Ask HN: Is AI overcrowded? Where should one look to be useful

I'm thinking about what to major in college and ultimately what to do with my life and have concluded that AI is one of the most important things one could work on and I also have quite an interest in it, I also find robotics fascinating. The only thing I'm concerned about is that it is extremely popular at the moment and possibly oversaturated. This brings up issues such as marginal effort. If the field really is overcrowded, then what difference am I really making by just joining in? Looking at the future and trying to predict anything with great success is incredibly hard, but it seems like the most important things for this century are AI, aging/bio/health, energy and development of space.

Now I could be completely wrong on this, but it seems like energy is being taken care of. Simple economics will create a demand for renewables and there does't seem to be a shortage of people working on solar, wind, water, storage, fission. Maybe fusion is worth looking at but I feel like the energy crisis will be solved by the others and fusion would be an added benefit. Biotech is an interesting area which I haven't considered, but it seems like computers / AI will be the major useful tool in this area also (this could be wrong, I've just read that biology is really data heavy). Space development is pretty broad, chemical rockets have been maxed out and the main differentiating factor between companies now is software (raptor engines pretty much hit the limit for chemical engines). And really any of the other fields mentioned (AI, biotech, energy) would be useful to work on if one was interested in space development.

Basically, I'm at a crossroads and really just want to be useful. As an experiment, if you could decide the trajectory of a freshmen interested in these things and wanting to have a positive impact, where would you suggest they go?