The idea that a field can be reformed by making it worse until it suddenly faces a reckoning and emerges much better is ... I don't know where people keep getting the idea that this might work. It has never worked in any field ever in history. The thing that can happen: fields gradually split into rigorous and non-rigorous camps. Like with evidence-based medicine, or chemistry/alchemy. Depending on the field, either…
All modern fields of study use quantitative measurement so splits in practice now would have to do with different ideas of rigor rather than rigor-nonrigor.
The problem of AI/ML isn't the blatant cheating but the way that the goal is often getting n% higher than soto on X benchmark. Just chasing benchmarks makes any connection to broad dubious imo. It might or might-not give you something practically useful but definitely give you something career-wise useful.
But just as much, when the field is just a giant race where no cares about any broader understanding, cynicism seems like a natural result. The ideal of academia, for all it's failings, is to give people some amount of space to speculate and explore wider vistas.
It would be good if X number of people had the space to explore a variety of visions of "AI" other than the dominant one. But despite the vast number of people being sucked into the field, my guess is this is getting harder, not easier. And, of course, the mere appearance of "rigor", of quantitative measurement, is not helping things, again in my opinion as someone of no authority at all in field.