Was a good read, but it's not really more complex than the easy stuff has been discovered, and it's the same thing everywhere.
The next generation of chip fabs (the factories that make computer chips) is expected to cost nearly $30B. That's 30 BILLION dollars for a single factory. Just to build it, it doesn't even account for running it 24/7 for several years. Of course they produce a ridiculous amount of chips but this is the same price tag as ITER, the international fusion experiment that was founded in the late 80's and has a huge international cooperation.
There are critical elements in the software behind artificial intelligence that cost billions of dollars to write, because it's built on generations of prior software and involved thousands of people working for several years. And of course that's also true of the hardware. Without generations of game-oriented computer hardware, at a cost of several trillion dollars of R&D and commercialisation over several decades, none of it would be possible.
The easy stuff has mostly been found, it's no longer really possible to just fool around and make a major discovery. Decades ago it was basically the norm, since it turns out that scientific discovery and technical progress is largely a brute force thing. It's also a lot easier to be in the first group to look for gold on unexplored territory, where it will sometimes sit right there on the ground. After a while, you need some damn heavy equipment to find, refine and process it. Still worth it, but a pick and a shovel won't do it anymore.