notice that little 'f' (or 't') everywhere?

January 16, 2011 6:00 AM

How (crowd) curation is making a comeback in search and how Facebook is using it to "remake whole industries."
posted by kliuless (27 comments total)

I'm Laughing Your Ass Off but only if I am within your social network.
posted by srboisvert at 6:10 AM on January 16, 2011

someone has had at least a tiny little sip of the kool-aid
posted by HuronBob at 6:31 AM on January 16, 2011

Facebook’s grand plan for the future: After the public presentation I join Zuckerberg... for 40 minutes he talks animatedly.... “If you look five years out, every industry is going to be rethought in a social way,” he says. “You can remake whole industries. That’s the big thing.”

This is a guy who thought beacon would be a once-a-century revolution in advertizing. He seems to think that just because he made a popular website he knows how to remake the world.

That isn't to say I don't think social networking doesn't have a roll in "the enterprise" Companies could actually save a lot of money by using facebook or google docs (which has a social element) instead of self-hosted email systems.
posted by delmoi at 7:25 AM on January 16, 2011

Facebook: You need to know what your friends and your friends of friends already know that you do not.

Jake just served Karen's Spitfire Roasted Chicken in Cafe World!
posted by scrowdid at 7:47 AM on January 16, 2011 [2 favorites]

On the one hand I think the author of the first link is 100% correct. Google is gamed out the wazoo and depending on what you want, be prepared to wade through a lot of clones of the Wikipedia page on the subject reformatted as a blog and with A SHITLOAD OF ADVERTISING down both sides and across the top. Compare that to MetaFilter, where, If I had a dollar for every time something obscure but interesting was brought to my attention here, and then someone at work mentioned it a few days later, my army of one off joke name sock puppets would be legion!

The problem is, there are things I'm interested in that my "friends" aren't. Hell, there are subjects on which I am probably one of the thousand most knowledgeable people in the world, but that has more to do with the fact that roughly 6893757638 people (more by time you read this) or unaware of the subject or wouldn't be interested if they were aware. I could tell you about them but it would just bore you.

I think using crowdsourcing to hide the internet's bigger messes from my search results is a good idea but Facebook is already busy trying to be a dating site, a photo-sharing site, an e-mail provider and an gaming server. It's hard to get excited about the news that they've sent Igor out to dig up the grave of web-rings and return with its brain.
posted by Kid Charlemagne at 7:49 AM on January 16, 2011 [2 favorites]

Google SEO is mostly for commercial stuff, where money is involved, but for other things it's not a factor and Google continues to work very well. So perhaps we need curration for things that are commonly SEO'd, like "shoes" or "pancake", but not for things like "William the Bastard". SEO is a form of curration, usually by a paid party, so the scale of the problem is within a size that a social network could probably take on. The trick is to find the SEO'd searches that need to be un-SEO'd through curration.
posted by stbalbach at 7:57 AM on January 16, 2011 [1 favorite]

Google is gamed out the wazoo and depending on what you want, be prepared to wade through a lot of clones of the Wikipedia page on the subject reformatted as a blog and with A SHITLOAD OF ADVERTISING down both sides and across the top.

1) Adblock
2) I havn't noticed this too much, but probably that's because, well first of all the adblock, but second of all most of the searches I do I'm looking for just random information which is probably on the wikipedia page for whatever it is I'm looking for anyway. If I'm going to buy something I'm pretty much going to go with Newegg or Amazon, and I don't really do a lot of shopping online anyway.
posted by delmoi at 7:59 AM on January 16, 2011

MetaFilter: there are subjects on which I am probably one of the thousand most knowledgeable people in the world
posted by hippybear at 8:09 AM on January 16, 2011 [6 favorites]

I am bewildered by the complaints about google search results. I fail to find what I am looking for on the first page of results about one time in thirty. The only thing I can think of offhand that I cannot find is information about radio, and that may be a user error.

I would be curious to see what types of things people fail to find with the google search engine.
posted by bukvich at 8:33 AM on January 16, 2011 [1 favorite]

I think the problem comes not with "William the Bastard" but a few levels further down, like what do we think the construction details are (type of wood, how did they attach the fletching, etc.) of the arrow that put out Harold Godwinson's eye (and brain)? Nobody is optimizing for that search, but once you get down there, garbage that isn't optimized for anything outweighs relevant results by a couple orders of magnitude.

It's fair to admit that I have a reputation for odd little niche interests in my odd little niche subculture, so maybe I just run into this more than the average person. (Today's agenda - What should the hinge look like and how should they be attached on a small panel and frame construction coffer chest ca. the 15th century?)

I want a T-shirt with a ying/yang design on them - in one half I want Hippybear's comment and in the other half I want, "Metafilter: I could tell you about it but it would just bore you.
posted by Kid Charlemagne at 8:55 AM on January 16, 2011 [4 favorites]

The problem with the Google search algorithm is not that you can't find basic information about your topic. It's that most of what you see on a lot of subjects is very superficial keyword farmed articles done for SEO linkage and ads (which I block). I'm noticing this more and more when I search on pretty much any subject that leads to something that might be sellable, although first-order results for nonsaleable things like obscure bits of medieval history are still OK. (On preview: Kid Charlemagne, if I weren't married I'd spouse you for your queries.)

My big takeaway from the first link was that they were right that Yahoo's decision to kill delicious was moronically stupid. It's too late to undo because all your "thought leaders" and tipping point people and whatever have moved elsewhere, but they had something there and managed to fuck it up spectacularly.

I've joined Quora but it's insufficiently socially sticky for my crowd so far, and for most subjects I'd still rather use AskMe to get curated answers. It's random but it's high-quality random.
posted by immlass at 8:57 AM on January 16, 2011

The author of the first post was just **really** excited to compare Google's search algorithm with a trading algorithm despite the inappropriateness of the metaphor because search is cooler than finance.
posted by sbenthall at 8:57 AM on January 16, 2011

Paul Kedrosky:

Any algorithm can be gamed; it's only a matter of time.

Any *process* whether algorithmic or human can be gamed. In fact to the extent that humans use a process at all, even a fuzzy, heuristic, subconscious one, there is probably some algorithm being used, albeit one vastly more subtle and complex than used by the likes of Google for ranking search results.

Whatever the field, whether it's search rankings, financial regulation, sports rules, there is mostly always going to be an arms race between would-be gamers and algorithm designers / regulators.

What you might hope for is that a sufficiently good algorithm makes gaming difficult enough to not be worth the effort, and that it will be mostly easier to play the game as intended.

That said Google search tends to have two issues. On the one hand for "major" topics, the search results gravitate to the "big" sites, like Wikipedia, regardless of whether they have the best material. On the other hand, for "niche" topics, it's quite possible for machine generated gibberish to outrank properly researched and coherently written articles.

So something based on what people I trust have recommended has at least a chance of being better. One reason why AskMe does a better job of answering questions like "What are the best books on X?" than Google does.
posted by philipy at 9:38 AM on January 16, 2011

When talking about quality of Google results, it helps to look at the different types of searches that people do (which are up for debate, but the list at "Four Modes of Seeking Information" works well enough): locating an item that you know exists somewhere, exploratory search with an idea of what you want, searching in a complex situation when you don't know what you need to know, and re-finding something you've seen. Another way to categorize searches is navigational (aka googling "facebook" or "brushes app"), informational (finding definitions and facts), and transactional (getting directions on a map, buying something). I find that Google gives me rapid great results for navigational and informational/definition/maps searches, but...

...well, for example, when looking for those links explaining search types, which are some mix of "re-finding something you've seen" and "informational/exploratory", I tried Google, gave up because most of the initial results were bloggy/SEO-advice stuff that weren't useful, and went to search on Delicious — because Delicious unconsciously specializes in being a library of geeky thoughtful links on topics interesting to people who work on the web, just like those articles. And after some refining of search terms, I found pretty much what I was looking for. I just take it for granted that Google won't always give me what I want, and it doesn't bother me, because I have my Delicious network (and Delicious in general) as a backup search engine. My Delicious network, in particular, is invaluable for deep "exploratory" searching — when I want to find out what a hundred people who I think are smart have bookmarked on topics I want to learn more about, like "polyhedra" or "The Dark Knight" or "new york restaurants". Or, as a former Yahoo person wrote a couple years ago, "If you ever search something you care about (I think of it as stuff about which I have some passion), you will most likely find better results on Delicious search than you will on Google or Y!...Google and Yahoo! treat you like a tourist, helping to point you to a place. Delicious, and its passionate users, direct you to places that help you do something meaningful."

In my little fantasy world, Google would use Delicious as the nucleus of a sensible and straightforward form of social bookmarking at scale. It'd be more usable and have precise and thoughtful private/public messaging so that a lot of people would save public bookmarks by choice, because it's fun and displays your knowledge and expertise (similar to how huge numbers of people choose to post to Twitter and Tumblr publicly). And you would get recommendations for good people to follow in your fields of interest, using collaborative filtering and clever algorithms to identify and expose good-quality relevant bookmarkers. You wouldn't be pushed to only follow friends, and asymmetric following would be socially acceptable. And more importantly, you'd be able to not only search the 500,000 bookmarks saved by your network, but also the 5 million bookmarks saved by all the networks of your network members — a specialized personally-valuable layer on top of the wild chaotic web. It's a fun dream to toy with. It's also fairly obvious and has been talked about before, so I'm sure people are working on some other way to do it.
posted by dreamyshade at 9:46 AM on January 16, 2011 [5 favorites]

This passage from the second link chilled me to the bone:

If Zuckerberg is to be believed, we are rapidly moving from a world where the web doesn’t know who you are, to a world where the web knows exactly who you are..."If you imagine a television designed around social, you turn it on and it says, ‘Thirteen of your friends like Entourage. Press play. Your dad recorded 60 Minutes. Press play.’” In other words, the world will be experienced through the filter of one’s Facebook friends.

I do not want to live in a world where everything I do is mediated and predicted by what my Facebook friends like.
posted by oozy rat in a sanitary zoo at 10:07 AM on January 16, 2011 [3 favorites]

I'll take these grand schemes more seriously when Facebook deigns to give its users a "Dislike" option.
posted by benzenedream at 10:33 AM on January 16, 2011

I do not want to live in a world where everything I do is mediated and predicted by what my Facebook friends like.

I don’t either, but this is not that, and the scenario described is not that "in other words," any more than plopping a friend’s FriendFeed into my RSS reader is that. My friends are my friends in part because they're interested in interesting things, and I'm curious about what they're curious about.

As long as I can say, “that’s great, but I don’t really care to watch 60 Minutes right now” and navigate to something I do care to watch, I'm o.k. with this sort of thing. If last night’s 60 Minutes is something dad likes to talk about a lot when I call him, I'm happy for the reminder. It’s the most benign part of the social networking experience.

I'm a lot more bothered by everything I do being mediated and predicted by what I like. That seems like the more dangerous outcome of constant online surveillance, and it’s why so many online communities become solliptical quagmires: Everybody’s there to get more of what they like, and avoid whatever they don’t like. It's where I sort of come down with Cass Sunstein.
posted by mph at 10:48 AM on January 16, 2011

> like what do we think the construction details are (type of wood, how did they attach the fletching, etc.) of the arrow that put out Harold Godwinson's eye (and brain)?

Kid Charlemagne this piques my interest as that is the type of use I have never bothered attempting with google. If I have a question at that level I go to a university research library with stacks of journals that go back way into the 20th century.

Now I wonder: did you ever in the past find google useful for this type thing and it is gone downhill since SEO guys tried to get their pages ranked? Or is it possible google was never useful for this and you are trying the wrong tool?
posted by bukvich at 10:54 AM on January 16, 2011

Can we stop using the word curation outisde of museum/collect/etc. contexts? Put down the thesaurus and go back to editing your links page.
posted by glip at 11:02 AM on January 16, 2011

Can we stop using the word curation outisde of museum/collect/etc. contexts?

I think we're stuck with it. Everybody decided "editor" was synonymous with "tyrannical old media gatekeeper."

On the bright side, I have a lot more confidence in tea shops that offer a curated collection of fine teas. Back when they were just "picked" or "selected," I had no idea what I was buying. Now I know they're probably pretty good. Same with yarn.
posted by mph at 11:17 AM on January 16, 2011 [1 favorite]

Facebook IS sometimes the better answer -- I'll post, "Hey, Locals, where's a good drycleaner? Mine closed," or "Does anyone have a body shop they trust? The estimate I got seemed high and I want a second opinion."

Yelp is of limited utility; Google pops me up every drycleaner in the area but not much else (and not always helpful for industries where web pages aren't common). My FB friends, otoh, can give me personal recommendations and even have discussions among themselves about which places they like and dislike.

Honestly, other than looking at pictures of other people's babies and pets (squeeeee!), this is the single most useful thing Facebook does, easily and fairly non-intrusively leveraging your LOCAL network for local information. I can hit up 150 people at one time instead of hoping that if I ask them one by one as I happen to see them around town, I hit on the one guy who has a super-strong opinion on drycleaning.

(Just the other day I was complaining about the shift in the garbage man's schedule and it turns out that the wife of the garbage company's exec VP is in my FB feed and she totally posted me this extremely useful information about why and when they shift schedule!)

But yeah, not so useful for general information. I have a lot of doctors and journalists in my feed, who tend to highlight interesting links in those areas, but I'll never learn anything about anthropology from my FB friends!
posted by Eyebrows McGee at 12:19 PM on January 16, 2011 [1 favorite]

this is scary enough:

If Zuckerberg is to be believed, we are rapidly moving from a world where the web doesn’t know who you are, to a world where the web knows exactly who you are.

I don't want anybody on the web to know I'm a dog. Ooops.
posted by oneswellfoop at 4:39 PM on January 16, 2011

It seems obvious that curation can not only be gamed, it is already being gamed. Those Oatmeal comics are a good example: low quality, but pushes every meme-button in sight to make sure it gets on Reddit's front page. And what about the proliferation of social media experts who teach companies how to increase twitter followers and get them to retweet content, how to make your brand go viral, etc. It seems like the thing that's supposed to rescue us from Google has already been compromised.

A second problem is the too-easy distinction between social and algorithms. Google's PageRank algorithm depends on users "voting" on relevance by linking from one page to another - isn't this already social, a kind of curation? And on the other side, sites like Reddit and Digg also depend on algorithms and the article holds up Twitter as an example of curation but they have their own trending algorithms.

All these problems illustrate how these systems depend on individuals uncritically accepting the role that has been inscribed for them: active, engaged, creative, but unpaid digital laborers whose social production is harvested by Google/Facebook/Twitter/etc. The profits are privatized and go to stockholders, but as soon as individuals attempt to seize the means of production and opportunistically also try to make a buck, it brings down the whole system. So an ideological program must be put into place to valorize free work by emphasizing the supposed nobility of intrinsic motivations like purpose, passion and autonomy as opposed to crude extrinsic greed. There is a kernel of truth here, that the system can't function if it were to compensate people for their labor, we have to be intrinsically motivated. A related part of the ideology of social media is that it simply offers people a venue to do what comes naturally to them: form connections, share pictures of cats, receive positive feedback and status from their community, self-actualization, etc. These references to the humanistic psychology and the utopian human potential movement of the 60s, and how they are used to inculcate forms of subjectivity that are perfectly compatible with and actually essential to the huge growth and profitability of the tech industry ought to give us pause.
posted by AlsoMike at 5:04 PM on January 16, 2011 [2 favorites]

The Oatmeal is gaming the system only insofar as it is something lots of people respond to. People like his stuff so it gets lots of hits—that's not a bug, that's a feature. If he was for instance only reposting comics from elsewheres but getting the same attention then it'd be analogous to the SEO spam Google results. The fact that he is doing it for the money (and not, say, from an altruistic desire to share art) is basically irrelevant.
posted by wemayfreeze at 1:28 AM on January 17, 2011

My friends are my friends in part because they're interested in interesting things, and I'm curious about what they're curious about.

I've always thought this was a crazy way to look at friendship. It is the fatal flaw in newsgroup-style topic-based discussion. My friends are my friends because I met them at some point (theoretically this could happen in cyberspace, but in my experience entirely in meatspace) and decided I liked them. We may have things in common, but those are incidental. Mainly, I live near them, which is why geographically organized social networks excite me, and I'll probably jump into them if they ever pick up a sizeable following.

Another datum: Almost every phpBB-style web forum has an "offtopic" subforum. I assume this is because there's nothing natural in having a forum, digital or otherwise, that is strictly about discussing one particular topic. This is the reason that, when I had a facebook page, I was friends with 15 other people who had exactly the same name as me, but I never messaged them or wrote on their wall. And they never wrote on mine. This is the reason that, when I am hanging out with my Model U.N. friends, we hardly talk at all about international diplomacy or parliamentary procedure.

I don't like the monetization of friendship at all, but I like even less the notion that, fifty years out, we might be friends chiefly with people who have the same idea of what should come up in a Google search of "cute kittens" or "cool gadgets," instead of people that, you know, we think are cool and fun to hang out with.
posted by LiteOpera at 5:18 AM on January 17, 2011 [2 favorites]

I've been thinking about this a bit, particularly buckvich's question, trying to put my finger on what, exactly, I think has changed. I think I was being a little unfair to Google, but more on that in a bit. I couldn't put my finger on what I felt was wrong until I started thinking back to how obscure information was circulated in the pre-internet era.

Doing research in a library is not easy and often required a bit of serendipity. I could give one or two personal examples, but the one that I have the most thorough story for is about a type of armour called a coat of plates. It's basically a leather jacket with steel plates riveted inside. This type of armour was largely forgotten until the city of Visby, Sweden was putting in a new sewer line in 1905 and found a mass grave left from the Battle of Visby. A book was ultimately written about the ensuing archeological digs and then promptly forgotten again. (Hey, this isn't shining - it's all about knights in shining armour!) Then, in the 1970's, a guy named Brian Flax tracked down these relatively obscure books and decided that it was the perfect SCA body armour since you could make it in reasonable amount of time without needing exotic tools or a ten year apprenticeship to a master armourer. The punchline is that Mr. Flax was essentially publishing an armouring zine at the time. Word spread and since then the humble coat of plates has started showing up in books on the subject like it was an important step in the evolution of armor between mail and fluted gothic plate and not just something they found in the mud in Sweden.

Twenty years or so later we had the internet! Now you didn't need to publish a zine on your obscure interest. Everybody and their dog could put up a web page showing off their work or research or bibliography or such on any number of free providers or at genericISP.com/~user. Sure, it was rich with animated gifs, bad grammar and colors that didn't play well together. But like things tended to stick to one another and if you were interested in a certain topic and could find one page on it, chances are you could link surf (remember the phrase "link surf"?) and find an amazing amounts of information for which curated really does seem like the right word.

At some point Google became everyone's defacto search engine. It seems like it happened all at once, but in retrospect, it was that much better than Alta Vista or Infoseek in the beginning. But like a small furry thing with a fast metabolism it shrugged off the impact of the dot com bubble while its big, slow rivals were dying left and right. (Hint: Have a business model that includes turning a profit and the good sense to reinvest some of that profit into not sucking.)

Add in high speed access and Web 2.0 (remember the phrase "Web 2.0"?) and pretty soon there were sites lining up to accept whatever you wanted to put on line - videos, text, a hundred megabytes of images. You didn't need a domain, even one of those cheesy /~user domains, much less dabble in the arcane voodoo of FTP (remember FTP?)

And the trap is sprung! The web is civilized. Common interest communities are centralized around one or two sites that chiefly contain links to things like Flickr, Photobucket and YouTube. People stopped putting a lot of effort into personal web pages, they just link to Flickr from a forum somewhere with a note saying something well thought out and meaningful like, "Hey, here's that project I was telling you all about!" Pages of useful links on obscure subjects seems like pulling off of the interstate to carve a notch in a tree so you can find your way back home.

So what we have is a giant bookshelf where great works are carefully preserved and indexed by emotionless machines who put the same degree of care into preserving and indexing the conversation I had the other day while waiting for an elevator.

Not that I think Facebook is going to fix that.
posted by Kid Charlemagne at 1:03 PM on January 17, 2011 [4 favorites]

An enthusiastic argument for curation (From the Edge FPP)
posted by wemayfreeze at 5:08 PM on January 17, 2011