TL;DR: Employees seduced by corporate jargon are terrible at making decisions, whilst a mind-reading milestone lets scientists watch movies inside a mouse’s brain. Meanwhile, the world’s…

I set up Claude Code agents early, stopped using them, then dug back in. Here's what the subagent architecture actually enables.

I’ve been studying the intersection of digital technology and office work for quite some time. (I find it hard to believe that my book, ​Deep ... Read more The post Why Hasn’t AI Made Work Easier?…

No longer do we have anyone to question what you’re trying to accomplish.

Watch Daniel Solove and Omer Tene (Goodwin) discuss the latest developments in privacy and AI law. Subscribe To Prof. Solove’s YouTube Channel Privacy Training Privacy Law Fundamentals Book The…

Dissociative Identity Disorder is a bitch. A dozen or so people are stuck in a single body, fighting for control over it without being able to reach out and restrain one another physically. It's not…

The new AI Era

This blog has been a part of my life for over fifteen years ago. When I first started, I was inspired by Patrick McKenzie’s article Don’t End The Week With Nothing. This post left a memorable…

LLM的默认输出是consensus：正确但平庸。Deep Research其实是Wide Research。我们找到了一种系统性方法，用个人认知上下文把LLM从consensus里强行扯出来。一年实验，有控制变量证据。

This was presented at the Prairie Postgres meetup in Chicago in March 2026. Your browser does not support iframes.

I built a TUI that interviews me and writes blog posts. Two phases — conversational interview, then section-by-section editing — running entirely on local LLMs.

The practice of writing code has been changing fast. Really fast. I–and a large portion of the industry–write code in a completely different way from a year ago. And by that I mean that I largely…

An LLM's default output is consensus: correct but mediocre. Deep Research is really Wide Research. We found a systematic way to pull LLMs out of consensus using personal cognitive context. One year…

Here's a hot take: any individual's given LLM usage, only extends the person's useful outputs by a factor of n, before giving the illusion of learning/creation/productivity/ideation beyond it by a…

This week, we sit down with two guests from Anthropic, Matt Samuels, Senior Product Counsel, and Den Delimarsky, a core maintainer of the Model Context Protocol, or MCP. Together, they unpack why MCP…

I volunteered to moderate a subreddit and immediately built tooling instead. This post walks through how Devvit endpoint restrictions forced a pivot from Claude to Gemini, and how that constraint led…

I’m going to tell you something that would have sounded absolutely insane five years ago: I’m running artificial intelligence on a computer the size of a lunch box, it works offline, my data never…

Deploying a TFLite Micro keyword spotting model on an nRF52840 from a single terminal session — no code written by hand, no hardware touched.

Right now, every way we talk about AI at work is helping define what ownership, competence, and judgment look like in the AI era. We should be more deliberate about that.

Josh talks to Luke Hinds, CEO of Always Further, about MCP and agent security. We start out talking about Luke’s new tool, nono which is a sandboxing tool that has AI agents in mind as a use…

A decade of information retrieval and three years of agent harness engineering, poured into a local-first knowledge system that avoids vector search entirely.

Explore the shift to AI-driven coding, the trust gap in GenAI, and how formal verification via Lean can replace manual review with mathematical certainty.

It almost feels like writing a journal entry detailing some sort of relapse on the road to an addiction I’m trying to break. It felt like something I knew I shouldn’t do, but was too…

On the temptation of LLMs, and why sometimes you still need to do the work by hand.

Dogfooding a code review tool through a planning-first AI workflow

Experience with AI-assisted coding varies widely across different projects, teams, and individual developers. How Orgs Actually Win With AI by Laura Tacho at Pragmatic Summit showed some interesting…

In recent months I noticed a shift in my information feed. More and more people are starting to write, the essay world is booming in quantity. Yet, few are well-crafted in the traditional sense. In…

Lately, I've been playing HARD with LLMs for robotics-related coding. The pace of development of these tools is crazy. I feel like if I'm not paying attention for two weeks, it becomes hard to catch…

Writing Skills That Agents Can Actually Execute banner

In what can only be described as sign of the times, one of the books I wrote, arguably, the most impactful one, Doing Math With Python is one of the works in the Anthropic Copyright Settlement. This…

<![CDATA[

ForgeCode now reaches 81.8% on TermBench 2.0 with both GPT 5.4 and Opus 4.6. The interesting part is not the score. It is what we had to change in the agent to make GPT 5.4 behave as reliably as Opus…

I present a system for communicating how AI is used in articles

If you&rsquo;ve read my previous posts on ChatGPT memory, Claude memory, and Clawdbot memory, you already know I keep coming back to the same question: how do these agents actually remember? Hermes…

Interactive treemaps showing LLM exposure for every occupation in the Netherlands and Europe, inspired by Karpathy's US job market visualizer

I’ve been doing a lot of research and prototyping lately to learn about new concepts and tech. This means lots of Rust code, low level memory management and optimization, complex math algorithms —…

Turning off web search and tools to see how Gemini analyzes the Iran war using only its training data.

HACRL proposes a new paradigm for reinforcement learning - instead of training models in isolation, multiple agents collaborate by sharing successful trajectories during training. This simple idea…

AI has turned solo developers into something new: part programmer, part supervisor of a tireless digital team. The productivity gains are staggering, but so are the risks. From debugging miracles to…

I love contributing to open-source projects. There is this insanely good feeling that I get when my changes get merged into the main branch. Dopamine goes through the roof when I see the number of…

Thinking big Remember February 2020, watching the covid virus spread overseas and wondering what would happen? Matt Schumer writes something big is happening: "I think we're in the 'this seems…

Some days I get in bed after a tortuous 4-5 hour session working with Claude or Codex wondering what the heck happened. It's easy to blame the model - there's so many options to choose from: They're…

I only made it to the morning of Build with AI at Imperial College London on Saturday, March 14, 2026, but honestly, that was enough to make the trip worth it. The event, put together by GDG London…

LLMOps Part 12: Understanding LLM fine-tuning, parameter-efficient methods like LoRA and QLoRA, and alignment techniques such as RLHF, DPO, and GRPO.

pPer Anthropic’s most recent article on a href="https://www.anthropic.com/research/labor-market-impacts"AI’s implications on the labor market/a, I belong to the most exposed occupation: computer…

It’s worth remembering the meaning of the luddites doesn’t lie in the fact that some of them were skilled machinists, somehow reluctant to rebel against the tech of the day. It’s that they smashed…

Could Claude be my Hyperbolic Time Chamber?

How a five-phase AI-assisted SDLC pipeline (SpecKit, Git worktrees, /done, Claude Poll, /fix-pr) enforces architectural discipline at speed without the AI sprawl.