AI is already being used by all manner of employees and workers, and it seems that politicians too are getting in on the action.
Sweden’s Prime Minister Ulf Kristersson has said that he uses AI for his work as Sweden’s highest decision-maker. “I use it myself quite often, if nothing else for a second opinion. “What have others done?” and “Should you think exactly the opposite?” — those kinds of questions,” he said.
Kristersson pointed out that he didn’t upload any documents onto the AI tool, such as political investigations, reports, motions and decisions. He said that he used AI as doctors did, to gain more perspective on issues.
The Swedish Prime Minister might be the first world leader to openly say that he uses AI to help with his job. However, he does seem to be skeptical about their security — he was quick to point out that he doesn’t upload any confidential government data to AI models.
There has been an instance of someone taking Kristersson’s idea to an extreme. Victor Miller, a 42-year-old librarian, ran as a mayoral candidate for Cheyenne, Wyoming last year. But unlike other candidates, he didn’t have any policies of his own, or make many promises — he simply said that he’d defer all major decisions as Mayor to an AI bot, and do what the AI told him. Voters, however, weren’t sold. Miller polled just 327 votes out of the 11,036 ballots cast, and lost in a landslide.
It does make sense for government officials to use AI tools. Many elected government representatives aren’t subject matter experts, and often would need to quickly get up to speed on various issues across tech, governance or geopolitics. Instead of relying on human experts, they can get inputs from AI models, which can help them shape their opinions and make decisions.
But this use-case is another compelling reason why making sure AI is truth-seeking and honest is so vital. Several AI models have been caught sharing biased points of view, or being overly sycophantic and agreeing with whatever the user says. AI is already being used in the governance of humans through human politicians — it’s more important than ever than its results are accurate and unbiased. It has only been a handful of years since AI went mainstream, but it already seems to having say in the lives of its human creators.