Seth Herd comments on Capabilities and alignment of LLM cognitive architectures

[Seth Herd]

I hadn’t seen your post. Reading it now.