Cognitive Computing: The Hype, the Reality

By on

The phrase ‘Cognitive Computing’ has been slowly, but prominently, edging its way into the mainstream business culture since IBM’s introduction to Watson in 2011. Watson was a unique response-capable computing system built to compete against humans on the popular game show, Jeopardy!. Inspired by this innovative approach, many companies have since followed suit and began to expand their research and investments into Cognitive Computing.

As a result, the “hype” around the subject has grown tremendously and caused a mixture of reactions ranging from enthusiastic to skeptical. In their presentation, Beyond the Hype: Cognitive Computing and Your Business, Your Job, Your Life, Sue Feldman and Hadley Reynolds do a fantastic job outlining the pros and cons associated with Cognitive Computing and give insight on how we can better adapt to its transition into modern society. Feldman and Reynolds are the co-founders and managing directors of the .

According to Feldman and Reynolds, as Cognitive Computing continues to evolve so does the impact it has on our jobs, devices, personal life, and the economy as a whole. Despite hearing things like McKinsey’s projection that state that 45% of all work that people are paid to do today could be automated with currently available technologies and that up to 90% of certain job classes will eventually be taken over by robots, it’s important to factor in all the positive aspects that have and will come of Cognitive Computing as well.

 “It is already making major breakthroughs in industries such as healthcare/healthcare insurance, banking and finance, lease management, and even assistance for the blind,” said Feldman.

Cognitive Computing is now being widely used within the healthcare industry to help better diagnose and treat cancer patients. A prime example of this discussed during the presentation, is being modeled by .

“This unique system is able to identify the type of cancer a patient has and then produce accurate results regarding the best possible treatment plan,” remarked Reynolds. The example brought up in the presentation consisted of an elderly man with a tumor asking questions about his illness, while also providing facts about his current physical health. The machine was then able to “root through” thousands of medical records, journals, and genetic profiles to find the best possible matching treatment for this specific patient.

The machine is also capable of finding CAT scans, clinical databases, X-Rays, regulations, and different medical insurance policies. The delivered results have proven to be extremely accurate and precise, thus making the machine a dependable source for all hospital staff. The history of medicine is a very vast subject that contains an abundance of different conditions and outcomes per each patient. Without a machine like this, human beings would never be able to produce as accurate or as well-researched solutions, said Feldman and Reynolds. “Essentially, the machine is capable of making better medical decisions therefore we are now capable of saving more lives.”

Another way Cognitive Computing is breaking its way into the Data Management industry is via Corporate Banking. Companies like CustomerMatrix are now implementing Cognitive Computing into their business structure to help bankers make strategically wiser investments.

 “These types of banking investment advisory systems are capable a wide range of useful functions. In order to make the best financial decisions, you need have vast knowledge of an assortment of different aspects including different companies, the news, marketplace, and the economy as a whole,” they said.

 “You then need to be able to match opportunities that may arise with the future business goals of your company,” said Feldman. With the use of a banking advisory system, this is all capable now. These systems give you a mental model, so to speak, on what will potentially work and not work with your financial institution. With this mental model, organizations then have the ability to much more accurately approach possible targets for investments in the future.

 “This unique application works continuously to monitor news, compare internal and external banking relationships, and then deliver alerts and insightful recommendations on where and how to approach the next best investment,” said Feldman.

Companies are also beginning to use Cognitive Computing in the Land Lease Management business as well.  Feldman discussed a unique example of the company ABBYY’s use of the new technology and how they use it effectively organize and sort through digital documents. As discussed, a Land Lease Management company generates a lot of paper per day. However, now that all that paper is being digitalized, “it’s important to that we have a reliable application in place and that can efficiently sort through all of it and produce accurate results,” said Feldman.

With an application like this in place, internal staff can easily find specific riders, leases, regulations, and business goal models. Using this method is quicker and more precise than having individuals try and sort manually through documents on a day-to-day basis. Saving time on this saves the company as a whole and increases overall revenue.

The last example brought up in the presentation by Feldman was how Cognitive Computing is now being integrated to better assist the blind. It helps the blind successfully navigate the physical world more effectively even if they have never been to that location before. It does this in quite a unique way too. It actually combines predictive models on how to navigate a physical space with voice recognition technologies. That’s not all that is applied, though.

“This advanced Cognitive Computing system combines an entire range of different technologies to perfect its precise nature,” remarked Feldman. Some others used include Machine Learning, computer vision, bar code readers, radio waves, dead reckoning, analytics, game theory, and various other dynamic and probabilistic models. This trustworthy system allows for the wider world to become much more navigable and opens the door of independence significantly for the blind.

Although the use of the Cognitive Computing can be extremely beneficial, it’s important to know that there is both a time and a place for it. Cognitive Computing extends computing to whole new range of different problems said Reynolds. The types of problems presented tend to be much more complex and human-like than the average non-cognitive system. These problems tend to comprise multiple different variables included, shifting data elements, and an ambiguous nature.

“There isn’t just one right answer that can be given,” said Reynolds. “In fact, multiple answers are preferred when using Cognitive Computing applications. The right answers then depend on much more accurate variable related to the context, what its purpose is, and who is presenting it.”

 All of those things are factored that need to be looked at when considering whether or not to use Cognitive Computing. If the results needed are repeatable and predictable then it’s probably best to consider other options. This also stands true if all the data is structured and numeric, when interaction isn’t actually necessary, and when a probabilistic approach is not desirable. If the current transactional systems are fully adequate and doing their jobs efficiently, said Reynolds, the use of Cloud Computing should potentially be reconsidered. “The value really needs to be justified in terms of cost, competition, and or improved productivity.”

As stated by Reynolds and Feldman, it is important to consider all the trade-offs when deciding on whether or not to implement a Cognitive Computing application. For example, does your organization want to lower the chances of returned exact matches or does serendipity work for your business model? Another tradeoff that should be heavily considered is speed of response. In how timely of a manner do you need the results for customers and or internal staff? Other trade-offs that most certainly should be considered include the impact of the data, thoroughness of type of data, and also the type of use (e.g. trend analysis, risk alerts, customer interaction.) “All should be carefully considered in the decision-making process.”

As Cognitive Computing continues to expand across the business world, more challenges will undoubtedly present themselves. It’s not often “the easiest pill to swallow” and can be hard for some to view it from the same perspective as enthusiastic computer scientists. Challenges will present themselves in both the marketplace and within the general population, said Reynolds.

In terms of the marketplace, Cognitive Computing is still fairly new so there is still going to be a lot of confusion. Trusted resources on the matter are not readily available, the overall terminology is still very unclear, and there aren’t yet a high number of skilled workers in the market relevant to the subject matter. It’s certainly going to take some time to fully accept this new genre of computing.

There are also of social and legal issues that stand in the way of accepting this new technology, said the presenters. As Cognitive Computing continues to expand, it will undoubtedly cause a lot of fear and unrest. The future is unknown when it comes to the impact this type of computing will have. Many individuals are concerned about their jobs and worry about the disruption this will cause. “There are also many concerns about data ownership, data privacy, and control of its use,” said Feldman. A lot will be questioned in terms of humans being replaced as a whole by robots. These are some questions and concerns the world just isn’t ready to face or answer quite yet.

Cognitive Computing has the potential to unleash a lot of benefits in society. However, we must keep in mind that there is always going to be a time, place, and industry to use it in. Feldman and Reynolds covered a host of different issues regarding Cognitive Computing in their presentation, but they also said it’s up to us to be responsible and properly analyze our job roles and how we can overall improve and or perfect them. As Cloud Computing progresses further into the future, more challenges and concerns will undoubtedly evolve. It’s then our job as humans to find the best way to adapt and accept this new and life-changing technology into our everyday lives.

Here is the video of the Smart Data Online 2016 Presentation:



Register for the Smart Data 2017 Conference Today (in San Francisco Bay Area Jan 30 – Feb 1, 2017)

Leave a Reply