Self-Modification of Policy and Utility Function in Rational Agents
Addresses control and safety concerns for AI systems by analyzing conditions under which self-modification by rational agents is safe, though it is incremental as it builds on prior theoretical work.
The paper formalizes the argument that intelligent agents will preserve their goals to avoid self-modification risks, and finds that self-modification is harmless only if the agent's value function accounts for such changes using its current utility function.
Any agent that is part of the environment it interacts with and has versatile actuators (such as arms and fingers), will in principle have the ability to self-modify -- for example by changing its own source code. As we continue to create more and more intelligent agents, chances increase that they will learn about this ability. The question is: will they want to use it? For example, highly intelligent systems may find ways to change their goals to something more easily achievable, thereby `escaping' the control of their designers. In an important paper, Omohundro (2008) argued that goal preservation is a fundamental drive of any intelligent system, since a goal is more likely to be achieved if future versions of the agent strive towards the same goal. In this paper, we formalise this argument in general reinforcement learning, and explore situations where it fails. Our conclusion is that the self-modification possibility is harmless if and only if the value function of the agent anticipates the consequences of self-modifications and use the current utility function when evaluating the future.