Dr. Rasmus Rothe is the founder of Merantix, a Berlin-based AI venture studio which addresses real-world problems through deep learning. He also founded HackZurich in 2014, which has become the largest hackathon in Europe.
With Merantix, you have set yourself the task of tackling real problems with the help of Deep Learning. What's it all about and where do you start?
AI has a high potential for innovation and a wide range of applications - from supporting a doctor when making a diagnosis to autonomous driving. This is exactly where Merantix comes in. As Europe's leading platform for the commercialisation of AI, we are setting up independent, disruptive companies which translate top AI research into marketable applications. Specializing in deep learning and computer vision, we work closely with European industrial and technology companies and employ scientists and engineers from renowned international universities.
Your projects are diversified. Which specific innovations are you working on?
The self-learning software developed in the field of healthcare partially automates the diagnostics of radiological images, thus taking over a large amount of the repetitive work and enabling the radiologist to focus on complex and challenging examinations.
As regards what is probably the largest future trend in road traffic, autonomous driving, Merantix has developed a test platform for self-driving cars in the Automotive field, which accelerates their development progress and can consequently prevent accidents.
MX Labs is developing its own AI applications for large companies in areas such as computer vision and natural language processing. In this way, we are also exploring new fields of application and partnerships for future spin-offs. Current examples include automated systems in production, pathology and space travel.
You have decided to found the company in Berlin. How do you rate Berlin as a centre for technological innovation, especially in the field of artificial intelligence? What distinguishes Berlin from other locations and makes the city attractive for talented people?
Germany offers a good infrastructure for research and development and thus enables a rapid transfer of technology and knowledge. In addition, Germany has numerous leading industrial companies with which we can enter into valuable partnerships, especially in the area of data and industry-specific knowledge. The goal is to establish "AI made in Germany" as an internationally recognized quality seal and to position Germany as an innovation driver in the commercialisation of AI.
What significance does Merantix have for the AI industry in Berlin and beyond?
Merantix attracts internationally highly qualified AI experts who move to Berlin and thus promote the AI location. In addition, our AI products not only open up new markets but also promote the domestic market. The Berlin location has the advantage that we can support the development of the German AI strategy in close cooperation with politicians. Through presentations, we try to further inform local society about the opportunities offered by AI.
Let's look into the future: What challenges, problems and opportunities do you see for the future of artificial intelligence in general? Which areas will have been transformed by Deep Machine Learning in a few decades?
AI works particularly well when there are high-quality data sets from which it can learn. This is not always the case in practical applications, and a lot of effort is required to process the data. Especially the so-called corner cases, i.e. rare data points - unusual tumours in automated radiology, complex traffic situations in autonomous driving - are particularly valuable for improving AI algorithms or testing them for errors. I see great opportunities in Germany, above all in industries in which we are traditionally strong: automotive (e.g. autonomous driving), health (e.g. automated diagnostics) and industry (e.g. automated quality control).
AI has developed enormously over the last few years and will soon become an indispensable part of our lives. What advice do you give to companies which approach you and want to focus more on the topic of AI in order to drive innovation forward?
Companies should analyse their business areas for AI application possibilities as early as possible. In addition, it is important to determine whether there are new business models which AI makes possible. Pragmatically implemented pilot projects can help to test the real potential before greater investments are made.