By Michael Karasick
As a computer scientist and director of one of IBM’s global research laboratories, I find it fascinating to trace the repeated patterns in the history of computing. Typically, the Next Big Thing spends years in incubation, either as military initiatives (the first electronic computers), consumer phenomena (the PC) or science projects (the World Wide Web). But, ultimately, these advances are adopted by business enterprises, where they’re deployed at massive scale to make organizations more efficient and effective—and, ultimately, to drive growth and dynamism in the global economy.
That’s going to be the pattern with cognitive computing, too. The technological threads that make up cognitive computing began to unspool in academic and government laboratories. Most of them have been incubating for years. IBM demonstrated the power of the latest advances in our Watson machine, which was created by IBM Research scientists, when it defeated two grand-champions on the TV quiz show, Jeopardy! But now we’re taking the technology to Main Street—where it has the potential to help transform industries ranging from financial services and energy to consumer packaged goods and retailing.
Today, at IBM Research – Almaden, we’re conducting a Cognitive Systems Colloquium aimed at forging a shared research agenda among companies, universities and governments, and aimed at demonstrating the power of cognitive computing in the enterprise. Join the discussion at Twitter #cognitivecomputing.
If you want to learn more about the new era, read Smart Machines: IBM’s Watson and the Era of Cognitive Systems.
Helping industries and companies adopt new technologies is a familiar role for us. Think about the Internet. In the mid-1990s, after Marc Andressen and his colleagues at Netscape introduced the first successful graphical Internet browser, the Net began to turn upside down the worlds of publishing and retailing. But it wasn’t until IBM and other leading enterprise technology companies got involved that large businesses across a wide swath of industries began to take full advantage of the Internet revolution.
Today, IBM’s software division is leading the charge for commercialization of Watson, as you would expect. It’s working with healthcare organizations to bring cognitive capabilities to bear not only on the practice of medicine but also on improving the way the healthcare payer systems works. The new Watson Engagement Advisor service helps businesses improve their interactions with consumers. And, announced just last week, the Watson cloud ecosystem makes it possible for businesses to embed the technology in an infinitely wide variety of services. The first three targets: shopping assistance, medical equipment procurement and personal health management.
But cognitive computing isn’t just about Watson. That solution combines natural language processing, machine learning and statistical methods to gather insights from a huge corpus of data and interact with humans in a question-and-answer mode. There will be many more solutions made by combining various cognitive technologies. They will address a tremendous variety of real-world problems and opportunities. In fact, within IBM Research alone, we recently conducted an informal survey and identified no fewer than 42 cognitive projects already underway. Many of them have huge potential in the enterprise.
Let’s explore the realm of marketing, for example. Tech visionaries and marketing gurus have been talking for years about the potential for marketing to individuals rather than larger segments of similar people. It’s called the “market of one.” But only now, through the combination of cognitive computing, big data, social networking and other new technologies is that promise at last becoming a reality.
IBM Research is working with a handful of forward-thinking retailers to push the envelope on marketing to the individual. Through the use of cognitive technologies, they can gather and organize vast amounts of information about consumer shopping and buying behavior. They can find patterns in the data that help them understand how consumers relate to their offerings, and with those insights they can do everything from planning their buying, inventory levels and staffing to crafting marketing campaigns and redesigning Web shopping experiences.
Consider this scenario: When you arrive at a shopping Web site, you’re greeted by a virtual personal shopping assistant that has a full profile of who you are, what you like, and your shopping behavior—and can combine it with an encyclopedic understanding of the shop’s offerings. You interact with the assistant through a spoken or typed conversation. The assistant will learn from its encounters with you. Perhaps the assistant will know from your history that you like to see demonstrations of products in action, and will offer to show you animated videos when you inquire about certain products.
Or, say you’re in a brick and mortar store with which you have an on-going relationship. The company’s customer-management system sends a special offer designed just for you to your smartphone. The choice of merchandise in the offer and the price are custom-tailored for you—based on what the cognitive system has learned about you through past interactions. The system knows that you like immediate gratification and arranges to have an out-of-stock item shipped to your home from a warehouse on the same day.
The shopping experience is as easy as shopping at Amazon.com—only better, more personal.
Cognitive technologies can also help marketers communicate with individual consumers in ways that will be most effective for them. Marketing e-mails and Twitter Tweets will pay off when they address the needs of the individual who receives them in language that they’re comfortable with. We’re developing a couple of solutions at IBM Research – Almaden that address this situation. One of them, called Mockingbird, analyzes the language and communications styles of demographic groups—millennials, for instance. Then, as a marketer is crafting a Tweet aimed at people in that cohort, the program analyzes what the marketer has written and suggests more effective words to use. Example: “sick” means “good” among teenagers in American youth culture.
The world’s technologists are just scratching the surface when it comes to bringing cognitive computing to bear on businesses. At IBM, we’re looking for partners in a wide range of industries and business functions to join us on this journey. It’s an historic turning point in the evolution of computing. Come make history with us.