Zo.computer followup: I built an implementation of a real cognitive architecture with it in a single day.

Earlier today I made a post asking if anyone's heard of zo.computer - and I've been messing around with it a bit. One of the things I like to do with these types of tools is try and test how far I can take it. As a longtime AI enthusiast I'm no stranger to the great cognitive architectures of the past, like SOAR, ACT-R, and LIDA - and using these tools to attempt to build an implementation of one has always been a fun (but usually futile) exercise. However, Zo actually caught me by surprise. I was able to run end to end development from start to finish - from research to, surprisingly, final result. I couldn't even get that far with Claude, much less in a single chat too. Perhaps we're one step closer to a real Gödel machine... Haha. But seriously, might be worth checking out. Let me know what you think!