There is no Wisdom in Crowds

There must be one really wise guy at Google ranking the trillion+ urls for everyone then.

I would like to see examples where a crowd acts wisely to do or produce something of value (other than predict its own behavior).

Your Google example is not a good one because the "crowd" you are referring to isn't making the decisions or the money.

There is a small group of engineers in a hierarchical organization that makes the decisions about how page rank works. They (and Google as a company) is rewarded dearly for this work.

This is a case of very smart individuals at Google using the irrational crowd to make a whole lot of money.

Google uses the wisdom of everyone on the internet. They count links as votes. That is an example and you are going through convoluted hoops to deny it.

Sure, some engineers write the code to extract the crowd wisdom, but they obviously don't provide the wisdom themselves. That comes from the crowd.

Now I think you are really just trying to bash the voting system here, and wouldn't cede any cases of crowdsourcing being legitimate (especially after having staked out the silly absolutism that there is no wisdom in crowds) but in case you aren't and you really want examples:

recaptcha - This is a program to digitize books while users fill out a captcha form. They are presented with garbled words scanned from books and type them to help digitize old manuscripts.

ebay feedback - This is a reputation system that used collective feedback to help users choose between safe transactions and potentially risky ones. The idea is that users with great feedback ratings from a diverse selection of the crowd are less likely to defraud you than ones without.

Amazon mechanical turk - A platform to buy and sell "Human Intelligence Units". Can be used for crowdsourcing or not.

Google Image Tagging game - two people play a word game where they tag an image till they use the same tag. Then this data is used in aggregate to categorize images for image search.

Now you excluded collaborative filtering, which is pointless in your criticism because you are criticizing the attempt to use the wisdom of the crowd on able2know, and that's pretty much the only thing it's being used for. So what's the point? Are you looking for examples of systems that don't do what this system is trying to do as an example of how it won't? Collaborative filtering does a great job at content recommendation when done right and here are some examples:

Netflix Movie Recommendations - Based on your rating of movies, it tries to recommend additional movies. They even have a million dollar contest for better ways to harness this crowd data. http://www.netflixprize.com/

StumbleUpon - Site recommendations based on crowd ratings.

Digg - News articles and internet links based on user popularity.

Last.fm - the Audioscrobbler music recommendation engine

Every good case of crowdsourcing is going to have a few smart individuals leveraging the wisdom of the teeming millions. What's your point?

Are you only looking for examples where everyone in the "crowd" directly participates in the product building?

Then try on your favorite open-source project for size.

Another good example: Wikipedia.

Nuff said.

Robert... you are giving me examples of how crowd data can be used to predict the behavior of a crowd. I concede the point. (Ironically, statistical analysis of lots of user data is closely related to my profession.)

You haven't given me any example of where the crowd is good for anything other than predict its own behavior.

Wikipedia is an interesting example... individual edits as a collection. However, wikipedia has the idea of "editors"... a group of individuals that control and direct the crowd. Even in wikipedia, the crowd by itself is a disaster.

Collaborative filtering as an example of the point I am making. Far from producing interesting results, it produces homogeneity, filtering out many unique and interesting results while the same old fluff rises to the top.

I can give you a perfect example right here. Go to the A2K home page, and select "Science and Technology" with the "sort by votes" option selected.

Yeah 'Crowd wisdom' is oxymoronic. Crowds can make choices that seem wise but the mechanism is not wisdom. It's much the same phenomena that makes evolution seem directed.

And that create order out of chaotic processes.

Nah, I'd argue that most individuals respond to negative political ads (like is the wrong word, you're not meant to like to them your meant to respond to them)



There are not infrequent cases of orderly crowds in high pressure situations, but I will grant you this may be an artefact of the culture of the individuals, or the qualities of the particular individuals in that crowd.


I concede your point, but.... You need somewhere to live, the population is rising, therefore demand is rising. Most housing purchase aren't a financial investment, although it is an additional selling point in the purchase decision.

I think I agree with your overall point but I really think the crowds are a product of the individuals who make them up. And just like individuals, crowds have varying 'quality'.

Let me rephrase your own account here. You read what a bunch of people said about a book. Judging by what you did next, you consider that that an adequate substitute for reading the book yourself. And what's more, you found what these people said through Google, which uses a wisdom-of-crowds approach to spot useful information on the internet.

For someone who finds the wisdom-of-crowds concept dubious, you put a lot of trust into it in your practical, everyday behavior.

(PS: The Wisdom of Crowds is an interesting, engaging, and intelligent book -- I highly recommend that you read it before dismissing its ideas as "dubious".)

the only wisdom i see in crowds is to avoid them.

1. Market economies. Reasonable people can disagree about the optimal proportion of markets, which employ the wisdom of crowds, and governments, which work through hierarchical planning. (But only do it well if tempered by a wisdom-of-crowds concept called democracy -- see below.) I don't think reasonable people can disagree that more or less pure market economies (Victorian England, Hong Kong) have historically worked much better than more or less pure command economies (Soviet Russia under Stalin, China under Mao.)
   
   
2. Democracy. Works much better than government by kings, politbureaus, and military juntas.
   
   
3. Science. The proverbial Republic of Scholars approach to science -- free speech, peer review, and all this, produces knowledge in a vastly more efficient manner than the process we had before: a hierarchical church, with a pope or the king of England at its top, deciding what's true and what's false. It was after these hierarchies lost control to crowds of fly-by night hobbyists, who pompously called themselves "natural philosophers", that alchemy turned into chemistry, and astrology into astronomy.
   
   
4. Language. Arguably the most powerful valuable thing in humanity. Pretty much all languages in the world were developed by the crowds who spoke it. There are exceptions, but they are either narrow and technical (like computer programming languages) or obscure (Esperanto).

How's that for examples?

Um... I think that's the whole point of crowdsourcing.

You don't let the crowd run off willy-nilly, you give the crowd something to do.

There's also the protein folding game, and the interesting skies thing where people look for unusual images from observatories.

Cisco has the billion-dollar project (or something). An internal wiki that they use to discuss and develop ideas. Then they pull what they consider the 10 best ideas, and hope that a few will actually work out.

It's like SETI-at-Home that uses your computer's downtime. A good crowdsource project uses people and their brains.

I agree, crowds left to their own devices are stupid. But crowds with a few tools and a little oversite can be hugely powerful tools.

I forgot to mention one of my favorite sites, threadless.com. Cool use of leveraging the masses.

Thomas,

Market economies are an interesting example-- the more "pure" they are, the more they tend to be dominated (and controlled) by a small group of people with an inordinate amount of power. This is perhaps why they work (although we haven't defined how to measure whether an economy is "working" or not). Whether it is the individuals with power making the decisions who produce "value", or whether it is crowd dynamics is perhaps a question of semantics.

Democracy is also interesting. First, it is clear that our Democracy is controlled by a relatively small group of powerful people. This is not only because of the influence of money and control of the media... the ability of a small group to control government is designed into the way our democracy works-- from legislative bodies to the electoral college.

It seems clear to me that a pure, crowd-controlled, democracy would be a disaster. In fact I have seen this at work on a smaller scale, where a group without a clear leadership was unable to make a rational decision (and ended in chaos).

You are completely off-base in regards to science.

Galileo (an individual) stood up against the pope. The crowd stood with the church. Isaac Newton, Albert Einstein, Charles Darwin... all stood out because they went against the conventional wisdom of the crowd (and it is a crowd mentality that still attacks Darwin).

There is a system in science which allows individuals to test new ideas. This is a strict set of rules and institutions that are the exact opposite of crowd rule... they let individuals prove the merit of their ideas-- whether they are popular or not.

I would think a crowd of people at say a concert is a completely different crowd than you would find on the internet.

I used to go to a lot of concerts actually and loved being right up in the mix. I’ve been to shows where my feet literally didn’t touch the ground for short periods of time because of the people surging around me. As long as no one got violent it was a good time. However, there were plenty of times where I was having a good time and someone or even a bunch of people would get hurt right next to me. I even remember one show where a guy ended up paralyzed.

I think the most dangerous crowd I was ever a part of was at the Tibetan Freedom Festival when it was in Downing stadium on Randall’s island. Two days of music, and by the end I think they estimated over 100,000 people. I just wiki’d it and they claim more like 50,000 but I remember someone addressing the crowd at the time and getting everyone excited by how many people were there. I could easily have remembered wrong, it was in 97’.

Anyway, the last act was the Beastie Boys of course. Their home town and I should have known what was coming. We were “close” to the stage at about 100 yards out heh. Getting any closer seemed impossible. There were endless people trying to get closer and closer and we felt pretty comfortable. That is until they came out on stage. I’ve never seen mass hysteria like this. My girlfriend literally clung to me as people surged the stage. We were being swept off of our feet, literally, and there was nothing we could do about it. There were so many people pushing. For the first time ever I was actually scared of what might happen. If we tripped we were almost certain to be trampled and it was my only thought. We were completely out of control of the situation and we were both seasoned concert crowd experts for the most part.

I’m sure it didn’t take long even though it seemed like forever, but the people up front starting pushing back. You could literally see the collective wave of force from people vying for room to breathe. Next thing I know there were people pushing through the crowd and screaming that someone had been hurt. People were panicking all around me which definitely ruined the moment for good at that point. I should say that I had a blast overall and the adrenaline surge from that was pretty amazing after we got out of there and felt safe again. It’s made me pretty leery of crowds of that size though. You just never know what’s going to spark an avalanche heh.

Anyway the point is that in my experience the only good that ever came out of those crowds was that no one was hurt.

Two points: One, whether this is true or not varies from market to market. In markets with great economies of scale (think TV networks), it's true. In markets without great economies of scale (think grain), you don't observe that. It's still pretty bottom up.

Apart from whether it's true or not, there is a second point: Is that concentration, taken by itself, a good thing or a bad thing? I would have argued that monopolies, multibillionaires, and the like are a bad thing, and that government generally improves the economy's workings by curbing the monopolies and other concentrations of market powers.


Again, we would probably disagree about the extent to which this is true. But more importantly, is this concentration of power good or bad? I would have argued that it's bad.

.
Unfortunately, I agree. Indeed one of the insights that turned me from a Green into a libertarian was that markets give you back most of the accountability of representative democracy, while retaining much of the bottom-up organization of grass roots democracy.


That's not true of the people who actively worked in science, and the process through which they worked in it.

I notice you didn't respond to the example of language. Where's your great individual there?

This is a REPUBLIC designed to be run by officials elected by the people thus eliminating mob rule.