Fable is being taken down? IMMEDIATELY?? www.anthropic.com/news/fable-m...
Who's Who of AI
Tim Kellogg
AI Architect | North Carolina | AI/ML, IoT, science
WARNING: I talk about kids sometimes
What they're sharing
Statement on the US government directive to suspend access to Fable 5 and Mythos 5 anthropic.com
When AI builds itself anthropic.com
An OpenAI model has disproved a central conjecture in discrete geometry | OpenAI openai.com
She won an exemption from using AI at her tech job. The Pope's remarks could fuel similar appeals. businessinsider.com
DiffusionGemma: 4x faster text generation blog.google
Encyclical Letter of His Holiness Leo XIV Magnifica Humanitas (15 May 2026) vatican.va
Anthropic co-founder Chris Olah's remarks on Pope Leo XIV's encyclical "Magnifica humanitas" \ Anthropic anthropic.com
wsj.com
Anthropic Walks Back Policy That Could Have ‘Sabotaged’ AI Researchers Using Claude wired.com
Project Glasswing: what Mythos showed us blog.cloudflare.com
Introducing Gemma 4 12B: a unified, encoder-free multimodal model blog.google
moonshotai/Kimi-K2.7-Code · Hugging Face huggingface.co
Articles & links
ALERT ALERT they’re posting about RSI on main www.anthropic.com/institute/re...
A general purpose (nothing special) internal OpenAI model solved one of the most famous previously unsolved problems in discrete geometry The solution involved far too many decisions for a human to feasibly explore openai.com/index/model-...
A software engineer in North Carolina won a religious exemption from using AI at work (unitarian universalist) www.businessinsider.com/worker-got-r...
Diffusion nerds are at it again — DiffusionGemma 26B-A4B unlike previous language diffusion models, this one doesn’t suck, and it’s very fast 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 →
The full encyclical (i haven’t yet read any of it) www.vatican.va/content/leo-...
here’s Olah’s full speech www.anthropic.com/news/chris-o...
Amazon researchers snitched to the US government about a jailbreak that triggered a letter forcing Anthropic to shut down Fable 5 www.wsj.com/tech/ai/anth...
Anthropic admits they fucked up www.wired.com/story/anthro...
Cloudflare talks about Mythos Preview - What makes it better than GPT-5.5 - Why they made their own harness - Safety considerations blog.cloudflare.com/cyber-fronti...
Gemma 4 12B Runs comfortably in 14GiB VRAM, but performs on par with Gemma 4 26B (which performed on par with gpt-oss-120b, which was up around o3) Also multimodal and awesome blog.google/innovation-a...
Kimi-2.7-Code: 1T A32B - 30% less reasoning tokens than K2.6 - A little worse than GPT-5.5 & Opus 4.8 but <20% the price huggingface.co/moonshotai/K...
Their own posts
Recent commentary
GPT-5.5 has abandoned "the smoking gun" for new terminology
Anthropic: “we can’t show raw CoT because IP” actual CoT: “omfg this guy is dumb as rocks. Who tf even asks questions like this?”
i guess this means Anthropic is building their own chips?
how likely is it that a global pause on AI development will work? as in, how likely is that we can get everyone to actually pause?
Someone posted a real Monet painting but claimed it was AI generated The comments look like a typical bluesky conversation
something we don’t talk about: even in AI, “artificial” intelligence — even that is not intelligent design we don’t “design” AI. we merely create a rich environment of data + compute and let the intelligence emerge spontaneously i wish philosophers/theologians would weigh in on that
talking to my mom about anti-AI in small town Trump-land. The big things they’re worried about: 1. jobs 2. destroying social structures & romantic relationships 3. skynet scenarios mostly general distrust of some unknown complex and possibly dangerous thing datacenters are not on the list
imo AI is closer to a raw material than a finished product, and most criticism just boils down to that the people getting most out of AI are builders who see potential
Opus and GPT-5.5 are both below 2T, DeepSeek can catch up to both, Fable will keep increasing the gap
people are framing Fablegate as an “AI pause”, but i don’t think any development is stopping or slowing
Their network
In Tim Kellogg's orbit
Center = Tim Kellogg. 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.