Ask HN: What research areas would you consider if you were to start a CS PhD?

I encourage you to think in terms of what is achievable thiout a CS PhD. What areas don't tend to be dominated by PhDs, and what areas are dominated by PhDs.

My conclusion: Theory, algorithms, cryptography, and machine learning.

There are tons of world experts in operating systems, graphics, databases, distributed systems, compilers, programming languages, computer architecture, security, and HCI without PhDs. In fact, I would argue some of these fields are dominated by non-PhDs (see: Ken Thompson, Linus Torvalds, etc.) I don't see a lot of non-PhDs doing cutting edge work in CS theory, algorithms, cryptography, or machine learning. Probably because these fields involve lots of theory and mathematics.

Just a thought.

HCI. The 'front-end' encompasses far more than just web dev, and some of it is quite interesting. VR? Smart cars? AI? People all need to interact with them.

Even standard web app problems can get quite complex on the UI. For example, recently I've had to use finite automona to verify that a combination of options on a form can result in all possible valid combinations. Since it was quite complex, I was worried that there was either: a) an invalid path in the options, or b) a valid but unreachable path.

Machine Vision and Robotics are at a crossroads where fundamental research is starting to cross over into industry, but the technologies are still far enough away from the mainstream that a Ph.D will give you a huge advantage.

That said, if you just want to do a Ph.D but are struggling to figure out what to do a Ph.D in, I would seriously re-evaluate why you want to do a Ph.D.

Maybe 'Crypto Smart Contracts', eg. Ethereum/BitShares- or 'Distributed Computing' eg. MaidSafe.

Why? Just personal preference, I think these are very interesting, & hold much potential.

Quantum computing!

Compilers, it's an area that has held my interest for at least a decade.