WebMCP By Google, Clearly Explained!
Everything you need to understand WebMCP and expose your site's actions to an agent.
Lire
Architectural Guardrails for AI-Generated Code
Consider a composite of a failure pattern that’s becoming increasingly common on teams that have scaled AI-assisted development past a handful of enthusiasts. A staff engineer named Priya opens a pull request. The PR is 340 lines and adds an endpoint that writes to the customer table directly, bypassing the internal customer service API. The […]
Lire
Understanding ChatGPT Work
OpenAI announced ChatGPT Work on July 9th, and have been furiously iterating on it ever since. It is an extraordinarily confusing and very powerful product. Here's what I've figured out about it so far. ChatGPT Work is actually two products The more interesting version of ChatGPT Work is the one that runs in the cloud. This can be accessed via chatgpt.com or through the ChatGPT mobile apps. Let's call it Work Cloud. If you install the ChatGPT desktop app - the app that used to be called Codex -…
Lire
Why KV Cache Stores K and V Vectors But Never Q?
A popular technical LLM interview question.
Lire
Introducing Hy4 Preview
Introducing Hy4 Preview New open weight text input (no vision) LLM from Chinese company Tencent today: 770B total parameters, 49B active parameters, 1M token context window, 1.56TB on Hugging Face. This is a big size increase from their previous Hy3 in July, which was 295B, 21B active, 256,000 context, 598GB. I recently started using model chat templates to better understand their capabilities. Here's Hy4's chat_template.jinja on Hugging Face, which includes this section: {%- if not…
Lire
Just a rumour of a bug is enough to find a security exploit these days
Just a rumour of a bug is enough to find a security exploit these days
Just a rumour of a bug is enough to find a security exploit these days Anil Madhavapeddy is a professor of computer science at Cambridge and a core maintainer of the OCaml compiler. In this somewhat alarming post he reports that security issues in OCaml projects are seeing evidence of attempted exploits within minutes of patches being shared for discussion: This normally takes a few days and a release within a week or two is reasonable. Within about ten minutes (!) this website was fielding…
Lire
Implementing a Siamese Network with Contrastive Learning
Full guide with code.
Lire
Tokens Aren’t Dollars
The following article originally appeared on Tim O’Brien’s Medium blog and is being republished here with the author’s permission. AI costs are easy to count and hard to understand, and judging effort by a token volume? While that might feel like a valid measure of value or complexity, it doesn’t capture the details that define […]
Lire
Writing really is an Emotional Rollercoaster
[[Writing]] has become my passion over time; I have built a whole company based on technical writing services (even though most say that’s impossible). After doing it for 11 years publicly (much longer privately), I still enjoy it every day. If I could have one dream, I’d probably imagine the book-author lifestyle, where I go to a faraway place or island and just write in my journal, articles from my vast notes system (~3 million words, or 2-3 books.
Lire
Breaking Claude Code Opus 5 Auto Mode
Breaking Claude Code Opus 5 Auto Mode
Breaking Claude Code Opus 5 Auto Mode Anthropic are putting a great deal of faith in Claude Code's auto mode for protecting their coding agent users against prompt injection attacks. They recently made that the default and have made bold claims about its effectiveness. Johann Rehberger is one of the most credible prompt injection researchers active today. He found an attack against auto mode which he claims works 80% of the time, by tricking Claude Code into downloading and uncompressing a zip…
Lire
KV vs Prefix vs Prompt vs Semantic Caching
...explained with best practices in production.
Lire
Writing with AI
A lot of professional writers say they never do it. And for writers who do, the consequences can be serious. Book contracts have been withdrawn. People have lost jobs. I have a contrarian view. I say AI is a medium, like painting, sculpture, photography, music, or language itself. If AI is a medium, it is […]
Lire
The Identity Crisis No One Planned For: Governing Nonhuman Agents at Enterprise Scale
For a decade, identity and access management meant one thing: governing the humans who log in. Employee joins, gets provisioned, gets a manager, gets a departure date, gets offboarded. That loop is well understood. What changed is that the fastest-growing population inside enterprise environments is no longer human, and the governance playbook written for people […]
Lire
Qwen3.8-Flash-Next
Qwen3.8-Flash-Next
Qwen3.8-Flash-Next Another open weights model from Qwen. This one is "a multimodal MoE model that also serves as an early preview of the architecture used in Qwen4". It's pretty big: 125B tokens, but only 6B active which means it gets a significant performance boost. I've been trying it out on a DGX Spark using these Unsloth quantized models. I'm still exploring the model - so far I've tried the 72.5GB UD-IQ1_S one (producing these pelicans) and the 78.9GB UD-Q2_K_XL (producing these). My…
Lire
![[Hands-on] Turn Scientific Figures Into Structured Data with Mistral OCR](https://substackcdn.com/image/fetch/$s_!swtY!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fff5a5719-b825-4c5e-8ebd-0ba1a961ff8f_1456x808.png)
[Hands-on] Turn Scientific Figures Into Structured Data with Mistral OCR
A full walkthrough of the extraction schema, with code.
Lire