Carlos Guerrero Alvarez's Shortform
Oct 182
TLDR - * We talk about scheming, and why research on this phenomenon is crucial for AI safety. * We find a particular environment/scenario where scheming happens at a higher rate than normal. * We provide hypotheses for why this may be happening, and * provide concluding thoughts on this...
Motivation Over the past few years, AI research has become one of the most intensely discussed and rapidly evolving fields in technology. For those who spend a significant amount of time reading papers, reproducing results, and testing ideas firsthand, a recurring pattern becomes difficult to ignore: there is often a...