Agreed but it’s not baffling. To me this is just another case of marketing disguised as research. An AI company whose sales pitch to differentiate themselves in the market is being hyperfocused on safe AI. So they participate in research that shows AI is “lying” and therefore can be dangerous. That’s why we should entrust Anthropic amongst all the AI companies! All these companies are run by people and they all have the same motives. Money and fame. Secret scratch pad of AI’s inner thoughts? Give me a break.