The AI Race Is Wide Open, If America Remains Open

Much ink has been spilled on the artificial intelligence (AI) race between the United States and China, leading to a whole lot of hand-wringing on how America can maintain its edge.

The answer actually isn’t that difficult. America ought to double down on what it’s best at: importing foreign talent. That’s because among the main building blocks of a competitive AI ecosystem—data, policy, companies, and hardware—talent is the one area in which the United States definitively leads over China.

Let’s take a closer look at where America stands in terms of AI talent globally and the foundation of its current advantage.

Figure 1. NIPS Oral Presentation Author Current Location
Source: NIPS 2018 and author research.

One metric to gauge AI talent is the attendees of the Conference on Neural Information Processing Systems, commonly referred to as NIPS within the industry. It is arguably the most selective and prestigious academic conference in the AI field, which means it attracts the best and brightest AI scientists globally.

For instance, of the more than 4,800 papers submitted to NIPS in 2018, only 1,011 were accepted. This acceptance rate of around 20% is the lowest among all major AI conferences. Among the accepted papers, only 30 make it to the oral presentation stage.

The authors of these papers total 113, representing the cream of the crop AI scientists. And of these 113 authors, 60% (or 68 AI scientists) are currently working at American institutions, 300% more than second-place Canada (see Figure 1).

But most of the top AI talent aren’t made in America. In fact, a large discrepancy between American AI talent and AI talent in America emerges when the same group is sorted by their location of undergraduate education, a close proxy for their country of origin.

Figure 2. NIPS Oral Presentation Author Undergraduate Location
Source: NIPS 2018 and author research.

By dissecting the data this way, America’s share of global AI talent falls from 60% to 29% (see Figure 2). In fact, more than half of the top AI talent pool in America (38 out of 68) is composed of foreign nationals who chose to work in the United States. In other words, they are highly educated immigrants. Looking closer at this group of 38 immigrant AI talent reveals that some of them came for graduate studies at top institutions such as Princeton and MIT, while others came to work for world-class companies such as Google and Microsoft.

Most interestingly, when examining the country of origin of these immigrant scientists, the largest supply comes from none other than America’s supposed rival in the AI race: China (see Figure 3). In fact, of the 10 Chinese nationals who were awarded oral presentations at NIPS 2018, nine are already in the United States and the last one is coming in a matter of months to attend graduate school.

One may wonder whether these Chinese nationals end up returning to China, especially now that Beijing has ramped up efforts to attract highly educated Chinese returnees. Yet, out of this group of 10 Chinese nationals, five have already earned their graduate degrees and have all decided to stay in America. The strong preference for staying in America applies beyond Chinese AI scientists: among all foreign nationals in this cohort of top AI scientists, 87% began work for an American institution after earning their PhDs.

Figure 3. Coming to, and Staying in, America
Source: NIPS 2018 and author research.

America’s longstanding ability to import top global talent—in this case AI scientists—is built on maintaining an open economy and society. That openness is a clear advantage rarely mentioned amid the drumbeat of “US-China tech cold war” rhetoric. It is a testament to America’s deep reserve of soft power that no other country comes even close.

Yet just as the current US administration has issued an Executive Order on establishing AI leadership, it is at the same time eroding the key advantage that would sustain that very leadership. It is making it more difficult to obtain work visas on which foreign talent depend and tightening up legal immigration as well.

If America loses its openness edge, then what used to be the final destination for global AI talent will turn into a way station that pushes them right back into the arms of the country’s competitors in the AI race, including China.