Markus Schmitt, Federal AI Association © Markus Schmitt

08 May 2019

“We need companies that live AI and don’t just mention it in marketing.“

One might think that the AI scene in Germany moved noticeably closer together in March 2018. A total of 24 companies whose business models are predominantly based on AI have joined forces to form the Bundesverband Künstliche Intelligenz (Federal Artificial Intelligence Association). With its roughly 50 members, the association aims to serve as a mouthpiece and lobby for the industry on the one hand and as an interface to federal politics on the other. #ki_berlin met Markus Schmitt, member of the KI-Bundesverband and owner of the start-up Data Revenue, and talked about research, financing and the future of the AI ecosystem in Berlin.

With your start-up "Data Revenue" and your position as part of the KI-Bundes­verband, you have an insight into both the AI ecosystem in Berlin and State politics. What observations have you made?

I am pleased to see that AI has gained political momentum as a focal topic. There is a good basic understanding of technology and meaningful programmes to support AI in Germany. Many people are dealing seriously and conscientiously with the topic.

Unfortunately though there is still a lack of assertiveness, sufficient courage and pragmatism as regards the details. Access to funding is still too costly for most small companies.

But I can well understand that there is a little reticence as regards this subject. In the foreseeable future, AI will be just one of many promising technologies - several of which will not receive the same attention. The ecosystem in Berlin is still expandable. There is still too little exchange between competent AI teams, and behind many of the "AI start-ups" there is too little AI.

The establishment of the Federal Board of Artificial Intelligence last year was welcomed both by politicians and by research and industry. What is the work like, and which goals has the association set itself for the next few years?

We help to network German AI start-ups, both with each other and with industry. For example, I am working intensively on networking with biotechnology and pharmaceu­tical associations. In addition, we are involved in topics that promote the AI scene in Germany - such as extending funding, defining the legal framework and increasing visibility in all directions.

A small reversal of the trend can be observed - AI is the subject of more intense public discussion. What role does technology play in political dis­course?

Unfortunately, too little professional competence is found in the discussions. And so there is a large focus in the discussions on moral, ethical and labour market issues - these are easier to discuss and also more effective for the media. However, a realis­tic view and topics such as training the right skilled workers and promoting and sim­plifying the framework conditions are not adequately taken into account.

So has the labour market not yet adapted to the requirements? What new steps are needed?

Many companies are still placing too much emphasis on theoretical knowledge, which is why doctoral students in statistics and mathematics are hired for positions in which talented, experienced programmers with a quick grasp of mathematics are rather needed.

In the end, this often leads to long delays in the development of AI systems and thus to frustration. With a realistic understanding of AI, companies would form and pro­mote the right teams more quickly.

Let's come to your start-up "Data Revenue": you offer tailor-made machine learning solutions. How are you positioned and which areas and cases are we talking about?

We are a team of developers who have specialized in the implementation of large AI applications. We recently built the recommendation systems at two German TV sta­tions, many of the AI systems at ImmobilienScout24, as well as forecasts in the energy market for municipal utilities. We are just starting out with AI for plant optimi­sation at two German automobile manufacturers. We also advise companies on the correct use of AI and help managers to better understand the technology, for exam­ple at Haniel and E.ON. However, our focus is increasingly on research and devel­opment topics for new drugs - where we help scientists automate their data analysis processes with AI.

There are a number of AI companies in Berlin - including your start-up. What makes the capital so important and what distinguishes Berlin from other AI centres?

I think Berlin has very many experienced developers. I also have the impression that Berlin universities were involved in this topic earlier than others. But I still don't have the feeling of a real "AI location". We still have a lot to catch up on and win over seri­ous AI teams. We need companies that live AI and don't just mention it in marketing.

The USA, and China in particular, are making immense efforts in the field of AI. What must politicians do to ensure that Germany does not fall behind in an international comparison?

We must make it easier for developers to set up technology-driven companies. To do this, we need more talents, more practical university courses and more technology-oriented investors.

Thank you for the conversation.