Dr. Tina Klüwer is a computer linguist and a qualified expert in AI. She heads the Künstliche Intelligenz Entrepreneurship Zentrum (K.I.E.Z.) and is a member of the German Chancellor’s Future Council.
| K.I.E.Z./Tanja SchnitzlerGermany has massively expanded AI research in recent years. Berlin is mecca of the industry. Unfortunately, the situation is different when it comes to technology transfer. There is still considerable room for improvement here: not enough AI research is actually finding its way into business applications. The Künstliche Intelligenz Entrepreneurship Zentrum (Artificial Intelligence Entrepreneurship Center (K.I.E.Z.)) aims to change that.
With support from the three Berlin universities and Charité university hospital, K.I.E.Z. primarily promotes the establishment of science-based AI startups. K.I.E.Z. offers continuous support to about 40 founding teams in various funding programs. The supply side has seen a huge boom in the range of promising AI solutions across all industries and value chain stages. Nevertheless, there is still great restraint on the demand side in Germany. SMEs in particular have been hesitant so far. While half of all companies with more than 2,000 employees have already introduced AI solutions, the figure is only 18 percent for medium-sized companies and just five percent for small companies. Triggered by the hype surrounding ChatGPT, the topic of artificial intelligence is now gaining momentum in Germany as well. More and more companies are contacting K.I.E.Z. to find out about the best way to get started with AI.
K.I.E.Z. is now also supporting established companies in strengthening their in-house AI expertise. Not everyone needs to become an AI expert, but basic processes should be clear to all. Only then can people truly begin to understand the level of detail required to make AI useful in the company. We recommend that companies first limit themselves to individual processes. This makes it easier to formulate KPIs and transparently determine what the AI project should actually achieve. Last, but not least: AI needs data. Very few companies will find the performance of large, pre-trained models will adequate for their needs. So anyone thinking about AI projects should first check whether they have already digitized their processes or whether they can collect digital data in the future. It is high time to scale up artificial intelligence usage. If you start with clearly defined use cases that have measurable benefits, you can quickly reap the rewards.
