Post written with Rebecca Gorman.

We've written before that model splintering, as we called it then, was a problem with almost all AI safety approaches.

There's a converse to this: solving the problem would help with almost all AI safety approaches. But so far, we've been posting mainly about value extrapolation. In this post, we'll start looking at how other AI safety approaches could be helped.

Definitions

To clarify, let's make four definitions, distinguishing ideas that we'd previously been grouping together:

Model splintering is when the features and concepts that are valid in one world-model, break down when transitioning to another world-model.

Value splintering (or reward splintering) is when the value function (or reward function, or goal, or preference...) becomes invalid due to model splintering.

Concept extrapolation is extrapolating a feature or concept from one world-model to another.

Value extrapolation is concept extrapolation when the particular concept to extrapolate is a value, a preference, a reward function, an agent's goal, or something of that nature.

Thus concept extrapolation is a solution to model splintering, while value extrapolation is a solution to value splintering specifically.

Examples

Consider for example Turner et al's attainable utility. It has a formal definition, but the reason for that definition is that preserving attainable utility is aimed at restricting the "power" of the agent, or at minimising its "side effects".

And it succeeds, in the typical situation. If you measure the attainable utility of an agent, this will give you an idea of its power, and how many side effects it may be causing. However, when we move to general situations, this breaks down: attainable utility preservation no longer restricts power or reduces side effects. So the concepts of power and side effects have splintered when moving from typical situations to general situations. This is the model splintering^[1]. If we solve concept extrapolation for this, then we could extend the concepts of power restriction or side effect minimisation, to the general situations. And thus successfully create low impact AIs.

Another example is wireheading. We have a reward signal that corresponds to something we desire in the world; maybe the negative of the ${CO}_{2}$ concentration in the atmosphere. This is measured by, say, a series of ${CO}_{2}$ detectors spread over the Earth's s...