Edit: see tptacek’s reply. He knows that corner of the world far better than I do. From my external observation there are two main career paths: the math side

Edit: see tptacek’s reply. He knows that corner of the world far better than I do. From my external observation there are two main career paths: the math side and the coding side (djb does both, but he’s djb). For the math side, a PhD is the most likely path to break in. If you are some kind of autodidact genius and were to publish a theoretical attack on an important crypto system, you could probably get involved without a PhD. On the coding side apparently the best way to break into writing secure code is to first break insecure code+. I don’t know if this is a reasonable filter or not, but it seems to exist. Some people are perfectly happy to stay on the break code side of things and never move to the write code side. If you think you might like this sort of thing check out the cryptopals ctf that should be floating around. + meaning find a side channel, some place where a nonce is reused, etc, etc. If you find a way to factor quickly in a particular ECC ring that’s the prior paragraph.