My partner Brad Feld wrote a post today crediting me with coining the term Intelligence Amplification. While I’d like to take credit for it, all I can take credit for is applying the term to what Brad had previously called Dynamics of Information. After Brad asked for a better name, I suggested Intelligence Amplification, a term whose origination dates back as early as the 1950s and has been attributed variously to William Ross Ashby, Vannevar Bush, Douglas Engelbart (as if the mouse weren’t enough), and various other pioneers of cybernetics, information theory and computer science.
I think this term applies very well to what is enabled when you combine elements of social networking, open-source knowledge (I rely heavily on Wikipedia for links in many of my blog posts), trusted relationships, folksonomy/tagging (see the wisdom-of-crowds at work with Flickr, Del.icio.us, Technorati, Digg, etc), collaborative filters, search engines and other tools that use the internet to coordinate human-to-human sharing of knowledge and information. These tools use algorithms to leverage human activities and human minds belonging to millions of strangers, and, increasingly friends and acquaintances, to help us find relevance in the flood of information we are trying to stay afloat in.
My undergrad degree at Stanford was called Symbolic Systems, and my concentration within the major was artificial intelligence. I left Stanford unconcerned that I’d need to worry about Skynet or any other mechanized intelligence (dystopic or otherwise) taking over the world within my lifetime. (Sorry, Ray Kurzweil.)
But while computer scientists might not bring us software that can pass the Turing Test anytime soon, and certainly not within the relevant time horizon for a venture capitalist, the latest generation of social media sites can do things for us that leave the latest-and-greatest AI efforts in the dust.
Take Flickr, for example. Late this past August, I was getting emails from friends and colleagues who had just returned from Burning Man, which had concluded less than 24 hours earlier. I went to Flickr, searched for the tag “Burning Man 2006” and was pleased to discover over 8,000 photos taken at the event. Try asking a hyper-sophisticated image analysis algorithm to find a picture of Burning Man 2006 in a (virtual) stack of photos, and you’ll get nowhere.
Our ability to produce information is growing exponentially, and this can be problematic. What do you get when you leverage internet applications to coordinate the clickstreams, hyperlinks, tags, actions, relationships and interests of billions of people? Hopefully the means for humans to synthesize information into knowledge exponentially faster than was previously possible.
Sure, we already rely on software for things like OCR and voice recognition that might be considered artificial intelligence applications in the classic sense, but we’re also beginning to rely increasingly on the intelligence amplification enabled by software that lets computers do what they excel at (fast computation, “perfect memory”, gruelingly repetitive tasks, statistical analysis, etc) and also leverages what humans are far better at (face recognition, voice recognition, any cognition, matters of cultural discernment, language generation, etc).
To paraphrase the immortal words of Nigel Tufnel, may our intelligence amplifiers one day “go to eleven”.