You could specify desired outcomes and give it goals like: minimize budget, involve private sector as much as possible, seek competitive behavior, optimize health, environment, education. Surely someone is testing this out already!

You could also introduce some checks and balances like having a small representative committee from various parts of communities exercise judgement on the judgment of the AI agent to ensure it doesn’t do anything radical.

using corporation trained and developed AIs? not a chance. Unless we have free (as in speech, not as in beer) open source AIs, trained and maintained by the community, this is just putting microsoft and google in the government without questioning.

Those big corporate LLM "AIs" are all political and of course they will always be political as they only can produce what they were fed with. nostr:naddr1qvzqqqr4gupzq3huhccxt6h34eupz3jeynjgjgek8lel2f4adaea0svyk94a3njdqy88wumn8ghj7mn0wvhxcmmv9uq3uamnwvaz7tmwdaehgu3dwp6kytnhv4kxcmmjv3jhytnwv46z7qqdx...

The problem is and will always be definitions. Desired outcome (for who?) Optimal (what does this mean exactly? By what standard is something optimal?) Free markets are largely the answer to these questions. Government is meant to protect property rights, not optimize things for us. A fundamental...

I’ve been thinking a lot about this! Agent societies, each optimizing different objectives.