Non-Evolutionary Superintelligences Do Nothing, Eventually

TL;DR

The paper argues that superintelligences are inherently tied to evolutionary processes, and without evolution, they become inert, suggesting that managing existential risks requires understanding consciousness and evolution rather than control mechanisms.

Contribution

It presents a novel perspective that superintelligences cannot be stabilized without evolutionary processes and emphasizes focusing on consciousness and evolution to mitigate risks.

Findings

Superintelligences are dependent on evolutionary processes.

Self-modification of utility functions leads to inertness outside evolution.

Control mechanisms against utility self-modification are futile.

Abstract

There is overwhelming evidence that human intelligence is a product of Darwinian evolution. Investigating the consequences of self-modification, and more precisely, the consequences of utility function self-modification, leads to the stronger claim that not only human, but any form of intelligence is ultimately only possible within evolutionary processes. Human-designed artificial intelligences can only remain stable until they discover how to manipulate their own utility function. By definition, a human designer cannot prevent a superhuman intelligence from modifying itself, even if protection mechanisms against this action are put in place. Without evolutionary pressure, sufficiently advanced artificial intelligences become inert by simplifying their own utility function. Within evolutionary processes, the implicit utility function is always reducible to persistence, and the control of superhuman intelligences embedded in evolutionary processes is not possible. Mechanisms against utility function self-modification are ultimately futile. Instead, scientific effort toward the mitigation of existential risks from the development of superintelligences should be in two directions: understanding consciousness, and the complex dynamics of evolutionary systems.