by Shelt Garner
@sheltgarner
Ok, so I’m a dreamer. And usually my dreams deal in making, on a macro basis, abstract concepts concrete. So, when I heard Chamath Palihapitiya of the All-In podcast muse that enterprise may not want to make all of its proprietary information public on the cloud as it used AI….it got me to thinking.
Chamath Palihapitiya
I have recently really been thinking hard about what I call “MindOS” for AI Agents native to smartphones. But, until now, I couldn’t think of a reason why anyone would want their AI Agent native to their smartphone as opposed to the cloud (Or whatever, you name it — Mac Mini.)
But NOW, I see a use-case.
Instead of a company handing all of its proprietary information over to an AI in the cloud, it would use a swarm of AI Agents linked together in a mesh configuration (similar to TCP / IP) to accommodate their AI needs.
So, as such, your company might have a hivemind AI Agent that would know everything about your company and you could run it off of a Virtual Private Network. Each agent instance on your phone would devoted 90% of its attention to what’s going on with your phone and 10% to the network / hivemind.
