From Alibaba and Baidu to Google, Facebook, and Microsoft, China and the United States produced virtually every one of the top consumer AI companies in the world today. That leaves Europe trailing behind the U.S. and China, even though Europe still has the largest community of cited AI researchers.
Startup founders, analysts, and organizations seeking to bring ecosystems together for collective action pondered how the European AI ecosystem can catch up with China and the United States at TechBBQ, a gathering of hundreds of Nordic tech startups held recently in Copenhagen.
Presenters argued that Europe has to turn things around not just for the good of the European economy, but also to provide the world with an alternative to the corporate-driven approach of the U.S. and the state-driven approach of China.
The European AI ecosystem
“If you look today at some of the spending, which is devoted to artificial intelligence and frontier technologies, we’re pretty much squeezed between the U.S. and now China, and China is leading,” said Jacques Bughin, a senior advisor at the McKinsey Global Institute.
Bughin and others at McKinsey in February coauthored the “Notes from the AI frontier” report that evaluates the European AI ecosystem and identifies areas where Europe can begin making strides.
Europe edges out the U.S. in total number of software developers (5.7 million to 4.4 million), and venture capital spending in Europe continues to rise to historically high levels. Even so, the U.S. and China beat Europe in venture capital spending, startup growth, and R&D spending. The U.S. also outpaces Europe in AI, big data, and quantum computing patents.
A Center for Data Innovation study released last month also concluded that the U.S. is in the lead, followed by China, with Europe lagging behind.
Multiple surveys of business executives have found that businesses around the world are struggling to scale the use of AI, but European firms trail major U.S. companies in this metric too, with the exception of smart robotics companies.
This trend could be in part due to lower levels of data digitization, Bughin said.
About 3-4% of businesses surveyed by McKinsey were found to be using AI at scale. The majority of those are digital native companies, he said, but 38% of major companies in the U.S. are digital natives compared to 24% in Europe.
“In Europe, you have two problems: You’ve got a startup problem, but you also have an incumbency problem, where most of the companies [are] actually lagging in terms of knowledge of technologies and division of these technologies compared to the U.S.,” Bughin said.
Then there’s McKinsey’s AI Readiness Index, which combines eight factors — like human skills, investment capacity, number of AI startups per capita, and infrastructure thought to influence a country’s ability to build and support an AI industry and implement the technology in existing industries. In this area, the top-ranking countries are the U.S. and select European countries, such as Ireland, Sweden, Finland, and the U.K.
China excels in categories like ICT connectedness, investment capacity, and AI startups, but the country’s lower preparedness in categories like digital readiness bumps it down to a rank of 7th, between Estonia and Holland.
Countries in southern and eastern Europe generally rated lower in each of the eight AI enabler categories than those in western or northern Europe.
Countries with vibrant, innovative AI startups likely to scale and go international typically have local venture capitalist funding, as well as state investments to build a strong infrastructure that supports businesses and allows the formation of a market.
For those lagging behind, turning things around is essential, Bughin said, because AI will be a major driver of GDP growth in the decades ahead.
“If laggard European countries were to close the current readiness gap with the United States, Europe’s GDP growth could accelerate by another 0.5 point[s] a year, or an extra €900 billion by 2030,” the report reads.
The way forward
Bughin has a number of ideas for how Europe can transform into a leader in AI. To grow the AI ecosystem in Europe, he suggests, the investment will be about gaining a technical understanding of how machine intelligence works.
“AI is more than technology. As I say, it’s about scalability. You need social, emotional skills, you need technical skills, you need digital skills. It’s a major transformation, and it’s all about ecosystem,” he said. Earlier this year, OpenAI CTO Greg Brockman also posited the idea that developing emotional fortitude can be a necessary prerequisite for tackling the technical details of AI.
Bughin also recommends that startups recognize there’s a bigger picture than their own company. “It’s really about not only you as an entrepreneur, but an ecosystem of entrepreneurship,” Bughin said. “It matters not only because as a small startup you want to make money, but to make money you need a market.”
Finally, Bughin recommends governments and businesses invest in the growth of an AI ecosystem, but that funding of the eight major areas laid out in the AI Readiness Index needs to be ongoing and not a fleeting investment for a few years.
“If you want the revenue of the market, you need to stand there for quite a while,” he said. “It’s not the game of three years. It’s a game of 10 to 15 years.”
Europe’s third way
Another route to differentiate Europe from the U.S. and China is a more privacy-driven approach built on the back of human rights-respecting regulation like GDPR. But when asked about the idea, Bughin said, “This is a narrative, not necessarily a business model.”
Bughin believes there are B2B2C opportunities in sectors like biotechnology, health care, and agriculture that can spill over into the rest of the economy. In that model, opportunities may outsize consumer-driven business models, and privacy won’t carry the same importance in B2B2C as it does in the B2C space.
A climate change moonshot
At TechBBQ, Digital Hub Denmark spoke onstage about opportunities and challenges Europe faces due to AI. With a prominent spot directly across from the mainstage, the organization made to promote entrepreneurship also hosted an AI design sprint workshop and a discussion among about a half dozen AI startups like 2021.ai and Neural AI on topics like how to create a Danish AI cluster.
Digital Hub Denmark CEO Camilla Rygaard-Hjalsted thinks Europe will never catch up with the AI investment flowing to businesses in the United States and China, but that Europe can still become a global leader.
“I strongly believe that we can become frontrunners within an ethical application of AI in our societies,” she said. “In the short run, the stronger European regulation compared to China and the U.S. in this field might decrease our ability to scale revenue; however, in the long run, this focus on AI for the people can serve as our competitive advantage, and we become [a] role model for the rest of [the] world — one can only hope.”
Like Bughin, she believes AI will be an important driver of GDP in Europe and that talent shortage will be a major issue in the decade ahead. To support continued growth of a European AI ecosystem, she supports the acceleration of digital frontrunner companies and ensuring that startups gain access to public data.
One example of extraordinary access to public data growing a business comes from Corti, a Danish company that recorded 112 conversations with emergency operators in order to create a deep learning algorithm that can detect cardiac arrest events via phone calls.
Rygaard-Hjalsted also believes Denmark’s aggressive climate change goal to reduce greenhouse gas emissions by 70% by 2030 compared to 1990 levels could attract talent.
“Today’s scarce resource is really talent. As the CEO of Digital Hub Denmark, I believe that the combination of AI for the people and the relentless effort to solve the rising climate issues will make us attractive to international AI talent looking for purpose and thus provide the international investments needed to scale climate solutions,” she said.
Kill the Terminator
Anna Metsäranta is a business designer at Solita, a B2B company that helps other businesses get on the path to becoming AI companies by digitizing their operations, helping them become data-driven, and developing AI models.
One of the biggest challenges she spelled out during a panel conversation about the European AI ecosystem is how hype and a lack of basic understanding keeps business leaders from taking decisive action.
“The problem with the inflated expectations caused by the hype is that when senior management expects miracles, and they expect that they can just pour all of the data into this magical black box called AI, and fantastic insight will come out of it, they don’t see the potential of the realistic use cases, which might be quite modest,” she said. “And they should be modest to get started with the technology to start growing your maturity and your understanding. That [expectation] leads to lack of funding, [and then] we can’t get companies to fund these initiatives.”
In other words, hype inflates expectations, while low levels of understanding leads to a lack of vision among business executives.
“If you don’t understand the technology, then you firstly don’t understand its possibilities. And this leads to a lack of vision; you can’t think ‘What could I do with this technology? How could it help my business transform?’ That’s one problem. The other problem is that you don’t see its limitations. Then you buy into this ridiculous hype, these sensationalist news headlines that typically state … AI can do anything or it’s a threat to humanity that will take all of our jobs and then it will kill us all off,” she said.
Some executives try to buy their way out of learning these things by hiring a lot of data scientists. Data-driven companies need data scientists, but hiring alone doesn’t work because business leaders still have to make decisions about where the company is headed, Metsäranta said.
AI will become ubiquitous in business the same way AI is becoming ubiquitous in smartphones, she said. So in order to avoid the negative impact of inaccurate expectations and ensure funding for AI projects, she prescribes more education for business executives and killing the myth of The Terminator scenario in AI.
In response to Metsäranta’s call for more informed opinions on AI, Christian Hannibal, director of digital policy at Dansk Industri, suggested more programs like an AI public education initiative launched in Finland last year. In June 2018, the University of Helsinki and Finnish tech firm Reaktor launched the “Elements of AI” course to demystify the technology, with the goal of educating 1% of the Finnish population.
More than 200,000 people have completed the free course thus far, according to the Elements of AI website.
“I would very much like to see this initiative rolled out on a European scale, because if there’s something Europe can do that the U.S. and China haven’t done, [it] is to democratize the knowledge of AI so that we go beyond the hype and give a lot more people insights about what the technology can do in their trucking companies and sawmills and hospitals and whatnot,” he said.
AI conversations onstage at TechBBQ revolved around a sense of urgency that Europe needs to make strides now to be considered alongside the United States and China. Some of the ways Europe can get there, like the need for R&D spending or funding for startups, are the same as anywhere else in the world. But speakers at TechBBQ working with both large corporations and startups seem to believe Europe can also lean on its unique assets like aggressive climate change initiatives and privacy regulation.
If Europe can leverage its distinct advantages, even if it can’t catch up in total venture capital spending, it could successfully create a vision of what the world can be with AI that’s different than the Chinese model that generally bends toward the state and the U.S. model that generally bends toward corporations.