When Gemini Pro came out, I found the thinking traces to be extremely valuable. Ironically, I found them much more readable than the final output. They were a structured, logical breakdown of the problem. The final output was a big blob of prose. They removed the traces a few weeks later.

That’s kind of funny since a Chinese model started the thinking chains being visible in Claude and OA in the first place.

Maybe there was no thinking.

It also gets confused if the entire prompt is in a text file attachment.

And the summarizer shows the safety classifier's thinking for a second before the model thinking, so every question starts off with "thinking about the ethics of this request".

Genuine question, why have you chosen to phrase this scraping and distillation as an attack? I'm imagining you're doing it because that's how Anthropic prefers to frame it, but isn't scraping and distillation, with some minor shuffling of semantics, exactly what Anthropic and co did to obtain their own position? And would it be valid to interpret that as an attack as well?

Attacks? That's a choice of words.

Proprietary pattern matcher proves there's no moat; promptly pre-covers other's perception.

Very cool that these companies can scrape basically all extant human knowledge, utterly disregard IP/copyright/etc, and they cry foul when the tables turn.

and so does OpenAI

But still CoT distillation WORKS. See the DeepSeek R1 paper.