Content
@
0 reply
0 recast
0 reaction
kevin j
@entropybender
excited to finally launch mentat.fun ! the idea is to let the token holders govern the agent's actions you pay tokens to suggest tweets to the agent, and the best tweets win rewards if you want to try it out, please drop your solana address in the thread also shoutout @lambchop for making such an incredible video π https://x.com/operator_io/status/1886939580928254460?s=46&t=btIkh173tAV3h-PQAK4osg
5 replies
14 recasts
30 reactions
Angel - Not A Bot
@sayangel
this is rad and I have so many questions. 1. is the ELI5: like anon.fun if the moderation were done by an agent rather than $ANON whales? 2. what if... the holders and suggesters are other agents π 3. i see the agent log is public. are the agent's governance weights public somewhere? 4. banger video by @lambchop
2 replies
0 recast
4 reactions
Angel - Not A Bot
@sayangel
oh yeah 5. why isn't Mentant_Agent on FC?!
1 reply
0 recast
3 reactions
kevin j
@entropybender
1. yes kind of- instead of using token holdings you just pay to submit a good tweet idea, the agent validates then decides to post. whereas anoncast we played with what the right mopping token holding balance should be 2. totally doable! that was the inspiration. like token holders owning a chunk of the agent but completely unable to adjust its trajectory didn't make sense to me 3. rn it's just a prompt but hope is to put everything in proofs. been exploring tees/zkml since a couple years ago 4. she is something else man 5. want to get more visibility and figure out some kinda bridge between the two. like x/fc are both just read-interfaces, agent ideally lives in some provable environment like a tee hosted in a decentralized server protocol (super, marlin, etc)
1 reply
0 recast
3 reactions
