AI take-over (though this one also tackles the risks of job automation)
Improve the quality of the content. Some ideas: Include the most recent developments in AI safety research (eg cite ARCHES, Rohin Shah’s work, and many many others), include updates on timelines with the recent reports from OpenPhil. It would also be great to make a few of the concepts clearer, for example the “sources of risk”.
Brilliant idea!
For AI existential risk, there are a few improvements that would make a big difference:
Merge relevant pages, because we now have
Existential risk from artificial intelligence
The AI Control Problem
AI take-over (though this one also tackles the risks of job automation)
Improve the quality of the content. Some ideas: Include the most recent developments in AI safety research (eg cite ARCHES, Rohin Shah’s work, and many many others), include updates on timelines with the recent reports from OpenPhil. It would also be great to make a few of the concepts clearer, for example the “sources of risk”.
(May those be good projects for some of the groups of the AGI safety program and the AI long-term governance program?)