New paradigms for Artificial Intelligence: An interview with AI pioneer Eberhard Schöneburg
There is no doubt that Artificial Intelligence (AI) can play a vital role in any industry where there is an opportunity to automate business processes. We spoke to Eberhard Schöneburg, award winning scientist and pioneer in AI about what this means for the banking sector, but also his take on emerging trends around the technology itself.
Eberhard, what would you say are the most promising applications of artificial intelligence in the banking industry and which will be the most relevant disciplines for banks?
Eberhard: In general, I see two areas that are interesting: the blockchain activities and AI – and then especially the combination of both. In my opinion, this has the most potential. In terms of disciplines, you can separate two areas in banking. One is anything connected to the front-end and the other one to anything that is going on in the back office. The requirements in both areas are very different. In the front end, I see many Natural Language Processing (NLP) activities and they have the highest return on investment. In the back office, things are much more diverse – you have legacy systems that have to be dealt with and a lot of optimization, automation of processes and so on. Many banking systems are ten, twenty years old and to replace them with modern AI is a major task.
Do you expect a significant leap of AI disciplines in the upcoming years?
Eberhard: That’s a good question. There are two fields, both of which I’m working in. The one is what we call Narrow AI applications in banking, the other one Artificial General Intelligence (AGI) and Alternative AI (AAI). In these last two disciplines, you will probably see some content leaps very soon that will solve problems that we have not been able to tackle in the past. However, the Narrow AI field is still dominating the market and will do so for a certain period, because it is a more proven technology. So the short term will be more focused on Narrow AI, Deep Learning, these kinds of things, but mid-term to long-term, it will be more advances technologies.
Could you share examples of AI predictions for the future?
Eberhard: Well, the current AI models all more or less derived from how the brain works. But there are more interesting models out there, especially for distributed systems: Swarm intelligence for example, looking at colonies like ants, understanding how the operate and work, how intelligence emerges. Even looking at bacteria, which have been around on earth for millions of year, much longer than our species. They can be extremely smart and self-organized, they work together. So having mentioned AGI and AAI, these are new paradigms for AI that will be exploited very soon.
Regarding banking, where do you see movement in terms of AI?
Eberhard: Banks are not known for innovation, are they? But what you can see in general, is that Asian banks or banks that have most of their business in Asia are much more aggressive and innovative than traditional European banks or even American banks. Just think about it… in the United States, when you want to make a payment, you usually still issue cheques. Written paper cheques! In Europe we have become more into automating financial transactions, but the key point is: The future doesn’t rely on traditional banks as intermediators. Messaging service providers, the Googles of the world, the Chinese players, they become banks. That’s a big threat and action is required now.
Speaking of threats and new players… which start-ups should banks have on their radar?
Eberhard: Generally speaking, I think banks would do a good job, if they knew what was out there and worked with FinTechs – not swallow them – but find a modus operandi to integrate their services. There are many promising approaches out there. One of the most interesting companies that I know of is an Asian firm, providing a platform for AIs in general. So instead of working with one AI, you work with this platform and use the service of a particular AI, but this AI itself then uses other AIs in the background that you’re not even aware of. I think that is a very cool and powerful approach, because it has a self-organizing effect and an underlying AI economy: These AI modules pay each other for their services. Other areas worth mentioning are everything around messaging systems and integrated AI in these messaging systems, but again, the Asian model there is a bit different from the European model, because in Asia you have much less privacy concerns, regulations and rules. So you can do things in AI, with AI in Asia, you would not be able to do in Europe.
Thank you for the interview, Eberhard!
Also watch our video interview with Eberhard Schöneburg:
