This strikes me as rather uninformed. Physics is far from a completely undiscovered realm. The standard model of particle physics (completed by the mid-70s) + general relativity (ca 1915) fits every experiment ever performed. Unexplained astronomical observations can be accommodated by additions which, while phenomenologically important, add nothing conceptually new. The remaining Holy Grail, quantum gravity, suffers no lack of proposed solutions; what it really does suffer from is a lack of ways to test any of them.
The sad reality of "fundamental" physics is, and has been for decades, that there is really very little that is both doable and worth doing. If somebody were to dump a Genius grant on me today, with the only condition being "work on whatever you want in physics for the next 5 years"... I would feel compelled to give it back. Because I really don't see anything sufficiently promising or interesting in physics that I would be willing to dedicate the next 5 years to it. (All those people you see screaming for more physics funding? What they are really screaming for is their "right" to keep comfortable upper middle class lifestyles, funded by the tax payers, in exchange for nothing of any value.)
Now compare that to what is going on in software. If "new login scripts" is the most interesting thing you can think of, you really need to broaden your horizons. Just look at the last few years of very rapid progress in deep learning. There is nothing comparable going on in physics. Things like thought vectors are giving me vertigo - my old skepticism about the feasibility of general AI has been seriously dented lately. Problems which once looked completely intractable suddenly look easy, and you start wondering how many more surprises are just around the corner, and how many - or rather few - years it might really take to get there.
And the potential? Suppose you managed to write down the ultimate, complete, fully quantized, elegantly unified, simple, concise, true Theory of Everything. There it is, an exquisite formula on a sheet of paper. It summarizes everything we know about physics. Everything around us follows from it. Wonderful. But... what do you do with it? Sure, stare at it, marvel at its ingenuity and the centuries of concerted effort needed to get there... but then what? How long can you stare at a sheet of paper and marvel at the history and elegance of an equation before it gets a little... old?
Now compare what AI, even just the narrow kinds emerging right now, can do, and the potential of true general AI. Now that would - will - obviously change the world.