A lot of the concentration comes from companies just gobbling others, and therefore being more large conglomerates than anything else. Alphabet could easily be 5 separate stock tickers if they felt like it, and most would probably be big enough to be in the index. You could say the same thing of Meta and Microsoft. Splitting AWS from Amazon would also give us two very real, top of the line companies.

So the concentration might be a lot less than it seems, just because a lot of the AI play might be as low risk as, say, when Meta sank so much money into VR, or Amazon decided that smart speakers were the future. A lot of profits are spent if it all fails, but the engine of the company is still sitting there, being the same money fountain it's been for the last decade. It's just that they are, in practice, investing on what would be a new company with the proceeds, instead of doing stock buybacks, providing massive dividends, or reinvest in the existing verticals.