OK, I will say this later and perhaps more polished-ly on my blog, but here goes:
I will refuse to review any academic paper that credits a chatbot in the acknowledgements. Why? For much the same reason that I would refuse to review a paper that studied conflict diamonds just because they were convenient.
The human cost is unacceptable:
https://www.theguardian.com/global-development/2026/feb/05/in-the-end-you-feel-blank-indias-female-workers-watching-hours-of-abusive-content-to-train-ai
The industry has no line against using the output of the Nazi CSAM generator and calling it "training data".
https://www.theguardian.com/technology/2026/jan/24/latest-chatgpt-model-uses-elon-musks-grokipedia-as-source-tests-reveal
https://www.theverge.com/report/870910/ai-chatbots-citing-grokipedia
Any marginal benefit that an individual scientist thinks they see is just saying, "Wow, the bus to the conference runs so much better on leaded gasoline."