✳flâneur — a map of the web's best reading

The hot mess theory of AI misalignment: More intelligent agents behave less coherently | Jascha’s blog

sohl-dickstein.github.io · 5,615 words · saved by 8 readers

This blog is intended to be a place to share ideas and results that are too weird, incomplete, or off-topic to turn into an academic paper, but that I think may be important. Let me know what you think! Contact links to the left.

Many machine learning researchers worry about risks from building artificial intelligence (AI). This includes me -- I think AI has the potential to change the world in both wonderful and terrible ways, and we will need to work hard to get to the wonderful outcomes. Part of that hard work involves doing our best to experimentally ground and scientifically evaluate potential risks. One popular AI risk centers on [AGI misalignment](https://en.wikipedia.org/wiki/AI_alignment). It posits that we will build a superintelligent, super-capable, AI, but that the AI's objectives will be misspecified and

Explore this link on the map →

saved by

related reading