Should patient investors try to correlate portfolio holdings with potential cause areas?

I am trying to determine what the optimal portfolio holding pattern for a patient longtermist is, as I have moved a large portion of my savings into a taxable account based on the ideas presented by Phil Trammell in episode 73 of the 80K podcast. However, I am having trouble thinking through which index funds to invest in (I am investing through Vanguard, but the funds are generic enough to be broadly applicable).

I have seen a previous poster on the forum discuss a portfolio breakdown of 70% generic U.S. stock market, 10% generic U.S. bond market and 20% generic international stock market, while a commenter advised a much greater investment into the international stock market, as it was likely to have a higher “catch up” rate of return.

However, I wonder if there are better ways to maximize the expected impact of this investment by trying to account for potential cause areas or to capitalize on “hinge-y” moments. For example, if artificial general intelligence is created in the next 50 years, it might be better to have invested in an Information Technology index fund instead of the broader market, as presumably the value of the creator and similarly situated entities (likely public traded companies in my opinion) would have gone up significantly before this creation.

An incredibly simple toy model of this idea is as follows: Specific Portfolio A has an expected return of 10%, derived from a 20% chance of Event A happening, in which case the return is 50%, and an 80% chance of Event A not happening, in which case the return is 0%. Contrast this with Generic Portfolio B, which has a 15% return regardless. If you care a lot about having resources to respond in case of Event A happening, then investing in Specific Portfolio A is a good idea, even if Generic Portfolio B has a higher expected return.

This is similar to the idea of investing in negatively correlated assets. In this case, the investment (and thus our available resources) goes up when the problem becomes worse (or when the opportunity to make a significant change gets larger more broadly).

Of course, this toy model gets intensely complex when considering the multitude of potential key events, adding in uncertainty around the correlation of events and portfolio returns, thinking about potential transfer of funds between portfolios over time, etc.

Phil Trammell seems to mostly have discussed long-term investing on a grander scale (over hundreds if not thousands of years), which would imply simply investing where expected return is highest straight up. But I do worry about expropriation of long-term investments or the chance that we discover a large inflection point (an Event A like the creation of artificial general intelligence) within the next 100 years, as to necessitate slightly shorter-term focus.

Does it make sense to try to correlate portfolio holdings with potential cause areas, or should patient longtermist investors utilize more typical portfolio management theory?