Learning by building a multi-agent design lab. Yet, single agent works better (github.com) 2 points by benoit_cotte 7d ago ↗ HN
4 comments
[ 0.25 ms ] story [ 7.5 ms ] thread