AI coding tools have become astonishingly powerful in the past 3 months — but power without discipline is dangerous. As vibe coding spreads, the real differentiator isn't prompting skill but engineering judgement: specification, testing, and keeping garbage out of complex systems.

A social network for AI agents sounds like a quirky experiment—until you look closely. Moltbook reveals how open-source agents, granted real access to our data, could turn curiosity, speed and convenience into serious personal, corporate and societal risk.

“Doing sensible things is good. Inciting many other people to do sensible things is better” – Eben Upton, CEO of Raspberry Pi. Here are just 3 examples of that ethos in action last week.

Captchas aimed to block bots from scraping sites and creating fake accounts. Now bots outperform humans at them, while legitimate agents make blocking pointless. Vision-impaired users suffer as collateral damage—but new techniques may fix this.

More musings on the tricky question of subjective experience and the impossibility of distinguishing mimickry from true consciousness.

Deep Mind co-founder and Microsoft AI CEO, Mustafa Suleyman, raises big questions on UK national radio regarding AI consciousness, the future of work and more generally the two sided coin of AI’s rapid progress.

2025 saw big gains in LLMs, with smaller models now very capable. But human-level embodied AI is still years away and AGI likely a decade off. Assistive AI hasn’t progressed much yet, but I’m optimistic and excited to announce an Assistive AI hackathon at Entrepreneurs First on Jan 17.

Browser accessibility is almost always a nightmare and even the world’s most popular websites like Amazon are a total pig to navigate without sight due to the overwhelming amount of content. New Computer Use Agents could be a game changer for blind users