What are we actually talking about when it comes to artificial intelligence?
Maranke Wieringa MA (Media and Performance Studies) and Dr Mirko Schaefer (Media and Performance Studies) talk about algorithms and artificial intelligence in the . In the podcast , Wieringa also talks about SyRi (System Risk Indication). Using algorithms, SyRI looked at neighbourhoods in order to profile citizens on the risk of them committing fraude with social services. This raises the question why it is so important to think about values and standards when developing or using an algorithm.
Schaefer: "An algorithm should be seen as a recipe of which you follow the instructions. A set of rules. Algorithms are a good tool for automating work processes, for example. But when using them, you have to be critical."
People work
Wieringa emphasizes: "You shouldn't think that an algorithm is neutral. An algorithm is made by people and trained with a dataset compiled by people, which is (unintentionally) not always complete (unintentionally)".
Social tension areas
If you have not thought carefully about these algorithms beforehand, you can unintentionally punish or disadvantage groups
"If you use algorithms in areas of social tension, you have to be careful," says Schaefer. "Think, for example, of algorithms that have consequences for at-risk groups, people with a migration background or the long-term sick. If you have not thought carefully about these algorithms beforehand, you can unintentionally punish or disadvantage groups."