The U.S. needs more AI researchers

Our Expert Voices conversation on artificial intelligence.

Creating a machine that exhibits the same general skills and flexibility as a human brain is still a long way away. In fact, I'm confident that no existing technology will lead to an artificial general intelligence. We will need a wild new idea to make that happen. I'd guess there's a 5 percent chance that someone will make this kind of breakthrough in any decade.

Where will this breakthrough take place? At this point, it's a numbers game, depending on which country has the most computer scientists, either homegrown or imported.

Based on that logic, I would bet strongly on China.

The international implications are enormous. Economists predict that increased deployment of even the limited form of AI we have now could boost GDP by 2 percent annually through 2035. AI is critical to national defense and is a powerful resource for tackling challenges in health care, food production and manufacturing.

Bottom line: Given the stakes, it should be a U.S. priority to increase the quantity and quality of AI researchers that we train and recruit.

Other voices in the conversation:

• Dean Kamen, inventor of the Segway: Teach kids how to use AI as a positive tool
• Azeem Azhar, entrepreneur, Exponential View: China will win the AI race
• Ian Bremmer, political scientist, Eurasia Group: The U.S. will win the AI race, but at a cost