We’ve been hearing a lot about artificial intelligence and machine learning lately. From the big screen to Netflix, AI is a popular theme in Hollywood. Its most exciting developments, however, are in the real world, where tremendous strides in creating human-like cognitive capabilities are tackling real challenges.
The goal is neither self-awareness nor autonomy. Rather, the forefront of cognitive computing is advanced systems that learn at scale, reason with purpose and interact with us naturally. The aim? To help humans make better decisions.
Digital images, sound and sensory data are just a few forms of unstructured ‘invisible’ data that computers can’t understand, and end up clogging resources and threatening growth. Enter cognitive systems.
Just as the steam shovel helped us solve the problem of scaling human capabilities, IBM’s cognitive system Watson enables us to use ‘invisible’ data and scale our human expertise.
But how does Watson see ‘invisible’ data? It uses natural language processing (the process to help computers talk to humans) and machine learning to reveal insights hidden in data. Watson understands, reasons and learns from the data, then interacts with humans to help us make decisions and scale our work.
Watson can understand data at astonishing speeds and volumes. In fact, it reads 800 million pages per second. It can reason to form hypotheses, make considered arguments and prioritise recommendations to help humans make decisions. It can look at a photo of a mole on a person’s skin and provide a hypothesis on the possibility of cancer, helping doctors personalise approaches to treatment.
And it never stops learning. Watson is trained, not programmed, by experts who enhance, scale and accelerate their expertise, and therefore Watson gets better over time.
Let’s look at InkWell, a program that writes haikus with your words and combines them with noted writers’ works to create a new poem. To do this a computer must understand emotion, tone, personality and word selection – something computer systems previously couldn’t do. But programmers trained InkWell with Watson’s Tone Analyzer and Personality Insights to analyse the words for emotion, word selection, personality and tone. The result is a perfect haiku that conveys emotion and tone, demonstrating how a cognitive system can understand language beyond statistics.
But where is AI heading? Right now AI is more about people querying machines. My dream is that Watson will ask us questions, giving computers abductive rather than deductive reasoning skills. Abductive reasoning will lead to conversation and dialogue with humans. And that in turn will lead to more creative thinking, because machine learning means cognitive computing systems will become smarter over time on their own. We’re on that path now, but much work is ahead of us.
In the meantime, we’ve given The Drum independent access to Watson to glean insights that have helped inform some of the features in this publication. But the questions is, what would you do with Watson?
David Kenny is general manager of IBM Watson which was used to help create the latest edition of The Drum.