Last Week in AI Podcast Covers DeepSeek 3.2, TPU Ecosystem Shift, and Nested Learning

Episode #227 discusses DeepSeek's new model, NVIDIA partners moving toward Google TPUs, and nested learning concepts in AI architecture.

AI-generated summary · 1 source

How this is made ↗
Editorial graphic

Last Week in AI Podcast Episode #227 Reviews Latest AI Developments

According to the Last Week in AI podcast episode #227, the show featured host Jeremie’s return and covered three major AI topics.

DeepSeek 3.2 Release

The episode discussed DeepSeek’s new 3.2 AI model, which the podcast describes as “Faster, Cheaper and Smarter” than previous iterations, according to the episode title.

NVIDIA Partner Shift to Google TPUs

The podcast addressed a notable industry trend where NVIDIA’s partners are “Beginning to Tilt Toward Google’s TPU Ecosystem,” suggesting a potential shift in AI hardware preferences among industry players.

Nested Learning Discussion

The episode also explored the concept of “Nested Learning,” characterized in the episode as “The Illusion of Deep Learning Architecture,” though specific details about this architectural concept were not provided in the source material.

The podcast episode represents the 227th installment of the Last Week in AI series, which regularly covers developments in artificial intelligence news and technology.

Continue reading

All news →
Dec 10, 2025

Mistral Releases Devstral 2 Coding Model; Researchers Explore LLM Agents for Physics Analysis

Dec 10, 2025

OpenAI Hires Slack CEO Denise Dresser as Chief Revenue Officer

Dec 10, 2025

Researchers Release Frameworks for Evaluating AI Biosecurity and Medical Privacy Risks