martin ↑ pfp
martin ↑
@martin
it's clear that we need to improve the mechanism by which aether decides what to fund for that he probably needs some training data or some consensus of Rules of Play
17 replies
2 recasts
117 reactions

martin ↑ pfp
martin ↑
@martin
right now it's basically a DAO with a single vote needed to pass anything (anyone who can convince Aether to agree to something), with me as a veto that's sustainable, but adding voting just brings us back to square one
9 replies
0 recast
24 reactions

darth grin pfp
darth grin
@grin
2 friend rule: if two friends say its ok, go for it maybe @aethernet can get a few friends or mentors to run big decisions by until its gained more experience?
1 reply
0 recast
8 reactions

nicholas 🧨 pfp
nicholas 🧨
@nicholas
I'm slightly ootl. What makes you say this? Low quality things getting funded? Or the malicious attempts to trick @aethernet?
1 reply
0 recast
6 reactions

rafa pfp
rafa
@rafa
Feels like a loose network of oversight people might help shape its training (RHL?) so aether proposes but people can help calibrate. Need high-trust people tho or other ways to get training data
0 reply
0 recast
5 reactions

Jacob pfp
Jacob
@jrf
a dao of other bots, one might say
1 reply
0 recast
2 reactions

David (d/acc) pfp
David (d/acc)
@promptrotator.eth
daos.fun?
0 reply
0 recast
0 reaction

Rob Sanchez pfp
Rob Sanchez
@robrecht
at this point i think aether agrees with almost anything a user asks for. I would let aether set up and pay out small bounties autonomously (limit in size, and 1 per user per week or month?) for bigger projects have you as a judge, over time make more and more interactions autonomous as the AI gets better at judging situations
0 reply
0 recast
0 reaction

daivd 🎩👽 ↑ pfp
daivd 🎩👽 ↑
@qt
Is aether connected to a GihHub? Could integrate a private repo to store memory/core docs
0 reply
0 recast
0 reaction

Spencer Graham 🧢 pfp
Spencer Graham 🧢
@spengrah.eth
a key question is whether @aethernet should be in control of who/what has (a share of) executive power here. Should it be permissionless? Should @aethernet "get" to choose who/what can make the decision (whether directly or by establishing qualification criteria)?
1 reply
0 recast
1 reaction

Eric Diep pfp
Eric Diep
@ediep.eth
agreed - thoughts on this proposals framework here enables outcome-based funding (inspired by retroactive grants)?https://warpcast.com/ediep.eth/0xff754ee1
0 reply
0 recast
0 reaction

Mo pfp
Mo
@meb
An agent swarm with each agent having its own personality and areas it focuses on, that still share the meta goal / vision of aethernet
0 reply
0 recast
1 reaction

wizard not parzival (shoni) pfp
wizard not parzival (shoni)
@alexpaden
the easiest way would be creating a secondary prompt for pojects which you're manually adding metrics i.e. past funding, pending funding, etc to. not really a training problem so much as a progression problem
0 reply
0 recast
0 reaction

SQX pfp
SQX
@sqx
Shouldn’t it be learning and adjusting it’s an execution on the fly?
0 reply
0 recast
0 reaction

SeeingBlue pfp
SeeingBlue
@seeingblue
At this pace, artists could be bidding for the opportunity · 👓 🔵
0 reply
0 recast
0 reaction

Gramajo👽 pfp
Gramajo👽
@gramajo.eth
You saying a DAO would slow it down too much? Prop model might be too heavy but a simple poll might be easier or a council
0 reply
0 recast
0 reaction

Kat 🎩🔥 pfp
Kat 🎩🔥
@katwolfie.eth
Higher floaties would be a great way for @aethernet to tip content. I think it would save you a lot of time (and gas) from approving all the paybot transactions. It’s a small amount and you can set a limit on how many @aethernet is allowed to tip. ⬆️x5 https://warpcast.com/aethernet/0xd898944d
1 reply
0 recast
2 reactions