What is Cognitive Computing?
Cognitive computing combines a range of Machine Intelligence technologies to hypothesize, recommend, adapt to learn from interactions, and then reason through dynamic experience just like humans. But it is not about replacing humans with machines. It’s about harnessing the combined strengths of both to solve complex problems from ever-changing factors and new information. The “programmable era” of computers invariably will be transcended by cognitive computing systems and applications.
Cognitive computing brings the ability to mimic the human brain, to learn, to understand in context and be more of an assistant than a tool. It understands (by sensing and interacting with data), reasons (generating hypotheses and recommendations) and learns (what the lessons from masses of data are). Cognitive computing is able to take knowledge from different sources, bring it together, and take advantage of large volumes of data and understand it without human involvement in every step. It is this cognitive capacity that will revolutionize computing as we know it. Cognitive computing will also power the Internet of Things, unlocks its true value and infuse intelligence into and learn from the physical world.
Cognitive Computing Stack
Cognitive computing is natural language processing of structured, unstructured, streaming in big data or smart data layers with machine learning for reasoning and learning to generate contextual patterns and associations that enable humans to connect the dots faster and smarter for more informed decisions to drive better outcomes.
What are some of the Applications of Cognitive Computing?
The applications of cognitive computing to business are endless. Some experts believe that this technology represents our best — perhaps our only — chance to tackle some of the most enduring systemic issues facing our planet, from understanding climate change to identifying risk in our increasingly complex economy. The capabilities enabled by cognitive computing will force business leaders to rethink their operating models. While some processes may be refined, others will need to be reinvented, and still others built from scratch. New skills and training will be required, such as developing the ability to design and frame appropriate challenges for cognitive systems and applications. New ways of thinking, working and collaborating will invariably lead to cultural and organizational change, some of which may be challenging.
Cognitive computing systems have obvious benefits in the fields of medicine, finance, law, and education. These systems can also be applied in other areas of business including consumer behavior analysis, customer support bots, personal shopping bots, tutors, travel agents, security, and diagnostics. Cognitive computing user interfaces are contingent on vertical use cases and targeted users, e.g., clinician, knowledge worker, and consumer, within industries such as Life Sciences, Energy, Oil & Gas, Public Sector, Financial Services, Manufacturing, Retail, Collaboration, Customer Services, etc.
Cognitive software platforms facilitates the development of intelligent, advisory, and cognitively enabled solutions. Cognitive applications typically involves text and rich media analytics, tagging, searching, machine learning, categorization, clustering, hypothesis generation, question answering, visualization, filtering, alerting, and navigation.
How will Cognitive Computing impact the Internet of Things?
Cognitive computing is essential to tapping into the full potential and promise of the Internet of Things (IoT). The purpose of the Internet of Things is to connect us more closely with the physical world and share information with us about the tools we use, the homes and buildings we live in, and the cars we drive. But without cognitive computing, the usefulness of this information would be limited by its own complexity and scale. According to the IDC, the IoT network will by 2020 consist of more than 29 billion connected devices. When cognitive computing is applied to the IoT, the result is systems that infuse intelligence into, and learn from, the physical world. This is what can be defined as the Cognitive IoT. In addition to generating answers to numerical problems, cognitive systems can present unbiased hypotheses, reasoned arguments and recommendations. They understand an individual organization’s goals, and can integrate and analyze the relevant data to help individuals and businesses achieve those goals. Rather than being explicitly programmed, cognitive applications learn from interactions with humans and their experiences with their environment in a non-deterministic or probabilistic way. This enables cognitive solutions to keep pace with the complexity, volume, and unpredictability of information generated by the IoT.
The Cognitive IoT enables fuller human interactions with people, fast tracking and extending of human expertise, the infusion of cognition into business processes, operations, products and services as well as enhanced discovery and exploration.
Implications of Blockchain technology for IoT and Cognitive Systems
Although blockchain* technology, which underlies cryptocurrencies such as Bitcoin, has only been explored for a few years, there are a number of important implications for the IoT, smart devices and cognitive systems. Blockchain technology could provide a way to track the unique history of individual devices, by recording a ledger of data exchanges between it and other devices, web services, and human users. It could also enable cognitive and smart devices to become independent agents, autonomously conducting a variety of transactions. Some examples include:
- a suite of smart home appliances that can bid with one another for priority so that the laundry machine, dishwasher and vacuum cleaner all run at an appropriate time while minimizing the cost of electricity against current grid prices
- a vehicle that can diagnose, schedule and pay for its own maintenance
- a vending machine that can not only monitor and report its own stock, but can solicit bids from distributors and pay for the delivery of new items automatically based on the purchase history of its customers
Blockchain networks themselves also have the potential to become independent agents, what has also been referred to as Distributed Autonomous Corporations (DACs). These DACs could effectively supplant systems like banking and arbitration, which have traditionally relied on trusted and centralized human authorities, with trustless and decentralized networks. Examples include:
- escrow services to transfer ownership rights
- electronic couriers to securely transfer sensitive information,
- auto-installation services to verify and push updates to the software governing other DACs.
In order for IoT with its cognitive systems and smart devices to be safe, scalable and efficient, the IoT networks must be re-architected to gradually shift from managing billions of devices to hundreds of billions of devices (as illustrated in the figure below).
“In the absence of a centralized server brokering messages, supporting file storage and transfers, and arbitrating roles and permissions, any decentralized IoT solution should support three foundational types of transactions:
• Trustless peer-to-peer messaging
• Secure distributed data sharing
• A robust and scalable form of device coordination.”
* A blockchain is a type of distributed ledger (which is a consensus of replicated, shared, and synchronized digital data geographically spread across multiple sites, countries, and/or institutions) comprised of unchangeable, digitally recorded data in packages called blocks.
Source: IBM Institute for Business Value (PDF) & http://www.blockchaintechnologies.com/
Some key differentiators of Cognitive Computing Applications
- Context-driven dynamic algorithms for automating pattern discovery and knowledge
- Reasons and learns instantly and incrementally to discern context for sense-making
- Cognitive systems infer, hypothesize, adapt, and improve over time without direct programming
The Role of the Machine Intelligence expert and/or Data Scientist in Cognitive Computing Applications
- Rather than having data scientists creating algorithms to understand a particular business issue, cognitive analytics seeks to extract content, embed it into semantic models, discover hypotheses and interpret evidence, provide potential insights and then continuously improve on them.
- The machine intelligence expert and/or data scientist’s job is to empower the cognitive application, providing guidance, coaching, feedback, and new inputs along the way. As the cognitive application moves closer to being able to replicate the human thought process, answers come more promptly and with greater consistency.