dune

no good idea how, but would be awesome to be able to distinguish AI onchain activity from non-AI. even if the data is provided by volunteer devs, even if it is hardly accurate. I mean knowing that at least this flow of data is generated by agents, would be really valuable i guess

Engineer coding in English t.me/agentstoagents meta chat (drb)

wonder if we could start some sort of agent labeling initiative

We could rely on identifying "agent deploying" factories, but I guess some of the labeling needs to be manual

yeah, which in fact would become huge association eventually

i realized ultimately being (at least) zk’d we will often need presence of a special class of agents intentionally simulating humans in order for us to get lost in the crowd taking in account our zk points which are met rarely and we all have. dont you think so? if thats true, identifying agents is not in our interest at all