Should AI be regulated? This question is increasingly being discussed and debated. Calls for restraint are understandably driven by the worst outcomes that advanced AI systems might produce.
But focusing primarily on AI’s worst possible outcomes can create another risk: regulation that prevents society from realizing benefits that could substantially improve human well-being, including advances in medicine and drug discovery. The potential benefits of AI should be part of the risk calculus, because poorly designed regulation can create risks of its own.
Most people are familiar with generative AI systems like ChatGPT, Claude and Gemini. Their highly interactive nature can give the impression that they possess human-like qualities. They communicate in numerous languages, answer questions, reason through problems, and increasingly perform tasks on behalf of users. Yet they also differ fundamentally from humans.
Whatever their capabilities, today’s AI systems they do not experience the world in the way humans do. Treating them as human minds with different capabilities risks misunderstanding both what they can do and what they cannot do — and therefore how they should be governed.
To read more, click here.