Convergence: Artificial Intelligence and Quantum Computing

CHAPTER 15Human Imagination and HAL

Erik Viirre, Professor in the Department of Neurosciences, University of California, San Diego

The technology of quantum artificial intelligence brings forth a variety of issues, with a key question arising around what quantum AI will actually be used for. As the technology will be used on a variety of problems that ultimately depend on setting variables and their operational ranges, the very large question of expertise in programming a quantum computer and then managing their operations also comes to the fore. Who will have access to quantum computers and quantum AI, and who will have access to humans that will be essential to these machines?

Closely related to the setup of quantum AI operations will be assessment of their results. Quantum computing systems by their very nature are nondiscrete and thus will have “error ranges.” Further, incorrect settings of these systems will result in their exponential power giving exponential failures. Who will measure reliability? Most important, any internal “examination” by the quantum computing system of its computational results will not be available using current technology.

Given these uncertainties, what are the ideal means of ethical management of these systems and their outputs? Of course, we already deal with uncertain, indeterminate, unexaminable quantum AI systems: humans. Our own imagination operations show us that while we might evince post-hoc explanations of our ideas, we really do not ...