David WeinbergerChatGPT on the couchIn which it acknowledges it just might be psychoticJul 24, 2023Jul 24, 2023
David WeinbergerGoogle Bard on why we should believe itYesterday I asked chatGPT about why it presents its answers with such confidence, especially since it also tells us to be careful about…Apr 25, 2023Apr 25, 2023
David WeinbergerchatGPT on why it pretends to know thingsI had a conversation with chatGPT this morning about its making factual claims while “understanding” that it has no knowledge of the world…Apr 24, 2023Apr 24, 2023
David Weinberger“ChatGPT, please improve The Gettysburg AddressThis morning I asked chatGPT to suggest revisions to a draft of an opinion piece I’m writing. About half of the ten it came up with were…Apr 1, 20231Apr 1, 20231
David WeinbergerTopic: The Supreme Court is hearing a case about whether section 230 exempts Google from…A thought from one of my favorite philosophers, Richard Rorty:Feb 22, 2023Feb 22, 2023
David WeinbergerinLevel Up CodingSemantic WordleThere’s a new version of Wordle called Semantle — not one that I “predicted” — that wants you to find the target word by looking not for a…Mar 28, 2022Mar 28, 2022
David WeinbergerinPeople + AI ResearchJust Different Enough: AI Recommendations, Newsfeeds, Jokes, and Civil SocietyIf “just different enough” were simply a quantitative measure of difference, recommendations would be easy. Alas, it is not.Apr 27, 20211Apr 27, 20211
David WeinbergerTV triumphs over theater. At last.This is the year in which the epic battle between theater and TV was settled. Television has won. Not everything, should, or will turn…Feb 9, 20211Feb 9, 20211
David WeinbergerinPeople + AI ResearchHashtags and ConfidenceWill we learn from ML to attach confidence levels to much of what we say? The rise of hashtags gives clues.Jan 21, 20212Jan 21, 20212
David WeinbergerinPeople + AI ResearchConfidence everywhereFor machine learning, expressing a lack of confidence can be a sign of strength.Oct 16, 20201Oct 16, 20201