LessWrong posts by zvi
LessWrong posts by zvi
zvi
“AI #133: America Could Use More Energy” by Zvi
1 hour 29 minutes Posted Sep 11, 2025 at 12:00 am.
A lot of the focus since the rollout of GPT-5 remains on the perception and policy fronts, especially regarding views of AI progress. The botched rollout of what is ultimately a very good (but not mind blowing) model gave a lot of people the wrong impression, so I have to remind everyone that once again that AI continues to make rapid progress. Meanwhile, we must also notice that OpenAI's actions in the public sphere have once [...] ---Outline:
Language Models Offer Mundane Utility
Productivity Puzzles
Language Models Don't Offer Mundane Utility
Huh, Upgrades
On Your Marks
Choose Your Fighter
Fun With Media Generation
Deepfaketown and Botpocalypse Soon
Unprompted Attention
Get My Agent On The Line
They Took Our Jobs
A Young Lady's Illustrated Primer
Levels of Friction
The Art of the Jailbreak
Get Involved
Introducing
In Other AI News
Show Me the Money
Quiet Speculations
The Quest for Sane Regulations
Chip City
The Week in Audio
All Words We Choose Shall Lose All Meaning
Hunger Strike
Rhetorical Innovation
Misaligned!
Hallucinations
Aligning a Smarter Than Human Intelligence is Difficult
0:00
1:29:17
Download MP3
Show notes
Even in quiet weeks like this one, there are noticeable incremental upgrades. The cost of the best video generation tool, Veo 3, went down by half. ChatGPT now offers conversation branching. Claude can directly edit files. Yet it is a good time to ask about the missing results. Where are all the AI agents? If AI coding is so good why aren’t we seeing a surge in GitHub repositories or iPhone apps?
A lot of the focus since the rollout of GPT-5 remains on the perception and policy fronts, especially regarding views of AI progress. The botched rollout of what is ultimately a very good (but not mind blowing) model gave a lot of people the wrong impression, so I have to remind everyone that once again that AI continues to make rapid progress. Meanwhile, we must also notice that OpenAI's actions in the public sphere have once [...] ---Outline:(
01:14) Language Models Offer Mundane Utility(09:44) Productivity Puzzles(12:31) Language Models Don't Offer Mundane Utility(14:13) Huh, Upgrades(16:43) On Your Marks(18:32) Choose Your Fighter(19:57) Fun With Media Generation(20:31) Deepfaketown and Botpocalypse Soon(26:11) Unprompted Attention(26:44) Get My Agent On The Line(28:03) They Took Our Jobs(31:37) A Young Lady's Illustrated Primer(36:24) Levels of Friction(38:41) The Art of the Jailbreak(40:13) Get Involved(40:29) Introducing(44:17) In Other AI News(48:46) Show Me the Money(52:26) Quiet Speculations(54:50) The Quest for Sane Regulations(55:45) Chip City(01:08:05) The Week in Audio(01:10:12) All Words We Choose Shall Lose All Meaning(01:11:28) Hunger Strike(01:14:41) Rhetorical Innovation(01:15:29) Misaligned!(01:18:49) Hallucinations(01:21:44) Aligning a Smarter Than Human Intelligence is Difficult(01:26:59) The Lighter Side ---
First published:
September 11th, 2025
Source:
https://www.lesswrong.com/posts/gDNCsJfJHp5YSbdhn/ai-133-america-could-use-more-energy
---
Narrated by TYPE III AUDIO.
---Images from the article:Apple Podcasts and Spotify do not show images in the episode description. Try Pocket Casts, or another podcast app.