The article “The trust collapse: Infinite AI content is awful” by Arnon Shimoni’ is brilliant.
His core point:
In 2024: A credible-looking cold email cost real money and human labor → you could trust most of them. Pattern recognition worked ~80% of the time.
In 2025: It costs ~0€ → you can trust none of them. Pattern recognition works ~20% of the time.
So prospects don’t even try to verify anymore. They just assume everything is noise.
If I want a newsroom or a platform to believe that my detection model will still work (and that I’ll still be around to fix it) when deepfakes get 10× worse next year, I can’t sound like ChatGPT with a Notion template.
Because in this space, trust isn’t a nice-to-have. It’s the product.
Time to invest in the trust funnel – because people out there are swamped with AI SDRs generating hundreds of messages.
Trust is still a human job. You need to show you’ve been showing up before you try to solve other people’s problems.
Anyone else in AI safety / detection / verification feeling this shift in outbound right now?
(Arnon’s full article – read it, it’s brilliant → https://arnon.dk/the-trust-collapse-infinite-ai-content-is-awful/)
Leave a Reply