Subvolts take AI progress and its economic impact discussed, from scaling to intelligence explosions. Dec 23, 2025 - Developers
- Learn about theoretical AI progress and scaling discussions relevant to model development.
- Business
- Understand economic implications of AI progress and potential delays in widespread benefits.
This is a solid, concise interview offering insights into current AI development and its potential future.
Thoughts on AI progress (Dec 2025)
Audio streams from Dwarkesh Patel. Your place is remembered on this device.
From Subvolts
Summary and analysis by Subvolts editorial, AI-assisted.
Dwarkesh Patel's 12-minute podcast episode "Thoughts on AI progress (Dec 2025)" explores AI scaling, the value of human labor, and economic diffusion lag. It also touches on goal-post shifting in AI development and the concept of RL scaling, concluding with a discussion on broadly deployed intelligence explosions.
This episode talks about how AI is getting better and what that means for jobs and the economy. It covers ideas like how much we're improving AI and if we're moving the goalposts on what AI can do.
This episode offers a concise look at current AI progress and theoretical discussions. It's for AI enthusiasts and researchers interested in the trajectory of AI development and its economic implications. The topics connect to ongoing work at frontier labs like OpenAI, Google Gemini, and Anthropic, as they push the boundaries of model capabilities and deployment strategies.
- HostDwarkesh Patel
- Show: Dwarkesh Podcast
- Title: Thoughts on AI progress (Dec 2025)
- Length: 12 minutes
- Topics Covered: What are we scaling?, The value of human labor, Economic diffusion lag is cope, Goal-post shifting is justified, RL scaling, Broadly deployed intelligence explo
Read the essay here.
Timestamps
00:00:00 What are we scaling?
00:03:11 The value of human labor
00:05:04 Economic diffusion lag is cope00:06:34 Goal-post shifting is justified
00:08:23 RL scaling
00:09:18 Broadly deployed intelligence explosion
This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit www.dwarkesh.com
- RL scaling
- Dwarkesh Patel
- Improving AI through reinforcement learning methods.
- Economic diffusion lag
- Dwarkesh Patel
- The delay in economic benefits from new technology reaching everyone.
1 h 371 h 201 h 37AI researchers debate how close we are to recursive self-improvement
WithJohn Schulman, Beren Millidge, Charlie O’Neill
2 h 21Questions people ask
- What are the main topics discussed in the podcast?
- The episode covers AI scaling, the value of human labor, economic diffusion lag, goal-post shifting, RL scaling, and broadly deployed intelligence explosions.
- Where can I find the essay mentioned?
- The show notes indicate an essay is available via a link, though the specific URL is not provided in the excerpt.
Page generated Oct 19, 2026. Summaries are Subvolts' own; the audio, artwork and notes belong to Dwarkesh Patel.