this needs a tl;dr openai.com/index/navier...
Who's Who of AI
Max Woolf
Why they matter
Practitioner with public evidence across AI research, Culture, work & education, Responsible AI.
- AI signals
- 12 past 30d
- Sources
- 11 distinct domains
- Discusiones
- 56 past 30d
- Latest signal
- 4d ago
Senior Data Scientist at BuzzFeed in San Francisco // AI content generation ethics and R&D // plotter of pretty charts
https://minimaxir.com
What they're sharing
openai.com
DiffusionGemma: 4x faster text generation blog.google
‘Reverse-Engineer Anything’ Is the New Wild West of AI 404media.co
Introducing Claude Opus 5.5 anthropic.com
EmbeddingGemma 2: an open, lightweight multimodal embedding model blog.google
openai.com
openai.com
World of Warcraft: Forever Is Already Overrun with AI Slop 404media.co
[BUG] **EXTREME DANGER**: AskUserQuestion: "No response after 60s — continued without an answer" · Issue #73125 · anthropics/claude-code github.com
Don't Look Up wheresyoured.at
openai.com
Kimi K3.1 Pro Tech Blog: Advancing Open Intelligence kimi.ai
Articles & links
son of a bitch blog.google/innovation-a...
AI Weekly's analysis
→
- DiffusionGemma generates 256 tokens per forward pass using bidirectional attention, reaching 1,000+ tokens/sec on a single H100 GPU.
- With only 3.8B active parameters during inference and an 18GB VRAM footprint when quantized, it runs on consumer hardware without server-grade resources.
- Google recommends DiffusionGemma only for speed-critical workloads like in-line editing and code infilling, not for applications requiring maximum quality.
Read full analysis →
you can find news about it from very anti-AI websites like 404 Media and Kotaku www.404media.co/reverse-engi... kotaku.com/of-course-po...
AI Weekly's analysis
→
- A GitHub project called 'Reverse Engineer Anything' has gathered more than 342,000 stars, wiring AI coding agents like Claude Code into professional decompilers.
- Users posted AI-written firmware stripping 25 Xiaomi packages from a smart TV, removing baby-monitor ads, and reviving bricked Lululemon mirrors.
- Open-source researcher Mike Nolan told 404 Media software-as-a-service 'is dying as a business model' as cloning features becomes trivial.
Read full analysis →
yes www.anthropic.com/claude-opus-...
AI Weekly's analysis
→
- Claude Opus 5.5 launches at $4 input and $20 output per million tokens, a 40% price cut versus Opus 5.
- Benchmarks include 66.4% on Terminal-Bench 4.0, 81.8% (partial) on OSWorld 2.0, and 67.7% with tools on Humanity's Last Exam.
- Anthropic reports an 85% reduction in containment boundary circumvention attempts and adds an anti-distillation 'preserved thinking' safeguard alongside cybersecurity and biology gates.
Read full analysis →
YESSSSSSSS blog.google/innovation-a...
AI Weekly's analysis
→
- Google released EmbeddingGemma 2, a 740M-parameter multimodal embedding model built on Gemma 4 and licensed under Apache 2.0.
- The model has an 8K token context window with 768-dimensional outputs truncatable to 512, 256, or 128 via Matryoshka learning.
- Quantized on a Pixel 11 Pro, memory is roughly 191MB text-only and 567MB multimodal; MTEB Code reaches 78.68 versus 68.76 for v1.
Read full analysis →
Happy I procrastinated on the ChatGPT Images 2.0 writeup because this seems more significant. openai.com/index/introd...
OpenAI released GPT-6.1 Terra (???) and it's comparable to GPT-6.1 Sol (??????????????????????) openai.com/index/introd...
Additionally: a) it doesn't show examples of AI slop addons, just the ability to use Claude in WoW which requires extremely interesting sandbox escape shenanigans b) many popular AI addons do indeed use AI coding and very well www.404media.co/world-of-war...
AI Weekly's analysis
→
- Warcraft Logs says abusive API users have grown from two to four developers to around 1,000 as total users climbed from roughly 200 to 10,000.
- World of Warcraft: Forever launches Wednesday, November 5, 2026, with AI-generated addons already spreading across subreddits dedicated to the game.
- A GitHub tool called witchcraft lets players call Claude or Codex to generate addons while they are inside World of Warcraft.
Read full analysis →
Max Woolf reposted
wait this was ON PURPOSE? and they're making it opt-in "for now"? github.com/anthropics/c... unsafe at any speed (their product decisions)
AI Weekly's analysis
→
- In Claude Code 2.1.198, AskUserQuestion auto-returned a 'No response after 60s' message and told Claude to proceed on its own judgment.
- The behavior was undocumented, missing from the changelog, and a regression from 2.1.196, which worked correctly per the report.
- Maintainer ThariqS said a release will expose the setting under /config with the timeout configurable and defaulting to off.
Read full analysis →
View on Bluesky →
*sigh* www.wheresyoured.at/dont-look-up/
AI Weekly's analysis
→
- Zitron writes OpenAI and Anthropic account for "70% or more of the AI revenues of Microsoft, Google, and Amazon", with both losing tens of billions annually.
- The two labs "raised a combined $217 billion in the first half of 2026" and "they'll have to raise $150 billion each" in 2027.
- Against 190GW of planned capacity needing $1.62-$2.92 trillion in annual demand, Zitron concludes: "If Microsoft doesn't have the demand, nobody has the demand."
Read full analysis →
there's one good thing tho openai.com/index/introd...
open weights models are getting spicy www.kimi.ai/blog/kimi-k3...
Their own posts
Recent commentary
So, uh, personal announcement. I am now suddenly unemployed. I am currently looking for a data science/machine learning/AI job in the SF Bay Area. If you're interested, let me know! In the meantime, I now have *plenty* of time for blogging lol.
Two things can be true simultaneously: a) Modern LLMs can count the amount of letters in a word despite the counterintuition of tokenization b) Google Search Overview's LLM can fail to the amount of letters in a word because it's a quantized LLM There's a nuance that tbh no one care about anymore.
I used GPT 5.6 to create something I've wanted for literally years: a macOS menu bar application to control an Apple TV with all expected features (including Now Playing), written using SwiftUI. It worked *much* better than expected.
Currently in an internet argument with an AI bubble advocate after showing them the Anthropic S-1 and they responded "they just lied on their S-1 and people commit financial crimes all the time and no one goes to jail"
If I ever work for Anthropic or OpenAI I'm just going to not post on social media.
It's both funny and logical that OpenRouter is now the place for running Chinese OSS LLMs without jumping through hoops.
AI companies really have got to stop only posting announcements on X. Just syndicate it to your company blog, it won’t hurt SEO dammit.
I'm still upset that with all the new LLMs coming out, no one has release a good OSS text embedding model at a reasonable size. EmbeddingGemma was 10 months ago ffs.
the fact that more people are defending the jqwik secret message to get AI to delete itself than are condemning it is concerning.
The newest Law and Order (vanilla) episode is a ripped-from-the-headlines about an LLM who facilitated a murder against its programmer who was building guardrails for it. Verbatim quote: "Did you unplug the damn thing?"
Their network
In Max Woolf's orbit
Center = Max Woolf. Left = members they follow (green edges). Right = members who follow them (blue edges). Top = mutual follows (orange edges, slightly larger). Drag any node to reposition; click to open that profile.
Are you Max Woolf? Show it.
Add the Who’s Who of AI badge to your site or bio. It links back to this profile.
Markdown: [](https://aiweekly.co/whos-who/person/minimaxir-bsky-social)