"Unsurprisingly to me, the tech boosters didn’t present any scientific arguments. I only saw unfounded assertions about Bayes’ Rule and Blackwell’s theorem"
This is a bizarre characterization of the extremely basic, sane, sensible, and scientific argument that has been made by many: if the concern from doctors is that one test is not enough to reach a determination about something, test more until you have adequate data to make a diagnosis or a determination that a more invasive or expensive procedure is justified. You can say "there's no substance to the idea that getting more data helps" over and over and it will still be wrong, little better than a lie. Data proponents have rightly pointed out that if doctors are concerned patients will worry about incidentalomas, you can use time series data to check growth rates and that would be a useful guide to deciding when something more invasive or expensive is worthwhile. Doctors could use composites of as many noninvasive tests as possible at the branching points of their decision trees before proceeding. Is the problem according to you or the medical establishment that the specific tests are at issue? That tests today aren't good enough to operate in that regime? Then the culture should strongly favor the development of new cheap fast noninvasive tests, and widely welcome any addition to the arsenal, so we can get there.
There is a kind of absurd barbarism to the idea that we have to live in medical dark ages, where "testing considered harmful" is a heuristic that shapes the culture and discourages progress. Dismissive, hostile attitudes from the medical establishment are actually bad, with long-term costs to people's lives. The majority of the discourse isn't even about the Midjourney machine, which is a completely unknown quantity for now, but that the total rejection of new diagnostic tools on the grounds that more data is bad will clearly lead to worse diagnostics in the long run.
