Intelligence.Log
2026-05-13
Extracted: 56 items. Sources: GitHub, Bluesky, X.
++ AI OVERVIEW ++
The AI job market remains a central tension, with a viral "Unethical Guide to Surviving AI Layoffs" and Naomi Saphra’s blunt observation that academics can easily "make ~10x the money" in industry sparking debate over career pragmatism versus research integrity. On the project front, a notable pivot toward practical tooling emerged with the "UI, not UBI" sentiment, signaling a push for tangible interfaces over abstract safety debates. Meanwhile, Angela Zhou critiques existential risk discourse for fixating on hypothetical futures while ignoring present harms, a theme echoed in Thomas Dietterich’s shared wisdom from Zey. The overarching trend is a collective recalibration: the community is grappling with how to balance lucrative industry exits, ethical engineering, and grounded risk analysis.
◆ Signal
Co-Starred · Last 7 days
Repos independently starred by multiple AI leaders in the week ending 2026-05-13. Stronger signal = more overlap.
antirez/ds4
×3 starrers▲ 8/10★ 8.1k
DeepSeek 4 Flash local inference engine for Metal and CUDA
grep TOPIC=
grep SOURCE=
sort --by=
GH
The Pi desktop app you want to use.
“Howcode is a desktop application for interacting with Pi, likely an AI assistant or coding tool. It provides a native UI for leveraging AI capabilities on the desktop, built with TypeScript for performance and reliability.”
GH
Objective-C port of the tokenizer in HuggingFace's swift-transformers
“This repository provides an Objective-C port of HuggingFace's swift-transformers tokenizer, enabling tokenization for LLMs in iOS/macOS apps. It bridges the gap between Swift-based tokenizer implementations and Objective-C codebases, making it easier to integrate transformer models into legacy or mixed-language projects.”
GH
“Space Doctor is a tool that helps manage and optimize disk space on Hugging Face Hub repositories. It provides insights into storage usage and assists in cleaning up unnecessary files.”
BSKY
This "Unethical Guide to Surviving AI Layoffs" by Mo Bitar perfectly captures the current moment www.tiktok.com/@atmoio/vide...
❤️ 55 Likes|[Deployment]
BSKY
BSKY
I don't really understand this. If you just want a job---at any career point as an AI academic, student, faculty whatever---you can make ~10x the money and work 9-6 without meandering chats with colleagues.
❤️ 8 Likes|
BSKY
Learned a lot from this! "UI, not UBI" Something concerning about a lot of "existential AI risk discourse" is its futuristic orientation about hypothetical harms, not about deeply engaging with existing systems to improve social protection.
❤️ 2 Likes|[Safety]
BSKY
😱 1 year ban from arXiv and no more tech reports anymore, must be papers accepted at a reputable venue. Wow, talk about taking action against AI slop science...
❤️ 19 Likes|[Evaluation]
BSKY
The UK’s state AI Security iIstitute findings on latest AI models: 1) Mythos is a big gain in cyber capabilities. But so is GPT-5.5 2) It is hard to establish an upper bound on Mythos/GPT-5.5, which appear to be limited by tokens used, rather than ability. 3) Capability doubling time is 4.5 months
❤️ 65 Likes|[Safety][Evaluation]
BSKY
I don't understand the path forward for Mythos releases. Google & OpenAI will have equivalent models, and they are approaching AI cyber risk guardrails differently, so they will presumably just release their versions without as much drama. How does Anthropic get out of the government approval path?
❤️ 31 Likes|[Safety][Deployment]
BSKY
So "don't feed the trolls" usually means "don't feed the trolls content to keep them going" but sometimes I think it maybe also means "don't feed the trolls to one's followers" (by e.g. boosting a particularly bad take for others to see).
❤️ 51 Likes|
BSKY
"Predictably, the huge spike in productivity that these companies claim their own AI products have enabled hasn’t resulted in more or better products, shorter work weeks, or better consumer experiences." www.404media.co/software-dev...
❤️ 196 Likes|[Evaluation]
BSKY
Hey it's the book's birthday! Happy 1st Birthday to our book @alexhanna.bsky.social :) The AI Con has been out in the world for a whole year. 🎂🎂🎂 thecon.ai
❤️ 108 Likes|
X
My most amusing interaction was where the model (I think I was given some earlier version with a ...
[LLM]
“DeepSeek Summary: Karpathy shares an amusing interaction with an early model version.”
X
I'm starting to get into a habit of reading everything (blogs, articles, book chapters, ...)
[LLM]
“DeepSeek Summary: Karpathy mentions his habit of extensive reading.”
X
2025 LLM Year in Review. 2025 has been a strong and eventful year of progress in LLMs. The following is a list of personally notable and mildly surprising 'paradigm changes' - things that altered the landscape and stood out to me conceptually. At the start of 2025, the LLM production stack in all labs looked something like this: ...
[LLM][Infra]
“DeepSeek Summary: Karpathy reviews paradigm changes in LLMs during 2025.”
X
Vibe coding is irresponsibly building software through dice rolls, not caring what code is produced
[Agent][Safety][Evaluation]
“DeepSeek Summary: Simon Willison defines vibe coding as irresponsible software development where code quality is neglected.”
X
A short note that the predictions that LLMs would favor 'boring technology' that's
[LLM][Deployment]
“DeepSeek Summary: Willison notes that LLMs may not favor boring technology as predicted.”
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: Agent can update its memory while running, either autonomously or via user prompt.”
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
[Infra][Agent]
“DeepSeek Summary: Agents increasingly require a sandboxed workspace to execute code and manage files.”
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 approach to building agentic systems has evolved significantly since ChatGPT's launch.”
X
When building agents, you need to iterate on production data much more than when building traditional software. You need to iterate on how
[Evaluation][Deployment]
“DeepSeek Summary: Agent development requires more iteration on production data compared to traditional software.”
X
Resource constraints are a beautiful thing. Survival instinct in a cut-throat AI competitive land
[Agent][Deployment]
“DeepSeek Summary: Jim Fan reflects on how resource constraints can drive survival instinct and innovation in the competitive AI landscape.”
X
Everyone's freaking out about vibe coding. In the holiday spirit, allow me to share my anxiety on the wild
[LLM][Multi-modal]
“DeepSeek Summary: Jim Fan comments on the hype around 'vibe coding' and shares his own anxiety about the rapid developments.”
X
It gives me a lot of comfort knowing that we are the last generation without advanced robots everywhere.
[Agent][Deployment]
“DeepSeek Summary: Jim Fan expresses comfort in being part of the last generation before advanced robots become ubiquitous.”
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][Evaluation]
“DeepSeek Summary: Jeremy Howard demonstrates Grok's bias by showing it searches for Elon Musk's opinion on Israel/Palestine before forming its own response.”
X
Here's what I would prefer to see:
[LLM]
“DeepSeek Summary: Jeremy Howard expresses a preference, likely related to AI or software engineering, but full content is truncated.”
X
reading "AI News" (previously Smol Talk) is probably the highest-leverage 45 mins
[LLM][Deployment]
“DeepSeek Summary: Recommends reading AI News as a high-leverage activity.”
X
MacStudio you ask? Apple Engineering's **actual** time spent on PyTorch support
[Infra][Deployment]
“DeepSeek Summary: Comments on Apple Engineering's effort on PyTorch support.”
X
ChatGPT seems to be **really** good for creative work and a solid starting point
[LLM][Multi-modal]
“DeepSeek Summary: Acknowledges ChatGPT's strength in creative tasks.”
X
anyone else feel burned out by a new AI breakthrough every week?
[Safety][Evaluation]
“DeepSeek Summary: Expresses burnout from the rapid pace of AI breakthroughs.”
X
Many people assume that LRM reasoning breaks down past a certain 'complexity' or 'number of steps'
[LLM][Evaluation]
“DeepSeek Summary: Chollet comments on a common assumption about reasoning in large reasoning models (LRMs).”
X
The 3rd edition of my book Deep Learning with Python is being printed right now, and will be in bookstores within 2 weeks.
[Deployment]
“DeepSeek Summary: Chollet announces the upcoming release of the 3rd edition of his book.”
X
I can now confess that I participated in the new #TronAres movie, playing myself I had a great time working with everyone especially Greta
[Multi-modal]
“DeepSeek Summary: Fei-Fei Li reveals she appears as herself in the movie Tron: Ares.”
X
congrats to OpenAI on winning the Turing Test
[LLM]
“DeepSeek Summary: Sarcastic congratulations to OpenAI for supposedly winning the Turing Test, reflecting on AI progress.”
X
American comedian before they are famous: yo what even are the minions,
“DeepSeek Summary: Phil Wang comments on American comedians' early material referencing minions.”
X
Some personal news: I recently joined Cursor. Cursor is a small, ambitious team, and they've created
[Tooling]
“DeepSeek Summary: Sasha Rush announced joining Cursor, a small ambitious team.”
X
Been reflecting a bit on the Harvard news. This paper from 2017 was ... Didn't realize at the time how lucky for us Americans to work with incredible people from around the world.
[Safety]
“DeepSeek Summary: Reflection on Harvard news and gratitude for working with international colleagues.”
X
Talk at Ray Summit on 'Building Cursor Composer.' Overview of the work from our research team.
[Tooling]
“DeepSeek Summary: Sasha Rush gave a talk at Ray Summit about building Cursor Composer.”
X
Classical Jensen math. Unidirectional bandwidth is topped at 450GB/s, and then there comes a protocol overhead of two digit percentage.
[Infra]
“DeepSeek Summary: Discusses bandwidth limitations and protocol overhead in the context of Jensen's math.”
X
Thanks to an awesome contribution from @omarnomad The Machine Learning Engineering Open book now can ...
[Tooling]
“DeepSeek Summary: Acknowledges a contribution to the Machine Learning Engineering Open Book.”
X
Good news! Ulysses Sequence Parallelism from the Snowflake AI Research and the Deepspeed ...
[Infra]
“DeepSeek Summary: Announces Ulysses Sequence Parallelism from Snowflake AI Research and DeepSpeed.”
X
To remind - this is the memory saving you get when enabling TiledMLP :) Left: normal memory ...
[Fine-tuning]
“DeepSeek Summary: Demonstrates memory savings from TiledMLP.”
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][RAG][Tooling]
“DeepSeek Summary: Philipp Schmid shares his reading of three recent technical reports: Moonshot AI's Kimi K2.5, Cursor's Composer 2, and Chroma's 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][Deployment]
“DeepSeek Summary: Philipp Schmid describes an experiment where an AI agent autonomously built a 0.8B model overnight by reading a repo and fetching training data.”
X
Random thought. We are going to be so much faster at creating and building.
[Deployment][Tooling]
“DeepSeek Summary: Philipp Schmid reflects on the accelerating pace of creation and building enabled by AI.”
X
AI is actually pretty good at ideas as well.
[LLM][Multi-modal]
“DeepSeek Summary: AI demonstrates capability in generating ideas, not just analysis.”
X
Had an interesting exchange with roon of OpenAI last night over whether super intelligent AI would actually be able to navigate organizational challenges.
[Safety][Deployment]
“DeepSeek Summary: Debate with OpenAI about AI's ability to handle organizational dynamics.”
X
From a pure 'do good for the world' mission perspective, having the acting like a solid personalized
[Safety][Deployment]
“DeepSeek Summary: Discusses AI's potential for personalized good, though snippet is cut off.”
X
Look what @alexhanna and I got to do! (Hang out with the cool kids ...) We're talking about the Turing Test, the grandmother of all tests for AI sentience. Joining us are AI researchers Alex Hanna and Emily M. Bender
[Safety][Evaluation]
“DeepSeek Summary: Bender and Hanna discuss the Turing Test as a foundational concept for AI sentience.”
X
For those playing along at home, here's a "AI is sentient!" argument bingo card.
[Safety][Evaluation]
“DeepSeek Summary: Bender provides a bingo card to debunk common arguments for AI sentience.”
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
RT @natolambert: A few facts, while the dust is settling. Ai2 still is... - releasing open models, folks want to,
[Infra]
“DeepSeek Summary: Naomi Saphra retweets a post about AI2 continuing to release open models despite industry changes.”
X
For the first time in almost a decade, I'm teaching a class on learning and control.
[Evaluation]
“DeepSeek Summary: Ben Recht announces teaching a class on learning and control after nearly a decade.”
X
I have a recommended reading list for Artificial Intelligence, and it hasn't changed since 2019.
[Evaluation]
“DeepSeek Summary: Ben Recht's AI reading list has remained unchanged since 2019, suggesting stable foundational knowledge.”
-- END OF LOG --
[STATS] 56 items · Filter applied
Powered by Horizon + DeepSeek