Todd Defren, who was cool enough to come into a class that I’m enrolled in today, has a post up about transparency and information that caught my eye. I’m not too familiar with Klout, so I’ll take his word on the issues with their algorithm.
The segment about Politifact and the Truth Goggles, though, that interested me. Todd cites the poll about FNC that’s been making the rounds and also links to the demonstrably false Romney ad that’s been making the rounds. He then excerpts this piece from Nieman:
That’s why Dan Schultz, a graduate student at the MIT Media Lab (and newly named Knight-Mozilla fellow for 2012), is devoting his thesis to automatic bullshit detection. Schultz is building what he calls truth goggles — not actual magical eyewear, alas, but software that flags suspicious claims in news articles and helps readers determine their truthiness. It’s possible because of a novel arrangement: Schultz struck a deal with fact-checker PolitiFact for access to its private APIs.
If you had the truth goggles installed and came across Bachmann’s debate claim, the suspicious sentence might be highlighted. You would see right away that the congresswoman’s pants were on fire. And you could explore the data to discover that Bachmann, in fact, wears some of the more flammable pants in politics.
Cool idea, eh? Todd thinks so:
How awesome would it be to have a trusted agent accompany your online reading, ensuring that you’re not unduly influenced by B.S.? “Are you sure you want to tweet that article link? Looks like the author is trying to pull a fast one. Click here for details.”
So what’s the big takeaway?
In a world in which everything is online and everyone has an agenda, the survival of Truth & Relevance will all depend on algorithms. And that’s scary.
End of rant. Hope you had a nice Turkey Day!
I think it’s cool, too… kind of. There are a couple of problems. Politifact gets to decide whether the statement is true or false. But what if Politifact is wrong, like they were about the Republican plan to “fix” Medicare? Admittedly, the statement “end Medicare as we know it” would have been more appropriate than “end Medicare,” and probably would’ve garnered a different rating, but Politifact’s assessment does still strike me as, to use Josh Marshall’s word, “credulous.”
Or what if Politifact forgets to adjust for inflation? (!!!!!) The point of this isn’t to shit all over Politfact, which is a useful website that helps with sifting through our pitiful national discourse. But Politifact isn’t the be-all and end-all. As Bob Somerby notes:
Politifact isn’t always right in its judgments; the last time we checked, no one is. In our view, the site labors under a bit of a conceptual strait-jacket. Politifact scores all statements on a scale defined by “true” and “false.” (Its standard categories include these: Barely true, half true, mostly true, false.) But in the real world, the most problematic public statements are almost never technically false; instead, they are grossly misleading. Professionals can thoroughly mislead the public without ever making any false statements. Indeed, Maddow would try to prove this point before this battle was done.
Exactly exactly exactly. When someone claims that half the people in the country don’t pay income tax, it is technically true but extremely misleading. When someone claims that the rich are paying more in taxes than they were thirty years ago, again, technically true but grotesquely misleading. Statements don’t have to be false to be bullshit and good politicians know that.
Also, I can’t help but imagine when some people get wind of this technology and use it to create a “fact-checking” mechanism based on FNC, Heritage, Cato, El Rushbo or some other equally “credentialed” conservative source of “information.” And that’ll just push us all farther down the rabbit hole of pre-established narratives and information.
So, yeah, cool idea in theory. In practice, not so much.