Intelligence.Log
2026-05-20
Extracted: 52 items. Sources: GitHub, Bluesky, X.
++ AI OVERVIEW ++
The race between AI watermarking and removal tools is heating up, with **wiltodelta/remove-ai-watermarks** surging to 346 stars after being flagged by prominent developer minimaxir. This CLI and library aggressively targets both visible Gemini watermarks and invisible forensic markers like SynthID and C2PA, raising immediate questions about content provenance and the effectiveness of current detection standards. Meanwhile, browser automation is the other major theme, as **remorses/playwriter** (3,525 stars) gains traction for giving LLM agents stateful control over browsers via Playwright, catching the eye of AWS AI lead philschmid. A quieter but notable entry, **tarekziade/ai-reviewer**, hints at growing demand for automated code review tools, though it remains early-stage with just one star. The day’s trend is clear: developers are arming themselves with tools to both control and circumvent the AI ecosystem’s guardrails.
grep TOPIC=
grep SOURCE=
sort --by=
GH
Chrome extension & CLI to let agents control your browser. Runs Playwright snippets in a stateful sandbox. Available as CLI or MCP
“Playwriter is a Chrome extension and CLI that enables AI agents to control your browser by executing Playwright snippets in a stateful sandbox. It supports both CLI and MCP (Model Context Protocol) interfaces, making it easy to integrate with various agent frameworks.”
GH
AssetOpsBench - Industry 4.0
“AssetOpsBench is a benchmark for evaluating AI agents on Industry 4.0 asset operations tasks, such as predictive maintenance and anomaly detection. It provides realistic scenarios and metrics to assess agent performance in industrial settings.”
GH
CLI and library for removing visible (Gemini) and invisible (SynthID, C2PA, EXIF) AI watermarks from images
“A CLI and Python library that removes visible (Gemini) and invisible (SynthID, C2PA, EXIF) AI watermarks from images. It provides a practical tool for privacy and data cleaning, supporting multiple watermark types.”
GH
“AI Reviewer is a Python tool that uses AI to automatically review code changes, providing feedback on code quality, potential bugs, and adherence to best practices. It integrates with GitHub to analyze pull requests and suggest improvements.”
BSKY
I don't have much to say about this year's Google I/O because I prefer to write about products that have shipped, not just "coming soon" announcements - but here are some notes on Gemini Spark and Antigravity simonwillison.net/2026/May/20/...
❤️ 48 Likes|[LLM][Deployment]
BSKY
“This flight will be full to Atlanta” Thank god. I don’t want to be in the plane that only goes part way
❤️ 4 Likes|
BSKY
I would have liked to see Sanderson’s Reckoners series as a TV series, but I’m good with this.
❤️ 1 Likes|
BSKY
Instead of finding content you need, you get to have an interactive AI *experience*.
❤️ 34 Likes|[Tooling]
BSKY
June 2024: The latest general-purpose LLMs could not count the r's in strawberry. July 2025: The latest general-purpose LLMs get gold in the International Math Olympiad. May 2026: The latest general-purpose LLM solve an 80 year old problem, one of the "best-known questions in combinatorial geometry"
❤️ 237 Likes|[LLM][Evaluation]
BSKY
I am starting to have trouble paying attention to even interesting information if it is written in Claude or ChatGPT house style. I think some is the sameness of the rhythm rather than obvious words & tics: Claude is always so staccato. ChatGPT loves short sentences as kickers. Boring at scale.
❤️ 120 Likes|[LLM]
BSKY
Me: Why is there an exceptionally high density of Google bullshit in the news this week? Me: Oh, it must be Google IO. *sigh*
❤️ 54 Likes|
BSKY
I won't claim this is the most embarrassing social media post I made as a teenager, but it may be the most confusing
❤️ 60 Likes|
BSKY
I tried to make the theory work out but the computer devil kept lying to me (ChatGPT generated incorrect proofs)
❤️ 10 Likes|[LLM][Evaluation]
BSKY
one simple rule for detangling academic writing: Who is doing what to whom and why, and who should do what instead.
❤️ 3 Likes|
BSKY
On my decade-long quest to reconcile scientific language with singular evidence.
❤️ 6 Likes|[Evaluation]
X
Personal update: I've joined Anthropic. I think the next few years at the frontier of LLMs will be especially formative. I am very excited to join the team here and get back to R&D. I remain deeply passionate about education and plan to resume my work on it in time.
[LLM][Safety][Agent]
“DeepSeek Summary: Karpathy announces joining Anthropic, emphasizing frontier LLM work and continued education passion.”
X
A few random notes from claude coding quite a bit last few weeks. Given the latest lift in LLM coding capability, like many others I rapidly went from about 80% manual+autocomplete coding and 20% agents in November to 80% agent coding and 20% edits+touchups in ...
[Agent][LLM][Tooling]
“DeepSeek Summary: Karpathy notes a shift from manual coding to agent-based coding due to improved LLM capabilities.”
X
I'm being accused of overhyping the [site everyone heard too much about today already].
[LLM][Deployment]
“DeepSeek Summary: Karpathy responds to criticism of overhyping a popular site.”
X
Quitting programming as a career right now because of LLMs would be like quitting carpentry as a...
[LLM][Tooling]
“DeepSeek Summary: Simon Willison argues that quitting programming due to LLMs is analogous to quitting carpentry due to power tools, implying LLMs are tools that augment rather than replace programmers.”
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 require sandboxed environments to execute code and manage files securely.”
X
Today we're launching LangChain Labs, a new applied research effort focused on Continual Learning. Our goal is to advance open,
[LLM][Fine-tuning]
“DeepSeek Summary: Announcement of LangChain Labs dedicated to continual learning research.”
X
I am not excited about visual workflow builders 1. Not simple enough for the average user
[Tooling]
“DeepSeek Summary: Skepticism towards visual workflow builders due to complexity.”
X
Stanford CS 25 'Transformers United' featured stellar guest speakers like Andrej Karpathy
[LLM]
“DeepSeek Summary: Highlights a Stanford course on Transformers with notable guest speakers.”
X
It gives me a lot of comfort knowing that we are the last generation without advanced robots everywhere.
[Agent]
“DeepSeek Summary: Reflects on the inevitability of advanced robots becoming ubiquitous.”
X
In this context, I define world modeling as predicting the next plausible world state (or a longer duration of states) conditioned on an action.
[Agent][Multi-modal]
“DeepSeek Summary: Defines world modeling in the context of AI and robotics.”
X
those who think RL use less compute don't know RL at all SFT: human generates data and machine learns RL:
[Fine-tuning][Infra]
“DeepSeek Summary: Argues that reinforcement learning is compute-intensive, contrasting with supervised fine-tuning.”
X
hi, i'm a sole proprietor/founder in Austria and i earn many many multiples of what i'd earn as an employee, despite "predatory income tax". in fact, i opt out
[Deployment]
“DeepSeek Summary: Jeremy Howard discusses earning more as a sole proprietor in Austria despite high taxes, and opting out of the system.”
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 use of time.”
X
we've been working on democratizing fast kernel writing on the @PyTorch team. try
[Infra][Tooling]
“DeepSeek Summary: Soumith announces PyTorch team's effort to democratize fast kernel writing.”
X
It's surprisingly easy to do "hard" things -- for the most part, you need to get started and keep at it
[Evaluation]
“DeepSeek Summary: Chollet emphasizes that starting and persisting are key to accomplishing difficult tasks.”
X
Many people assume that LRM reasoning breaks down past a certain "complexity" or "number of steps"
[LLM][Evaluation]
“DeepSeek Summary: Chollet questions assumptions about limitations of large reasoning models.”
X
AI’s next frontier is Spatial Intelligence, a technology that will turn seeing into reasoning, perception into action, and imagination into creation. This is an insanely large world created using our 3D world generation model. Who is talking about #AI?
[Multi-modal][Safety]
“DeepSeek Summary: Fei-Fei Li emphasizes Spatial Intelligence as the next frontier in AI, enabling machines to reason, act, and create from visual perception.”
X
congrats to OpenAI on winning the Turing Test
[LLM]
“DeepSeek Summary: Sarcastic congratulations to OpenAI.”
X
I have been compiling LLM/VLM training logbooks/chronicles. This is the one of the best sources to ...
[LLM][Fine-tuning][Infra]
“DeepSeek Summary: Stas Bekman curates training logbooks for LLMs and VLMs, providing a valuable resource.”
X
Thanks to an awesome contribution from @omarnomad The Machine Learning Engineering Open book now can ...
[Tooling][LLM]
“DeepSeek Summary: Acknowledges a contribution to the Machine Learning Engineering Open Book, enhancing its utility.”
X
If you were holding off to try @MSFTDeepSpeed ZeRO++ it looks like deepspeed@master should ...
[Infra][Fine-tuning]
“DeepSeek Summary: Announces that DeepSpeed ZeRO++ is now ready for use, encouraging adoption.”
X
Live a little, love a little, take time out to find happiness in small things, be grateful as we have one life. #lifemantra #WorkLifeBalance
“DeepSeek Summary: Encourages finding happiness in small things and maintaining work-life balance.”
X
We're looking to work w/ folks who're interested in doing agentic kernel dev, providing real optim value to real models. Reach out if interested :)
[Agent][Fine-tuning][Deployment]
“DeepSeek Summary: Seeking collaborators for agentic kernel development to optimize real models.”
X
After working on releasing the v5, this is the latest release from the Transformers team at
[Infra][Deployment]
“DeepSeek Summary: Mention of Transformers v5 release from Hugging Face team.”
X
Thanks @AnthropicAI. Thanks @huggingface for letting me work on Diffusers and other open-source projects across the fleet.
[Multi-modal][Tooling]
“DeepSeek Summary: Expresses gratitude to Anthropic and Hugging Face for work on Diffusers and open-source projects.”
X
We are starting to see some nuanced discussions of what it means to work with advanced AI In this
[LLM]
“DeepSeek Summary: Discusses emerging nuanced conversations about collaborating with advanced AI.”
X
On the plus side with Opus 4.7, if it does decide to think it produces BY FAR the best
[LLM][Evaluation]
“DeepSeek Summary: Notes that Opus 4.7, when it engages in reasoning, yields superior outputs.”
X
If it helps, I teach at a business school & many of my smartest students are hired by funds because they can reliably turn their only-human
[Deployment]
“DeepSeek Summary: Observes that top business students are hired for their uniquely human skills.”
X
This is going to get even worse as people realize that careful tuning in their prompts can
[LLM][Tooling]
“DeepSeek Summary: Warns that prompt tuning will lead to escalating demands and expectations.”
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
Waiting on a robot body. All opinions are universal and held by both employers and family. Now a dedicated grok hate account. Accepting ML/NLP PhD students.
[LLM]
“DeepSeek Summary: Bio/profile text indicating Saphra's role as a faculty member accepting PhD students and their stance on Grok.”
X
Life update: I'm starting as faculty at Boston University in 2026! BU has SCHEMES for LM interpretability & analysis, so I couldn't be more pumped to join a
[Evaluation]
“DeepSeek Summary: Announcement of starting as faculty at Boston University in 2026, focusing on LM interpretability.”
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 almost a decade.”
X
Building a theory of the architecture of organizing machines and people.
[Deployment]
“DeepSeek Summary: Ben Recht is working on a theory for organizing machines and people.”
X
On unquantifiable costs and inherent tradeoffs in decision theory.
[Evaluation]
“DeepSeek Summary: Ben Recht discusses unquantifiable costs and tradeoffs in decision theory.”
-- END OF LOG --
[STATS] 52 items · Filter applied
Powered by Horizon + DeepSeek