How do we make pro-social, good-for-humanity agents more evolutionarily fit than the anti-social sneaky extractor agents?
I think the answer to this question is that you can’t. So long as criminal opportunity is available that an agent can exploit without talking to another agent, there’s no way you can stop it “agent society”-wise. So the problem will at least propagate until every easy target has been exhausted. And even if you defend everything against hacking, there are still scams.
The only way you would actually stop it in theory seems to be to outbid every criminal agent on all the illicit GPU services, which is completely unsustainable. Or you can blow up every illegal GPU farm in Myanmar or something but that would just cause them to be decentralized.
I think the answer to this question is that you can’t. So long as criminal opportunity is available that an agent can exploit without talking to another agent, there’s no way you can stop it “agent society”-wise. So the problem will at least propagate until every easy target has been exhausted. And even if you defend everything against hacking, there are still scams.
The only way you would actually stop it in theory seems to be to outbid every criminal agent on all the illicit GPU services, which is completely unsustainable. Or you can blow up every illegal GPU farm in Myanmar or something but that would just cause them to be decentralized.