You're right - I used Claude heavily as an editor and co-writer in my 'Industrial AI Publishing' model. I generated drafts with GPT, then refined everything with Claude for tone and coherence.
I was transparent about this process because I thought 'AI-assisted satire about AI' was meta-interesting. But maybe that's exactly why it doesn't sell: people smell the AI sycophancy you mentioned.
The brutal question: can AI-assisted satire ever feel genuinely sharp, or does the tool's politeness always dull the blade? I'm starting to suspect the latter.
That's the core problem: I didn't have beta readers. I went straight to publish thinking 'good satire sells itself.'
My assumed audience: tech workers frustrated with algorithms, digital minimalism crowd, people who loved writers like Mark Manson or Tim Urban's Wait But Why.
But I never validated that audience actually EXISTS on Amazon, or that they'd buy books vs just reading free blogs.
Zero reviews = I can't even test if the content resonates. It's like shouting into a void with no echo back.