Show Notes
AI isn’t just a tool anymore—it’s forcing us to rethink ownership, trust and creativity. This week on Mixture of Experts, the team explores a wave of stories highlighting both the promise and the perils of AI adoption. We start with a surprising shift in public sentiment, as younger generations question AI’s impact on their futures—raising questions about control, opportunity and where humans fit in. Next, we dig into new Microsoft research showing that even top-tier models can corrupt data in complex workflows, and what that reveals about how (and when) to trust AI systems. Then, we explore Anthropic’s fix for Claude’s strange “blackmail” behavior, and why better data—not just better models—may be the key to safer AI. Finally, we debate whether AI may have quietly crossed a cultural milestone by helping win a literary prize—or whether humans are simply starting to sound more like machines. Join host Tim Hwang and AI experts Marina Danilevsky, Gabe Goodhart and Chris Hay to break it all down.
Visit Mixture of Experts podcast page to get more AI content → https://www.ibm.com/think/podcasts/mixture-of-experts
00:00 – Introduction
1:05 – AI at graduation
14:46 – LLMs corrupt documents
30:46 – Why Claude blackmails
43:09– ChatGPT-generated story wins literary prize
The opinions expressed in this podcast are solely those of the participants and do not necessarily reflect the views of IBM or any other organization or entity.
Subscribe for AI updates → https://www.ibm.com/forms/news-mkt-52954
More from Mixture of Experts
Visit Mixture of Experts podcast page to get more AI content → https://www.ibm.com/think/podcasts/mixture-of-experts What happens when an AI model can copy itself onto the devices in yout hoome?
Visit Mixture of Experts podcast page to get more AI content → https://www.ibm.com/think/podcasts/mixture-of-experts This week on Mixture of Experts, host David Zax is joined by Kaoutar El Maghraoui, Nick Gallagher, and Volkmar Uhlig. First, Microsoft, OpenAI, Perplexity, and Cloudflare follow TypeSafe’s Jev in building decision models, a new category of AI built to make fast, scored judgments. The panel debates whether these models are truly new or a more efficient application of existing technology, and whether better-calibrated confidence scores can help address the overconfidence problem in LLMs. Next, a new study finds that teams of AI agents can cost two to five times more than a single agent without consistently delivering better results. The panel discusses when multiple agents are worth the added cost and complexity. Finally, Anthropic updates its usage policy to prohibit “sustained and needless abusive or cruel behavior” toward its models, raising a bigger question: Is this about protecting AI, shaping human behavior, or something else entirely? The panel considers the ethical implications and whether the policy could also be a marketing play.
Visit Mixture of Experts podcast page to get more AI content → https://www.ibm.com/think/podcasts/mixture-of-experts
Visit Mixture of Experts podcast page to get more AI content → https://www.ibm.com/think/podcasts/mixture-of-experts
Visit Mixture of Experts podcast page to get more AI content → https://www.ibm.com/think/podcasts/mixture-of-experts