Fake news, and dangerous ways to detect them
“Have you ever clicked a link on Facebook, been taken to a site you’ve never heard of and wondered
whether what you were reading was real, fake news or satire? Of course, you have…”
“[And] Even to a trained eye, some sites require a double-take or a close look at the finer details to tell if you’re really on the website you expected to be on. Fortunately, Eyeo, the company behind AdBlock Plus, aims to combat this internet gray area and clear things up for everyone with its new browser extension TrustedNews. Here’s how it works” (continues at Cnet.com, pointing out that “TrustedNews requires no effort to use, so there’s no reason not to use it.”
Labeling Fake News? Hmm…
Almost ten years ago, I myself wrote about how wonderful it would be if ” we could configure computers to always show, in real time and almost automatically, third-party proofs or refutations of everything politicians say on TV!”
This TrustedNews browser extension seems quite close to what I was asking for, back in 2009. But the ways TrustedNews is proposed, or commented, make me worry about how it could gloriously backfire.
TrustedNews requires no effort to use, so there’s no reason not to use it? The same could be said about Facebook, but there are plenty of reasons to NOT use Facebook. Or cigarettes, for that matter.
I also saw people making really worrying comments or suggestions. One is this:
[just labeling a site as fake news or biased would not be enough! Adding an] “automated link to an opposit view could be useful”
Smart, uh? Until you realized that adding such a useful function means that any report about vaccines, fossils or satellites orbiting Earth should include automatic links to novax, creationism or flat-earthers. Giving undeserved exposure to crap is exactly one of the factors that created the current climate (pun intended). Then I read this other gem:
“Why do i not trust proprietary software deciding what constitutes fake news?”
with a browser extension like TrustedNews, its software license counts very very little. Surely not more than the data fed to it. Garbage in, Garbage out (or even hidden censorship, in extreme cases). Face recognition software trained only with faces of white people failed to detect people of color as humans. If 100% of Facebook code were Free/Open Source Software, Facebook would still be as harmful as it does today. In general, as I have been saying for +10 years now, you can run the perfect police state using only “free as in freedom” software. So go ahead and install TrustedNews if you want, but think twice, every time it gives you an opinion, before trusting it.
I wonder how TrustedNews would classify that Cnet article. What do you think? Write me at firstname.lastname@example.org to let me know, or to arrange speaking about these topics!