Crowdsource THIS, Buddy!

     |    Wednesday June 20th, 2012

I’ve got a friend in Bonn who likes to mess with me. The other day I got a message from him: “I just got you one of these,” followed by an Amazon URL. I clicked it and saw something called a “mankini,” all done out in the colors of the German flag. The first, last, and only time I’d previously seen one of these, essentially a jockstrap with shoulder straps, was on Sascha Baron Cohen, who wore one in, I think, Borat. “Ha ha,” I wrote the joker back. “Very funny.”


And then I had a horrified thought. I once bought a camera from Amazon. I once bought a printer from Amazon. Amazon tries to sell you more stuff when you buy from them, and they personalize it. About three days after I’d bought my camera, I got an e-mail suggesting I’d like to buy…a camera. Same with the printer. It also works when you only browse something: a friend, looking to buy an espresso machine, found one that cost about $5000 on Amazon and said “Of course, I’m not looking for one of these,” and I clicked it, and sure enough a couple of days later, I got an e-mail suggesting other very expensive espresso machines. So far, thank heavens, I’ve been spared the “Customers who looked at this also…” e-mail for the mankini because, well, maybe nobody’s actually buying those things except idiotic comedians. I hope the algorithm doesn’t know me well enough to know how stupid I’d look in one.

You didn't think I was going to picture a mankini, did you?


But this thought was still in my head when I read an article from the Wall St. Journal last week about crowdsourcing, which is the human equivalent of the Amazon suggestion machine. The article mentions two cases where asking a large, random crowd to make a final decision went wrong. The first was when the struggling Pittsburgh Symphony Orchestra asked people to pick soloists to perform a concerto, and, in the end, none of the young performers was actually good enough to do it. The PSO had to backtrack and do what it always had done: pick one themselves. In the second instance, the Walters Art Museum in Baltimore, which had its public pick everything from the title of the exhibition to the objects in it, with, if the Journal is to be believed, fairly weird results.


I’m not at all sure where the idea that this was such a great idea came from in the first place, although in one instance I can think of, it predates the Internet, or the widespread use of it, in any case. Way back in 1979, Tim and Nina Zagat started the Zagat Survey, which was a guide to New York restaurants. The Zagats asked all their friends to help out, and soon had a wealth of data on a lot of restaurants. Tim and Nina had exactly as many votes as anyone else in the survey: one apiece. Then they asked the people who’d bought their first book to help out, and increased the number of informants hugely. There was a format, of course, with respondents asked to rate various qualities of the restaurant (food, service, atmosphere, etc.) and write a short paragraph, which the editors then excerpted for an annoying summary in the book which was laden with quotes, something like this “quaint” Midtown steakhouse serves “generous” to “humongous” portions of aged beef in a “traditional” “old-time” New York steakhouse atmosphere. As the company got more famous, stickers saying “Zagat Listed” appeared in restaurant windows, which, to me, meant the rubes had already been there and you should go elsewhere. Today, trumpets itself as “the world’s original provider of user-generated content” and now offers tons of different locations internationally, and was acquired by Google last year.


Tim & Nina Zagat


So here’s my problem: if it was 1979 and I knew the Zagats, or at least had some idea of who they were and what kind of friends they were likely to have, I’d have had a place to start in evaluating their recommendations. Once they expanded the pool, I remember asking myself why I should buy their book. Who were these people, anyway? How can we trust their decisions? And now that there are sites like Yelp (sorry that that comes up Austin; I must’ve dropped a cookie in there) don’t even put the Zagat’s restrictions on content, why should anyone even bother to mess with them? If my girlfriend owns a restaurant and we have a bitter break-up, it’s easy as pie to get my friends to go to Yelp and leave horrid reviews with details of how Manny in the kitchen spits into the food and so on. And what’s awful is that people believe this — more than they do an individual critic. Not that, as the cookie on that link shows, I haven’t used it myself, but that’s often because the only link for a restaurant these days is one of those “user-generated content” sites. And there’s a little sense to it: if everyone tells you not to order the spinach, it’s probably wise not to order the spinach, because the people who contribute to those sites are the kind who’ll order the spinach just because everyone says it’s terrible just so they, too, can add to the anti-spinach chorus. (And if I know someone in town, I might just ask them: “Is it true the spinach at Joe’s is awful?” “Yeah, I don’t know why they keep that on the menu.” Okay. Now I really believe it).


Saag paneer, good spinach


The thing is, this is where democracy can become tyrannical. If my bitterness shuts down my ex-girlfriend’s restaurant, who benefits from that? When, for that matter, does a “crowd” become a “mob,” and how do we tell crowdsourced from mobsourced? And in a society where we all have our areas of expertise, isn’t it actually good to have an elite? If you’re a professional photographer and I want to buy a camera, I’m going to believe you about stuff I don’t even understand when it comes to buying one, even if the crowd doesn’t agree with you, and I’ll be happy to advise you on things that you might not know as much about as you’d like because that part of your brain is occupied with photography. Elites are good, because everyone’s an elite about something (or so we hope). That’s why plumbers get big bucks.


There’s an even more terrifying dimension to this that I don’t really want to get into, that I found in a post over at Mick Farren’s blog (don’t worry about the warning: someone’s got it in for Farren). It mentions, among other things, that 50% of Americans believe that the Bible is correct in all of its teachings, yet half of the graduating high school seniors believe that Sodom and Gomorrah were man and wife and ten percent of adults think Joan of Arc was Noah’s wife (yeah, it took me a while to get that one, too). So as America tilts more and more towards theocracy, it turns out that huge numbers of people who want it that way don’t have a clue what’s in the Bible! That’s the kind of crowdsourcing that really scares me.