If you work at OpenAI or Anthropic, how do you lay the foundation for social agents and collaboration? Make it grounded, fun, accessible. And maybe start by replacing every first and last initial profile photo with a pixel character.
This leak showcases Claude's upcoming release and a signal they're entering the social agent and collaboration space.
The last few weeks I've been deep on digital agent teams, or agent squads. We need a new term for it. Think: a digital agency of IC characters that manage many agents but represent you. The IC characters are made up. Said another way, your Claude or OpenAI account isn't really you, it's more of a made up persona. The closest analogy I have rn is running a small digital agency.
If you've known me for a while, or remember a previous adventure I built, think Mobile Cavalry. An agency with multiple digital teammates, and those teammates manage other agents. I'll write more on this later, but going back to the leak. It sorta stopped me in my tracks for 10 minutes. Reshaped my thoughts on where we're headed.
If you have any doubts about Anthropic or OpenAI trying to own consumption and generation, don't. They will, and it will be relentless.
Fear not. There's plenty of space. The trick I'm focused on is building outside of these core areas. Which at first glance seems impossible, until it hits you square in the face.