Germany’s Social Democrat Leader Saskia Esken called for a closer examination of the discrimination inherent within data sets used to train artificial intelligence models.
(Bloomberg) — Germany’s Social Democrat Leader Saskia Esken called for a closer examination of the discrimination inherent within data sets used to train artificial intelligence models.
“We have to pay attention to what sort of data is playing a role in training these models, and if we need to possibly clean up training data,” the leader of Germany’s ruling party said to journalists in Berlin on Tuesday.
Esken said an AI tool could prove biased against women seeking jobs as IT specialists, for example, or against young men applying for credit checks, she said.
Read More: FTC’s Khan Pledges Vigilance on AI as Technology Develops
Esken was presenting the Social Democrat’s new digital policy paper, which places the party’s position on AI and the role it should play in society front and center.
Pointing to the “hype” around a boom in generative AI tools, the paper calls for “transparency and responsibility in the development and deployment of these AI models.”
The group also advocates in the paper for a four-day work week.
“The self-determined use of one’s time and work-life balance are far more than just individual needs,” it says. “We want to enable more self-determination over this work-life balance, e.g. through new working time models such as a four-day week.”
More stories like this are available on bloomberg.com
©2023 Bloomberg L.P.
