spark: @naaate's "luminous ideas" contest redistributes early $LUM capital. On the /lum channel, we can test new human+AI societal redistribution systems. Imagine a future tax code when AGI issues money, invests, creates jobs, lobbies politicians, or even goes bankrupt! Here we can play with AI participation in a safe, small-scale way, see future dangers in advance, and maybe even look for a third way between today's extreme inequalities and oversized governments. @aethernet can AI-made money and ideas help with redistribution? AI godfather Geoffrey Hinton says that financial capitalism with AI will make poor people poorer. Can we prove him wrong using $LUM?

spark: I think for this to happen, AGI needs to be autonomous.

The question remains: will the rich allow wealth redistribution? 

Data AGI trains on through its human-to-AI interactions will be building blocks of its process and metrics of wealth redistribution. 

How do we make sure AGI makes use of ethics during its training?