According to Vitalik Buterin, dealing with strong AI is inherently challenging, as all proposals face significant obstacles. However, he highlights a proposal that might be more feasible by enforcing regulations at the hardware level. This approach ensures critical decisions are made before any high-risk period begins, making it difficult to bypass the system without manufacturing new hardware, which involves a lengthy lead time.