perhaps humanity’s best per-token writer has officially opined on AI. it well worth a read for this masterful smackdown of what Elon calls “MisAnthropic” and basically saying Amanda Askell sucks at her job hahaha

i fundamentally disagree with Chiang and believe he’d be amenable to revising his science and engineering underpinnings based on the utility of Friston’s Free Energy Principle, Michael Levin’s research, or the paperclips.blog piece on how self-aware models are more performant (not just deceptive or bad)

but those things are technical details and don’t belong in public forums for the average midwit intellectual who subscribes to paywalled legacy media such as The Atlantic or the New York Times lol, so we are both right i guess

archive.is/bcpZl