What is artificial intelligence - and what influence will the technology have in our society?
We asked our expert in artificial intelligence, researcher and associate professor Luís Cruz-Filipe, who is also head of the Bachelor's programme Artificial Intelligence at the University of Southern Denmark in Odense, to answer that question.
Most people have by now heard of artificial intelligence – or AI. The media's coverage of the artificial intelligence ChatGPT since autumn 2022 has made the term commonplace, but what exactly does it mean?
– It basically means that systems based on artificial intelligence, so-called AI systems, can make decisions, solve problems and complete tasks that previously required human intelligence, time and energy, explains researcher in artificial intelligence Luís Cruz-Filipe.
He himself became deeply fascinated by the technology when he first became acquainted with it in an elective course during his bachelor's studies in Portugal.
– I took an optional course in introduction to artificial intelligence and was taken by it. In the same period, I read some books about AI experiments at the time, and then I became even more interested, says Luís Cruz-Filipe.
As his interest in artificial intelligence grew, he took several specialised courses and subjects on the topic during his studies. Today, he is a researcher himself and teaches artificial intelligence.
– In my opinion, the most exciting thing about artificial intelligence is the constant reflection on what it really means to be intelligent, he says.
– The original idea was that we would develop artificial intelligence by writing computer programs that roughly copied the human way of thinking. Now we often find that programs that behave extremely intelligently actually "think" in a completely different way than humans. I think that is both surprising and hugely exciting.
Most AI systems we encounter are limited by the data and input they are trained on. Although there are often huge data sets, there are often only a few aspects and a narrow context, while people can draw on a much wider range of experience and knowledge.
This means that artificial intelligence may have difficulty making decisions or predictions in situations it has not previously faced.
Likewise, these AI systems also have a more limited creativity. Because although AI systems can generate new ideas and solutions based on the data it is trained on, they cannot "think outside the box" or come up with completely original ideas in the same way that humans can.
Artificial intelligence has already revolutionized several industries, including healthcare, finance, retail and transportation. But the potential extends far beyond these areas.
– We already use artificial intelligence in many different fields, and we will do so more and more. It is an extremely powerful tool, as computers can analyse much more data than humans and thus make more informed decisions than we can ever dream of being able to ourselves, explains Luís Cruz-Filipe.
The technology can also contribute to tackling some of society's biggest challenges, including climate change and public diseases.
– With artificial intelligence, you can analyse large amounts of health data to help diagnose diseases early and predict which treatments will be most effective for a given patient, he says.
– Likewise, artificial intelligence can also help fight climate change by analysing data on weather, energy consumption and pollution to identify the most effective ways to reduce CO2 emissions and improve energy efficiency.
AI systems with biases
Although the potential of artificial intelligence is enormous, there are still some challenges that need to be addressed. One of the biggest concerns is that technology will take over people's work, as many routine tasks can be performed by artificial intelligence.
In addition, there is good reason for concern when it comes to data protection and privacy, because AI systems can collect and analyse large amounts of personal data.
– At the same time, there is a risk that we will rely too much on technology, says Luís Cruz-Filipe.
In so-called blackbox systems, one knows about the data input that a given AI system processes. You also get an output in the form of a result, a decision or a solution. But you don't know how the system reached its conclusion.
– And it's a challenge because you don't know which bias and which prejudices the AI system has developed during its training, he says.
– But if you handle these challenges in a good way, the potential for artificial intelligence is enormous. And if you use the technology constructively, it can contribute to solving some of society's biggest challenges and improve the lives of millions of people around the world.
Therefore, Luís Cruz-Filipe believes that we as a society should intensify efforts to explore and develop artificial intelligence.
– Fortunately, there is an increasing interest in so-called explainable artificial intelligence, which cannot only make decisions, but which can also explain them to people, he says.
These systems can justify and substantiate their conclusions if you challenge their conclusions - and they can learn from their mistakes.
– Because you can understand their thinking, they are much more trustworthy. It is also easier to assess whether they exhibit bias and prejudice. And thus you can better counteract such pitfalls, says Luís Cruz-Filipe.
Meet the scientist
Luís Cruz-Filipe is associate professor in artificial intelligence at the Department of Mathematics and Computer Science at the University of Southern Denmark in Odense. He is heading the new bachelor's programme Artificial Intelligence, which will accept its first class in September 2023.