“A spate of outages, including incidents tied to the use of AI coding tools”, right on schedule
Some with “high blast radius”
Lire
Prompt Caching in LLMs!
A case study on how Claude achieves 92% cache hit-rate.
Lire
Claude Code, Claude Cowork and Codex #5
It feels good to get back to some of the fun stuff.
Lire
Anthropic sues US government, with good reason
As I wrote yesterday, Dario Amodei is no saint, but I fully support his company’s new lawsuit against the US government.
Lire
Import AI 448: AI R&D; Bytedance's CUDA-writing agent; on-device satellite AI
If Ukraine is the first major drone war, when will there be the first major AI war?
Lire
A Practical Guide to Evaluation of LLM Apps (Part C)
Understanding evaluation of conversational LLM systems, toolcalls, tracing, and red teaming.
Lire
There are no heroes in commercial AI
When it comes down to it, Dario Amodei isn’t all that much different from Sam Altman
Lire
Why I Still Blog — and Why the Future of Blogging Is Connected
Why I Still Blog — and Why the Future of Blogging Is Connected
I’ve been online twenty years, and blogging for ten of them. This is the story and lessons learned of blogging online for a decade. It goes beyond blogging topics and includes note-taking (workflow), how to write well as well as the medium in which writing works best, and also the format in which writing works long-term such as writing in open formats and methods such as vim motions to navigate and edit like a surgeon.
Lire
Anthropic Officially, Arbitrarily and Capriciously Designated a Supply Chain Risk
Make no mistake about what is happening.
Lire
Don’t trust Generative AI to do your taxes — and don’t trust it with people’s lives
“The problem comes down to how A.I. chatbots are fundamentally designed”
Lire
AI #158: The Department of War
This was the worst week I have had in quite a while, maybe ever.
Lire
Gemini 3.1 Pro Aces Benchmarks, I Suppose
I’ve been trying to find a slot for this one for a while.
Lire
Git for Data Applied: Comparing Git-like Tools That Separate Metadata from Data
Git for Data Applied: Comparing Git-like Tools That Separate Metadata from Data
Continuing from Part 1, where we learned what git for data is, how the architecture and use cases work, how you can achieve git-like functionality with different approaches, and how the key is to avoid moving data as much as possible to keep state that can be referenced and rolled back to, but at the same time saving cost by not duplicating all data every time you create a new branch.
Lire