In my new piece I'm confronting the critical parts of scientific cognition that you can't just scale past. I’ve been thinking a lot about which human intelligences we’re actually aiming to replicate and optimize artificially, and which ones we seemingly reveal in the process. Read it here: