New AI models, token minimization and IBM’s new sub-1nm chip

Learn more about IBM’s new chip technology → https://research.ibm.com/blog/sub-1nm-node-chips

IBM debuts world’s first 0.7nm chip technology! This week on Mixture of Experts, Sascha Brodsky sits down with Huiming Bu, VP of Silicon Technology R&D for a special segment diving into a major semiconductor breakthrough. Then, host Tim Hwang is joined by Abraham Daniels, Gabe Goodhart and Martin Keen. Our expert panel debriefs the new AI models: Sakana Fugu, which reports higher performance than Claude’s Mythos, and Z.ai’s GLM-5.2. Next, we discuss Google DeepMind’s recent collaboration with A24 and what this means for the future of AI in the entertainment industry. Finally, we’ve talked tokenmaxxing and now there is a new phenomenon: tokenminning. What’s the most important outcome with using AI, and how should we measure it? All that and more on today’s packed Mixture of Experts!

00:00 -- Introduction

1:11 -- IBM's new sub-1 nm chip

14:02 -- New AI models: GLM-5.2 & Sakana Fugu

29:48 -- AI in the film industry

38:41 -- Tokenminning

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.

Tokenmaxxing is dead, learn more here → https://www.ibm.com/think/insights/tokenmaxxing-dead-long-live-valuemaxxing