Monthly Roundup: Deregulation, Hardware, and Inference Scaling

Ben Lorica

Ben Lorica and Paco Nathan on Immigration & Regulatory Policy, Hardware, and Model Insights.

Subscribe: Apple • Spotify • Overcast • Pocket Casts • AntennaPod • Podcast Addict • Amazon •  RSS.

This is our monthly conversation on topics in AI and Technology with Paco Nathan, the founder of Derwen, a boutique consultancy focused on Data and AI. This episode begins with an analysis of potential shifts in US immigration policy and its impact on skilled tech workers, followed by a look at the contrasting regulatory approaches to AI in the US and EU. The conversation then explores the need for balanced deregulation, ensuring it doesn’t favor specific companies or compromise safety. A deep dive into the AI chip market reveals Nvidia’s continued dominance, while also highlighting the potential of AMD and other players. The episode also examines DeepSeek’s innovative approach to model development through resource constraints, and the implications of reasoning-enhanced models on inference scaling. Finally, the discussion touches on the rise of multimodal models and their potential impact on end-users, as well as the need for better multimodal understanding.

Subscribe to the Gradient Flow Newsletter

Audio version

 

Interview highlights – key sections from the video version:

 

Recommendations:

 

Related content:

If you enjoyed this episode, please support our work by encouraging your friends and colleagues to subscribe to our newsletter: