Simulation expectation
Teruji Thomas (Global Priorities Institute, University of Oxford)
GPI Working Paper No. 16-2021, published at Erkenntnis
I present a new argument that we are much more likely to be living in a computer simulation than in the ground-level of reality. (Similar arguments can be marshalled for the view that we are more likely to be Boltzmann brains than ordinary people, but I focus on the case of simulations.) I explain how this argument overcomes some objections to Bostrom’s classic argument for the same conclusion. I also consider to what extent the argument depends upon an internalist conception of evidence, and I refute the common line of thought that finding many simulations being run—or running them ourselves—must increase the odds that we are in a simulation.
Other working papers
Staking our future: deontic long-termism and the non-identity problem – Andreas Mogensen (Global Priorities Institute, Oxford University)
Greaves and MacAskill argue for axiological longtermism, according to which, in a wide class of decision contexts, the option that is ex ante best is the option that corresponds to the best lottery over histories from t onwards, where t is some date far in the future. They suggest that a stakes-sensitivity argument…
Three mistakes in the moral mathematics of existential risk – David Thorstad (Global Priorities Institute, University of Oxford)
Longtermists have recently argued that it is overwhelmingly important to do what we can to mitigate existential risks to humanity. I consider three mistakes that are often made in calculating the value of existential risk mitigation: focusing on cumulative risk rather than period risk; ignoring background risk; and neglecting population dynamics. I show how correcting these mistakes pushes the value of existential risk mitigation substantially below leading estimates, potentially low enough to…
Will AI Avoid Exploitation? – Adam Bales (Global Priorities Institute, University of Oxford)
A simple argument suggests that we can fruitfully model advanced AI systems using expected utility theory. According to this argument, an agent will need to act as if maximising expected utility if they’re to avoid exploitation. Insofar as we should expect advanced AI to avoid exploitation, it follows that we should expected advanced AI to act as if maximising expected utility. I spell out this argument more carefully and demonstrate that it fails, but show that the manner of its failure is instructive…