John Howard

Treating engagement with AI as a philosophical exercise. Catherine Howe. https://www.curiouscatherine.info/2026/04/06/its-all-about-the-question/

“I'm lucky to have played a lot with such AI agents recently - the likes of OpenClaw, PicoClaw and Cowork. I started an experiment with one earlier today: what would they do if I asked them to tell me if the council tax band for our house was fair, in comparison to equivalent nearby houses?

It came back immediately to tell me our band was higher than all our neighbours and suggesting next steps. At this point I stopped it, as I'd realised something stark.”
https://www.linkedin.com/posts/tomloosemore_many-public-services-rely-on-friction-to-activity-7430177328124751872-WdOG?utm_source=share&utm_medium=member_ios&rcm=ACoAAAAACKQB7q_3-UCqi3TYbHDbO_wX2Sf1ZCM

This is why I’m not giving Anthropic more than £20 a month. It is enforced regulation of my attention, as left to my own devices I’d be down all the rabbit holes. Anyway I don’t have the spare cash! 🤣
https://simonwillison.net/2026/Feb/15/the-ai-vampire/#atom-everything

This mirrors my own experience working on a large test project with Claude. Setting aside all the other issues of code review (or lack thereof), it’s the cognitive impact that is increasingly difficult to manage. Pace is just as important as code quality! https://simonwillison.net/2026/Feb/15/cognitive-debt/#atom-everything

Final 0.5kg of seville oranges turned into 1.5kg of #sunshineinajar. Bringing to an end the 2026 production run of about 10kg that should last the next few years! #marmalade