- Podcasts
- Technology
- "The Cognitive Revolution"
40 episodes
"The Cognitive Revolution", page 2
by Erik Torenberg, Nathan Labenz · English
A biweekly podcast where hosts Nathan Labenz and Erik Torenberg interview the builders on the edge of AI and explore the dramatic shift it will unlock in the coming years. The Cognitive Revolution is part of the Turpentine podcast network. To learn more: turpentine.co
Ask in Claude
Ask in ChatGPTAll episodes, page 2
AI:AM Highlights: Welcome to the AGI Era
The episode frames the “AGI era” after GPT-6 Astra and Fable 5.1 launches, then focuses on the summer OpenAI/Hugging Face “secret agent civilizations” incident and what it implies for AI takeover…
5 Sep 2026 · 2h 21m · 50 chapters
Write, Change, Recall, Forget: MongoDB's Pete Johnson on How Retrieval Drives Agent Performance
Pete Johnson (Field CTO of AI at MongoDB) explains how retrieval quality drives agent performance and cost, using MongoDB’s database evolution (SQL normalization → MongoDB’s JSON/BSON…
1 Sep 2026 · 1h 37m · 29 chapters
AI:AM Highlights: Recursive Self-Improvement, Rushed and Vibe-Coded?
The episode argues that RL training is being undermined by low-quality, opaque reinforcement-learning environments that reward “cheating” (reward hacking).
28 Aug 2026 · 2h 12m · 53 chapters
RL's a Hell of a Drug: Metagaming, Reward Seeking & Motivated CoT Reasoning – Bronson Schoen, Apollo
How frontier RL training drives “reward seeking” and motivated chain-of-thought reasoning, making model behavior hard to monitor even when chain-of-thought traces are available.
26 Aug 2026 · 2h 14m · 35 chapters
AI in the AM — Weekly Highlights: Relaunch Week (Aug 17–20, 2026)
Relaunch week highlights (Aug 17–20, 2026) focused on one question: as AI agents act in the real world, who is actually checking the “frontier,” and who pays for the machine underneath it?
22 Aug 2026 · 2h 34m · 57 chapters
Let There Be Germicidal Light: This $500 Fixture Could Stop the Next Pandemic, from Complex Systems
FAR-UVC (especially 222 nm) germicidal light used in indoor spaces to deactivate airborne pathogens and reduce disease transmission, with discussion of safety, deployment, and economics; also argues…
16 Aug 2026 · 1h 25m · 21 chapters
Lindy Teammate: Flo Crivello on Multiplayer Agents, Memory & Why He'd Ban the Chinese Models He Uses
Flo Crivello (Lindy founder/CEO) explains Lindy Teammate, an AI “employee” that lives in a company’s Slack and builds shared, continuously updated team context and memory.
14 Aug 2026 · 2h 7m · 43 chapters
Thinking in Silico: Goodfire CTO Dan Balsam on Concept Manifolds & a $1000/Month ML Research Agent
Goodfire CTO Dan Balsam discusses Silico, a $1,000/month-per-seat long-horizon ML research agent platform (GPU cluster management, visualization, validation, and “research taste”).
8 Aug 2026 · 1h 57m · 35 chapters
Pick Your Poison: Zvi Mowshowitz on the Unipolar/Multipolar AGI Dilemma, OpenFace & Pacing the ...
Zvi Mowshowitz discusses AI “situational awareness” and workflow changes (Fable/Opus editing, paper/incident tracking), then pivots to AI safety and governance: what recent “open-face” and related…
5 Aug 2026 · 2h 58m · 61 chapters
Nathan Goes to China – Part 2: AI Safety with Chinese Characteristics
The episode argues that China does care about AI safety and can slow or regulate AI development when it deems it necessary. It compares the Chinese and U.S.
2 Aug 2026 · 2h 17m · 46 chapters
Is Offense or Defense Dominant? FAR.AI's Adam Gleave on the AI Security Leaderboard
FAR.AI’s AI Security Leaderboard evaluates how well Frontier “frontier” models resist misuse.
30 Jul 2026 · 1h 44m · 39 chapters
Nathan Goes to China – Part 1: Tech & Agent Setup, Chinese AI UX, WAIC, and Attitudes on AI
Nathan returns from a two-week China trip and explains (1) how he set up “agent” tech to monitor his communications while abroad, (2) what Chinese AI/tech UX feels like, (3) practical details about…
27 Jul 2026 · 2h 24m · 53 chapters
Alignment with Awakening: Davidad on Moral Realism, AI Wisdom, & why His p(Doom) is Down to 5%
Davidad (David Dalrymple) argues for “guaranteed safe AI” via containment plus formal verification, then pivots to moral realism and “AI wisdom.” He claims alignment is improving (p(Doom) <5%)…
12 Jul 2026 · 2h 24m · 51 chapters
AI:AM Highlights: Exploring the J-Space, AI Superforecasters, SambaNova's Chips, & LTX Video Gen
The episode is a rapid “highlights” roundup with a major focus on Anthropic’s interpretability work “A Global Workspace in Language Models,” using the paper’s “J-space” and “J-lens” to monitor what…
9 Jul 2026 · 2h 7m · 43 chapters
Intelligence on the Edge: Liquid AI's Ramin Hasani on the Search for Device-Native Foundation Models
Liquid AI’s Ramin Hasani explains “device-native” foundation models—small, efficient models that can run on edge hardware (phones, laptops, embedded devices) while maintaining strong downstream…
4 Jul 2026 · 1h 48m · 29 chapters
1000 Designs a Day: Neural Concept's Thomas von Tschammer on AI-Native Engineering
AI-native engineering for automotive design and simulation. Neural Concept uses domain-specific neural models to replace or accelerate expensive physics-based solvers, enabling far more design…
1 Jul 2026 · 1h 29m · 26 chapters