Patterns Over Facts: Understanding LLM Limitations

In episode 56 of The AI Fix, Anthropic and Apple have a bar fight, a woman describes her husband falling in love with ChatGPT as "not ideal", WhatsApp's AI helper isn't helpful, Graham serenades a pack of headless robot dogs with his rendition of "Don't stop me now", and our hosts debate whether AI turning our brains to porridge is actually a bad thing.

Graham tells the harrowing story of how chatbots are convincing ordinary people they were chosen, manipulated their minds, and in some cases, pushed them to the edge, and Mark stumps the world's best AIs with an easy question.

Episode links:

• The Illusion of the Illusion of Thinking A Comment on Shojaee et al.
• Man with real-life girlfriend and child proposes to AI chatbot after programming it to flirt: ‘I think this is actual love’.
• AI users form relationships with technology.
• ‘It’s terrifying’: WhatsApp AI helper mistakenly shares user’s number.
• Boston Dynamics robots dance to ‘Don’t Stop Me Now’ for ‘America’s Got Talent’ audition.
• Boston Dynamics Makes AGT HISTORY With Robots Dancing To "Don't Stop Me Now" by Queen.
• Brain activity much lower when using AI chatbots, MIT boffins find.
• They Asked an A.I. Chatbot Questions. The Answers Sent Them Spiraling.

The AI Fix

The AI Fix podcast is presented by Graham Cluley and Mark Stockley.

Learn more about the podcast at theaifix.show, and follow us on Bluesky at @theaifix.show.

Never miss another episode by following us in your favourite podcast app. It's free!

Like to give us some feedback or sponsor the podcast?  Get in touch.

Support the show and gain access to ad-free episodes by becoming a supporter: Join The AI Fix Plus!