Artificial Intelligence—General Purpose Technology

AI algorithms are the foundation of the development of speech-based systems and chatbots, they help to predict production downtimes and support doctors in recognizing and treating diseases. AI algorithms save time and money. More often than not, they even make better decisions than their human counterparts.

Sometimes, people ask whether or not AI is only a hype, because methods like neuronal networks have existed for decades. While this observation is correct, I am nonetheless convinced that AI is—unlike in many digital transformation projects—not only a buzzword.

The reason why AI is experiencing such a boom is because the corresponding framework has improved immensely over the last couple of years. Now, data needed to, for example, train neuronal networks is abundantly available. Computing power and memory space are cheaper than ever before. What is more, for the first time ever, there are many free toolkits and software libraries which greatly simplify and improve the development and quality of applications.

Artificial intelligence is more than just a tool for the automation of existing processes. In the future, the creation of new data-based business models will be the focus.

AI investments

A strong indication for its ever-increasing importance is the growing figure of investments in AI technology. From 2011 to 2017, the investments in AI start-ups have grown to over 15 billion U.S. dollars (CB Insights, 2018). By 2020, experts expect the number to have grown even more to 70 billion dollars (OECD, 2017). This shows that it is a growing trend to invest in AI.

However, these investments in AI start-ups are not evenly distributed. 48 percent of them have been coming from China, followed by the U.S. with 38 percent. The rest of the world only accounts for 14 percent (2 billion dollars) of investments in new AI companies.

Even though Germany is home to many excellent AI scientists, the country is not even the frontrunner in Europe. Great Britain is number one, where start-ups like Darktrace (cybersecurity) and Graphcore (AI chips) have become financial unicorns, meaning they are valued at over one billion dollars.

The German government now wants to take countermeasures with an "AI master plan". However, despite the excellent AI training at universities, the economy has not really jumped on the bandwagon so far. "Germans don't recognize the world-changing power of artificial intelligence," warned AI professor Jürgen Schmidhuber years ago. The money is there, but although Google is investing half a billion dollars in an AI lab, German companies are not. Little has changed to this day.

While Americans tend to invest here and there without a clear strategy in mind, China’s investments are more goal-oriented. The government invests in few, but already developed start-ups. As a result, five of the seven most valuable new companies are from China.

Right now, a lot of money is being pumped into building the AI companies that are expected to achieve the major technological breakthroughs in medicine, agriculture, industrial production, and infrastructure for the Chinese economy in the coming decade.

While the Americans spread their investments widely, the Chinese invest specifically in a few, but already mature start-ups. As a result, five of the seven most valuable start-ups today come from China: Sensetime is at the top with a valuation of $4.5 billion, followed by Cambricon ($2.5 billion) and Megvii/Face++, CloudWalk and Yitu ($2.5 billion each) in third to seventh place.

If anything, Germany's gap to the leading AI nations is widening. In our opinion, it is high time for a constructive debate on AI if we do not want to lose touch again with a future technology in which we are actually among the pioneers.