5 Countries Leading the Way in AI
These Are The Countries That Publish the Most Research on AI
January 8, 2018, Bruno Jacobsen
That artificial intelligence (AI) is growing faster than ever before is no surprise. Since 2010, it has grown at a compounded annual growth rate of almost 60%. Our Future Proof blog has already discussed many of the consequences of this growth. But while we can go on about what AI will mean to our future, who exactly will create this AI. Or, better yet, what are the countries leading the AI revolution?
When it comes to AI, not all countries are the same. Here are the top 5 leading countries based on the number of research papers published every year.
The 5 leading countries in AI research
1 - China
While not too long ago China was though of as a manufacturing country, the country now intends to be a leader in many fronts. In our blog, we've talked about how much they are pushing for renewable energy. AI is another area the Chinese consider of utmost importance.
According to the Times Higher Education, in the period between 2011 and 2015, China published over 41,000 papers on AI. That's almost twice as much as the US number.
The Chinese government stands strongly behind AI adoption. Last year, they announced their intention to become "a principal world center of artificial intelligence innovation" by 2030. Then there are companies like Tencent, Alibaba and Baidu. From e-commerce to self-driving cars or search engines, AI will play a fundamental role in their success. Combined, they are worth around US$ 1 trillion.
2 - United States of America
In terms of papers published, the US comes at an undisputed second place. In fact, both China and the US are miles away from other countries. Between 2011 and 2015, the US published almost 25,500 papers, according to the same source.
On top of that, the US ranks as the top country with the most AI companies. With over 1000 companies and US$10 billion in venture capital, the US is likely to become an AI superpower. Then there's companies like IBM, Microsoft, Google, Facebook, and Amazon. Not only do they publish a significant amount of papers, but they also invest heavily in AI.
America's pool of scientific knowledge combined with its business market power will allow it to stay on top.
3 - Japan
According with the Times Higher Education rankings, Japan stands in third place, with about 11,700 papers published. Indeed, this is not surprising. With an ageing population and decreasing workforce, AI will play a vital role in the Japanese economy. Even now, about 55% of work activities in Japan could be automated. With current technology. Its manufacturing sector, according to the HBR article, has a 71% automation potential. In the US, that number stands at 60%. And in office and administrative work, the difference is 16% to 9%.
With plenty of research into AI, a decreasing workforce and a high automation potential, Japan is likely to continue right at the top. Its long-standing willingness to invest in technology may also prove key.
4 - United Kingdom
The UK is not much behind Japan, though. In fact, when it comes to published research papers on deep learning, it has already passed Japan. With close to 100 published papers, the UK became number 3 on the topic. As for total published papers on AI, between 2011 to 2015, the number was 10,100 - slightly behind Japan.
And the UK is no stranger to AI. Remember when Google's famous DeepMind beat Go's grandmasters last year? Well, DeepMind Technologies Limited was founded in 2010, in Britain. According to the Financial Times, DeepMind is today a world leader in AI. It employs 250 researchers, from mathematicians to neuroscientists.
5 - Germany
Finally, the 5th country with the most published researched papers on AI is Germany. Between 2011 to 2015, the number stood at nearly 8,000. Germany, like China, also plans to become a leading hub for artificial intelligence. According to an FT article, Germany's Max Planck Society, two technical universities, and its leading exporting state are combining their artificial research intelligence together with companies like Porsche, Daimler, and Bosch. The Cyber Valley, as they call it, is the result of this, and it has even received support from Amazon, who plans to open a lab there.
Germany, like Japan, is also experiencing a working population decline. What's more, it too has a high automation potential, standing at 47.9%. Its strong industry capabilities, combined with powerful companies and good education make it a fertile ground for AI.
But do the number of research papers matter that much?
Published research is not everything
While the number of research papers can be a good indicator of the amount of research done, it's not the only metric. Indeed, the greater your population, the more likely you are to publish more.
Another good metric is the field-weighted citation impact of these papers. As the Times Higher Education explains, the index takes into account the citations each paper receives, adjusting for year and impact. When we look at this, we get quite different results.
Here are the top 5 countries or autonomous regions in research impact. A field-weighted citation impact of 1 represents the average. Anything above that means that the citation impact is above average.
Hong Kong (2.00)
United States (1.79)
According to this index, Germany and the UK stand in 8th and 10th place respectively. Mainland China and Japan do not make it to the top 10.
And the top universities with the highest field-weighted citation impact? The Times Higher Education research shows MIT is the undisputed number 1. With an index of 3.57, it is far above second-placed Carnegie Mellon University (2.53). Nanyang Technology University (2.51), University of Granada (2.46) and University of Southern California (2.35) complete the top 5.
Research and its impact matter, but other things do too
We saw here what countries are leading the AI research. But research and its impact say little about a country's success at adopting the technology. Some countries, given their market power and investments in technology, will inevitably become leaders. Countries like the US, China, Japan, and the UK fall under this category. But it's up to other countries fight for a spot at the top.
A word of caution
Why would countries want to lead the AI revolution? AI is not just about improving society and developing strong economies. It can also be a powerful military weapon, give rise to new industries, and shift the global balance of power. In Superintelligence, Nick Bolstrom warns of the dangers of an all-powerful AI. As countries near its development, even small advances will confer significant advantages.
In fact, if we experience an exponential growth in AI capabilities, we should hope that AI is used to create a stronger and more united world. If not, it might be a seriously destructive tool, for weaker or stronger nations.