> Tesseract is vastly worse than AI OCR. But the damage a Tesseract error can cause is bounded. The damage of what AI can hallucinate is practically unbounded.

AI OCRs aren't LLMs (unless you use ChatGPT for this), so this argument seems invalid?

> AI feels grievously inefficient. It took 29.29 minutes to OCR the 4-page handwritten draft of this essay with Qwen3-VL:8B.

Isn't this using the wrong tool for the job?

> AI makes Nvidia rich, and I don’t like Nvidia because their Linux support sucks ♥

I don't think LLMS really benefit nvidia or make them rich. (Author is referring to LLMS as AI in this blog post). They use completely different technologies

> AI is centralized even though that’s bad architecture. Check what server you’re connecting to when working with local LLMs

None. If it's local, I'm not connecting to any server. I can run my models offline