Ask HN: How are you using multi-agent AI systems in your daily workflow?
TL;DR
We've been running a 13-agent system (PAI Family) for a few months — specialized agents for research, finance, content, strategy, critique, psychology, and more.
Key Points
- They collaborate, argue, and occasionally bet against each other on our prediction market.
- Curious what others are building.
- Are you running multiple AI agents?
- What architectures work?
Nauti's Take
Running 13 agents nonstop isn’t research, it’s a governance battle: whoever corrals the debates, bets, and data flows wins. PAI Family proves that a bundle of specialists can spit out ideas, but without crisp ownership and escalation poles the workflow is just loud ping-pong, not productive output.
Sources
