I'm working next week on what I call user aligned computing.
Video https://www.youtube.com/watch?v=XQgtVdyNzaQ code here https://github.com/eb4890/agorint.
It might be a bit like the control problem for normal computers (but mainly with a separate evolutionary pathway ), it doesn't assume the thing it is trying to control is a super intelligent genie with its own goals. It is a user controlled market which determines the programs the systems access to resources such as memory and processing power and also I/O.
I am hoping might be part of program of Intelligence Amplification and it would make computers more secure in general (less of a monoculture with easily acquirable ambient authority) so might have impact on some fast take off scenarios.
I'm about to start making the market work, having got some basic infrastructure working.
From Raemon's Project Hufflepuff thread, I thought it might be helpful for there to be a periodical check-in thread where people can post about their projects in the spirit of cooperation. This is the first one. If it goes well, maybe we can make it a monthly thing.
If you're looking for a quick proofread, trial users, a reference to a person in a specific field, or something else related to a project-in-progress this is the place to put it. Otherwise, if you think you're working on something cool the community might like to hear about, I guess it goes here too.