Originally posted on August 09, 2006
Will Collective Intelligence Enable End-Users to Fix Their Own Computers?
I posted some comments a few days ago about Web 2.0 and the fact that I think a lot of the talk about collective intelligence is off base. I want to expand a little more on my reasoning against the concept of collective intelligence.
I think that collective intelligence is so compelling to people because you can extrapolate such grand potential futures even from humble beginnings. Everyone is talking about how the original intent of the internet as a grand, centralized nerve center of cerebral activity is finally approaching reality. Who wouldn’t want that? To connect the brainwaves of millions and billions of people to form a collective super-intelligence capable of solving world problems, curing diseases, and creating knowledge as yet impossible.
To me, the whole thing is starting to sound a lot like the story of the tower of Babel; everyone trying to reach the stars with bricks and mortar. The whole thing will probably end up pretty much the same, as well.
The CPUs of thousands of computers can be linked together to process things in the blink of an eye (ie Google search). The hard drives of thousands of computers can also be connected to offer almost unlimited data storage. The computing analogy has been extended to form the argument for collective intelligence. “Just like thousands of computers linked together” so the argument goes “humans will be able to process problems that would have been impossible for a single person.”
At first, this analogy bears itself out with examples like Wikipedia and other repositories of human knowledge. The web itself is the prime example. There’s so much information on the web that no one person or organization could have created such an extensive database of knowledge.
That, however, is where the analogy stops. Collective knowledge is one thing and collective intelligence is quite another. While humans may be able to combine all of their knowledge into a single repository, intelligence remains fiercely individualistic.
The central problem with the concept of collective intelligence is that humankind’s intelligence is encapsulated in billions of separate and segregated skulls. In computing, thousands of CPUs can crunch efficiently over the same problem at the same time without repeating cycles; one CPU crunches one aspect of the problem while another CPU crunches another. All the CPUs don’t have to know the whole problem, because the whole process is tightly controlled and organized in a hierarchical system. Humans don’t have this luxury. Human CPUs can’t work together efficiently because they all end up processing the same whole problem instead of efficiently dividing the processing. Human intelligence cannot be collective because it is divided by walls of bone on every side. Knowledge can be shared, but intelligence cannot. I can tell you what I know, but I can’t help you process it the same way I do. Einstein could not enable me to come up with relativity even if he tried. Knowing what Einstein knows could help me to understand relativity, but thinking it up is another matter. Along the same lines, it would be ridiculous to think that Einstein could have come up with his theory earlier if I had been there to help him process the problem.
Just look at history. What inventions or accomplishments have been made via collective intelligence? In almost every case, even if a lot of people contributed to the whole, the epiphanal moment was in the brain of a single person, not 1000 people thinking together.
IQ points cannot be added together. If a person with an IQ of 90 and a person with an IQ of 130 get together to think about a problem, the knowledge may be greater between the two of them, but the max output will still be about 130 not 220 (that is, if the dummy doesn’t get in the way).
I know that this will be a major downer for a lot of people. It sure would be fun to use the smarts of thousands of people on some of the world’s problems, but unfortunately, we will just have to settle for collective knowledge instead. Toward this goal, we’ve made a lot of progress. It’s a heck of a lot easier to find information than it was a few years ago.
To close out my argument, and to make a plug for my company, I’d like you to think about the example of support juxtaposed against the discussion of collective intelligence. Just a few simple questions for the support guy (or gal):
- Do you think that you will one day not be needed because people will be so collectively intelligent that they will solve their own problems?
- Do you think that you will one day not be needed because people will have the answers to their IT problems at their fingertips (aka Google groups)?
- How many users must be assembled before their collective intelligence enables them to install MS Project Server without your help?
- Are users smarter together or separately?
I feel like our market is secure.
