Am just thinking about this now, so not high confidence.
I'm thinking a total market cap index of all publicly traded companies in the world -- unless some stock exchanges are really sketchy for some reasons (I'm aware "sketchy" is ill-defined).
Also maybe a fraction (10%?) stored in a total market cap of deflationary currencies (decentralized ones I suppose, as for centralized ones it's generally harder to tell if they will become inflationary) as they will presumably become more in demand, in the short term as the economy might transition to them, but also in the long term as the population grows and so those the demand for currency (although if minds start running faster, I think that would decrease the value of currencies as they would be traded more often).
Maybe worlds in which reanimation is possible is correlated with other technological development, and so that might also inform some investments, but not sure which ones, and that starts adding a lot of complexity, at least while we don't have large prediction markets.
I'm also thinking those investment don't need to be liquid at all, and REITs often have a higher return because they are less liquid than stocks (AFAIU); so if this is true, then maybe having a lot of investments in REITs would also be beneficial. (2021-05-08 update: this is also true of other investment vehicles, like loans)
My intuition was in the opposite direction: in worlds where technological development is quick, you are very likely to have the funds last long enough anyway, so you should optimise for worlds where the development is slow. There are effects in both direction, so the answer likely depends on how big your initial fund is compared to the costs.