“The next frontier” is where those with money want it to be, more or less. It was the government contracts plus a few scientists that started all that computer stuff going on now . And today the government has virtually everything it needs, except maybe more data, more control on information. The public face recognition tech is already good enough, imagine what the military can do. Also, it can’t be the next big thing, because the application is too narrow.
But the government is not the only player anymore. As @voice_of_raisin said, there are Fortune 500. But it doesn’t really matter if something will become “obsolete” or not (maybe it will, we don’t know), because new industries, as @guzey already stated.
What does matter, is the people, who own those companies. They are extremely wealthy, and they are going to die pretty soon (relative to the future possibilities), if nothing changes. There are already some big players investing in biotech. Alexey, correct me if I’m wrong, but I believe there are no problems with money in a bio research at all .
So, combining those points, my best answer for this thread would be something like user-friendly quantified self tech, with maybe a little more biology going on. It has everything for everyone: data for the government/corporations, longevity/extreme lifestyle quality boost for individuals, AI/NN for those on a hype train, and it will be a part of the IoT market, which is a very promising one, too.
The next company to combine all those things and make it relatively accessible will win. Neuralink? And, oh man, since everyone likes dopamine addiction so much: if you add some virtual reality/general entertainment stuff to it… What, Neuralink again?
 The first high tech company (which started the Silicon Valley) was the Shockley Semiconductor Laboratory, which was a division of Beckman Instruments, which was and is, basically, a scientific/medical devices company. And Arnold Beckman—brilliant guy—was, mainly, a chemist. A large part of his inventions, surely, were sponsored by the government—as stated in Other secret projects, Computers and automation on his wiki page.
 Well, except there’s a problem with people, maybe? I am probably biased, because I don’t know shit in bio, but I also don’t think there are enough resources to get started easily in this field at all. Furthermore, it is a highly academical one -> higher education/academia is in a crisis now -> the progress in bio is even more delayed.