The Myth Of Crowdsourcing… Or Misunderstanding Crowdsourcing?
from the it's-the-latter dept
A bunch of folks have sent in this silly opinion piece at Forbes, claiming that crowdsourcing is a myth. The reasoning? Because there are individuals in the crowd. Except… um… did anyone say anything different? Of course there are individuals, and the point of crowdsourcing isn’t that everyone in the crowd is equal, but that they each get to contribute their own special talents, and something better comes out of it. Every example that the guy dismisses as not crowdsourcing — Wikipedia, the Netflix prize, open source developing, etc. — actually does involve crowdsourcing. The problem is that this guy defines crowdsourcing in his own way — that if any individual contributes a greater amount, there’s no more crowd. Say what? The fact that a few determined individuals help craft a specific Wikipedia page, doesn’t take away from the fact that it’s the overall crowd that made Wikipedia so useful. It’s many of those determined individuals together who made the entire Wikipedia so useful. He then goes on to mock the Netflix prize… even though it disproves his entire thesis:
The Netflix contest is a prime example of individual virtuosity at work. One team was clearly in the lead and then a consortium of teams that had worse performance joined together and combined their innovations to create an algorithm that won the contest. For most of the contest, individuals toiled to figure out a solution. At the end, a consortium was formed. None of the invention happened through a crowd.
The problem — yet again — is that this guy (a consultant) seems to again be confused about the difference between invention and innovation. Yes, some individuals came up with different ideas. But what made the Netflix prize so interesting was that they weren’t able to really achieve the necessary breakthrough until they collaborated. That’s what pushed them over the edge. That’s what added that real value to take it to the necessary next step. A crowd is always made up of individuals. That different individuals contribute different parts doesn’t change the value of crowdsourcing at all.
Filed Under: crowdsourcing, individuals, innovation
Comments on “The Myth Of Crowdsourcing… Or Misunderstanding Crowdsourcing?”
Yeah, this guy missed the boat
“individuals toiled to figure out a solution. At the end, a consortium was formed.”
Would that not be delegating a task to a large diffuse group? Is that not the definition of crowdsourcing?
And yes, while one page on Wikipedia may not be crowd sourced, Wikipedia sure is. But that’s like saying that one word isn’t crowd sourced but the book is, it’s splitting hairs.
Re: Yeah, this guy missed the boat
Interesting that Forbes saw the fact that “individuals” made up a crowd as a problem. In fact, the math behind the “wisdom of crowds” DEMANDS that each member of the crowd act as an individual, then those individual recommendations are consolidated.
If the members of the crowd acted in some way that were co-related, then the “wisdom” falls out of the equation.
The key to crowd wisdom is that each person may be wrong, but if they are guessing something independently, then the wrongs will cancel out, and the “correctness” will start to emerge. As soon as one person’s guess is dependent on a prior person’s guess, bias is inserted, and the results are compromised.
Taken to the subject of elections, we say “we get the leaders we deserve”. Actually not. Our elections have extreme co-dependence. We have two years of endless news cycles, polls, and the “vote for this guy because he’s a winner”, and “a vote for this gal is a wasted vote because she can’t win”. Each of these serve to bias the independence of the individual vote. Thus, the crowd is NOT wise. In fact, the way we elect, and the electoral college, seems to be specifically designed to avoid the “wisdom of crowds” and instead allow bias to be introduced and amplified.
fun stuff.
Not only is crowdsourcing a myth, crowds are a myth. They’re just a lot of individuals grouped together. Forests are myths, too. They’re just trees.
Re: Re:
The first definition of a crowd is “a large number of persons gathered closely together”. The dictionary also points out that the people in a crowd may be joined by some common reason or response to a common stimuli. Therefore, a crowd cannot be a myth, as it is not an entity in and of itself; it is a description of the commonality of those who make it up.
The collaboration within a crowd is a completely different matter than one person’s own ingenuity; it is the resultant synergy of such a collaboration that makes crowdsourcing such an effective and efficient means of solving a problem or attaining a goal shared by those making up the crowd.
Maybe this is just my slavery to TV, but another interesting misunderstanding of crowdsourcing, this time by turning it into an indefensible strawman then proceeding to beat the ever-loving shit out of it, was the latest episode of House. Basically, uber-nerd game developer gets brought in with a mysterious illness, refuses treatment from his doctors, just wants them to do the treatments suggested by an online community, obviously to disastrous results.
Obviously not worthy of a Techdirt writeup, but I thought it was interesting the way they completely ignored the nature of crowdsourcing, and how anyone in their right mind would never refuse medical treatment in favor of crowdsourced opinions and at best would use it as a supplement.
Re: Re:
Wellll… there is a difference between croudsourcing and idiot sourcing….
Re: Re:
I spoke to a doctor about something smiler to that a little while ago. It’s not that the internet will give you bad information (sometime it will but not always) but that it gives you all the correct information. It takes someone trained in medicine to determent what part of all that correct info fits with what’s wrong with you.
I didn’t see that episode of House (don’t watch that show at all) but I would lay odds on the writers just making fun of the internet. Possibly based off a real life event, possibly based off of their fear that the Internet will replace them (I’d bet the former).
Re: Re:
It would be more relevant if a large team (perhaps a CROWD?!) of doctors that all had the information a personal doctor would have on a patient made their best guess as to how best treat the patient and they would vote upon the safest and best treatment to start with or list of ways to diagnose the issue.
Re: Re: Re:
It would be more relevant if a large team (perhaps a CROWD?!) of doctors that all had the information a personal doctor would have on a patient
I think that’s broadly the idea behind a large clinical trial, isn’t it ? As opposed to the previous method of a Doctor taking clinical decisions based on his own experience.
Any one patient is an anomaly, but taken as a huge mass they allow you to infer an overall benefit.
Of course, the problem is that this still won’t say if the drug would cure YOU specifically.
Re: Re:
One of the primary problems with the patients crowdsourcing idea was that there was no possible way for him to post every pertinent piece of information that the team already had. There’s also the fact that while a crowdsourced solution is usually correct, when it’s wrong, it’s horribly wrong (as was the case in the episode of House).
Personally, I would trust my attending over a crowdsourced answer any day. I may suggest the crowdsourced answer to my doctor, but if he/she knocks it down, then that’s that, end of discussion.
Re: Re:
I DO accept medical decisions from a crowd. Kaiser Permanente HMO uses a hybrid approach to health. (I should stress this is my opinion, I cannot speak for Kaiser).
They look at your data, and then compare it to the reams of data they’ve seen in the past. Expert systems designed by multiple MDs expertise then recommend guidelines for courses of treatment. The system’s suggestions are filtered either by your primary MD, or by a onsite panel of MDs for more serious decisions.
As such, the treatment is decided by much more than one idividual doctor. I’m pretty sure Freakonomics talks about this, and about how the results are better than what could be provided by a single MD.
Of course, it sounds like the House case didn’t use crowdsourcing among MDs, but rather online geeks like us. Good luck with that, Mr. Laurie!
Crowd Sourcing
Crowd Sourcing in concept is probably a good idea, it has a lot of positive aspects to it.
The problem comes into play with it when the final product is assembled and deployed. Take for instance a lot of Open Source stuff (Not all but a lot). Open Source products have a large following because generally the product is good. The problem I have, and the biggest reason most of my Servers are still Windows is in the deployment and support of the product.
Finding information on a given product is difficult and in a lot of cases going directly to the open source group is out of the question as well. I recently had a problem with a product Ias evaluating and really wanted to use in production.
I went to the developers for support and was snubbed, basically told to go somewhere else because I didn’t have the foundation knowledge of Linux.
Then I went to the end users forum where all I got was a bunch of links to sites that had little information about my given problem, but had more links to someplace else and so on and on and on.
Pretty sad really because the lack of information for a beginner cost the open source system what would have eventually evolved into a full time position working for me on this system. I ended up deploying a windows product that was a lesser product just because I could find people to help with it.
So my tyrant here is that crowd sourcing actually created a better widget, but the very crowd that created the widget is holding back it overall potential because they all felt they were too good to help someone use their product.
Re: Crowd Sourcing
In response to Michial Thompson, but generically applicable to the topic at hand…
So, you succeeded in crowd sourcing the software but failed in crowd sourcing the support? That how I read your rant.
I too got my start with Linux solving a problem, and I too went to the Internet for help. Today I make my living building and developing Linux solutions.
Just because you crowdsource does not mean you will succeed. Like anything else, a effective group needs a good leader. You got lucky because the first crowd (software development group) had a good leader, or the software you think is good would not be present. The second crowd (support) did not have an effective leader.
You can either find a leader or be the leader of the crowd. You chose to be the leader of the second crowd and did not work effectively as that leader. You could have failed to do the proper work of a leader such as finding the proper crowd or properly defining the problem or the desired goal. Maybe you failed to provide the correct crowd the proper data to define the problem.
Any group or crowd is only as effective as its leader(s) allow(s) or facilitate(s) it to be.
As an aside (rant of my own) I find it mildly entertaining how that if I provide a free product, you think I should provide my service free and on your terms. Maybe I do not want you as a customer of my free product.
I have refused to provide (monetary cost) services to some potential customers because my cost (headache, time) will out weigh the costs they pay me, even at my commercial rate of $120/hr. Other customers I have helped at no monetary cost, what you think of as free. The cost to me (time), was worth the benefit. I profited in knowing I helped a good person or cause.
When I write software I put in a certain amount of time (fixed cost). The more times that product is sold or provided to another person, the more that fixed cost per person approaches zero. There is no significant marginal costs in giving you a copy of what I already wrote.
Support has little fixed cost. Support is almost entirely marginal. The marginal costs may drop as I improve FAQs, add knowledge bases, educate more people, improve the product, but at the end of the day support costs time over and over and over again. You want my support, the next guy who uses the software wants my support and so on and so forth.
While in seeking (positive) exposure for my software I am willing to provide some regular support, I have expectations (costs to you, even if not in terms of money). You expect the software to work, and if it does not I expect you to be useful in helping us (you and I) solve the problem. If you complain but do not provide me what I need to solve your issue, than I do most of the work and you get most of the benefit.
I am more willing to give you (free of monetary cost) support if you give me value (data to help make the product better for everyone) in return. However there are customers not worth having and not worth trying to make happy. I learned long ago that it is Ok to fire a customer (whether charging money or not). I do not need every customer that exists, only enough to provide adequate profit and benefits I desire (money, recognition, reputation, good feelings, thanks, etc…).
My time has great value even if I do not charge you money for such.
/rant
All of that said, you can also crowdsource the Internet and get help with my or any other product. But then the pressure is on you to be an effective leader, or find one, to make an effective group/crowd to provide your support.
Secondary rant, for most Linux related (and/or open source) developers, it is not their goal for world+dog to use Linux or their product. It is more about being part of a community (crowd!) to solve a problem and/or fulfill a need in such a way that *everyone* in the community/crowd benefits.
If you want the full benefit of the community’s work, then become *part* of the community do not leech off of it.
/second rant
Enjoy,
Woody
Re: Crowd Sourcing
If you went to *application* developers and were asking for assistance with *platform* questions (which, from your own description, seems to be what happened), then I’m not surprised they weren’t willing to help you.
How much free help do you think you would get from a Windows application vendor if you were to come in and start asking them questions that showed you didn’t have any grasp of how Windows works and hence couldn’t understand any of their answer to your questions?
Woody covered it all pretty well, but if you want and need a paid level of support… that’s what open source companies like Red Hat, Canonical, Oracle and Novell are for.
Fun time.
There are good reasons crowd sourcing works…As well as a few restrictions to it.
As with Netflics..If you want to know the MOST rented movie..great site.
The First thing you must understand is IF’ the crowd KNOWS what they talk about.
AS they recently did on TV, HOUSE’, when the NET trys to tell a patient WHAT may be wrong with him..
Bringing resources TO a crowd source is the best part. you must KNOW that your Crowd has a source of knowledge based on the Subject, or has access available to UNDERSTAND and make a Asserted answer.
Re: Fun time.
As with Netflics..If you want to know the MOST rented movie..great site.
It’s a bit better than that. Or Amazon certainly is. Haven’t used netflix since I left the USA in 2005.
(“people who liked that also liked…”).
hmmmm....
278 note/entry) Crowd Source the busines plan for the new media distribution system. Include Standards, Creative Commons, Accounting standards, etc
Found another gem …
Engaging crowds
Here’s a good alternative view of crowds. Sometimes you involve crowds not so much as a collective source of good ideas, but to encourage them to buy into the final result.
http://mashable.com/2009/09/02/attract-your-crowd/
I then wrote my own thoughts on the above article and how I think the direction of music is not toward the tribe model, but more toward the community model, where the fans are the show and the artist serves more as the community organizer.
http://brandsplusmusic.blogspot.com/2009/09/people-formerly-known-as-fans.html
i think the point dan woods is trying to make is that only shakespeares can write plays, not committees. quite like nicholas carr, he disses the concept of crowdsourcing because of the fear that individual talent will not get its due in the current frenzy to crowdsource everything.
like most things, the truth is somewhere in between woods’s paranoia and masnick’s disdain.
I'm Finally
I'm Finally Beginning to Understand...
So crowdsourcing DOESN’T involve hive minds or collective consciousness? Glad we got that cleared up.
Crowdsourcing
It appears to me that there is something really odd behind the entire crowdsourcing schtick. Why is it that every single time Forbes does it the sites and names “selected” by the “crowd” are always those entirely unknown to me …or anyone I know?