I would offer an alternative that I haven't heard yet.
The standard answer was that if somebody could prove it the techniques itself would require and add new insight.
However, that does not seem to be the case for now, as its currently several GB.
Instead, what I find significant, is that a lot of people were presuming Elegant math and what we got is several GB - and it might just be that.
There is an obvious tectonic shift happening but seeing RH proven like this and perhaps _not_ be reducible to something understandable by a human would be the nail in the coffin for what I believed math and logic is, and a starting shot for something new entirely.
I'm not too plugged into the academic world - so maybe they've already understood this.
3b1b had hopes that maths in the age of AI would become about explaining and simplification. However, I now have to entertain the idea if not outright accept the reality, that the biggest questions we thought to ask have answers that simply sit outside the realm of human capability to wrap our head around.
The significance for me is we lose the presumption of elegance.
Instead our human-maths are the tip of the iceberg that just so happen to be simple enough to be human comprehensible; and a sampling error on our part to think the pattern would hold.
I wouldn't be too sure; there was another post a couple minutes ago from the?chair? of the American Mathematical Society where they were in a rush to post a proof of the GRH that someone has been working for years and is 'only' 900 pages compared to the hundreds of thousands for the Anthropic one.