4 ms·
> is generally about as on the level as their contemporaries This points to another pretty obvious problem with NewsGuard, though: flagging happens at the site
by Bartweiss 8y ago
> is generally about as on the level as their contemporaries
This points to another pretty obvious problem with NewsGuard, though: flagging happens at the site level instead of the article level. It's obvious why that's necessary, but the result is a complete failure to deal with any source where quality varies widely. Fox's written reporting is sometimes quite good, but Glenn Beck's old videos are still posted under the same domain. The result is that NewsGuard happily puts a big green checkmark above a video declaring that the US is the only country in the world with birthright citizenship.
In the same sense that labels like "no added sugar" can actively mislead by giving the appearance of health, I worry that NewsGuard is going to warn people away from Onion articles while encouraging them to accept outright lies without due diligence.
- netsharc 8y agoInteresting idea, Truth as a Service. You don't have to use your own critical thinking, just take what the machine says! /s
- kbenson 8y agoWell, right now a lot of people use pundits for this already, so it's not really that new to people. It's hard to gather and interpret all the evidence and make a reasoned opinion. Given all the things we form an opinion on daily, it's actually probably impossible. Much easier to defer to an expert for most things. Unfortunately, "expert" is a loaded term here, and people aren't always getting what they think as the people we've deferred our reasoning to have their own agendas, or try to extend their reach to items they are not experts on. Then they're just pundits that have an opinion about everything, because their job is no longer to be as correct as possible on what they can, but to have an opinion on everything.
- derefr 8y agoI mean, that's what Certificate Authorities are. ...Though, that being said, the original idea behind X.509 certificates was quite different: you [the domain sysop] would use your critical thinking to pick a root CA you trust, and configure your machines to only use that single root CA; and then that root CA would either directly sign [code, server, client] certs they trust, or would cross-sign other CAs' root certs to indicate that they trust them to delegate to them. Essentially, a domain (like a corporation or university) would have a business relationship with a particular CA, who would "do their critical thinking for them" (and would also be legally liable if they failed to do that critical thinking.) For X.509, we pretty much immediately moved away from that world, though, to a world where each application (e.g. OS HTTP API, third-party web browser, etc.) is acting as a root CA, you endorse that root CA by installing/using the app, and then the app devs make decisions on what to trust/delegate trust to. Which is kind of what's happening here: you use your critical thinking to install an app/OS, whose devs are then your "truth source." Sadly, though, this isn't a business relationship where they can be sued for failure to deliver quality content through that channel. (For contrast, where a customer is paying a business for "truth", see: investment newsletters.)
- Barrin92 8y agoJournalistic institutions themselves already employ vast fact-checking tools and services internally to verify articles and information, and you already indirectly rely on them. This is only sensible, critical thinking is helpful but is not the same as checking factual statements. And it's a good thing that fact checking exists as a service (now apparently also in independent fashion), because nobody has the time, knowledge or capacity to literally dig through every piece of journalism they read, on every topic.