Watson is a cognitive computing system developed by IBM and named after the company’s founder, Thomas J. Watson. The Watson supercomputer is a “question answering” system able to process natural language and other unstructured data such as videos, pictures and text. By text mining vast amount of raw, unstructured data quickly, Watson is able to provide a range of confidence-scored responses and recommendations to questions about highly complex situations. In addition, Watson can identify new insights and patterns hidden within these large volumes of data. Currently, Watson is one of the few systems in development that can closely replicate how humans think and reason. With ability to “converse” with humans about a variety of complex subjects, Watson represents a new paradigm in artificial intelligence and machine learning.
Watson in comprised of the following hardware and software:
- 90 IBM Power 750 servers
- 2,880 POWER7 processor threads
- 16TB of RAM
- SUSE Linux Enterprise Server 11
- Apache UIMA (Unstructured Information Management Architecture) and Hadoop frameworks
- IBM DeepQA
In total, Watson is capable of processing up to 500 gigabytes (500 books) of data per second. To put this processing power into perspective, Watson can read up to 15,000 medical research papers in approximately 15 seconds, whereas doctors may only read up to half a dozen per month. It is difficult to underscore the amount of possible applications this technology has in data-driven industries such as health, commerce, financial services, law and education. There are already a number of businesses and industries using Watson in applications for increased customer insight and engagement.
On one hand, Watson seems to have the potential to replace humans for a variety of tasks, especially of the low-level administrative type. Customer services representatives, the administrative assistant at your doctor’s office, and traffic court judges come mind when thinking of jobs that could be handled by cognitive computing systems. However, the reality is that we may never achieve a level of cognitive computing and artificial intelligence that allows us to fully replace humans with machines, and if we do that technology is still decades away.
IBM has seemed to embrace this reality and actually promotes Augmented Intelligence rather than Artificial Intelligence. Their position is that AI and machine learning can better benefit humanity by augmenting, rather than replacing, human knowledge, skills and idiosyncrasies. It is an interesting position on IBM’s part considering all we have been hearing about for years is how machines will eventual replace humans at many everyday tasks.
Since winning the game show Jeopardy in February 2011, IBM has been attempting to successfully commercialize the Watson platform. Currently, IBM Watson is available as a set of APIs and a collection of over thirty Software-as-a Service (SaaS) offerings for developers to tap into. With the explosion in popularity and application of technologies such as artificial intelligence, machine learning, predictive analytics, and speech recognition, Watson’s time is now.
To Learn More: