On the Undecidability of Artificial Intelligence Alignment: Machines that Halt
This addresses the AI alignment problem for ensuring safe and controllable AI systems, proposing a foundational shift toward intrinsic architectural guarantees, though it is incremental in applying existing theoretical results.
The paper proves that the inner alignment problem for arbitrary AI models is undecidable by reducing it to the Halting Problem using Rice's theorem, and argues that alignment must be built into AI architectures rather than imposed post-hoc, with examples illustrating halting constraints.
The inner alignment problem, which asserts whether an arbitrary artificial intelligence (AI) model satisfices a non-trivial alignment function of its outputs given its inputs, is undecidable. This is rigorously proved by Rice's theorem, which is also equivalent to a reduction to Turing's Halting Problem, whose proof sketch is presented in this work. Nevertheless, there is an enumerable set of provenly aligned AIs that are constructed from a finite set of provenly aligned operations. Therefore, we argue that the alignment should be a guaranteed property from the AI architecture rather than a characteristic imposed post-hoc on an arbitrary AI model. Furthermore, while the outer alignment problem is the definition of a judge function that captures human values and preferences, we propose that such a function must also impose a halting constraint that guarantees that the AI model always reaches a terminal state in finite execution steps. Our work presents examples and models that illustrate this constraint and the intricate challenges involved, advancing a compelling case for adopting an intrinsically hard-aligned approach to AI systems architectures that ensures halting.