Canadian here. Government sticking its nose into AI bias in human rights issues like law enforcement powers and employment standards … OK. But regulating search algorithms … feels worlds apart in importance and maybe too far.
back
2 comments
While I'd generally agree and do not expect the government to do this correctly (no government succeeds with tech it seems), I do think there's a discussion to be had about just how search algorithms intersect with human rights issues.
The results you're fed are based on a profile of you, and how those results differ for people of different race, creed, economic background, etc. absolutely can become a human rights issue.
An extreme example of this might be: An indigenous user performs a search for how to get healthcare and the "AI" / "Algorithm" (whatever term you want) buries resources on much later pages of search results that are otherwise shown for caucasians. Replace "how to get healthcare" with "apply for a scholarship" or other similar concepts.
In practice I suspect nothing that awful is happening today but we do see all kinds of systems like this being only as good as the training data.
I feel exactly the same way. There’s a huge amount of common sense that’s being missed, and I can’t understand why our government is trying to get there before anyone else.