samek.fyi

Cattle not pets, for software developers: The shift from handcrafted servers to immutable infrastructure taught us that mutability is the sworn enemy of understanding.

I’m trying to get into a new habit.

There’s an old business idea, usually attributed to Michael Gerber, about the difference between working in your business and working on your business.

This is a list of AI benchmarks I’m watching.

I had an deck in Anki, a flashcards app, with an annoying problem that I had caused when I imported the cards.

I’ve been doing a bunch of keeping up with and experimenting with AI.

Whenever a new model is released, I read the benchmarks.

I tried Dia, a new Chromium-based browser with built-in AI features from the creators of Arc.

Simon Willison released tool calling in his llm CLI.

Google released an impressive improvement to Gemini, which now tops the WebDev Arena Leaderboard.

Kagi benchmarks LLMs on a private benchmark.

I’m Brian Samek.

I keep seeing descriptions of Model Context Protocol (MCP) that are either vague metaphors or get too specific too quickly.

Deep Research (three separate products!

A few days ago I wrote that Perplexity is better for quick searches.

o3 and gemini 2.5 pro being a single Elo point apart on LMArena with style control feels very steam-engine time.

Simon Willison writes that “AI assisted search-based research actually works now.” o3 and o4-mini agentic research is truly impressive, as are ChatGPT’s and Gemini’s Deep Research.