A terrible tragedy happened this week in Oslo. A single Norwegian named Anders Behring Breivik, possibly acting alone, took the lives of almost 100 people in a singular act of violence. This was the worst act of non-wartime violence in Norway’s history, producing twice the fatalities per capita than September 11th. Our heart goes out to the entire community of victims in and around Oslo, who will no doubt be suffering from this for years to come.
While the news was still unfolding, the social media sphere was abuzz with speculation about what was happening and why. A single “terrorist expert” by the name of Will McCants reported that this was the act of an Islamic terrorist, citing a post by Abu Sulyaman al-Nasser on an extremist Jihadi forum. The Guardian and The New York Times quickly picked this up, and before long, the Sun had gone to press with the headline “Al-Qaida Massacre, Norway’s 9/11”.
The only problem was, Al-Qaida had nothing to do with the attacks. The “analyst” was purely speculating and the “source” was a well-known liar and braggart to those even passingly familiar with the small world of terrorist intelligence groups.
While the mainstream press was running with the headline, however, the Twittersphere had already identified the false source, discredited the faux analyst and made short work of the Al-Qaida claim. They were also propagating and interpreting live reports from Oslo 20 to 30 minutes before mainstream sources were on the scene, with much greater depth and nuance than any of the major outlets.
Without making light of this tragedy in any way, what does this tell us about the way intersubjective collective intelligence systems work in the real world? What lessons can we draw from it to help identify and make sense of tragic surprises like this one, after they occur, or even while they are occurring, in the future?
There were several aspects of this event which describe successful collective intelligence (CI) systems:
Many Eyes: There were thousands, if not hundreds of thousands, of discrete minds and bodies focused on this event. Each had their own vantage point, background, perspectives and level of expertise. Some were on the ground in Oslo. Some were reporting live from Doha. Some were in their lounges and offices in London. But this vast number of perspectives, each with a unique point of view, was able to collect information, interpret it, and communicate it. Such a “depth of bench” is a key component of intersubjective collective intelligence systems.
Distributed & Multi-Sourced: Unlike the traditional media, these eyes were all participants in the unfolding interpretation of the event and active creators of what it meant. I personally had dozens of different trusted sources thinking about this, which I backed up with emails to friends in Oslo, texts to friends in the Middle East, and calls back to the States. This combination of geographic, personal, professional and political sources served to both generate a large amount of unique data, as well as filter it, cross-check it, and make sense of what it meant. Distributed, multi-source, active perception and interpretation is another key aspect of successful CI systems.
Real Time: There was practically no lag between the perception and analysis of what was occurring. Analysis was constant and on-going, even when I had to log off to cook dinner for my family or focus on other things. The massively parallel, synchronous nature of CI systems is part of what enables real-time sensemaking to occur. By leveraging the hundreds of thousands of different perspectives at play, CI systems are able to help develop a more rapid and comprehensive understanding of what is occurring, while it is occurring.
Multi-Scale: Dave calls this property “disintermediation”. It means that, through a combination of Tweets, emails, phone calls and web searches, I was able to access and contribute information at multiple levels of granularity and resolution. I wasn’t forced to rely on the BBC’s interpretation of events, but also wasn’t forced to rely only on the large volume of small Tweets filtering across my screen. A successful CI system allows for interaction with the data from the smallest scale to the highest scale, easily enabling comparison and transition through scales. Although this was done manually in this case, the sum result of hundreds of thousands of people performing similar multi-scale filtering and clustering created an extremely high resolution picture of both the events and various perspectives on their meaning and cause.
Based on Relationships: Like Twitter, all successful CI systems are built on top of networks of human relatedness. I have a carefully cultivated network of friends and colleagues, both online and off, who I trust to help me make sense of the world. I trust their input more than others because I have evidence to believe they have something to offer, more so than an anonymous media outlet or random Tweet from the web. So although CI enable total accessibility from multiple sources, they depend utterly on the trust of human networks to succeed. That is how I help evaluate the veracity of conflicting information, how I decide who to send what information to, and how I help compose my view of the world. No automated system without this layer of human relatedness and trust could ever hope to replicate the functions of a human-driven intersubjective CI system.
Transparent: The last factor tying this all together is the transparency of the system. The fact that all this could occur made it trivial to identify the source of the Al-Qaida claim, the source of his source, and his background and trustworthiness. The combination of deep analytical horsepower, real time reporting across scales and total transparency enables rapid self-correction and error-discovery. Although the mainstream media did ultimately correct its Al-Qaida claim, and sub sequentially changed its description of Breivik from “terrorist” to “attacker”, it did so at a glacial pace compared to Twitter and only after millions of people learning about this for the first time had formed erroneous first impressions of the event.
All of these factors, as demonstrated by Twitter and other real-time sources during this terrible event, suggest ways that intersubjective collective intelligence system may play in our lives. Could such a system have prevented these attacks? Of course not. Even total surveillance could never have identified and predicted a specific event such this. But such systems can provide us with far more effective pattern matching and awareness building that, if linked to response teams, could someday help to enable more rapid intervention thereby, we hope, saving lives. The fact that Twitter and other media also helped to play a sensemaking role could have, possibly helped to moderate the emotions and violent reactions that such a terrible, tragic events often inspire.
But whither wisdom? Will such comprehensive intersubjective collective intelligence systems make us more wise? Will they help prevent the kind of hatred and insanity that drove Breivik to enact this tragedy in the first place?
I argue no; we should not confuse real-time environmental awareness with greater wisdom, compassion or decision-making ability. No amount of technology or omniscient awareness will help us become better parents, neighbours or human beings. But they can, perhaps, empower us to valorise the best parts of our humanity and moderate the worst, even in terrible situations such as this. It was for this reason that I tweeted and retweeted the Mayor of Oslo’s famous words that, “Norway’s response will be more democracy and more humanity”.
We may not always have the luxury to respond this way. But when we do, we should use every tool at our disposal to do so.
Thank you all for a great two weeks guest blogging here. I hope you have enjoyed these thoughts. Thanks for letting me share them with you and please feel free to get in touch to continue the conversation.