Thinking Like a Cognitive Computer: a Technology Mash-up
Data Mining. Natural Language Processing. Pattern Recognition. Computer Vision. Machine Learning…
You’ve probably heard at least one of these technologies mentioned at a tech conference or in an industry pub. Each of these is a powerful tool on their own, but put them together and you have one of the most exciting developments in technology today: Cognitive Computing.
Put simply, Cognitive Computing tries to combine various technologies to emulate as closely as possible the way humans think or approach a task. This can be anything from selecting a matching shirt for your pants, to determining the best course of action for a patient in the ER.
For those of us in the agency world, let’s take a look at a simple parallel on how the various elements of Cognitive Computing would mimic what happens in your brain.
Imagine that you’re in a conversation with a colleague who has a concern about the treatment of a brand’s logo. This concern would act as the trigger for a sequence of events inside your brain as you try to address it. Here’s what that sequence might look like:
As your colleague is talking with you about the concern, you dig through your brain looking for previous experiences to identify any previous situations regarding the use of the logo.
Natural Language Processing
During your conversation, your colleague mentions that he wants the logo to “pop”. Your brain can place this word in context within the current conversation and interpret the reference to some visual treatment of the logo relative to the rest of the creative (i.e. not that the logo should literally pop or burst into hundreds of pieces).
After digging around in your brain, you remember 3–4 previous executions of the logo and start to identify certain patterns in the way it has been treated. With these patterns identified, you recognize that by “pop,” your colleague might mean brighter contrast to the rest of the creative.
You take a look at the current creative and identify that it doesn’t fit with what you concluded doing pattern recognition.
After identifying, analyzing, and eventually solving the problem, your brain updates the previous information you had on how to treat the logo so that next time you work on it, you know how to make it “pop.”
This example is pretty straightforward, but it should give you an idea on what Cognitive Computing tries to mimic when attempting to emulate the way we think.
In addition to being an amazingly powerful tool to solve problems, cognitive computing can also be used for creative purposes. For example, IBM Watson collaborated with 20th Century on the first AI-directed movie trailer.
Cognitive Computing is the natural evolution of our love for technology mash-ups. It will surely continue to evolve as more capabilities are combined and integrated to create new and exciting applications that will eventually be part of our everyday lives.