How much wisdom is there in the "wisdom of the crowds"?

[eeik]

I hear it repeated with dogmatic insistence in numerous different circles as if it were one of the great universal truths of the human "organism".

 

Economists, software creators, strategists, et al, ... the wisdom of the crowds must be consulted or captured to make the best decision possible. Augur for instance as a new decentralized prediction market bakes this notion into their brief product intro video. The problem to me however is that it doesn't seem to hold up even initially to philosophical inquiry since crowds don't exist as anything other than collections of individuals. To channel a bit of Jon Rappoport, the notion of accepting this premise is counter to the power and agency of the individual which is a common theme in our creeping socialistic tendencies as modern 2016 Western democracies.

 

The idea that a collection of sample guesses from 100 people is better than 1 expert may sound sexy to the disempowered with their critical thinking capacity diminished, but universality of this notion doesn't hold true in my mind. One hundred Walpiri people (https://numberwarrior.wordpress.com/2010/07/30/is-one-two-many-a-myth/) that have never seen an M&M, guessing how many M&Ms in a one liter jar, vs an M&M plant manager well versed in Quality Assurance and Packaging seems like a sensible null hypothesis to me. The validity of any said "result" of the "one, two, many" Walpiri people's guesses would seem to bias towards luck, question phrasing, and sample size variation vs anything to do with "wisdom" of the crowd's assessment. Even if one sample out of a series of samples gets the answer "right" what can we objectively say about the result other than to compare it to the infinite monkey theorem (https://en.m.wikipedia.org/wiki/Infinite_monkey_theorem)?

 

Is it "wise" to give this notion much credence? What do we even mean by wisdom here?

 

Is this symptomatic of our increasingly socialistic and/or cultutally relative Western world? Is this word "wisdom" another victim of newspeak, being morphed before our eyes by its association with biased common denominator crowd consensus?

 

Has anyone given this topic any thought? Does anyone else see the irony of me asking this question of a biased crowd in the first place...?

[Sima]

if here is a group of 100 people, there is a bigger chance that 1 of them can come with a good idea, or that there is a doctor between them. But i wouldn't let 100 random people to make a surgery at once :DDD 

[vahleeb]

You can never underestimate the collateral political gains through this dogma.

 

Sustaining that there's a "wisdom of the crowds" inadvertently props up the "democratic" political process as a "wise one".

 

Yes there is a case where if you have a crowd that is reasonably informed on a topic, the probability that more people will come up with the correct answer than those that come up with the wrong answer is significantly large. Where the whole construct goes off the rails is in assuming a crowd will chose a valid answer regardless of their levels of instruction. Not only is there no consistent empirical evidence to support this theory, but the reasoning for it is highly shaky.

[AccuTron]

I'm reading Tolstoy's War And Peace.  (It picks up the pace about halfway in for those of you sitting on the fence about reading it.)  I'm in the part about Napoleon making his way into Moscow.  The crowds of Russians, be they noble or serf, run amuck with rumors and convictions, virtually none of which are true.  We have modern communications, but I really don't see much difference in the result in how crowds behave (even when dispersed over the internet).  

 

From an evolutionary perspective, where it's all rocks and dirt, plants and other animals, rain or lightning, any member of the human (or other species) that arrives with information for the group is probably correct, in that it boils down to seeing other animals, or natural phenomena, and they either are there, or they aren't, even if the scale of something is mistaken.  So we have a survival imperative to believe whoever brings us information.  In our modern human world, this is like a car without steering, but in the natural world, this worked.  

 

Various species engage in deception, ask your dog where he hid something.  At a tribal level, wherever that emerges in the primate lineage, any doubt is referred to a trusted tribal leader, presumed to know better, and even if it's from experience, the tribe believes in the leader's ability almost by magic.  Again, in the natural world, this worked.  But now, tribal is "I only vote red" and "I only vote blue."  So the crowd will believe this red or blue person as tho' they have magical powers.  Alas, in our civilized world (starting with ancient civilizations), this person is almost always lying.

[labmath2]

Yes there is a case where if you have a crowd that is reasonably informed on a topic, the probability that more people will come up with the correct answer than those that come up with the wrong answer is significantly large. Where the whole construct goes off the rails is in assuming a crowd will chose a valid answer regardless of their levels of instruction. 

 

What is that based on? If we had 100 people informed on a topic, and each individual had a 60% chance of coming up with the correct answer (assuming no one individuals answer affects another) then that maybe true. However, there is no reason to assume this. If their answer is more like a bell curve with most people falling somewhere in the middle, then you are likely to get the same answer by selecting a random person in the group. The more people you have, the chances that they are all equally informed also diminishes. Wouldn't you be better off simple finding one person who is an expert on the subject? Isn't that why specialization works?

[Mister Mister]

I don't think there is inherent wisdom in crowds, not if there is no check against mass irrational delusions.  Similarly, a very intelligent person does not always know best, when there is no feedback from the crowds.  But I think this example is talking about a free market, in which the desires and resources against the crowds, are in a feedback with the ingenuity and efforts of entrepeneurs, which does yield results with greater "wisdom" than any one person, i.e. a central planner could determine.

[TheRobin]

Maybe I'm confusing it with another term/concept, but from what I remember the idea is that if you have a large enough group you'll have some people in it who are well informed on whatever topic you're asking about. So while everyone else just generates random noise (statistically speaking) you'll have the few experts who will tip the result in the correct area, hence why it works (statistically). 

[rickheere]

I see it like this.
Of course, having someone who knows the correct answer to the correct question is always the best option. But when the question for example is, "how do we allocate all resources in the economy", the question becomes to complicated to answer. Even if it is possible to identify all the variables, new once have popped up by the time you finished. But "how do I allocate the resources that are available to me" is a question people deal with there entire lives. That makes it easier to deal with for one person.  

 

And, it is also not the case that the goal is to reduce an answer from the "wisdom" of the crowd. A functioning system is the goal. 

[vahleeb]

Maybe I'm confusing it with another term/concept, but from what I remember the idea is that if you have a large enough group you'll have some people in it who are well informed on whatever topic you're asking about. So while everyone else just generates random noise (statistically speaking) you'll have the few experts who will tip the result in the correct area, hence why it works (statistically). 

 

I don't think you're confusing it. That seems like the theory I remember reading as well. The one objection to it is what RoseCodex pointed out in the post above that you cannot vaccinate against mass delusions so if the topic you're investigating is vulnerable to a mass delusion then the "specialists" would easily be overwhelmed and their signal would be lost in the madness.

 

To sum up again, crowd intelligence (from the theory) can only work if the subject it is faced with is dispersatory enough that the common man opinion will be an inert scatter allowing you to see the direction coalescing around the signal from the specialists. Whenever you submit a subject that is "controversial" all you get is a popularity contest much like the democratic system of voting .

[eeik]

Whenever you submit a subject that is "controversial" all you get is a popularity contest much like the democratic system of voting .

 

This, to me, is the most pragmatic way to summarize what I've heard thus far from others and why I've been feeling so curious about this topic to begin with. Far too frequently, the problem space that "wisdom of the crowd" believers would have you accept their argument for involve these more controversial topics ... those that involve a broad aesthetic preference or vague representation of an abstraction vs. epistemologically grounded correctness. Thank you for sharing your thoughts on the topic everyone!