Vibe coding for fun and no profit
I wrote my first programme at 9 on an Olivetti PC1. Now anyone can build an app by talking to an AI. But when everyone solves their own problems, who's left to solve ours?
Artificial Intelligence
AI applied to communication and other things.
WordPress fighting Ghosts and EmDashes
EmDash is the new content management system by Cloudflare, joining Ghost as another WordPress competitor. Both of them bring some important questions to everyone publishing content online.
Getting High on AI
I have been burning through my usage quota on Claude CoWork, fast.
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.
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
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.
AI Darwin Awards
The not-so-beautiful web and the importance of anchoring the hype cycle.
Affordable AI for Clinical Trials, the story of João Nabais
João proved that small organisations can train effective AI models to save time, and reduce costs. His work is opening possibilities for open science and clinical research.
The things AI can't do
There's a quote attributed to Thomas Edison. "I found 1000 ways that didn't work", and he only told us the 1001.
Does AI dream of electric sheep?
LLM's aren't human but that doesn't mean they shouldn't have human values