Take a deep breath and join Alex and Emily in AI Hell itself, as they take down a month's worth of hype in a mere 60 minutes.This episode aired on Friday, May 5, 2023.Watch the video of this episode on PeerTube.References:Terrifying NEJM article on GPT-4 in medicine“Healthcare professionals preferred ChatGPT 79% of the time”Good thoughts from various experts in responseChatGPT supposedly reading dental x-raysChatbots “need” therapistsCEO proposes AI therapist, removes proposal upon realizing there’s regulation:https://twitter.com/BEASTMODE/status/1650013819693944833 (deleted)ChatGPT is more carbon efficient than human writersAsking disinformation machine for confirmation biasGPT-4 glasses to tell you what to say on dates, "Charisma as a Service"Context-aware fill for missing data“Overemployed” with help from ChatGPTPakistani court uses GPT-4 in bail decisionChatGPT in Peruvian and Mexican courtsElon Musk’s deepfake defenseElon Musk's TruthGPTFake interview in German publication revealed as “AI” at the end of the articleYou can check out future livestreams at https://twitch.tv/DAIR_Institute. Follow us!Emily Twitter: https://twitter.com/EmilyMBender Mastodon: https://dair-community.social/@EmilyMBender Bluesky: https://bsky.app/profile/emilymbender.bsky.social Alex Twitter: https://twitter.com/@alexhanna Mastodon: https://dair-community.social/@alex Bluesky: https://bsky.app/profile/alexhanna.bsky.social Music by Toby Menon.Artwork by Naomi Pleasure-Park. Production by Christie Taylor.
In this episode, we talk to Mar Hicks, an Associate Professor of Data Science at the University of Virginia and author of Programmed Inequality: How Britain discarded Women Technologists and Lost its Edge in computing. Hicks talks to us about the lessons that the tech industry can learn from histories of computing, for example: how sexism is an integral feature of technological systems and not just a bug that can be extracted from them; how techno-utopianism can stop us from building better technologies; when looking to the past is useful and when it's not helpful; the dangers of the 'move fast and break things' approach where you just build technology just to see what happens; and whether regulatory sandboxes are sufficient in making sure that tech isn't deployed unsafely on an unsuspecting public.