100% this. ML has been very hyped for a while now and having it was seen as a badge for the company. To be fair, ML is not also something that was historically central to a degree, so many people wanting to get into AI, even good engineers, did not have the background in it. This too is changing though, but the hype and the lack of an experienced pool doesn't help.
I don’t think the background is really that important tbh.
From physics I have a good theoretical grounding in how ML works (optimizing a cost function over a high dimensional manifold to reconstruct a probability distribution, then using the distribution for some task) but I personally find actually ‘doing ML’ to be rather dull.
