Intelligence.Log
2026-05-11
Extracted: 60 items. Sources: GitHub, Bluesky, X.
++ AI OVERVIEW ++
Today’s discourse was notably split between high-stakes AI biography and the ongoing tension between open and proprietary models. Marc Lanctot kicked off a new book recommendation, praising *The Infinity Machine* on Demis Hassabis and DeepMind, signaling sustained interest in the personalities behind the superintelligence race. Meanwhile, Emily M. Bender shared a positive note on a screening at SIFF, subtly reminding the community of the cultural and ethical conversations happening outside of code commits. On the GitHub front, while the hockey post was a welcome distraction, the real trend remains the community’s deep dive into foundational AI narratives and the philosophical debates they ignite.
◆ Signal
Co-Starred · Last 7 days
Repos independently starred by multiple AI leaders in the week ending 2026-05-11. Stronger signal = more overlap.
antirez/ds4
×2 starrers▲ 7/10★ 2.7k
DeepSeek 4 Flash local inference engine for Metal
grep TOPIC=
grep SOURCE=
sort --by=
GH
A community trust management system based on explicit vouches to participate.
“Vouch is a community trust management system that uses explicit vouches to grant participation rights, aiming to reduce spam and abuse. It provides a decentralized approach to access control based on social trust rather than centralized moderation.”
GH
Developer blog for PyTorch
“A developer blog for PyTorch, providing insights into the development process, new features, and technical decisions. It offers a behind-the-scenes look at one of the most popular deep learning frameworks.”
BSKY
What a game by the @canadiens.com winning 6-2 vs. Buffalo to take a 2-1 lead in the series! #gohabsgo youtu.be/mswVDkbb7J0?...
❤️ 0 Likes|
BSKY
Ok #booksky, time to post about the third of the recent biographies: "The Infinity Machine: Demis Hassabis, DeepMind, and the Quest for Superintelligence". I ♥️ LOVED ♥️ this book! I've been recommending it to pretty much everyone who is curious to know about the history of Google DeepMind. 🧵 1/N
❤️ 13 Likes|[LLM]
BSKY
BSKY
New TIL: I figured out how to use my LLM CLI tool in a shebang line, which means you can write executable scripts in English, or hook up more complex scripts with a snippet of YAML template - til.simonwillison.net/llms/llm-she...
❤️ 114 Likes|[Tooling]
BSKY
This is excellent. I particularly like the definition of the "Zombie Internet", which starts: "It’s people talking to bots, people talking to people, people creating “AI agents” and then instructing them to interact with people. It’s people using AI talking to people who are not using AI [...]"
❤️ 72 Likes|[Agent][LLM]
BSKY
Pretty wild I got my PhD 4 years ago to the day. I feel very lucky that I got to do it and make my switch into AI. Lot's of people today in AI are underselling the value of going through the process of a PhD.
❤️ 48 Likes|
BSKY
Excellent explainer video by @fryrsquared.bsky.social on the risks of AI agents. We shouldn’t make the mistake of thinking current limitations will necessarily persist. As we’ve seen for years now, the capabilities of frontier AI models are continuously increasing. www.youtube.com/watch?v=WnzR...
❤️ 15 Likes|[Agent][Safety]
BSKY
BSKY
hivemind what are your recommended reads for getting out of the chicken-egg cycle of co-design ?
❤️ 0 Likes|[Agent][Infra]
BSKY
The quantification trap, or how well-intentioned attempts at intersubjectivity breed institutionalized postmodern optimization. (Part 2 in my series on what I wish I had more clearly articulated in The Irrational Decision.)
❤️ 17 Likes|[Evaluation]
X
Judging by my tl there is a growing gap in understanding of AI capability. The first issue I think is around
[LLM][Safety]
“DeepSeek Summary: Karpathy observes a growing gap in understanding AI capabilities, starting with a first issue.”
X
I'm starting to get into a habit of reading everything (blogs, articles, book chapters,…)
[LLM]
“DeepSeek Summary: Karpathy shares his habit of extensive reading to stay informed.”
X
My most amusing interaction was where the model (I think I was given some earlier version with a
[LLM]
“DeepSeek Summary: Karpathy recounts an amusing interaction with an early model version.”
X
By training LLMs against auto
[Fine-tuning][LLM]
“DeepSeek Summary: Karpathy discusses training LLMs using auto-generated data or objectives.”
X
A short note that the predictions that LLMs would favor 'boring technology' that's once you attach them to a good coding agent harness at least
[Agent][LLM][Tooling]
“DeepSeek Summary: LLMs attached to coding agents may favor boring technology, challenging earlier predictions.”
X
I'm beginning to suspect that a key skill in working effectively with coding agents is developing an intuition for when you don't need to
[Agent][Tooling]
“DeepSeek Summary: Key skill for coding agents is knowing when to intervene.”
X
This may be the best guidance I've seen anywhere on writing a really good commit history. My ideal commit combines
[Tooling]
“DeepSeek Summary: Recommends best guidance for writing commit history.”
X
In the hot path as the agent is running. The agent can decided to (or the user can prompt it to) update its memory as it is working on the core
[Agent][Tooling]
“DeepSeek Summary: Memory updates can happen in the hot path of agent execution, either agent-initiated or user-prompted.”
X
TL;DR: More and more agents need a workspace: a computer where they can run code, install packages, and access files. Sandboxes provide this
[Agent][Infra]
“DeepSeek Summary: Agents increasingly require sandboxed workspaces for code execution and file access.”
X
Your harness, your memory ... The “best” way to build agentic systems has changed dramatically over the past three years. When ChatGPT came out,
[Agent][LLM]
“DeepSeek Summary: The best practices for building agentic systems have evolved significantly since ChatGPT's release.”
X
We launched LangSmith Agent Builder this week as a no-code way to build agents. A key part of Agent builder is it's memory system. In this
[Agent][Tooling]
“DeepSeek Summary: LangSmith Agent Builder enables no-code agent creation with a focus on memory systems.”
X
Resource constraints are a beautiful thing. Survival instinct in a cut-throat AI competitive land
[Agent]
“DeepSeek Summary: Resource constraints drive innovation and survival in competitive AI landscape.”
X
I've been a bit quiet on X recently. The past year has been a transformational experience.
[Agent]
“DeepSeek Summary: Jim Fan acknowledges a period of silence due to a transformative year.”
X
Everyone's freaking out about vibe coding. In the holiday spirit, allow me to share my anxiety on the wild
[Agent]
“DeepSeek Summary: Jim Fan comments on the hype around 'vibe coding' and shares his own concerns.”
X
It gives me a lot of comfort knowing that we are the last generation without advanced robots everywhere.
[Agent]
“DeepSeek Summary: Expresses comfort in being part of the last generation before widespread advanced robotics.”
X
Here's a complete unedited video of asking Grok for its views on the Israel/Palestine situation. It first searches twitter for what Elon thinks.
[Safety]
“DeepSeek Summary: Jeremy Howard demonstrates that Grok searches Twitter for Elon Musk's opinion before answering about Israel/Palestine.”
X
Here's what I would prefer to see:
[Agent]
“DeepSeek Summary: Jeremy Howard expresses a preference for an unspecified alternative approach.”
X
reading 'AI News' (previously Smol Talk) is probably the highest-leverage 45 mins
[LLM]
“DeepSeek Summary: Soumith recommends reading 'AI News' as a high-leverage activity.”
X
we've been working on democratizing fast kernel writing on the @PyTorch team. try
[Infra][Tooling]
“DeepSeek Summary: Soumith announces efforts to democratize fast kernel writing in PyTorch.”
X
Sometimes we forget that NVIDIA wins because it's a software company.
[Infra]
“DeepSeek Summary: Soumith points out that NVIDIA's success is due to its software, not just hardware.”
X
It's surprisingly easy to do "hard" things -- for the most part, you need to get started and keep at it
[Agent]
“DeepSeek Summary: Chollet emphasizes that starting and persisting makes hard tasks easier.”
X
Don't miss David Ha @hardmaru's keynote at @ALifeConf #ALIFE2021 on 'World Models and Attention for Reinforcement Learning'!
[Agent]
“DeepSeek Summary: David Ha gave a keynote on world models and attention for reinforcement learning at ALIFE 2021.”
X
Dario is wrong. He knows absolutely nothing about the effects of technological revolutions on the labor market.
[Safety]
“DeepSeek Summary: LeCun dismisses Dario Amodei's claims about AI job loss, asserting ignorance of historical labor market impacts.”
X
The emergence of superintelligence is not going to be an event. We don't have anything close to a
[Safety]
“DeepSeek Summary: LeCun argues that superintelligence will not appear suddenly, contrasting with doomsday narratives.”
X
It seems to me that before 'urgently figuring out how to control AI systems much smarter than us' we need
[Safety]
“DeepSeek Summary: LeCun questions the premise of controlling superhuman AI, suggesting more fundamental issues come first.”
X
Very excited to share @theworldlabs 's latest research work RTFM!! It's a real-time, ...
[Multi-modal]
“DeepSeek Summary: Fei-Fei Li announces RTFM, a real-time research work from The World Labs.”
X
congrats to OpenAI on winning the Turing Test
[LLM][Evaluation]
“DeepSeek Summary: Max Woolf sarcastically congratulates OpenAI for supposedly passing the Turing Test, reflecting on AI milestones.”
X
Some personal news: I recently joined Cursor. Cursor is a small, ambitious team, and they've created
[LLM][Deployment][Tooling]
“DeepSeek Summary: Sasha Rush announced joining Cursor, a small ambitious team.”
X
Wager established. Jonathan Frankle (@jefrankle) stepped up to my Transformer long bet.
[LLM]
“DeepSeek Summary: Sasha Rush established a bet with Jonathan Frankle about Transformers.”
X
today i woke up to a living version of a phd student's nightmare. a new paper in my inbox: a detailed reproduction of a paper i wrote
[Evaluation][Safety]
“DeepSeek Summary: Sasha Rush received a reproduction paper of his own work, calling it a PhD student's nightmare.”
X
I have been compiling LLM/VLM training logbooks/chronicles. This is the one of the best sources to
[LLM][Fine-tuning]
“DeepSeek Summary: Compiling LLM/VLM training logbooks/chronicles as a valuable resource.”
X
Thanks to an awesome contribution from @omarnomad The Machine Learning Engineering Open book now can
[Tooling][Deployment]
“DeepSeek Summary: Machine Learning Engineering Open Book updated with contribution.”
X
If you were holding off to try @MSFTDeepSpeed ZeRO++ it looks like deepspeed@master should
[Infra][LLM]
“DeepSeek Summary: DeepSpeed ZeRO++ now available on master branch.”
X
Modern art. Artist: PyTorch memory profiler Model: Llama-8B The piece on the left is the
[Tooling][LLM]
“DeepSeek Summary: PyTorch memory profiler visualization of Llama-8B training.”
X
Together w/ the community, our initiative of profiling Diffusers pipelines & potentially improving them is going very strong
[Infra][Deployment][Tooling]
“DeepSeek Summary: Sayak Paul highlights community-driven profiling of Diffusers pipelines to improve performance.”
X
My presentation at the @PyTorch Conf EU is now live. It's an exciting piece given its emphasis on how we make Diffusers play quite well w/ `torch.compile`
[Infra][Deployment][Tooling]
“DeepSeek Summary: Sayak Paul presented at PyTorch Conf EU on integrating Diffusers with torch.compile for performance gains.”
X
I read three technical reports from Moonshot AI's Kimi K2.5 paper, Cursor's Composer 2 report and blog post, and Chroma's Context-1 write-up
[LLM][Tooling]
“DeepSeek Summary: Philipp Schmid read three technical reports: Kimi K2.5, Cursor Composer 2, and Chroma Context-1.”
X
Told an AI agent to read the autoresearch repo and build a version for QMD. Get training data from tobi/qmd github. Went to sleep. Woke up to a 0.8B model
[Agent][Fine-tuning]
“DeepSeek Summary: He automated model building by instructing an AI agent to clone a repo and train a 0.8B model overnight.”
X
Random thought. We are going to be so much faster at creating and building.
[LLM]
“DeepSeek Summary: He believes AI will drastically accelerate creation and building processes.”
X
So much work is going into faking continual learning and memory for AIs
[LLM][Safety][Evaluation]
“DeepSeek Summary: Critiques the effort to simulate continual learning and memory in AI systems.”
X
Had early access to GPT-5.4 and Pro. They are very good. One fun illustration of progress
[LLM][Multi-modal][Deployment]
“DeepSeek Summary: Early access to GPT-5.4 and Pro shows significant progress.”
X
Image is of the 1990s Microsoft writing assistant character Clippy with its eyebrows raised positioned in.
[Safety]
“DeepSeek Summary: Bender posted an image of Clippy, likely to critique or satirize AI hype.”
X
New preprint! Everyone loves causal interp. It's coherently defined! It makes testable predictions
[Evaluation]
“DeepSeek Summary: Naomi Saphra announces a new preprint on causal interpretability, emphasizing its coherent definition and testable predictions.”
X
Ok, I wrote this up (link below)
[LLM]
“DeepSeek Summary: Naomi Saphra shares that she has written up something, presumably a blog post or paper.”
X
#throwback coz it's finally the day again!!! #HellOnWheels back on AMC 9/8c tonight! It's gunna be so intense.
[Deployment]
“DeepSeek Summary: Angela Zhou promotes the return of the TV show Hell on Wheels, expressing excitement.”
X
Revisiting Sutton's Bitter Lesson in the wake of GPT-5.
[LLM]
“DeepSeek Summary: Revisits Sutton's Bitter Lesson in the context of GPT-5.”
X
X
For the first time in almost a decade, I'm teaching a class on learning and control.
[Evaluation]
“DeepSeek Summary: Announces teaching a class on learning and control after a long gap.”
-- END OF LOG --
[STATS] 60 items · Filter applied
Powered by Horizon + DeepSeek