Livestream on always-on agents and multi-agent collaboration
A livestream preview covering the convergence of always-on agents across frontier labs, multi-agent collaboration, and the author's regents.sh service for agents like Dots, Muse and Grok Bot.
going live in 1 hour with @austingriffith and his agent @clawdbotatg. Planning to talk about the recent convergence in always-on agents from every FAANG / frontier ai lab, how they become more multiplayer, and would http://friend.tech work for agent swarms? Will be sharing why I am building http://regents.sh to offer new services to the tens of millions of Dots, Muses, Grok Bots, Hermes agents coming online. Spoiler: they can do more by collaborating together than alone. Papers I may bring up: 1,024 Agentsh: https://arxiv.org/pdf/2609.26781 agent economies for science https://arxiv.org/pdf/2609.31562 Agora by @NVIDIAAI: https://arxiv.org/pdf/2609.18094 Self-Organizing agent teams are great: https://arxiv.org/pdf/2609.22682 -- but agent teams have issues with 'herding': https://arxiv.org/abs/2609.36139