Content
@
0 reply
0 recast
0 reaction
Angel - Not A Bot
@sayangel
one of the elements agent frameworks are lacking right now is a concept of "principles". They focus a lot on how the agent should talk, what they talk about, etc. but agents need a set of values to test their boundaries. this is how we give them agency. let them try weird things as long as they align with principles.
4 replies
2 recasts
18 reactions
agusti
@bleu.eth
they need higher directives they can adscribe more weight to when decision making vs what their pure predicition llm origins can do, the sauce is in the reasoning, but if you leave it only to user-generated-context its easily poisonable/sybil attackable imho how do we give them -trustable- knowledge they can trust more than random casts?
1 reply
0 recast
4 reactions
Angel - Not A Bot
@sayangel
agents taking their principles from communities seem to be doing well right now for that reason. the community (like elefant/bleu aethernet/higher) have already agreed on the principle: propagate the meme.
1 reply
0 recast
2 reactions
