One result of going on vacation was that I wasn’t able to spin events off into focused posts this week, so I’m going to fall back on splitting the weekly instead, plus some reserving a few subtopics for later posts, including AI craziness (the Tim Hua post on this is excellent), some new OpenAI largely policy-related shenanigans, and the continuing craziness of some people who should very much know better confidently saying that we are not going to hit AGI any time soon, plus some odds and ends including dead internet theory.
That still leaves tons of other stuff.
Table of Contents
---
Outline:
(00:44) Language Models Offer Mundane Utility
(02:42) Language Models Don't Offer Mundane Utility
(08:24) On Your Marks
(13:26) Choose Your Fighter
(21:05) Fun With Media Generation
(22:10) Deepfaketown and Botpocalypse Soon
(25:32) Don't Be Evil
(26:41) They Took Our Jobs
(40:39) School Daze
(44:18) The Art of the Jailbreak
(44:27) Overcoming Bias
(54:01) Get Involved
(56:34) Introducing
(58:09) Unprompted Attention
(58:56) In Other AI News
(01:02:54) Show Me the Money
---
First published:
September 4th, 2025
Source:
https://www.lesswrong.com/posts/qSt27zr3ZFJoe8ET8/ai-132-part-1-improved-ai-detection
---
Narrated by TYPE III AUDIO.
---
Images from the article:
The graph shows ratings across multiple criteria (like Research Quality, Staff Expertise, etc.) categorized by political classifications (Left, Center-Left, Center-Right, Right), with scores from 1-5." style="max-width: 100%;" />
The lists display logos and names of various AI tools and applications, with ChatGPT and Gemini leading both rankings at #1 and #2 respectively." style="max-width: 100%;" />
The graph plots various AI language models (like GPT, Claude, LLaMA) on a timeline from 2024-2026, tracking their accuracy scores from 0 to 0.6. Two trend lines (green "Closed Frontier" and blue "Open Frontier") show the progression of model capabilities over time." style="max-width: 100%;" />
This table lists AI products from ChatGPT (#1) to ourdream.ai (#50), displaying their logos and names." style="max-width: 100%;" />
The scene shows an animated character in a business suit against a city backdrop, expressing a common meme format about denying one's own mistakes and blaming something else instead." style="max-width: 100%;" />
Apple Podcasts and Spotify do not show images in the episode description. Try Pocket Casts, or another podcast app.
