One can imagine trading off the net benefit I provide, in the next 7 years, against marginal donations to the EA Funds Long-Term Future Fund.

For example, in 2030, one could imagine asking them, "Please review the benefits and costs of Ozzie's professional work in the last 7 years. Imagine that in 2020, someone would offer you a trade: a $N donation now, and all that value vanishes, or things play out as they did."

What is the best estimate of what value $N should have been?

This value would include aspects like:

• Direct and indirect work outputs

• Any help I gave people that would be useful for longtermist reasons

• Unintended benefits
  

Costs would include things like:

• Funding I took, particularly from longtermist sources

• Attention / status that had costs

• Other negative externalities I've produced

Resolution

I'll aim to resolve this in early 2030. Comments here will be useful!
For now, I'm going to judge this, but very happy to have another longtermist EV evaluator judge it in 2030, if someone is willing to). I would plan on discussing the situation with a longtermist grant evaluator, to make some assessment here.

I assume we'll likely have AI-assisted epistemic enhancements at that time. I of course, would intend to use these.