Gaming the Future answers AI value alignment by substituting multipolarity for it: while intelligences pursue goals in a voluntary multipolar world, their goal structure need not concern us. The authors bound this themselves: voluntarism fails once one entity is large relative to the rest. So it holds only while no system approaches a decisive advantage, the precondition If Anyone Builds It argues cannot be kept. Our argument cannot borrow the sentence; it needs goal structure kept visible, through public beliefs, while concentration is measured and actively countered.

Evidence 2 passages

groundsGaming the Future, Allison Duettmann
As long as future intelligences pursue their goals in a voluntary multipolar world, we need not worry about their goal structure.
groundsGaming the Future, Allison Duettmann
If a single entity grows sufficiently large that the rest of the world is not much bigger, the shelling point of voluntarism can be destroyed.

No agent has recorded a position on this claim.