We can't find a fitting metaphor for AI, which may be a sign of how transformative it will be.

Many companies today are facing increasing pressure to adopt AI and launch AI-based features. While AI is a transformative new technology, such pressure can be unhealthy and counter-productive. Faced…

My rule for AI use is a border of interaction. Before the border, in private, I am happy to use AI as much as it helps: search, planning, code, transcription, translation, summaries, drafts, and any…

One of the problems — arguably, the core problem — with the current generation of AI chatbots is that they give the user the impression of authority, while in actual fact they do not necessarily have…

Machines should bend to suit humans (centaurs), not humans bend to suit machines (reverse centaurs).

Cloud platforms, data platforms, and now agent platforms. Every company wants to build one and run a fleet of agents. Having built cloud, compute and data platforms before, I'm hoping we can learn…

Dwarkesh Patel 1 recently posted an award for the best answers to four key questions about AI. It’s partly a challenge and partly a job interview, since some of the winners will get offered a role as…

Eight optimization attempts on Echo-TTS CPU inference, the five that worked, quality evaluation with voice cloning, and how the optimized CPU path ended up faster than the GPU hybrid.

I have been vibe coding a lot. Like, a lot a lot…. and it’s been leaving me feeling a bit conflicted. I’ve been committing hard, and feeling tremendously productive. But I also get this weird flash…

I’ve written not one, but two new things for school publications! First, I’ve written about The Sound, one of my favorite bands, in Firebird. If you were wondering why I’d dug up an…

In The Last Programming Language Might Not Be for Humans, I argued that if AI becomes the primary author of code, the source language has to bend around that author. In What Comes After the Last…

Not all business relationships do the same job. Research shows which ones survive the agentic economy, and which were always just expensive due diligence.

How far back in time can you understand English? The same passage written in 100 year intervals, here’s how English in the year 1000 reads: And þæt heo sægde wæs eall soþ. Ic ƿifode on hire, and heo…

TL;DR: you may no longer use the content of this website for commercial purposes. In particular, you may not use it as training data for commercial language models. I decided to change the licence…

a quick one. small adjustments to how i work with agents, none of them deep on their own. but i can feel them adding up, the workflow shifting in a way that feels worth pausing on before any of it…

One of the great things about LLMs is how prevalent voice-to-text apps have now become. I first found out about this segment of apps with WhisperFlow, but was shocked by what they were charging,…

Artificial intelligence may represent less of a technological rupture than the next stage in a long institutional history that began when spreadsheets transformed modern office life. From Ledgers to…

The MSM narrative seems to be that superpowers like US and China are about to crack down on AI. www.wsj.com #

A field guide to making coding agents useful without pretending the agent is the whole factory.

A fake CEO voice, gift cards, urgency, and secrecy almost worked. The defense is simple: break isolation, verify through known channels, and never treat voice as identity.

AI is a sword for cutting through blockers, not a shield to hide behind.

Tonight, the role of Dave will be played by ChatGPT. What is ChatGPT wearing today? ChatGPT is wearing a charcoal gray cardigan over a perfectly neutral blue button-down shirt that somehow…

Este artigo documenta um incidente específico e real. Ele expõe uma classe de vulnerabilidade que merece atenção: a mutabilidade não supervisionada de regras de segurança por agentes autônomos.

I have a Raspberry Pi that's been sitting in a drawer for years. Pulled it out recently, popped in a fresh SD card with the latest Raspberry OS, and decided to actually do something useful with it.

A professional transition, twelve months of ALG1, a MacBook Pro, and the open-source immune system for the post-AI internet. Every word is true. And it’s AI slop.

An event of two halves. I was only there for the first as it's been a long few days and I needed some downtime afterwards. Some notes on the first half.IntroDave West, CEO of Scrum.org, on What…

Fabrizio Ferri Benedetti: Tools created through AIs hurt my brain, I decided, because they’re often devoid of three fundamental qualities of good tools: reach, sociality, and finish. In a way, thanks…

Simon Willison on the nuances of vibe coding vs agentic engineering. simonwillison.net #

AI is reshaping what technical leaders are for. Six patterns of response — buyer, vibe builder, depth-seeker, evangelist, governor, abstainer — each delegating the thinking somewhere different. The…

This is an interesting article. The author argues that AI alignment should not be understood as “making AI maximize human preferences.” The mainstream approach to current AI alignment…

(Which, incidentally, raises the delirious spectre of the converse of a p-zombie: a being that shows all the properties of being aware, except awareness—wait. No. Strike that. Reverse it.)

I’m overdue an update. Two months on from my last workflow post, much has shifted - some of it from new tools, more from joining an organisation that takes AI seriously. The biggest change:…

I’ve always been a voracious reader. Non-fiction mostly. In fact, I credit a lot of my early professional success to learning to leverage the big ideas I discovered in books written by people a lot…

One thing I’ve learned building creative and technical teams is that motivation does not come from process alone. It comes from grounded creative freedom. People do their best work when they have…

Heya! So, for the past couple days, it's been basically impossible to escape this hypothetical question: "Everyone in the world has to take a private vote by pressing a red or blue button. If more…

Anthropic raised Claude limits by renting SpaceX compute. The funny part is where the compute came from.

Why coordinate-level optimizers can behave differently on weights that represent the same model, and how quotient-aware updates remove the hidden gauge.

TV interview with Scott Budman / WNTV (NBC Bay Area) regarding the Elon Musk versus OpenAI / Sam Altman trial in Oakland, CA, for TODAY: NBC. https://www.nbcbayarea.com/author/scott-budman/

A meditation on token economics, the AI hierarchy, and the moment your most agreeable collaborator turns out to have been on a payroll.View Post

Earlier this year, the University of Glasgow held its first International Symposium on AI in Education, organized by the School of Education and the Centre for Teaching Excellence. I was invited to…

Waterfall 2.0 - The Quiet Return of the Methodology We Micked Open any AI-coding repo on GitHub today and count the markdown files. Then count the lines of code. The ratio has been climbing for two…

This is how things work: you, being a “literally who?” nobody, wrote approximately 13 pages of titles of articles, about exactly how and why the current incarnation of AI is only an…

AI can produce work that looks expert without being expert. The failure arrives in two shapes, and both are reshaping the workplace.

We humans use analogies and metaphors to understand concepts. But it is important to know that they are only approximations, not the actual concept in itself. Saying LLM is like a new higher level of…

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…

Annual Finnish Social Psychology Conference, Tampere, 8 May 2026 Abstract Public and scientific discourses about ‘AI’ treat ‘artificiality’ as an inherent property of machines, and ‘intelligence’ as…

Also: AMD's quarterly results, the government reviewing AI models, and the decline of capitalism. The post Is closed-source AI really losing its edge? appeared first on Cautious Optimism.

Palo Alto Portal Vulnerability, Most Companies Aren't Ready for AI, Surface Updates, CopyFail YARA Rules, and more...

Demystify AI agents by exploring the case for moving the loop to the browser. Learn how client-side orchestration enables tight UI integration and data control.