The text below was published on 13.3.2023 in YLE news. Read the original article here: https://yle.fi/a/74-20021837 NOTE! the original article contains media content that is not visible here!

What is artificial intelligence?

The question might be expected to be easy for a computer science professor, but Matti Tedre says the whole word is problematic.

– Artificial intelligence is a very vague term. It depends on the decade what is considered artificial intelligence, says Tedre, who works at the University of Eastern Finland.

Immediately after that, he admits that he too speaks of artificial intelligence.

The reason is that if he starts explaining deep neural networks, language and diffusion models, only a few fourth or seventh graders would understand.

And among others, for them Tedre talks about artificial intelligence.

You encounter AI everywhere

Professor Tedre leads GenAI research, which aims to strengthen children's and young people's ability to understand artificial intelligence and machine learning.

Understanding is needed, because in today's world it is difficult to avoid technologies that utilize artificial intelligence.

AI feeds us images and videos on social media, selects search engine results and customizes targeted ads. In other words, AI influences how the world appears to us. The image of hippos taking selfies was created using Midjourney AI by entering a simple command into the application.

Tedre hopes that children will learn to understand and control their own data flows.

– It's good to understand what information is collected about you and what it's used for. This helps if you know how artificial intelligence and machine learning systems work.

Researchers develop teaching methods

In educational institutions, teaching the basics of artificial intelligence is complicated by the fact that AI is also a new thing for a large part of the teaching staff. This is why one of the goals of GenAI research is to develop teaching methods for schools that make it easy to get acquainted with artificial intelligence.

– We work closely with schools. The research includes 12 schools that we visit, says university researcher Henriikka Vartiainen from the University of Eastern Finland.

The research group is also multidisciplinary. GenAI research includes researchers in computer science, education and law from the University of Eastern Finland, University of Helsinki and University of Oulu.

Getting to grips through practice

Difficult things are learned most easily by experimenting and doing them yourself. This also applies to artificial intelligence.

Matti Tedre and Henriikka Vartiainen have led classes for schoolchildren in which children and young people get to design and experiment with building a classifier. A classifier is one of the key technologies of artificial intelligence.

– For example, self-driving cars need a classifier to recognize traffic signs. The car must be taught that this is a speed limit sign and that one is a triangle, Tedre says.

During classes, students built classifiers using Google's Teachable Machine program. The result was working applications, such as an app that recognizes colors for the color blind.

– One group of girls developed an app that recognizes cheerleading moves. They filmed all the moves themselves and taught them to the app, Tedre says.

Problems also become visible

Building your own artificial intelligence and testing existing applications helps you understand the problems associated with AI.

– When asked to create an image of a prime minister, AI offers a picture of a gray-haired man. From this it's easy to start a discussion about whether the image matches your own idea of what a prime minister looks like, university researcher Henriikka Vartiainen says. If the data used to teach AI is biased, the bias is visible in the result. When Midjourney AI is asked to create an image of a nurse, it provides an image of a woman.

This way we get to the bottom of algorithmic bias, which is one of the biggest problems with artificial intelligence. Since most of the prime ministers in history have been middle-aged men, the algorithm doesn't know how to correct the bias but assumes this is still the case.

– We have had very deep ethical discussions with children and young people in classes and workshops, Vartiainen says.