advertisement
advertisement
HANNOVER MESSE 2019, 01 - 05 April
switch to:
Prof_Sepp_Hochreiter_Portrait

Prof. Dr. Sepp Hochreiter

Don’t throw away your lead

Professor Dr. Sepp Hochreiter, Head of the Machine Learning Institute at Johannes Kepler University in Linz, is a leading expert on Artificial Intelligence (AI). His ground-breaking research into deep learning has been applied by Google, Amazon and Facebook. Hochreiter also works in the automotive industry , e.g. for Audi.

Artificial intelligence is booming on the West Coast of the USA (Google, Amazon, Facebook, etc.), in China (Baidu, Alibaba, etc.) and in South Korea (Samsung, Naver, etc.) And in Europe? In Germany? Dear Mechanical Engineers, I’m very concerned: don’t throw away your technological lead! We should not be guided by Google and Baidu, despite their positive public image. Instead, Europe should place emphasis on artificial intelligence in the mechanical and plant engineering sector. Our goal must be a smart lathe, and not a new smartphone. The user should be able to communicate with his machine: “Watch out, you are going too fast!” “No problem,” says the machine. “I’ve been given a new type of lubricating oil and can work faster.”

But we must act now. Companies in Silicon Valley tell me, “We can do the engineering ourselves or buy in external expertise.” As it stands now, the mechanical engineering sector is blind. Unlike Facebook and Apple, it does not maintain permanent contact with its customers, nor does it analyze customer data – because most customers do not want this. AI will lead to revolutionary successes in the medical field, but we must bear in mind that patient-related information and genetic data are much more sensitive than production data.

“Where do we need to take action? In relation to lathes, dishwashers or power drills. Many other countries are now capable of building good cars.”

Prof. Dr. Sepp Hochreiter

Machine manufacturers and users need to wake up and pay attention. Every company must have its own data officer who can analyze information and define processes. Small enterprises should join forces to collect data, generate common benefits, share profits and obtain/exploit experience with AI. We already train artificial experts at universities – for example, at Linz, where we are inaugurating a new degree course in AI. The aim is to produce engineers – not for the Googles of this world, but for the mechanical engineering sector – engineers who are equipped with the necessary data and marketing strategies for new business models.

And then we need to invest – in infrastructure, 5G and fast networks. Companies lack computing power (both internal and external) as well as GPU clusters linked to storage systems via high-speed connections. Without a high-performance computing (HPC) strategy it will be impossible to keep pace with the competition. This is Europe’s bottleneck, alongside the shortage of experts for AI, data science and algorithms.

Where does the competition come from? From the USA, certainly, but also from China and South Korea. These countries spend billions on AI projects, with special emphasis on medical technology and biotechnology. Where do we need to take action? In relation to lathes, dishwashers or power drills. Many other countries are now capable of building good cars – unfortunately.

Are we afraid of AI? We have no reason to be. AI is like dog breeding. Humans are capable of eliminating weak AI. AI does not have an intrinsic will to survive. It is a tool in service of humankind. It will be well-disposed towards us.

advertisement
advertisement