The Global AI Race – Which Country Is Winning?
“AI Competition Is the New Space Race,” said Bloomberg a few days ago in an article they published which highlighted how things are heating up in the global AI race. (As far as media outlets go, Bloomberg still manages to produce quality articles without
political spin or first-sentence typos.) The article pretty much said that artificial intelligence (AI) has a long way to go towards being commercially mature, and that countries around the world are scrambling to not be left behind. What sparked our interest the most was a report produced by Stanford University that the article mentioned, titled AI Index 2018, which attempts to “track, collate, distill, and visualize data relating to artificial intelligence.” Essentially, a bunch of people a lot smarter than us did loads of research that we can turn into an article without having to ask our underpaid team of MBAs to do research work over the holidays. All we need to do is just say the following to avoid any pesky cease-and-desist letters:
Yoav Shoham, Raymond Perrault, Erik Brynjolfsson, Jack Clark, James Manyika, Juan Carlos Niebles, Terah Lyons, John Etchemendy, Barbara Grosz and Zoe Bauer, “The AI Index 2018 Annual Report,” AI Index Steering Committee, Human-Centered AI Initiative, Stanford University, Stanford, CA, December 2018.
Since you will all want to spend time during the holidays with your families and mistresses, we pored over every single word of this 94-page report to pull out the most interesting bits that you can casually mention to your boss – when you go back into that horrible office after your holiday break – to convince him or her that you’re a thought leader and ahead of the game when it comes to your knowledge of artificial intelligence. Let’s get started.
Artificial Intelligence Papers
One way to measure research is by looking at academic papers that are being published on any given topic using a database like Scopus, the largest abstract and citation database of peer-reviewed literature. A search of papers on Scopus relating to artificial intelligence shows that 28% were affiliated with European authors, followed by China (25%), and the U.S. (17%):
However, we all know that quality is better than quantity. Turns out that U.S. AI authors are cited 83% more than the global average.
Other findings showed that AI papers in China are more focused on Engineering and Technology and Agricultural Sciences, while AI papers in the U.S. and Europe tend to focus on Humanities and Medical and Health Sciences. Makes sense, as China has to feed 20% of the world’s population. Speaking of which, the Chinese government produced nearly 4x more AI papers than Chinese corporations. In the U.S., the proportion of corporate AI papers was 6.6x greater than the proportion of corporate AI papers in China, and 4.1x greater than that in Europe. Still, academic papers dominated each region as a percentage of the total:
- ~14,000 or ~92% of total AI papers in China
- ~9,000 or ~81% of total AI papers in the U.S
- ~16,000 or ~88% of total AI papers in Europe.
So, we might conclude that the U.S. is doing better quality research, the Chinese are producing more papers than any other country, and the volume of artificial intelligence research has increased dramatically in the past 15 years.
Artificial Intelligence Patents
One way to measure intellectual property is by looking at patents, a task that’s much easier said than done. Due to the long lead time between filing and publishing, the report looked at the period of time between 2004 and 2014. Findings showed that in 2014, about 30% of AI patents originated in the U.S, followed by South Korea and Japan, which each hold 16% of AI patents. Of the top inventor regions, South Korea and Taiwan have experienced the most growth, with the number of AI patents in 2014 nearly 5x that in 2004.
Just because you publish lots of papers doesn’t mean you’re taking appropriate steps to protect your technology. However, not everyone agrees with the above findings. One example is Tsinghua University’s China AI Development Report, released summer 2018, which shows that China has the greatest number of total published patents relating to AI, followed by the U.S. and Japan. The authors of the report we’re reviewing state that they welcome dialog as to how we might all agree on a single way to measure this difficult metric.
Artificial Intelligence Focus Areas
The report then went on to look at capabilities by region, and several themes were highlighted that we’ve discussed before. Robotic process automation is pretty popular in Europe (companies like Blue Prism) and the Americas (companies like Automation Anywhere). In China, computer vision is quite popular. That’s evident by the fact that the world’s biggest and best computer vision companies are Chinese. When it comes to industries and functions, AI is being used heavily in telecom service operations (companies like Pindrop or Afiniti.) There is also heavy usage of AI in retail marketing and sales, topics we’ve touched on in numerous articles like 8 Artificial Intelligence Startups Improving CRM or 6 Artificial Intelligence Startups Customizing CRM. In financial services, risk management is a key focus, with AI being used to monitor things like compliance and fraud. Automotive manufacturing is also seeing quite a bit of AI usage, and we’ve looked before at some of the intelligent industrial robots that operate in this space.
Other Interesting Tidbits
The report goes on to mention some other interesting bits, some of which we’ve highlighted below:
- China’s AI talent isn’t going anywhere. 78% of Chinese AI researchers are staying within the country’s borders.
- Only around 30% of articles in the media on AI speak positively about it
- The Chinese are installing a huge number of robots
- Government mentions of AI – particularly Canada, the U.K, and the U.S. – have skyrocketed in just the past two years
Lastly, we’ll leave you with these notable “human-level performance milestones” that AI achieved in 2018:
Diversity in AI
For some reason, the report is liberally peppered with mentions of “diversity” with respect to just one category – gender. Any competent hiring manager or company founder could care less about whether or not their team members sit or stand when they go to take a piss, but rather that they are competent individuals who play well with others. At any rate, we’re mentioning this because we would expect these bright minds from Stanford University to do a much better job of trying to measure diversity than just telling us the gender breakdown of various categories of people. Next time, here’s what they should do.
Start by figuring out the population of qualified candidates, people trained in any aspect of AI – community college, Ivy League university, online training certifications – in every location around the world. Then, slice-and-dice it by gender, race (no, you don’t get to say half the world is “Asian,” you break it down by EACH country,) sexuality, economic background, hair color (we need more gingers in AI), political affiliation, religion, height, and attractiveness (ugly people need AI too.) Once you’ve thoroughly vetted the entire population of qualified AI workers, then you are in a position to do the difficult work, which is to then analyze each role and location and compare it to the available pipeline of candidates. Even then, what you uncover is irrelevant because you didn’t include the most important category. Competence. Founders and hiring managers want the best people for any given job. If that means that 94% of truck drivers happen to be male or 80% of cashiers happen to be female, so be it.
After reading this report, we realized that we’ve also been doing some research on who is winning the global race towards artificial intelligence superiority. Throughout this past year, we’ve sent our MBAs gallivanting across the globe to look at some of the top artificial intelligence startups in various countries and regions, and taking every chance possible to take the piss out of people using every cultural stereotype we could think of. Here are some of those pieces:
- Top-10 Artificial Intelligence Startups in Ireland
- Top-12 Artificial Intelligence Startups in Sweden
- Top-11 Artificial Intelligence Startups in Finland
- The Top-10 Artificial Intelligence Startups in Korea
- Top-10 AI Startups in Australia and New Zealand
- Top-10 Artificial Intelligence Startups in Hong Kong
- Top-10 Artificial Intelligence Startups in Canada
- Top-10 Artificial Intelligence Startups in India
- The Top-10 Russian Artificial Intelligence Startups
- Top-10 Artificial Intelligence Startups in Germany
- 10 French Startups Using AI in Healthcare
- The Top-10 French Artificial Intelligence Startups
- Top-10 British Artificial Intelligence Startups in the UK
- And of course, The 10 Biggest Artificial Intelligence Startups in The World
We learned that in most cases, countries have recognized the need for a national strategy that addresses how they plan to develop globally competitive AI capabilities. As we go into 2019, we will continue to look at how more countries are claiming their place in the global AI race, a race that seems to be much longer than previously thought.
Despite what the pundits say, FAANG stocks (Facebook, Apple, Amazon, Netflix, Google) don't give you real exposure to AI. Read about 7 stocks that give you true pure-play exposure to AI in our guide to investing in AI healthcare companies, freely available to Nanalyze subscribers.