News for hackers. Stuff that matters.

A comment on Dutch gov't wants to ban officials from using AI software. Parent comment

Re:Dutch gov't wants to ban officials from using AI software by vanderZwan on Thursday November 16, 2023 @10:19AM (#38290597)
As another Dutch person I'm actually disappointed that this is purely limited to generated images or text, given that the Childcare Benefits Scandal[0] also involved machine learning algorithms that encoded institutional biases against parts of our society (among many other structural issues, but let's focus on the one relevant to this topic).

But then again, I no longer live in the Netherlands so maybe I missed an earlier memo that the government decided to no longer do that kind of thing. I highly doubt it though.

[0] https://en.wikipedia.org/wiki/Dutch_childcare_benefits_scand...

[ Parent | on HN ]
Re:Dutch gov't wants to ban officials from using AI software by creesch on Thursday November 16, 2023 @02:52PM (#38294522)
> also involved machine learning algorithms

I could be wrong, but afaik it involved "just" algorithms not machine learning. Which is a meaningful distinction, as traditional algorithms can actually be audited much easier in regards to how they make calls.

Which is also (although with limited success) what they have been focussing on with things like the algorithm register (https://www.digitaleoverheid.nl/overzicht-van-alle-onderwerp...)

[ Parent | on HN ]