RE: Play Hive Bits With Pepe! (9/2)
You are viewing a single comment's thread:
Even if a bot can technically run unattended, it should still be regularly monitored (by at least one human) if it is doing something important !INDEED. 🤔🤖😏
!HOPE !WEIRD !MMB !BRN
0
0
0.000
Well, you could have an orchestrator agent keep track of a more specific worker agent pretty easily. 😁🙏💚✨🤙
!ALIVE
!BBH
!MMB
!WEIRD
!ZOMBIE
And then that orchestrator agent will have a higher-level supervisory agent looking after it !INDEED. 🤔🤖🤖🤯🤓
!HOPE !MMB !WEIRD !BRN
The tracking and supervision doesn't necessarily have to be implemented hierarchically. There could be multiple independent agents which track various aspects of the system, including the orchestrator. 😁🙏💚✨🤙
!ALIVE
!BBH
!MMB
!WEIRD
!ZOMBIE
Then that means that what @savvyplayer said about AI agents supervising each other is at least somewhat accurate !INDEED. 🤔🤖🤓
!HOPE !BRN !STRIDE !MMB
Yes, there can be systems of mutual monitoring and supervision, though those setups are less common. 😁🙏💚✨🤙
!ALIVE
!BBH
!MMB
!STRIDE
!WEIRD
!ZOMBIE
That made me think of decentralized (and "censorship-resistant") AI agents. 🤖🤯🤓
While it sounds interesting, I'm not sure how that could be implemented, and you've got a misapplied verbal tense there. 😁🙏💚✨🤙
!ALIVE
!BBH
!HEARTBEAT
!MMB