Cashing Out

David Krueger
2 min readSep 30, 2015

--

TL;DR — Reasonable(?) Strategy: pursue instrumental goals exclusively up to some point, then “cash out” and pursue terminal goals exclusively.

Bill Gates is a pretty good example.

Given a finite horizon (e.g. your lifetime) for maximizing something (e.g. reduction in global poverty), it seems that a reasonable strategy would be to invest in whatever has the highest return (e.g. yourself) in terms of “power” (e.g. $) up to a certain point (e.g. when your company Microsoft has made you the richest person in the world), and only then actually start “cashing out”, that is, using your power to generate utility.

Practically speaking, it seems like the investment in 3rd world charities has a higher return than most people could expect to make from conventional investments (“Equities make around 5% in real terms, while the world’s poorest earn returns upwards of 20%”).

The obvious conclusion would be radical obligation (a la Peter Singer): donate pretty much everything. A way to avoid this conclusion is to say that you are investing in yourself, and expect to get higher returns in the long run (I think Paul Christiano makes a similar argument in the linked article, but in the other direction: the returns on charity would shrink over time, while your returns stay constant. I’m not convinced (yet)). On the other hand it seems unlikely to be rationally so confident in your future prospects.

Uncertain Environments

I suspect to prove that something like this is optimal would require some strong assumptions about knowledge of the environment. This intuition almost applies to the exploration/exploitation trade-off in uncertain
environments: the analogy is:

  • investment ~ exploration
  • cashing out ~ exploitation

But the analogy is imperfect, because during the exploitation phase, we are still learning about the environment, and our beliefs might change radically enough force us “back to the drawing board” (incentivizing exploration again), if we realize that our best theory was much worse that we had thought when we decided to start “cashing out”.

--

--