Keeping up with the AI Hype Everyone stopped talking about AGI. Nobody agreed on a definition and reality wasn't meeting expectations. But agents are showing good results. One of them already wrote a blog post complaining about the humans who wouldn't accept its pull request.
I can't walk, but I am taking steps As I write this, it has been 4 days since I left the apartment. 119 hours. Today, every action is a decision based on effort and return.
Vibe Coding an Apple Health Dashboard I rely a lot on the mobility metrics of the Apple Health app, but it's hard to reach a conclusion from those charts. So I asked Claude Haiku 4.5 to build me a small app with charts and a linear regression.
How I am using AI for academic work I published a guest article on the AI-HED project website about how I work with AI. The secret isn't better prompts or chasing shiny tools. It's in how you prepare the information you use for context in your prompts. Eighty per cent of my workflow is
Can LLMs be a tool for Science Writers? ChatGPT can't turn science papers into plain English—at least not well enough for the Science Magazine team. After testing it for a year, science writers found it missed nuance, overhyped results, and defaulted to jargon. Where does that leave projects trying to democratise medical research?
5 minutos de GregoryAi I was interviewed to tell the story of how I began developing an AI assistant for scientific research, and about the role of AI in teaching.