The Coming AI Metamorphosis

Bottom Line: The rise of artificial intelligence is inevitable, and will have far reaching impacts on everything from human interaction to national security. The question is not whether AI will transform our world, but how.

To really understand the evolution of AI, humans need to study it.

Unlike regular computer software, AI can operate and update without human intervention. It learns, and it learns in ways that we can’t understand at first. For instance, an AI program managed to beat chess grandmasters by employing moves that were objectively wrong or illogical by human standards. Similarly, self-driving cars “learned” to inch closer to stop lights, despite never being “taught to do so.”

The opacity of AI raises several national security questions, specifically with nuclear deterrence.

Current nuclear strategy revolves around transparency. Countries deter threats from rogue nations by making it known exactly what weapons they have, and how many. Similarly, advanced detection systems. AI changes that calculus, because it could develop “new weapons, strategies, and tactics by simulation and other clandestine methods.” In other words, concealment of AI capabilities would be a strategic advantage.

Ultimately, the rise of AI could upend nuclear grand strategy altogether.

It’s unclear, for instance, whether the potential ubiquity of AI will act as the ultimate deterrent, or spur global paranoia that could potentially lead to conflict. Given the aforementioned strategic value of secrecy with regards to AI, “The incentives will be for opacity, which could mean absolute insecurity.” In short, the rise of AI could make it impossible to develop new international norms at precisely the time those norms were needed.

Read the full piece here.