Artificial intelligence should have a say in difficult medical decisions or the question of who is invited to an interview. The Swiss population thinks so.
In Switzerland, people are open to artificial intelligence (AI) and want to involve it more. Markus Christen and his research team found this out. He is an AI expert at the University of Zurich’s Digital Society Initiative.
AI should have a say in who gets the job
Around 1,500 people from all over Switzerland were interviewed in the representative survey. About two-thirds of respondents want artificial intelligence to be involved, including on sensitive issues such as dropping out of therapy, getting a mortgage, or releasing a prisoner.
Legend:
AI is increasingly finding its way into our lives.
REUTERS/Yves Herman
The more serious the decision and the more artificial intelligence decides such things alone and fully automatically, the more critically the Swiss population views it. If the AI helps to identify fake news, then that is not a problem, for example.
AI ignoramuses, women and the political left are more skeptical
On the other hand, when it comes to who gets invited to an interview or who gets a loan, these are momentous decisions. In these cases, it is more acceptable if the AI simply makes recommendations. The AI should be included, but ultimately a human must decide.
Those respondents who know nothing or very little about AI or digital topics, i.e. around a third, are more skeptical about the new technology. Around two thirds of the Swiss population are more familiar with the topic and are more optimistic. According to the survey, women and the political left tend to be more skeptical about AI.
Trust more important than knowledge
The survey was conducted at the beginning of last year, before the time of ChatGPT. Study author Markus Christen admits that ChatGPT could have certain effects on the data and results. Now, for the first time, people are becoming aware of what AI can do. Even though we’ve had AI around us with Google and Siri for a long time, says Christen.
Legend:
ChatGPT has brought AI to a wider audience.
EPA/WU HAO
Nevertheless, the study remains relevant. Because knowledge about artificial intelligence does not influence the acceptance of AI that much. According to study author Markus Christen, the much more important factor is whether we trust the companies that use AI.
Banks and tech companies have a hard time
According to the survey, trust in AI is highest in healthcare. When hospitals use artificial intelligence, the Swiss population accepts it more than when a bank does.
Legend:
Instruments of the robot-assisted Da Vinci surgical system.
KEYSTONE/DPA/Robert Michael
Next to the banks, the tech companies enjoy the least trust when dealing with AI, for example when they use AI in social networks. The AI technology is usually provided by these technology groups, even if it is later used by hospitals.