Artificial intelligence - opportunities / risks

Why do we, as specialists in travel expense management, deal with artificial intelligence?

Before we take a closer look, we would first like to clarify our motivation as to why we deal with artificial intelligence (AI).

It's basically quite simple. On the one hand, we keep hearing or reading quite exciting things about artificial intelligence, which either trigger enthusiasm or fear. Enthusiasm was created, for example, when an AI, Alpha Go, won for the first time in the most complex game GO against the best human player in the world. Move 37 has become synonymous with (supposed) AI superiority. Or Alpha Fold 2, which created a sensational breakthrough in biology (predictions of protein structures, folds).

Worries or fears arise when it is revealed that Ais are sometimes used to predict crime or even to hire employees, which perpetuates old prejudices, racism or misogyny.

Whether AIs are good or bad, what they will be able to do in the future or where the limits are, is certainly still a matter of debate.

However, we have noticed that many people are following this discussion and/or have already formed an opinion about it. At the same time, people are almost completely unaware of how AI is already used by all of us on a daily basis and as a matter of course. This happens when we run searches on the internet, are on social media platforms, perhaps have a text translated at DeepL or with all our customers when they use our SaaS solution for travel expense reporting.

There are certainly AI specialists who could explain all the intricacies of AI in a much more in-depth and differentiated way. Our motivation is to provide you - our customers, interested parties and followers - who are relatively certain not to be AI specialists, with an initial introduction to the topic. Because one thing is certain for us: AI will increasingly affect us directly or indirectly in all areas of life.

Where to start?

That is a really difficult question. As we wrote above, we don't want to dive into the depths of the Rabbit Hole here, but rather give our readers a first approach. At the same time, we think that in order to deal with the topic of AI in a relaxed manner, a few very basic points should be understood.

What do we actually mean by AI? There are certainly some very different views on this. From our point of view, AI means all areas of information technology (and possibly robotics) that set themselves the task of generating intelligent behaviour by means of computer programmes. This can be a selection, the recognition of patterns or correlations, the control of a concrete action, e.g. of a robot, etc.

But even with this view, the question immediately arises: What does intelligent mean? Does it mean human-like intelligence? Or is a similar one sufficient? Or can it be quite different?

Let's take an example. How do I stack different objects on top of each other? If I have a few cubes, the task is quite simple for us. For a child, however, it is not. He will gradually find solutions by observation and then trial and error. From our point of view, that is definitely

Intelligence. There is a problem (or a task), a problem-solving process and a solution. And optimally, during the problem-solving process, a model is created that enables us to solve this problem again and again in the future without having to start from scratch.

Or do you have to start from scratch every time you submit a travel expense claim to analyse how to deal with it? Of course not.

Almost all AIs in use today work according to a very similar principle. There is a specific goal that is given. And there is data and models. The AI learns with the help of data until the desired results are achieved.

In our view, the way AI learns also explains why the first waves of AI research did not produce the results promised at the time. This in turn has led to the fact that even today some people are convinced that AI will not prevail in the long term or will again fall short of the promises. (There is talk here of a next AI winter).

On the contrary, as you can easily see from our article, we are convinced that AI will be one of the most important technologies of the future, which is why it is worth getting involved with it. But why should this time be different? AI research as a whole has made enormous progress in the last 10 years. And in our view, this is because two very significant points have changed from the first waves of AI:

• Computing capacities have expanded exponentially. Above all, the possibilities for parallel computing play an important role here.
• Enormous amounts of data are available in many sub-areas.

Do I have to become an AI specialist to survive in the future?

Were the first sections too technical or confusing? Don't be put off.

Who among us knows how exactly an engine or the brakes in our cars work? And yet we just get in and use them, whether by driving ourselves or being taken to our destination in a taxi. In exactly the same way, we use various AIs.

At the same time, we feel safer when we have at least understood basic principles of techniques etc. So how does a car work in general, or what criteria does a company use to recruit its employees, etc.?

In this respect we can reassure you. It will certainly be the same with AI in the future as with all other technologies before. At first we don't understand anything and are just excited or a bit scared (depending on our disposition), then we use it and gain our first experiences and at some point we don't want to give it up. (Some still remember the first mobile phone/Blackberry/Smartphone?).

And a few more analogies come to mind:

• “Cars will never catch on.”
• “Speeds in trains of more than 35 km/h are deadly”
• “The telephone is a rich man's game”

There are always technologies that have the potential to change our society from the ground up. In our view, AI is one such technology. There are certainly critical issues related to AI, such as the protection of our individual rights, data protection.

To whom does the data actually belong to and who is allowed to use it and how do we decide this? And at the same time, there are huge opportunities in this technology to help us solve fundamental problems, first and foremost climate change. Or in the health sector, where AI is already being used in diagnostics, among other things. Then there are the many business applications that make our lives easier, such as the use of AI in our software for digitising your receipts.

How we use them is up to us and certainly requires a broad discussion. However, we can only take part in this discussion in a profitable way if we deal with the topics and have at least a basic understanding of them.

Outlook

For us, AI is a super exciting topic and we are already using the technology for the benefit of our customers. Perhaps this blog post has helped to make the topic of AI more accessible to you and has made you curious.

As with other topics, we are starting a small mini-series with this post. So there will be more blog posts on AI to follow, in which we want to deal with topics such as:

• White Box and Black Box
• Symbol-based AI and neural networks
• AI in our application (travel expense accounting; digitisation of receipts)
• etc.

Feel free to send us feedback or make an appointment for a free exchange of experiences. We will be happy to share our expertise with you.

https://www.openmind-travelmanagement.com/kontakt

To make sure you don't miss anything, feel free to follow us on LinkedIn or facebook.

https://www.linkedin.com/company/openmind-gmbh

https://www.facebook.com/openMINDtravelmanagement