Who is leading the world in AI?

The United States emerged the clear leader, with China some way behind in second place. The UK held a significant third-place lead on the rest of the pack.

Since then, the world has changed. Has the AI landscape changed with it?

Today, we publish the 2020 edition of the Index and release the updated rankings.

China rising

China is catching up with the USA and could surpass it in a decade if it maintains momentum. The country already has a vast pool of data, with rising talent, research and AI companies. 

This year, improvements in AI research, infrastructure and talent, coupled with a commitment in Beijing to invest heavily in artificial intelligence, saw China’s Index score rise from 52 to 62 out of 100, the benchmark given to the world-leading USA.

While most of the 62 nations in our Index have published national AI strategies to guide public investment in AI, Beijing’s is the most ambitious. 

The exact sum China has earmarked to spend on artificial intelligence is unknown and hotly debated – but, on any measure, our Index shows that it clearly outstrips the USA. This commitment appears to be paying off. China improved its position across our 143 indicators, with particularly notable progress in AI research:

• China’s academics are gaining clout. In 2020, yet another Chinese university joined the Times’ list of top 100 computer science universities. 
• Chinese researchers are also rising in prominence on the h-index, a measure of academic influence. The total number of citations from high-achieving Chinese computer science academics jumped by 67 per cent over the course of the year. This compares to a 10 per cent rise in those from US authors. 
• The number of Chinese academic AI papers accepted by the IEEE – a body which sets AI standards and also publishes a number of influential AI journals – now out-do those by US academics by a factor of seven. 
• China is granting more AI patents than ever. China overtook the US in terms of AI patents granted around two years ago and has been pulling further ahead ever since.
• The number of Chinese users on coding platforms like Stack Overflow and GitHub are on the rise. This year we found 16,300 contributions to open source AI projects from Chinese profiles on GitHub, a platform for sharing code, up from 13,400 this time last year.

The UK has slipped back

The UK remains an AI powerhouse, ranking third in our Index behind the US and China, and above countries of more comparable size.

It is a base for world-class research institutions and pioneering companies like DeepMind, which this week announced it had cracked the mystery of how proteins go from being a long string of amino acids to structures that drive all life on earth. The discovery could lead to transformations in all areas of biology.

Despite the UK leaving the EU this year, and the looming prospect of a no-deal Brexit, the first 11 months of 2020 saw 50 per cent more funding in AI than the same period last year, representing substantial growth in a difficult climate. 

However, our prediction a year ago that the UK could be usurped is on track to come true. Countries which were some way behind the UK are now nipping at its heels.

We already knew that the UK had dedicated relatively little cash to AI. In 2018, the UK government’s AI strategy promised a £1bn five-year spending plan, including £300m for new AI research. Yet Germany has pledged €4bn for AI, in addition to a €50bn “future-focussed” tech fund which will include AI funding. 

Meanwhile, France has committed €1.5bn for research in AI to be spent by 2025. South Korea will invest 2.4 trillion won ($2bn) into AI research by 2022 and has committed an additional trillion won ($820m) to support the AI semiconductor industry, an essential part of AI innovation.

Overall, the UK began to lose its advantage in the past year through incremental changes across the board:

• Supercomputing power in the UK – needed to run big AI systems – is not keeping pace with that of other nations. Last year, it had 11 of the world’s top 500 supercomputers; now it has 10. 
• This year, UK universities on the Times Higher Education Top 100 Computer Science Universities dropped an average of over three places, while Queen Mary University dropped off the list entirely in 2020. 
• The diversity of UK AI professionals is down, from 19.1 per cent to 18.9 per cent, according to a survey by Kaggle, a machine learning and data science community. 
• UK-based academics published fewer papers in certain influential AI journals than their peers from China and the US this year, losing the UK points in our Index. Papers by the UK’s top 1,000 academic researchers on computer science were cited less often. 
• People across the world have flocked to “Massive Open Online courses” – sites like DataCamp that teach online classes – but the UK has seen less of a spike than other countries. 
• China has filed a greater proportion of overall AI-related patents, pulling away from the UK on this measure. 

Who’s up? Israel, Netherlands and Finland

The Global AI Index is not all about the superpowers: in 2020, we saw small, AI-focussed nations rise up the rankings to nip at the UK’s heels. 

Israel raced up from 12th to 5th place, the Netherlands from 16th to 7th, and Finland from 14th to 11th.

Israel is one of the most AI-intense nations in the world. With 3 AI start-ups per 100,000 people (compared to 0.5 in the USA), there’s nowhere you are more likely to bump into an AI entrepreneur. In the last year, Israel also improved in our measure of AI talent, with more software downloaded that enables people to code using the language ‘R’ and more contributions to open source projects on GitHub. 

The Netherlands, which moved from 16th to 7th, has the highest number of data scientists per million in the workforce, the third highest for the same measure of data engineers, and the seventh highest per-capita number of informal AI gatherings organised on Meetup.com, a community-building platform. Despite its small size, it has the third most 5G networks. 

Finland gained a world-class supercomputer this year, but individual efforts also boosted its rank. Coding became a favourite pastime for Finns, who downloaded AI-related software via coding languages Python and R. They also saved more changes on coding platform GitHub. All this points to increasing AI development. 

The Covid effect

The first year of our Index has been dominated by news of the Covid-19 pandemic. It’s unclear what can be attributed directly to the pandemic, but young startups appear to have been hit particularly hard. 

This time last year, the UK had 529 AI startups – defined as an AI company less than three years old – listed on Crunchbase, according to our analysis of the companies database. That figure now stands at 338. Other countries have seen similar drop-offs in startup numbers. 

This said, when taken as a whole, the 62 countries in our Index are still pulling in similar levels of private funding for AI as they did in 2019. Several countries have seen a fall in investment, but these don’t necessarily match those that have been worst affected by the pandemic. 

Both the US and China, for instance, have seen a similar drop-off in funding, despite the US having much higher rates of coronavirus. 

The year to come may prove equally disruptive to The Global AI Index Rankings. The USA will have a new government with a different perspective on its role in the world. As European nations focus more on AI, the UK’s Brexit transition period with the EU will come to an end, with unpredictable outcomes for the country’s AI scene.

---

FAQs

A fuller description of our methodology is available on our Intelligence website.

What is an index?

An index is a ranking built from a careful selection of different measurements around a central topic or theme. Here, the index ranks countries on the basis of their capacity for artificial intelligence.

Which countries are we covering?

We limited our analysis to 62 countries, most of which had published some sort of national strategy on AI setting out future plans.

What’s included?

We grouped these data points into three clusters: innovation, implementation and investment. Within those are seven key categories: research, development, talent, operating environment, infrastructure, commercial ventures and government strategy – all of which contribute to overall AI capacity. We weighted the factors by their significance, and gave a preference to human and intellectual capital, as well as investment, as these as critical drivers to building capacity.