About Whizzy Ideas

AI news and articles by Marko Balabanovic

Hi, I'm Marko Balabanovic. A long time ago I studied AI, in the 1990s when things were very different. I've enjoyed seeing it develop. Even recently, noone guessed that machines trained to predict the next word in a sequence would be so ridiculously useful, surprising, terrifying and powerful. Nowadays I am doing my best to empower nation-scale health research at Our Future Health but still keeping up with AI. This site, blog.whizzyideas.com, is a place where I'm collecting the interesting AI news I see and collecting thoughts and experiments.

I post a few times a month, usually a roundup of things I've found thought-provoking, and occasionally longer pieces when something deserves more than a paragraph. Topics that keep coming up: AI safety and alignment, the philosophy of what these systems are, AI in creative work, and increasingly agentic engineering and coding.

You can also find me on LinkedIn

Here's the most recent 5 posts:

• 29 Mar, 2026 Agentic Engineering: 5 lessons from a 4-day AI experiment
• 07 Mar, 2026 Military LLMs; Stop Killer Robots; LLMs going to great lengths to cheat; AI re-implementing open source libraries
• 22 Feb, 2026 Transformation of work; OpenAI in a pickle; Training customer support agents with the Diplomacy board game
• 15 Feb, 2026 Gas Town satire; AGI is already here; Consciousness explained; Claw on a Pi
• 08 Feb, 2026 Views from the maker of OpenClaw and the author of Claude's constitution; New York in isometric pixel art; Fun robots and security vulnerabilities