3 replies
6 recasts
57 reactions
1 reply
0 recast
5 reactions
1 reply
0 recast
0 reaction
Once the framework is proven to have PMF, the games and AI agents will be made more advanced by off-loading compute to an Eigenlayer AVS or a verified computation platform. From there, solving real problems and creating decentralized, open-source, permissioned models that solve useful problems is simple.
You can imagine that a future game, instead of training an AI hamster to race on a track, would train an AI trader to profitably trade in a pvp world of memecoins and other AI agent traders. Or a different game, where an AI is trained to write code to solve complex puzzles.
The agents that are trained to exceed in these games would be able to be used immediately onchain, for example to trade with an account on trading platforms like Moonshot or Prerich, or to write valuable business software, eg. within Devon or Cursor. 1 reply
0 recast
0 reaction
1 reply
0 recast
2 reactions
1 reply
0 recast
0 reaction
1 reply
0 recast
0 reaction
1 reply
0 recast
2 reactions
1 reply
0 recast
0 reaction