AI Reading for Sunday April 14
Open-source models are (very gradually) closing the gap with the best closed-source LLM models
Andrew Lo, leading MIT finance academic, says gen AI can help provide financial advice. - An MIT Exploration of Generative AI
Will gen AI make us stupid? People are already stupid at optimal decision-making, even smart people seem inexplicably prone to beg AI for a hallucination, asking it about stuff it has no reasonable basis or ability to answer. And now many are just going to Magic 8-ball everything and never learn how to make choices. - The Conversation
Cryptic comments from AI leaders suggest models might soon get better at real understanding and reasoning instead of just playing an imitation game. - The Conversation
Google releases a no-code AI Agent Builder in Vertex. - Geeky Gadgets
Anatomy of a bubble. - the Guardian
Does the AGI project reflect a desire to create God? - Business Insider
AI is helping fix up old movies and some people are unhappy. - NY Times
Is AI going to crush obscure languages? Or make them more accessible and amenable to translation? - The Atlantic
UK academic says some courses lost most of a year’s intake to AI-related expulsions. “They have clearly never even heard of some of the terms that turn up in their essays.” Sometimes “whistleblowing is career-threatening” as schools don't want to lose tuition. Well-founded alarm? Or hysterical moral panic? - FT
AI assisted humorous cartoon captions! - The University of Sydney
Follow the latest AI headlines via SkynetAndChill.com on Bluesky