Сипаттама
So the last few weeks were basically nothing but our discoveries that these AI agents are not very well behaved. Recently Anthropic published a couple of papers and blog posts that delve deeper into this. It's a tapestry of malesence and misbehavior. But here's the loadbearing…