Participation Literacy

CI means many things to many people. Here, it refers to the capacity of human communities to evolve towards higher order complexity and integration through collaboration and innovation.

George Pór’s definition of collective intelligence above uses words and phrases as communities, evolution, “higher order complexity”, integration, collaboration and innovation. Collective intelligence is useful as metaphor in the Web 2.0 discourse. I am going to use the concept based on the words above. Collective intelligence, in this context, is thus something created in evolving communities on the Internet, which through integration, collaboration and innovation creates higher order of complexity, an understanding, experience, and intelligence larger than the sum of the participating users. A large group of people talking right into the air is not especially intelligent thus the community’s intelligence increases relatively to how well the software is able to manage these voices, how well the software manages to harness the sum of the intelligence of these people..

Two of the most noticeable examples of collective intelligence are the highly commercial Amazon.com and the open access encyclopaedia Wikipedia. In January 2005 Wade Roush wrote the following in Technology Review:

Wikipedia is the world’s newest, largest, most varied, most participatory, and most controversial encyclopedia. It is composed and edited entirely by volunteer netizens; as of November 2004, there were some 29,000 “Wikipedians” writing for it in 109 different languages. The site’s massive archive, including 380,000 articles in English alone, puts even Britannica to shame. If you don’t see an article addressing your passion for miniature-teapot collecting, don’t fret. Just write one (Roush, 2005).

The screenshot from Wikipedia 2006-02-02, below, shows a massive development for 2005. The number of articles has thus gone from 380,000 to 945,000 in one year.

One of the first Web 2.0 companies, Amazon.com figured out how to use the collective intelligence of hundreds of thousands of users, getting them to provide free reviews of books and gaining significant competitive advantage in the process. Amazon.com was founded by Jeff Bezos in July 1994. He was an investment banker who left New York and moved to Seattle with the idea of creating an online bookstore (Frey, 2004).

Amazon is a commercial business with the main goal of selling as many products as possible. But Amazon is also a community of literature lovers, music freaks, textbook users etc – more about Amazon in detail below. These communities have evolved from a few participants in the beginning to hundreds of thousands.

When discussing collective intelligence in a Web context, it might be useful to divide it into two separate phenomena in praxis: the Amazonian form of collective intelligence and the Wikipedian. Both forms have vast possibilities. The Amazonian form builds on a large amount of people participating with small pieces of knowledge. These pieces are treated by the CI machine to give the participant other pieces of knowledge in return, relating to their own knowledge. Their knowledge expands and makes them able to feed the system with more threads of knowledge. The Wikipedian form of collective intelligence is more precise and therefore more vulnerable. One participant may feed the CI machine with large, seemingly objective, and for the system noticeable and important pieces of knowledge. Other participants are then expected to interact with this knowledge either by using it, discussing it or changing it. The rationale behind includes the idea that this piece of knowledge will be enhanced as time goes on, and as more and more people invest their time and knowledge in it.

The Wikipedian form is by far the most discussed and criticized. The main critique is about the following question: can we trust this piece of information? The question is more than relevant. I am a big fan of Wikipedia, but since I never have trusted traditional encyclopaedias either, nothing is really new. Since information and knowledge are contextual, one single piece of information is very lonely. Adding more sources gives a bigger context and more trustworthy information, even if the information is contradictory.

In the Amazonian form, the physical CI machine has a more profound and complex role because the CI machine’s algorithms visualize and in a way enhance the collective intelligence. Noone expects the information pouring out of an Amazonian CI machine to be objective or true in the same sense the information in an encyclopaedia suggests. Thus the truth value depends more on expectancy than something inherent in the system.

The whole Web can be viewed as an example of collective intelligence. “Much as synapses form in the brain, with associations becoming stronger through repetition or intensity, the Web of connections grows organically as an output of the collective activity of all Web users” (O’Reilly, 2005). Several of the new Web companies have a deep understanding for the potential of the hyper linking features of the Internet. One of these is Google. They revolutionized the search engine market, with their PageRank technology. Before Google, search engines ranked their hit pages based on factors such as title, meta-information, headers, number of words etc. This, Web 1.0, kind of page ranking gave unnaturally high ranking to irrelevant pages, and the other way around. For Google it is not the page in itself that sets the rules for the page ranking, it is how the context valuates that page (It is probably possible to manipulate Google’s rankings also, but it is much more difficult). If I for example search for Volvo, the hits in Google are 31,200,000. On top of that list are Volvos official pages because they have more pages linking to them than pages lower down the hit list. The Internet community creates a ranking complexity, just by doing what they normally do in their daily lives. An equivalent situation in the physical world would be if every person’s footsteps suddenly gave marks on the streets. The most visited restaurants would then have more footsteps in front of their door than other restaurants.

Another example of collective intelligence is Ebay. Ebay’s about page says: “eBay is The World’s Online Marketplace®, enabling trade on a local, national and international basis. With a diverse and passionate community of individuals and small businesses, eBay offers an online platform where millions of items are traded each day”. Ebay’s competitive advantage is due to its critical mass of buyers and sellers, but it is not only about quantity. Ebay lives on word of mouth. Every time someone buys something at Ebay, that person is asked to write if s/he is positive, neutral or negative. It is also possible to write something more in detail. This evaluation also works in reverse; the seller can evaluate the buyer. Every buyer can therefore look at the seller’s aggregated evaluation. Thus both the buyer and the seller can feel reasonably assured that their business partner is honest.

Collective intelligence is a new way of looking at information and knowledge. If I wonder what an API (Application Programming Interface) is, I could search Encyclopaedia Britannica Online for an answer. This would be the Web 1.0 (and still relevant) way. I tried this and got no answer relevant to my search question: API.

Instead I performed the corresponding search in Google: define: API. I got about 20 relevant hits. The total list was about 25, but 5 of them were other denotations of the word API such as American Petroleum Institute.

A quick look at the URLs in Figure 5 probably raises suspicions in most researchers. The hit list from the Google define search shows an array of definitions from sources with questionable creditability, at least at this quick look. None of the 20 hits in the whole list have the credibility of for example Encyclopaedia Britannica Online. Yet we have 20 definitions and most of them are different even though there is a core of truth in them, or if you like, a core of similarity. One day perhaps a CI machine will be able to harvest this truth in a quite reliable way, but until then it is up to the user to be that CI machine. Acting as a CI machine I scan these 20 definitions, and as my mind registers the differences and similarities in the meanings, my mind builds an algorithm, which puts an aggregated meaning together, representing an approximate of all those definitions. We could also explain this as a hermeneutic process spiralling down to some kind of similarity core in those 20 definition texts.

I always use definitions as feeds into my hermeneutic machine. One sole definition is not worth much, even if the definition is created by men or women in power within their field. A definition should never be treated as a standard, like the XML standard, but as feeds by the power of the masses. Of course, the collective intelligence increases not only by quantity; quality is also an important factor. Humans have always been CI machines, aggregating and reconstructing information, the novelty lies within the power of ICT (Information and Communication Technology). A well crafted set of algorithms, together with databases and powerful software/hardware will perhaps rewrite the map of intelligence. Intelligence with the human as blueprint might be the perfect pair together with collective intelligence based on masses of different human voices and powerful CI machines to handle all data.

The last story about collective intelligence I will tell in this section is the information redundancy in the blogosphere. Blogosphere critics often say that the blogging community is an echo chamber. The echoes consist of the word of mouth. One blogger writes something. Another blogger believes that text to be relevant and therefore quotes the original text in his/her own article - and so on. The result is a wide array of texts echoing in a blogosphere. This echo chamber corresponds to the researchosphere and is not a bad thing at all. This is collective intelligence at work, filtering out the most relevant information (according to the group) in a wild torrent of voices. In a way, the echo chamber corresponds to Google’s PageRank, where a Web page gets higher rank in the Google hit list if it has more pages linked to it, than the pages further down in the hit list. The blogosphere is also similar to Web of Science, a science Web service, which creates an aggregated index of researchers refereeing each other in scientific journals.

Several Web 2.0 companies have tried to structure these choirs of voices. One example is Digg. You could call Digg a bookmark flag service. It works like this: you find an interesting page on the Internet; you add this page to Digg’s database. It, so to say, lands on the bottom of the Digg repository. When users find it interesting, they click on the digg button. The digg button displays how many users clicked it. For every user clicking it, the value aggregates with 1 and when enough users have clicked it, the bookmark rises one level in the repository. The algorithm also takes into account how new the bookmark is. The fifteen bookmarks floating around on the highest level of the repository have between 50 and 1000 clicks. There are bookmarks further down with several thousand clicks, but they are older. Digg can be viewed as some sort of anti gravitation chamber where things are floating vertically depending on the weight created by the number of clicks and how new things are.

CI means many things to many people. Here, it refers to the capacity of human communities to evolve towards higher order complexity and integration through collaboration and innovation.

George Pór’s definition of collective intelligence above uses words and phrases as communities, evolution, “higher order complexity”, integration, collaboration and innovation. Collective intelligence is useful as metaphor in the Web 2.0 discourse. I am going to use the concept based on the words above. Collective intelligence, in this context, is thus something created in evolving communities on the Internet, which through integration, collaboration and innovation creates higher order of complexity, an understanding, experience, and intelligence larger than the sum of the participating users. A large group of people talking right into the air is not especially intelligent thus the community’s intelligence increases relatively to how well the software is able to manage these voices, how well the software manages to harness the sum of the intelligence of these people..

Two of the most noticeable examples of collective intelligence are the highly commercial Amazon.com and the open access encyclopaedia Wikipedia. In January 2005 Wade Roush wrote the following in Technology Review:

Wikipedia is the world’s newest, largest, most varied, most participatory, and most controversial encyclopedia. It is composed and edited entirely by volunteer netizens; as of November 2004, there were some 29,000 “Wikipedians” writing for it in 109 different languages. The site’s massive archive, including 380,000 articles in English alone, puts even Britannica to shame. If you don’t see an article addressing your passion for miniature-teapot collecting, don’t fret. Just write one (Roush, 2005).

The screenshot from Wikipedia 2006-02-02, below, shows a massive development for 2005. The number of articles has thus gone from 380,000 to 945,000 in one year.

One of the first Web 2.0 companies, Amazon.com figured out how to use the collective intelligence of hundreds of thousands of users, getting them to provide free reviews of books and gaining significant competitive advantage in the process. Amazon.com was founded by Jeff Bezos in July 1994. He was an investment banker who left New York and moved to Seattle with the idea of creating an online bookstore (Frey, 2004).

Amazon is a commercial business with the main goal of selling as many products as possible. But Amazon is also a community of literature lovers, music freaks, textbook users etc – more about Amazon in detail below. These communities have evolved from a few participants in the beginning to hundreds of thousands.

When discussing collective intelligence in a Web context, it might be useful to divide it into two separate phenomena in praxis: the Amazonian form of collective intelligence and the Wikipedian. Both forms have vast possibilities. The Amazonian form builds on a large amount of people participating with small pieces of knowledge. These pieces are treated by the CI machine to give the participant other pieces of knowledge in return, relating to their own knowledge. Their knowledge expands and makes them able to feed the system with more threads of knowledge. The Wikipedian form of collective intelligence is more precise and therefore more vulnerable. One participant may feed the CI machine with large, seemingly objective, and for the system noticeable and important pieces of knowledge. Other participants are then expected to interact with this knowledge either by using it, discussing it or changing it. The rationale behind includes the idea that this piece of knowledge will be enhanced as time goes on, and as more and more people invest their time and knowledge in it.

The Wikipedian form is by far the most discussed and criticized. The main critique is about the following question: can we trust this piece of information? The question is more than relevant. I am a big fan of Wikipedia, but since I never have trusted traditional encyclopaedias either, nothing is really new. Since information and knowledge are contextual, one single piece of information is very lonely. Adding more sources gives a bigger context and more trustworthy information, even if the information is contradictory.

In the Amazonian form, the physical CI machine has a more profound and complex role because the CI machine’s algorithms visualize and in a way enhance the collective intelligence. Noone expects the information pouring out of an Amazonian CI machine to be objective or true in the same sense the information in an encyclopaedia suggests. Thus the truth value depends more on expectancy than something inherent in the system.

The whole Web can be viewed as an example of collective intelligence. “Much as synapses form in the brain, with associations becoming stronger through repetition or intensity, the Web of connections grows organically as an output of the collective activity of all Web users” (O’Reilly, 2005). Several of the new Web companies have a deep understanding for the potential of the hyper linking features of the Internet. One of these is Google. They revolutionized the search engine market, with their PageRank technology. Before Google, search engines ranked their hit pages based on factors such as title, meta-information, headers, number of words etc. This, Web 1.0, kind of page ranking gave unnaturally high ranking to irrelevant pages, and the other way around. For Google it is not the page in itself that sets the rules for the page ranking, it is how the context valuates that page (It is probably possible to manipulate Google’s rankings also, but it is much more difficult). If I for example search for Volvo, the hits in Google are 31,200,000. On top of that list are Volvos official pages because they have more pages linking to them than pages lower down the hit list. The Internet community creates a ranking complexity, just by doing what they normally do in their daily lives. An equivalent situation in the physical world would be if every person’s footsteps suddenly gave marks on the streets. The most visited restaurants would then have more footsteps in front of their door than other restaurants.

Another example of collective intelligence is Ebay. Ebay’s about page says: “eBay is The World’s Online Marketplace®, enabling trade on a local, national and international basis. With a diverse and passionate community of individuals and small businesses, eBay offers an online platform where millions of items are traded each day”. Ebay’s competitive advantage is due to its critical mass of buyers and sellers, but it is not only about quantity. Ebay lives on word of mouth. Every time someone buys something at Ebay, that person is asked to write if s/he is positive, neutral or negative. It is also possible to write something more in detail. This evaluation also works in reverse; the seller can evaluate the buyer. Every buyer can therefore look at the seller’s aggregated evaluation. Thus both the buyer and the seller can feel reasonably assured that their business partner is honest.

Collective intelligence is a new way of looking at information and knowledge. If I wonder what an API (Application Programming Interface) is, I could search Encyclopaedia Britannica Online for an answer. This would be the Web 1.0 (and still relevant) way. I tried this and got no answer relevant to my search question: API.

Instead I performed the corresponding search in Google: define: API. I got about 20 relevant hits. The total list was about 25, but 5 of them were other denotations of the word API such as American Petroleum Institute.

A quick look at the URLs in Figure 5 probably raises suspicions in most researchers. The hit list from the Google define search shows an array of definitions from sources with questionable creditability, at least at this quick look. None of the 20 hits in the whole list have the credibility of for example Encyclopaedia Britannica Online. Yet we have 20 definitions and most of them are different even though there is a core of truth in them, or if you like, a core of similarity. One day perhaps a CI machine will be able to harvest this truth in a quite reliable way, but until then it is up to the user to be that CI machine. Acting as a CI machine I scan these 20 definitions, and as my mind registers the differences and similarities in the meanings, my mind builds an algorithm, which puts an aggregated meaning together, representing an approximate of all those definitions. We could also explain this as a hermeneutic process spiralling down to some kind of similarity core in those 20 definition texts.

I always use definitions as feeds into my hermeneutic machine. One sole definition is not worth much, even if the definition is created by men or women in power within their field. A definition should never be treated as a standard, like the XML standard, but as feeds by the power of the masses. Of course, the collective intelligence increases not only by quantity; quality is also an important factor. Humans have always been CI machines, aggregating and reconstructing information, the novelty lies within the power of ICT (Information and Communication Technology). A well crafted set of algorithms, together with databases and powerful software/hardware will perhaps rewrite the map of intelligence. Intelligence with the human as blueprint might be the perfect pair together with collective intelligence based on masses of different human voices and powerful CI machines to handle all data.

The last story about collective intelligence I will tell in this section is the information redundancy in the blogosphere. Blogosphere critics often say that the blogging community is an echo chamber. The echoes consist of the word of mouth. One blogger writes something. Another blogger believes that text to be relevant and therefore quotes the original text in his/her own article - and so on. The result is a wide array of texts echoing in a blogosphere. This echo chamber corresponds to the researchosphere and is not a bad thing at all. This is collective intelligence at work, filtering out the most relevant information (according to the group) in a wild torrent of voices. In a way, the echo chamber corresponds to Google’s PageRank, where a Web page gets higher rank in the Google hit list if it has more pages linked to it, than the pages further down in the hit list. The blogosphere is also similar to Web of Science, a science Web service, which creates an aggregated index of researchers refereeing each other in scientific journals.

Several Web 2.0 companies have tried to structure these choirs of voices. One example is Digg. You could call Digg a bookmark flag service. It works like this: you find an interesting page on the Internet; you add this page to Digg’s database. It, so to say, lands on the bottom of the Digg repository. When users find it interesting, they click on the digg button. The digg button displays how many users clicked it. For every user clicking it, the value aggregates with 1 and when enough users have clicked it, the bookmark rises one level in the repository. The algorithm also takes into account how new the bookmark is. The fifteen bookmarks floating around on the highest level of the repository have between 50 and 1000 clicks. There are bookmarks further down with several thousand clicks, but they are older. Digg can be viewed as some sort of anti gravitation chamber where things are floating vertically depending on the weight created by the number of clicks and how new things are.