Among the ballot papers in the Brighton Pavilion constituency, one candidate stands out. He is not a real person. The candidate is named after its creator, businessman Steve Endacott. The avatar “AI Steve” was created with artificial intelligence (AI) and answers questions about everything from queer rights and immigration to housing challenges. – What we do is create a representative you can talk to 24 hours a day, 365 days a year, says Endacott to Reuters. On the website of “AI Steve”, voters can talk to the AI-generated avatar. This is a screenshot of the front page of the website. Will use the conversations in politics But a computer program cannot appear in the national assembly. If “AI Steve” gets enough votes, the man behind it, Steve Endacott, will be the representative for the district. He claims he will vote for proposals from voters, who have spoken to the KI version of him. – It discusses politics with you. It also gives information back to me about what people want, says Endacott to the news agency. A cardboard figure of “AI Steve” stands with its creator, businessman Steve Endacott. Photo: Muvija M / Reuters Endacott, who is running as an independent representative, believes the avatar will make the election more democratic. It’s just a stunt, says UK expert Erik Mustad. Erik Mustad is senior lecturer at the University of Agder. He has little faith that the British will have any particular interest in “AI Steve”. Photo: Trine Andersen / NTB scanpix – It seems like a desperate attempt to show that he has money and wants influence, says Mustad to news. Hundreds of candidates stand for election in Britain on July 4, few receive national attention. Mustad believes the KI politician is only trying to buy it. – We have seen it before in the UK that a number of rich people want to gain power and influence, either through a legal system or through a political system. AI-generated politicians in the future? People who want to know about Endacott’s politics can talk to the avatar. They get the answers right away. – Politicians are quite inaccessible to most people. Being able to talk to an avatar like “AI Steve” is a funny thought, says digital ethicist Leonora Onarheim Bergsjø at Østfold University College. Associate professor at Østfold University College Leonora Onarheim Bergsjø thinks the KI politician fits the times. Photo: Brother Kvammen Bjerke / news She sees potential in “AI Steve”, but points to several challenges in the future use of AI-generated avatars by politicians. Trust, credibility and climate footprint are examples of this. – If everyone in the population is to talk to politicians’ avatars, we will have an enormous climate footprint. Then we have to ask the question whether this is really where we should use energy. The avatars will also be influenced by the data it is trained on. It is a myth that technology is ethically neutral, Bergsjø explains. – The avatar will appear more refined in its opinions, but it is just as charged with values as any politician and the experiences that a human being has. For accessibility, “AI Steve” is excellent, but when it comes to integrity it won’t work, she states. – It’s about trust. There must be an actor who can take responsibility and say sorry when something goes wrong. It is about integrity and many other things than just being able to give good answers. Published 30.06.2024, at 21.06
ttn-69