When media and technology experts discuss AI and the world order, the conversation typically revolves around the US and China. In Kai-Fu Lee’s new book, called ‘AI Super-Powers’, the subtitle includes just China and Silicon Valley. Outside of Deep Mind that was founded in London, every other major AI player is from either China or the US. It’s logical at this point in history as these two countries control the infrastructure – the talent, capital, and markets. Is there still hope for Europe?
This era of AI is built on deep learning. In short, deep learning is based on huge data models with multiple layers of non-linear processing units. They are vaguely modeled on biological processes where each neuron has a unique input and output. Thus, they are called neural networks, but are vastly more simplistic than brain structures. The reason markets are so critical to deep learning is because the vast data sets needed for this type of learning can only be created by large homogenous markets.
But there is still hope for European countries to be major competitors, or even leaders in this field. AI is in its infancy. Eventually, there will be diminishing returns from increasing data size. A key example comes from child development. When we teach children to recognize shapes, they only need to see a small number, certainly not thousands like machines need. Researchers like Thorpe et al. estimate that a child learns all 10-30 thousand object categories in the world by six years of age. In computer science this is known as one-shot learning.
Without a strong effort Europe risks being left behind in the next industrial revolution.
In other words, knowledge can be induced from a very small number of examples. There is every reason to believe that artificial intelligence will be able to perform the same tasks given the right structures. The research is encouraging.
Europe and the next era of growth
Because of Europe’s smaller individual markets, focus on new machine learning paradigms like one-shot learning could give opportunity to leapfrog current paradigms. Similar to how many Chinese skipped land line telephones and personal desktop computing and went straight to smart phones. There is an emergent opportunity for leading companies in Europe to lead the next era of growth. There are several requisite actions that governments and companies need to take to support new AI paradigms.
- Embrace technological change
European companies lag behind US companies in technology adoption.
- Increase funding for research, both public and private
What Europe lacks in VC funding needs to be made up by governments.
- Support innovative companies
With more public grants, tax benefits, and loans.
- Lead the dialogue
Educate the public and policy makers on technology.
- Focus legislation on supporting innovation rather than impeding innovation
GDPR, though great for individuals, makes it harder for companies to adopt technology.
Leading European nations must embrace the change and focus on actions.
Don’t become a surveillance state, do look for the middle ground
It is no coincidence that both China and the US have largely left AI unregulated. In fact, the strength of the Chinese market for AI is also due in large part to collective embracing of technological change while embracing data transparency. Clearly, Europe does not want to become a surveillance state. But, there is a middle ground where attitudes towards data can be more relaxed for the benefit of progress. Without a strong effort Europe risks being left behind in the next industrial revolution.
The potential payoffs and pitfalls are huge. McKinsey estimates that the impact to European GDP could be up to 3.6 trillion Euro by 2030. This would be an increase of almost 30% simply from catching up to the US AI frontier. Furthermore, the distribution of benefits would likely favor northern European countries such as the Netherlands due to digital readiness, innovation, and human skills. After all, Europe is home to more developers than the US.
Clearly, Europe does not want to become a surveillance state. But, there is a middle ground.
Ultimately, Europe has a history of prescient market timing. Take watches as an example. In the 1800s, Swiss watches were mass-produced, low-quality time pieces, before they were industry leaders. When Japanese watch makers introduced quartz time pieces in the 1970s, Swatch used the technology to out-style the Japanese. At the same time, mechanical movement went luxury and spawned a high-end watch trade. Fast fashion is another example of this marketing dynamism where companies like H&M and Inditex have revolutionized the speed clothes make it from runway to store. The Netherlands is an agricultural powerhouse despite having very little arable land due to technological leadership. There are many more examples.
AI will likely follow a similar trend. Leading European nations must embrace the change and focus on actions.
Robin Youngs keynote during the NRC Live-event ‘Winning with AI’!
“It is not a battle, it is a war. And Europe is going to lose’. Robin Young, CEO of Seedlink in Shanghai, indicated the state of affairs on the world stage during the NRC Live event ‘Winning with AI’ on April 11th 2019. As a half-Chinese, half-American and frequent visitor to the European continent, he is convinced that the market ultimately dictates. What does this mean for our future? Will the whole world get a Social Credit System as is now being rolled out in China? And, is that actually as problematic as we in Europe often think?
Are you ready for a world where AI is a must-have for every company? During the two-day businessclass ‘AI for managers’ you will learn how to apply AI into your company.
With a 15-year career in business operations and international marketing management behind him, Robin founded Seedlink in 2013. He applies his knowledge and passion for language technology, human resources, organisational equality and diversity. Before that, he was responsible for the publication and viral marketing for the top of the mobile game, 调教三国 (Tiao Jiao Jiao San Guo).
Today, as founder and CEO of Seedlink, Robin has successfully integrated various technologies – such as AI, NLP and machine learning – with human resource management. Seedlink develops leading technology from Shanghai that helps companies predict employee behavior and hire talent by analyzing natural human language. By building the core values of equality and diversity, Robin, assisted by his team, uses this breakthrough solution to challenge traditional performance assessments and human prejudices.