Machine translators have made it easier than ever to create error-plagued Wikipedia articles in obscure languages. What happens when AI models get trained on junk pages?
If you build the infrastructure for a certain thing to happen, you’re responsible for the thing. For the same reason we hold facebook accountable for the rise of the far-right, we should hold WikiPedia accountable for this stuff. Infrastructure is never neutral.
That is a completely unfair comparison. For starters, Facebook is a for-profit advertising company and Wikipedia is a community-driven encyclopedia and should be judged by different standards
Second, both admins and users can edit Wikipedia when there’s a problem. Everyone is “responsible” for fixing it - or at the very least equally at fault
Next, the content in question. Facebook was (rightfully) given hell for hosting gore, CSAM, adult porn, etc. Things that are immoral, illegal, or outright dangerous. The offending content on Wikipedia is bad translations.
Lastly, the bigger issue is always enforcement of said content. Facebook was made aware of the problem users/pages/uploads and slacked off on doing anything. These Wikipedia pages have very low traffic and weren’t getting reported. And even with reports, Wikipedia then has to consult with people who speak the rare language.
They’re similar problems of vastly different scales
Not exactly the same. I don’t blame facebook for the rise, its just a place to post and share… I blame the algorithm that facebook created and keeps updating to enhance and expand those bubbles while pushing users to outrage and divide them into bubbles that empower and embrace conspiracies, right/alt-right, and other extreme viewpoints. Same thing with X/Twitter.
Wikipedia doesn’t have any such algorithm. They don’t have a team dedicated to pushing people to those extremes (or anything at all).
If you build the infrastructure for a certain thing to happen, you’re responsible for the thing. For the same reason we hold facebook accountable for the rise of the far-right, we should hold WikiPedia accountable for this stuff. Infrastructure is never neutral.
That is a completely unfair comparison. For starters, Facebook is a for-profit advertising company and Wikipedia is a community-driven encyclopedia and should be judged by different standards
Second, both admins and users can edit Wikipedia when there’s a problem. Everyone is “responsible” for fixing it - or at the very least equally at fault
Next, the content in question. Facebook was (rightfully) given hell for hosting gore, CSAM, adult porn, etc. Things that are immoral, illegal, or outright dangerous. The offending content on Wikipedia is bad translations.
Lastly, the bigger issue is always enforcement of said content. Facebook was made aware of the problem users/pages/uploads and slacked off on doing anything. These Wikipedia pages have very low traffic and weren’t getting reported. And even with reports, Wikipedia then has to consult with people who speak the rare language.
They’re similar problems of vastly different scales
Not exactly the same. I don’t blame facebook for the rise, its just a place to post and share… I blame the algorithm that facebook created and keeps updating to enhance and expand those bubbles while pushing users to outrage and divide them into bubbles that empower and embrace conspiracies, right/alt-right, and other extreme viewpoints. Same thing with X/Twitter.
Wikipedia doesn’t have any such algorithm. They don’t have a team dedicated to pushing people to those extremes (or anything at all).