Sure do wish that people would just stop blindly reprinting a company's claims of AI use, without asking what exactly they mean.
Like, rad as hell if this works. But when they say AI, do they just mean an algorithm? At this point, the term has become virtually meaningless, almost becoming synonymous with "computer technology".
Didn't seem that complicated or at least that was the way it was explained, just a matter of checking whether the child was looking more at faces, eyes or objects. Does seem like something some AI could do.
The point is that the 'tech' removes the long and expensive 'proper' medical diagnosis that is currently required in any reasonable system. It is replaced by a set of commercially validated parameters that have not necessarily been tested. AI is as dangerous as poorly trained corrupt doctors in league with corrupt officials.