AI has become a fieldthat is increasingly being seen as a global race. Countries are making effortsto develop AI and get an edge, enterprises are also racing to pool AI talent,data advantages.
The global AI competition is different from any other race given that the level of innovation varies substantially from country to country. Generally speaking, academia has contributed to the majority of innovations, and contribution from governments mainly lies in procurement, not internal R&D.
With the development of digital economy, the share of commodities in global trade has shrunk, while the share of digital services has climbed. It is expected that, by 2025, digital sector would create around 50 percent of economic value.
Against such a background, countries ranging from China, New Zealand to Finland and the US are working on national strategies for AI development to foster domestic talent and prepare to claim a position in AI value chain in the future, no matter about social programs or automation on labor markets.
However, the nature of AI race is yet to be seen. It is not likely to be limited to any single field, and the most crucial determinant will be how governments would regulate and monitor AI applications. The US, China and some other players do not share all ideas about privacy or data, and also diverge on their visions for the 21st century.
Until now, in governments’ assumption, the country who is first to reach the finish line can capture the majority of AI’s potential value. However, the thing is not whether the assumption is true, but whether a nationalized strategy is necessary.
To frame the matter in strictly national context is to ignore the way AI develops. Whether countries share data set could determine whether machine-learning algorithms develop biases towards specific country. And whether specific kinds of chips be considered as proprietary technology and restricted to the hands of certain countries or enterprises could determine how far innovation can go at the global level. Given these possibilities, there is a good reason to worry about the bad effect fragmentation of national approaches could have on the growth of digital economy.
In addition, under the current situation, talent pool is rather limited against the wide national AI projects. Though the bank will expand, the competitiveness needed for AI-driven economies could change. For example, cybersecurity will need a greater pool of experts.
AI developers who are working out of R&D centers and universities labs have found their exit plans. And corporations are driving up the price for researchers, all this contributes to a widening gap in talent between leading enterprises and “others”. And the market is already markedly concentrated as top tech companies have access to massive data that smaller participants and newcomers cannot get.
The composition of global trade is changing, suggesting that, the bulk of economic value will not emerge from goods and services, but from data regarding them. Therefore, those having access to global data flows will get the lion’s share.
It is important that key AI players and governments consult and coordinate to make sure the technology is a public good for the whole world, and used in a safe, fair and responsible way.