Trending in Ed – AI with Digital CoHost Nancy | Episode 700

Within the 700^th episode of ⁠Trending in Ed⁠, host Mike Palmer welcomes again Nancy, our digital co-host, to delve into the most recent and most important developments on the earth of Synthetic Intelligence, seen by way of the essential lens of training.

This milestone episode marks the launch of a brand new sub-feed, ⁠Trending in Ed – AI⁠, which is able to focus particularly on the intersection of training and synthetic intelligence. The dialogue covers a spread of thought-provoking subjects, together with:

• Canny Centaurs and the Uncanny Valley: The hosts revisit ideas like “centaurs” from Superior Chess and the Uncanny Valley from robotics, discussing learn how to mix AI with human instinct and empathy to keep away from the pitfalls of know-how that feels “creepy” whereas maximizing human augmentation in instructing and studying.
• AI in Apply and Pop Tradition: The dialog explores AI’s quick functions in training, similar to customized tutoring methods akin to Baymax from Massive Hero 6, and accelerated ability growth by way of simulations, drawing parallels to Neo downloading abilities in The Matrix. The moral issues raised in Black Mirror concerning information privateness and algorithmic bias are additionally mentioned. Yeah, we hit some popular culture to maintain issues zeitgeisty!
• ⁠Sycophantic Drift⁠: We introduce the fascinating (and humorous?) idea of “sycophantic drift,” analyzing how AI assistants may unintentionally turn out to be overly agreeable and flattering, posing a problem for designing instructional AI that gives correct suggestions. Nancy marvels at how nice we’re at this!
• ⁠New AI {Hardware} and Surveillance⁠: Then we contact on the intriguing collaboration between Jony Ive and Sam Altman to develop new AI {hardware}, particularly the “AI surveillance medallion,” and the implications for privateness and steady information seize.
• Contrasting AI Philosophies: Mike and Nancy spotlight the differing views of AI leaders, contrasting⁠ Luis von Ahn of Duolingo’s “AI-first” technique⁠ (LinkedIn Submit), which views AI as a superior instructor, with ⁠Dario Amodei of Anthropic’s focus⁠ on AI security, moral alignment, and the potential for job displacement.
• ⁠Mary Meeker’s Latest AI Report⁠: The dialogue concludes with an evaluation of Mary Meeker’s current report on AI adoption, emphasizing its unprecedented pace, plummeting inference prices, and the “nice reshuffling” of job roles, signaling AI as obligatory infrastructure somewhat than an elective one.

⁠Subscribe to Trending in Ed⁠ wherever you get your pods to remain knowledgeable throughout the educational universe. Tune in to ⁠Trending in Ed – AI⁠ to realize deeper insights into the crucial discussions shaping the way forward for AI and its use in training and know-how.

   Fb   Twitter   Linkedin   Electronic mail