POWER > Trends > IBM Research

Cognitive Systems Mark the Third Era of Computing

IBM researcher Hendrik Hamann examines an array of wireless sensors that detect environmental conditions. Photo by Jon Simon/Feature Photo Service for IBM

Imagine computing systems that can see, feel, smell, hear and taste—and then help people make decisions based on those sensory inputs. According to IBM researchers, such systems aren’t all that far-fetched. In fact, according to Paul Bloom, IBM CTO for Telecom Research, it’s likely to happen within the next five years.

IBM’s view is that an entirely new era of computing—cognitive computing—will take advantage of the flood of data available today and how it’s processed. In this introduction to a six-part series on IBM’s 5 in 5—five innovations that will change our lives in the next five years— Bloom describes how cognitive-based systems will augment decision-making processes, whether it’s shoppers buying something online, doctors more accurately diagnosing medical issues or consumers choosing healthier foods.

Q. Could you give us a brief overview of cognitive computing?
We call this the third era of computing. The first one was basically the calculator. The second, which we’re in today, involves programming to tell computers what to do. The third wave is an intersection between neuroscience, supercomputing and nanotechnology.

Q. Is this an attempt to mimic the brain?
The brain is architecturally like a computing system. It has memory, processing and communications, which are the key ingredients of any computing system. But the brain has some very interesting characteristics. It weighs less than three pounds. It uses less than 20 watts of power. It actually operates as a massively powerful distributed processor, and we’ve been working on this type of technology in the laboratory.

Several of the main things that are different about the brain, however, are that it’s event-driven, meaning it reacts to things in the environment, it’s very adaptable and, the real key here, is that it learns. As it receives and processes more information through the five senses, it learns what it and the body should be doing based on the surrounding environment.

It’s been estimated that the processing power of the brain is akin to 100 million processors. Think of Watson, which is an early stage of cognitive computing. It has more than 2,800 processors. Now consider all of the connections within the brain, which have been estimated at a quadrillion or so, with all of the pathways between neurons, synapses and axons. This is analogous to computers and their processors, communication interconnects and memory.

What we’re doing at IBM Research is essentially creating a model of the mind by reverse-engineering the brain. We’ve already modeled the brain of a worm and a bee and now we’re working on modeling the brain of a monkey. Eventually, we’ll move on to humans, and we estimate that by 2018 we’ll have enough compute power to do a pretty good job at that. So, although we’re in the early days of this whole new era of computing, I think we’ll start seeing specific brain functions that can be modeled.

I think it’s important to stress that we’re not building these systems to replace human beings. Rather, this should be seen as an augmentation of human capabilities. These systems will be able to see things humans can’t or see things better than humans can, and the same with the other senses. They’ll be able to collect information from many different sources and use heightened sensing capabilities to, for example, provide improved healthcare.

Jim Utsler, IBM Systems Magazine senior writer, has been covering the technology field for more than a decade. Jim can be reached at jjutsler@provide.net.

comments powered by Disqus



2017 Solutions Edition

A Comprehensive Online Buyer's Guide to Solutions, Services and Education.

IBM Systems Magazine Subscribe Box Read Now Link Subscribe Now Link iPad App Google Play Store