" as a way of signaling extreme confidence that AI will cure cancer. While I…

I've been getting weird e-mails for quite some time now, and this got me thinking about the future of the internet. LLM-generated e-mails are becoming harder to spot, and some of them are landing in…

This content is from In 30 Minutes Guides. Here’s an AI newsroom example from the news service I’m developing. The original version by the Tbot (translation bot) had a strange chart and…

AI access and AI ability aren't the same thing. The real measure is leverage with verification: how much competent work you can hand off and still trust.

Humans are far less random than they think. When asked to pick a number from 1 to 10, we tend to favor 7, sometimes choose 5, and avoid the extremes. LLMs (Large Language Models) learn from…

I was a guest earlier this week on the Ezra Klein show, discussing the current AI moment. As ever, I was focused on the underlying financial and funding architecture, the implicit assumptions, and…

Using an agent doesn’t make you an agentic engineer though, and you are ready for this conversation.

The Wizard of Oz experiment has always been one of my favorite Lean Startup techniques.

The machine runs 24/7: writes the code, files the tickets, fixes the bugs, reviews the PRs, deploys, monitors, finds the leads, writes the outreach, reads the stats. The developer is resting, and…

After listening to a recent Ezra Klein podcast with Sherry Turkle ( youtube , nytimes ), I think creating a sense of awkwardness when interacting with an LLM chat interface is not currently possible.…

Why assigning trust by action, not by agent, is the design system work nobody's doing yet

Sorry for this almost provocative title. Anyone who works in tech and regularly reads about it has probably heard about a new tool called REA - Reverse Engineer Anything . The name pretty much…

The average developer needs to accept that the latest tier of foundation models writes better code than them. Most developers who think they are above average are not. It's time to learn how to be…

Judgment to the model, precision to the tool, decisions to humans

Let me write about AI, tech, and the internet once again. Today, I chatted with some online friends. On the off chance that those online friends may see this, this is not a targeted attack or insult…

Everything seems broken in the software development world right now. I’ve been thinking about this for a while, but it has been escalating very rapidly, especially in the last few weeks, so…

Every now and then I see a story like this and I imagine the alternate universe where biological computers became the next big thing instead of LLMs. They're building a "biological data center"…

Why is AI writing so bad? It can one-shot an entire app, but if I ask it to help me write a 1000-word essay, it is riddled with non sequiturs and a lack of coherence. I fed it the outline, key ideas…

On June 16, 2026, SpaceX announced it would acquire Anysphere, Inc, the parent company of Cursor, for $60 billion in stock.  A lot of people argued that Cursor wasn’t worth anywhere near that…