I saw this at the bookstore and was intrigued by having a book that walks through the math of neural nets and machine learning. I have a very conflicted relationship with modern AI to say the least.…

🔊 Listen to this post Tonight I found myself thinking about a comet. Not a metaphorical one. Halley’s Comet, the old celebrity of the sky, the bright visitor that swings through our neighborhood so…

How I achieved to ship code to production in the last two months without writing a single line of code by hand, with zero production incidents, zero rollbacks, and zero downtime.

In one month, the US government switched one frontier AI model off worldwide, rationed a second to about a hundred approved companies, and held a third back before launch. One government now decides…

How I run a solo AI consulting business from Slack: a fleet of single-purpose agents with the Claude tag as the meta-agent above them — reaching every surface, opening PRs, and maintaining the other…

AI gets you to a working draft fast. The trouble is what the missing fifth contains, the edge cases and operational reality that used to teach engineers their judgment, and who builds that muscle now.

The second wave of making my studio's work citable: AI installations and tools as research outputs, and the Zenodo metadata that keeps each record honest.

Today's top 5 papers from arXiv covering AI, machine learning, NLP, and computer vision

The constraint that launched a thousand reasoning models may have just been optional.

My Vue MAD 2026 talk, written up. What an agent actually is, and the three things every Vue project needs to get more out of AI: context, feedback loops, and discoverability. The twist is that none…

Current-generation AIs are particularly good at finding and generating corresponding patterns. This can be useful when classifying images as cats or dogs (classifiers) or creating new sentences or…

The honest math on building your own AI box in June 2026, who it actually makes sense for, and why timing matters more than people think.

In January of this year, the Conference Board of Canada, now Signal49 Research, projected that 555,000 jobs in Canada would disappear by 2030 because of the widespread adoption of artificial…

When I say AI is too expensive normal people tend to look at me like I'm crazy. ChatGPT has a pretty generous free tier. For those who choose to pay, $20/mo gets you essentially unlimited AI for…

Some personal hot takes from AI: Engineer Miami follows...1. Software development is a dead-end profession because anyone can be a software developer now. 2. Anyone can use Cursor or any other tool…

Reasonable people

AI approval gates: reversibility-tiered human-in-the-loop design with four health metrics that prevent reviewer atrophy at machine speed.

I’m a massive nerd but I thought this was a really fun post. These security tilted models are very aware and know how to interact with/exploit the evaluation harness, the juice is about halfway…

Cross-posted from my Intentional Teaching newsletter. The other day on Bluesky, education professor Jon Becker shared a few highlights from a panel he attended featuring high school students talking…

The Project of AI is a world-building endeavor, wherein those who fund and develop AI systems both operate through and seek to sustain networks of power and wealth. Janet Vertesi, Alex Taylor, Ben…

It’s been an incredibly long few weeks, and as a result my previously-planned Hater’s Guide just isn’t possible within what little time I have left in this week, which is why…

In my previous post, I talked about the cautionary aspect of AI hallucinating its own hallucinations. There’s a deeper element … More When AI Performs Causality Instead of Practicing It…

I did a lengthy interview a little while back, and it is now out. Folks may find it a helpful restatement of some of my views on the current situation, from historical parallels to the AI paroxysm,…

And: OpenAI's delayed IPO, worker discontent, and Chinese AI! The post California’s about to learn that you can’t tax a moving target appeared first on Cautious Optimism.

perche’ non sono sorpreso ? il modello culturale e’ Biff Tannen Grok AI is for porn. Two former xAI employees estimate that well over half of all Grok traffic goes to pornographic images,…

On May 22, June 3, and June 8, 2026 , Anthropic published three cyber research posts that looked like different stories. One was about exploit benchmarks. One mapped malicious AI use to the MITRE…

Download Transcript | Keywords: Virtual Assistants; AI; Wake Words; Summons; Turn design ‘Wake words’ such as “Alexa”, “OK Google” and “Hey Siri” have become commonplaces within everyday talk,…

徐贲： 学术共同体真正应当追究的是责任（responsibility），而不是来源（origin）。因为来源从来都是混合的，而责任必须是明确的。

Standardizing knowledge for the future of AI agents My last piece on Google’s Open Knowledge Format (OKF) was one of the most popular I’ve ever published. Since then, I have been heads down working…

Let me short-circuit the flames. He wasn’t using AI, but my attempts at trying to rid myself of AI slop in my feed reader flagged him as the worst offender. Is he? No. It just points out how hard…

AI tools make one software failure mode much easier to miss: a bad product decision can now arrive wrapped in a working...Read More... Source

I have admittedly taken a softer line on some “AI” gen since our clients use it, and combined with our awareness of how desktop publishing unfolded, and where Medinge Group’s own thinking is heading,…

A couple of months ago I moved to a new team at work. The team had been running for a while before I joined, and there was a lot of context I didn't have — issues, discussions, strategy docs, repos…

UI execution is being outsourced, teams are shrinking. The industry is more of a mess than it looks. In uncertain times, shrinking the team and increasing efficiency through AI is precisely what will…

How structured logs helped Codex fix a year-old bug, and why logs are the key to unlocking autonomous workflows.

The useful part is not giving agents more context. It is making your research, notes, and sources available again in the next session.

OpenAI reveals its first custom AI chip, IBM extends Moore's Law, Anthropic accuses Alibaba of stealing Claude, and $27M spent on one congressional race.

The Endgame We live in interesting times. Like many people, I work in an industry that has become absorbed by an existential angst - the robots have invaded, proved surprisingly capable, and it is…

Short note on trying local open-weight LLMs across Qwen-Code, Codex, and Claude Code harnesses.

AI does not make creative work disappear. It moves the work toward framing, judgment, iteration, and responsibility.

Where the entire discussion about AI isn't about how great it is, but how biased and destructive it can be. I know, I'm shocked too.

You might have read recently on this blog that my procurement preferences for hank.parts are basically EU,(self hosted) open source, UK/CH,Rest of the world,in this order. This article is a…

In part one of this little series I introduced TasmanianDevil, a standalone, offline PII detection and de-identification engine for .NET. We saw it on its own - detecting and validating PII,…

I almost shipped it. The integration was finished. Cattitude, our own CAT tool, had full support for Lara, the AI translation engine from Translated that is replacing the old ModernMT. The SDK was…

A short musing on what I find wrong with so-called second brain/brain augmentation tools

A few months ago I introduced AgentGuard, a library for declarative guardrails and safety controls for .NET AI agents. One of the rules it shipped with from day one was PII redaction, but back then…

Locally-running generative AI has made considerable jumps in quality in the past three years. I think it is finally time to evaluate such models in terms of how they can help researchers do their…

AI took away the constraints that brought discipline to MVPs. You must impose them yourself.

In folklore, a genie grants wishes without judgment — it gives you exactly what you asked for, whether or not it is what you needed. The danger was never the genie. It was the wish. Anthropic’s…

I’ve been using an AI orchestration framework I built called Mozart.You can check it out here: https://github.com/jstuart0/mozart-orchestrationThe idea behind Mozart is pretty simple. Instead…