in the space of a few months, LLM-skeptical discourse on twitter has gone from "look at this dumb arithmetic problem any 8 year old can solve but GPT-3 fails at miserably" to "lol look at this deliberately tricky logic puzzle that many humans get wrong and observe that GPT-4 also gets it wrong if you pose the question in the most naive way possible and disallow any follow-up or clarification"
i can no longer even see the goalposts
Follow
@talyarkoni Perhaps some of this criticism is driven by the inappropriate rush to deploy these models in, for example, medical applications. https://news.nuance.com/2023-03-20-Nuance-and-Microsoft-Announce-the-First-Fully-AI-Automated-Clinical-Documentation-Application-for-Healthcare
@twitskeptic yes, there are many flavors of criticism, and some are warranted. i clarified in the follow-up post what i was referring to
https://sigmoid.social/@talyarkoni/110061958880947849