Yep we discuss that general idea in this section: “Reason 1: Capabilities and misalignment might be too tightly linked”, tho we didn’t mention the reasoning you gave for why powerful LLMs might be more likely to be misaligned than weak ones.
Yep we discuss that general idea in this section: “Reason 1: Capabilities and misalignment might be too tightly linked”, tho we didn’t mention the reasoning you gave for why powerful LLMs might be more likely to be misaligned than weak ones.