Subscriptions tech digital media and capitalism
I don’t care if you wrote something with Claude, ChatGPT or Ouija Board: it’s your work product. Comport yourself accordingly, and don’t waste my time. There are no lossless transformations of natural-language text. The post It’s your work product appeared first on Joseph Bou-Younes s Website .
https://gruhn.me/blog/2026-08-03/ Too often I ask a question in Slack or leave feedback under a merge/pull request or argue with friends in a WhatsApp group and get back: Claude said: [giant response verbatim] Please don t do this. I mean, I ve done this. But I ve been on the receiving end too many times now. This is [ ] The post This absolutely sends me appeared first on Joseph Bou-Younes s…
This is very interesting. The more time I spend in Claude Code, the more I thought that coding agent harnesses were the perfect sort of software to be open sourced. The blossoming popularity of Pi seemed to reinforce this. But then… Simon Willison, Better Models, Worse Tools The post Coupling models and harnesses appeared first on Joseph Bou-Younes s Website .
AI is going to help humanity by revealing just much human time is spent on bullshit work. Consulting giant Accenture is trying to figure out how to stop non-technical workers from blowing through companies’ AI token budget on trivial tasks like converting PDFs to presentation slides, according to leaked audio obtained by 404 Media. Across [ ] The post Understanding the amount of bullshit work…
The post I was there appeared first on Joseph Bou-Younes s Website .
South Shore news: AI-generated newsletter has paying audience (Boston Globe) This feels like the future. AI agents, directed by assignment editors, gather local intelligence and context. Other agents draft and edit stories. Humans handle quality control and accountability. Meanwhile, separate software tracks traffic and prominence, feeding signals back to adjust agent behavior — optimizing for [ ]…
With models that can run for days, NLW argues we ll all have to up-level ambition — and become token-efficiency optimizers who match models to use cases. Nate B. Jones s framing: most of us have nothing that s ever taken even an hour on AI, so the scarce skill is imagining tasks worth handing to a model [ ] The post The new skill: task imagination appeared first on Joseph Bou-Younes s Website .
Jevons paradox is coming to knowledge work. By making it far cheaper to take on any type of task that we can possibly imagine, we’re ultimately going to be doing far more. The vast majority of AI tokens in the future will be used on things we don t even do today as workers: they will [ ] The post Jevon’s Paradox and AI appeared first on Joseph Bou-Younes s Website .
The best, shortest and sweetest explanation of how an LLM works. The post The math behind an LLM appeared first on Joseph Bou-Younes s Website .
Today, no amount of model training is too much. No price for that training is too high. The builders bet is that inference will be orders of magnitude cheaper in 5 years. Build for that. Be Netflix or YouTube, not Qwest. This year, American tech companies will spend $300 billion to $400 billion on artificial [ ] The post Be YouTube, not Qwest appeared first on Joseph Bou-Younes s Website .