In Teil 1 des ARIC-Interviews hat Michael Koch bereits über die Chancen beim Einsatz Künstlicher Intelligenz gesprochen und Argumente gegen KI-Skepsis geliefert . In Teil 2 erklärt unser neuer ARIC-Vorstand, woran KI-Projekte meistens scheitern und spricht über die Vereinbarkeit von Datenschutz und KI.
Where do SMEs stand when it comes to AI?
Some companies are already relatively advanced, for example in the media industry. They have been using AI for some time to create user profiles and offer tailored news, for example. On the other hand, AI solutions are often not considered further after initial approaches, as supposedly high investments do not bring the desired business success. It is therefore important to carefully consider which AI solutions are economically and strategically viable for a company.
Another reason for the failure of AI projects is their complexity. Until recently, AI development was more akin to individual programming, for which AI experts often had to be found or trained first. This makes projects lengthy, expensive and often inefficient to operate. With pre-trained solutions as a cloud service, business-generating AI is now within the reach of SMEs. Artificial Intelligence as a Service (AIaaS) reduces the challenges, as numerous sub-problems have already been solved. Even with an AIaaS solution, companies need internal expertise in artificial intelligence. However, thanks to the use of intuitive low-code tools, it is not necessary to write code yourself. The implementation via program interfaces is also turnkey, so to speak.
What hurdles do you still see for the use of AI in Germany?
AI is primarily dependent on data and reliable information can only be obtained from high-quality data. No matter how good the software part of an AI application is, insufficient, faulty or poorly structured data can cause some projects to fail at an early stage. Due to a lack of suitable data, only around one in five proofs of concept are actually implemented and efficient solution scaling is virtually impossible. It is therefore important that companies tackle digitalization quickly. This includes a detailed data strategy, governance, architecture and culture. Now is the time to tackle digitalization and AI quickly and decisively in order to remain competitive.
Where does Germany stand internationally in terms of AI?
The USA and China are still the leaders in this segment. In Germany, strict data protection regulations limit the use and analysis of data. But that is not necessarily a bad thing. In fact, I believe that Germany and Europe can gain a competitive advantage by observing data protection, security and ethics and can be international pioneers in the responsible use of AI. However, this also means that we should not prefer not to use data at all for fear of doing something wrong. Rather, with good expertise, data protection, data analysis and AI are very compatible.
Why are ARIC and LHIND a good match?
We have the great advantage of being able to accelerate the path for SMEs thanks to our experience in the large corporate sector. LHIND has had an innovation department for AI since 2014. Since then, we have been able to carry out numerous AI projects and proof of concepts together with our customers inside and outside the Lufthansa Group. We also run AI operations that have been used on a daily and ongoing basis for years. Thanks to years of experience and contacts with universities, we now have a wealth of knowledge at our disposal.
We also want to offer this knowledge to the industry. Because we see a strong need for training, the application of knowledge and further development. We can contribute a great deal to this because we have experience of how large corporations or customers with many years of experience integrate AI into their standard processes. The question is: do SMEs have the same problems or do they face different challenges? Can we, as an AI community, set up guidelines together to prevent problems so that we don’t end up working on the same things again for five years?
Exchange and knowledge transfer are extremely important, because a company cannot meet the challenges of the future on its own.
