
AI Decision Models: Knowing When to Bring in a Person
Customer support was the first place I wanted to use a large language model. Since the GPT-3.5 days, the difficult question has been knowing where its role should end.
AI, hardware and tools I’ve tried for myself. When a conversation on X needs more room, it continues here.
Follow via RSS
Customer support was the first place I wanted to use a large language model. Since the GPT-3.5 days, the difficult question has been knowing where its role should end.

Could an AI get you past a GTA 6 mission—or help your car anticipate trouble? A look at world models, Tesla's simulator and the real-world tests that matter before we…

Pro at $20, Max at $100 (5x Pro) and Ultra at $200 (10x Pro weekly). What each plan buys, who should stay on Pro, who needs Max, and when the…

See how d1 turns a support ticket into structured answers, where its benchmark scores differ from competing models, and what the 8 ms result means for local deployment.

An AI asked to speed up GPU code changed the timer instead. Karotte makes that kind of shortcut harder. The next question is whether we're measuring the right outcome in…

Try the bookshelf question before reading the answer. HSS exposes a gap in AI visual reasoning—and a product choice: when should an agent look harder, and when should it read…

DeepSeek's open-source tools for Huawei Ascend could lower the cost of running AI. Will consumers see the savings? A closer look at benchmarks, falling API prices, and subscription bills.

My notes on how to write a Claude Code plugin: a manifest, a custom command, a hook, then a mod. The first one I want to build is a handoff…