At least I'd know I have a job waiting for me, supporting "legacy" code still running in production.
In the Dyson swarm scenario, assuming constant population, the productive capacity per person is 2500 times higher than the current global output, and a random act of rounding-error charity would give you more than Musk, Bezos, Zuckerberg combined, and it won’t matter if you have nothing (unless your sense of self worth exists only by the size of your bank account).
In the stagnant scenario, you might be OK with property, if you can be sure it will be maintained in the meanwhile.
In a decay scenario, I wouldn’t even want to guess what’s resilient, but houses will be fairly redundant due to oversupply.
But those scenarios were deliberately chosen as illustrative outliers; there are options between them, and I would do no better with making the right choices there — a limited AI that makes one sector redundant would destroy any investment I made in that sector while also making the output of that sector too cheap to bill separately, much as digital computers have done to the human profession of “computer”.
I don’t think it’s possible to forecast economics on those scales. If I was cryonically preserved and woken up in 2122, I would be equally willing to believe we were declining or stagnant for the reasons you give… or that we had solved AI and von Neumann replicators and that the planet Mercury had been fully dismantled and used to construct a Dyson swarm whose descendants had already done the same to a similar planet around each of the nearest dozen or so stars.