I still don't think top hiring bars will lower if supply increases ( and they shouldn't). ML is a hard field with a lot of complicated math - you can definitely teach a focused individual to use it in a month or two and hire them - but, on paper, they're still not going to stack up against masters/PHDs from top schools who understand all the math involved, are able to expand the toolset as needed, have proven experience tackling research problems, etc. Just my two cents.
You're right. Simply increasing supply of candidates will not increase demand from employers for those candidates.
However, if ML continues to become more and more important for businesses to remain competitive (and there seem to be a lot of indications that it will), you can bet that demand will increase and then, eventually, supply. Perhaps you could even argue that that is what's driving the current increase in supply.
Either way, companies will probably eventually feel a lot of pressure to look beyond academic credentials if they want to actually build up teams focused on those technologies.
I definitely agree; something like that may happen, though, I'm not sure how sustainable or successful that would be. I remember someone posting a famous diagram awhile back about the adoption curve of technologies, and how with ML we may be approaching an unsustainable glut that will even out soon. We'll definitely see how these things play out over the coming years.
