I would say just the opposite. There isn't a business too small to benefit from ML/AI. That is, assuming you acknowledge that there's more to ML/AI than "deep learning". Not everybody needs a deep neural network. Sometimes you just need linear regression or a random forest.

If you come at it from that point of view, the knowledge you gain from taking Andrew Ng's machine learning course is enough to create value for companies / organizations.

A lot of the discussions on this topic here on HN seem to be based on an assumption that you have to be doing cutting edge original research to be useful. I think that's very far from the case.

Now the problem is, does the mom and pop bakery on the corner know that they could benefit from ML? And perhaps the more salient question is "are they looking for somebody to do ML for them"? The answer is probably "no" in both cases, so you might have to do some work to sell to that market. But the value you can create is real. Help them optimize production so they throw away less bread on slow days, etc. and you're talking direct business impact.

I suspect that there will be, for a long time, a wide continuum in terms of how ML/AI skills can help create value. Which means there will be a lot of ways to leverage this field from a career standpoint.