Who hasn’t heard the term “cognitive computing” so far? Most of us in the IT industry most likely heard, read about it, or already started working with the cognitive systems for various practical use cases. Would you like to have a quick taste? Then, keep reading! I simplified this complex topic as much as possible in this article.
The purpose of this article is to briefly introduce to you what cognitive computing entails, its current progress in the industry, the value proposition for its necessity, and my personal observations and thoughts on trends and future plans.
Cognitive computing is made up of a combination of emerging technologies, processes, and approaches in cross disciplines: cognitive and computer science. It is being built on long term research and development work in Artificial Intelligence (AI). Cognitive computing, as a strong extension to AI, is the simulation of our thought processes in computerised models. From a scientific discipline perspective, in a nutshell, cognitive computing emerged by combining related attributes of cognitive science dealing with natural intelligence and computer science dealing with artificial intelligence.
Many systems in the industry and research settings can display knowledgeable behaviour. These behavioural elements are the simulation of human thought processes using computerised models.
The value of cognitive computing can be realised by three major capabilities in computer science:
a) Self-learning systems using pattern recognition
b) Data mining and analytics
c) Natural Language Processing
These three key points can enable the creation of autonomous computer systems that can solve business, scientific, academic, and other problems with minimal input or intervention from human beings.
I ask two critical questions while assessing the merits of cognitive transformation:
1. What is the significance of this era needing new solution approaches?
2. What made it a new milestone and steppingstone in our technological transformation?
These two broad questions pop up three key considerations in my analysis:
1. We are stuck in our technological progress!
2. Emerging technologies in computer and cognitive science can offer new capabilities to extend our current capabilities to produce novel solutions.
3. Computers finally started mimicking human thinking to some limited extent.
Whilst there are some skeptics, some of us are about to reach the consensus that cognitive computing can be a real game-changer for our productivity and effectiveness to address the growing and sticky world problems.
Transformative leaders keep emphasizing that we need new approaches and more productive ways in agility to meet our growing demands on this aging planet. The prime premise is articulated as “today’s problems cannot be solved with the technologies of the past”. This emphasis resonated with the computer and cognitive scientist hence a strong thought leadership prevailed globally.
The ultimate goal of cognitive systems is the capability to solve complex and time-consuming problems without human intervention. This statement has tremendous implications on our technological development and consequentially in economies. We attempt to bring the technology to such a state that it is almost autonomous in solving our difficult problems faster and more efficiently than we perform as human beings.
From emerging technologies perspective, the key contributions to cognitive computing have been in developing research in deep machine learning, neural networks, big data analytics, data mining, predictive analytics, smart embedded objects, mobility, and natural language processing. We have been dealing with these technologies over decades however now the fusion of these technologies in an integrated manner has started creating synergistic outcomes for the progress and rapid growth of cognitive computing.
Even though there is a massive debate going on in various forums, in my humble opinion, artificial intelligence and cognitive computing are two different fields from developmental and scientific discipline perspectives. However, they are closely and tightly interrelated.
Whilst the focus in cognitive computing is that even though computers learn by themselves, the process is still managed and controlled by human beings. By looking at from the same perspective, AI literature envisages that augmented intelligence of machines can surpass human capability hence they can be autonomous. This premise has been a controversial topic and AI progress rightly or wrongly created fear in society. We need to learn more about the implications of emerging technologies, increase our knowledge experimentally across disciplines, and gain new experiences in mass collaboration to deal with this fear and leverage the augmented intelligence of technological offerings.
These comparative observations in my interactions with the thought leader collaborators in the industry and academia created an interesting sentiment that I wanted to share here. Whilst AI pessimistically scare the people for losing control over the machines, cognitive computing is displaying an optimistic acceptance for its usefulness without dominating and controlling our lives.
To simplify this view metaphorically, AI is like an aggressive boss entering our lives as a fearful controller, whereas cognitive is considered like a docile personal assistant who facilitates the process, correct our errors gently, improve our tasks, solves our problems faster, and makes our jobs and lives easier.
From a technical point of view, cognitive systems acquire knowledge from the vast amount of data turned to processed information using various analytical techniques such as descriptive, prescriptive, predictive and semantics. Big data analytics have been the key enabler of generating information and turning it to knowledge.
The cognitive systems transform content into context using confidence-weighted responses and supportive pieces of evidence. For example, the machine learning algorithms in the systems refine the way they use patterns with multiple repeat loops. This specific way of processing data can enable guessing new problems hence can generate new solution models. This approach is characterised as deep learning which contributes to the functional requirements of cognitive systems.
So far AI, as a growing discipline in computer science, achieved many intelligence marks in various fragmented forms. For example, there is a growing body of knowledge in expert systems, neural networks, virtual reality, and robotics. The challenge has been integrating these fragmented pieces for coherent capability and value propositions. Based on this premise, cogent speculative ideas prevailed that a cognitive approach must leverage AI technologies and integrate them for productive outcomes and value propositions.
From a historical point of view, computers demonstrated fast calculations and data processing. However, they struggled understanding natural language or image recognition. With the introduction of AI, they started having these new capabilities. With integrated capabilities introduced by cognitive computing, computers now can learn in a way mimicking human learning.
From the industry point of view, as you may notice from the media, there is a tremendous focus on cognitive computing. One of the recent remarkable developments in cognitive computing was achieved with the implementation of IBM’s Watson. It is practically used in industry as a support system for example by doctors. It helps collate the deep knowledge around a patient’s condition, history, link them to established journal articles, medical best practices, and diagnostic tools, by analysing these factors in an integrated manner, it rapidly provides informed advice to health professionals.
Other leading technology organisations such as Google, Microsoft, Samsung, Amazon, and Apple made a considerable amount of progress with cognitive computing. For example, most of us familiar with and have been interacting with Siri, Cortana, Bixby, Alexa, and Google Assistance using natural language day by day.
There can be limitless use cases for cognitive computing. Some areas which are leveraging the capabilities of cognitive computing are cybersecurity, insurance, governance, education, climate, financial, manufacturing, and medical solutions.
The future of cognitive computing can be outlined in three key words discovery, engagement, and decision. In the near future, cognitive systems are expected to be able to simulate how the brain truly works. Imminently, they can help us understand complex issues that we couldn’t grasp before, manage complex risks with informed decisions, gain insight for solving our most difficult problems. These can be achieved with cross-discipline studies and tight integration among multiple disciplines towards fundamental shared goals and objectives of humanity.
From a practical point of view, cognitive computing also interrelates with Cloud Computing and uses the Cloud service model efficiently as its enabling hosting infrastructure. It leverages Internet technologies and contributes to further development of IoT (Internet of Things). In the near future, we can see the creation of a new integrated platform of Cognitive, Cloud, IoT, Mobility, integrated with Analytics and Big Data ubiquitously in the workplace, homes, schools, shopping centres, banks, entertainment centres, and everywhere else that we can imagine.
This is a topic close to my heart, with a strong interest in cognitive science as a technologist, hence I serve as an advocate of the topic by creating awareness in all walks of my life. I believe that it is time to embrace cognitive computing not only for its business value but also for potential benefits that it can offer to our society as we have been yearning and fantasizing as depicted in science fiction books and movies for many decades.