IBM is fond of Watson and speaks about its computer system as if it were a real person. The machine itself (it lives in Astor Place, NY and in San Francisco too) is built to mirror the same cognitive learning process that we humans have i.e. it makes decisions in the following sequence: observe, interpret, evaluate and then decide.

The computer understands ‘meaning & intent’

One of the core concepts driving Watson is its ability to interpret questions spoken in natural language. The firm’s latest developments inside the big blue glass cabinet see Watson and the IBM-branded Watson Developer Cloud service gain additional intelligence in the form of advanced language understanding as well as speech and vision services.

This is a computer that is capable of extracting meaning from unstructured text, video, photos and speech. IBM has subsequently constructed cognitive APIs (software connection points) and tools for software application developers to code Watson intelligence into new applications.
Is that interesting? Well look at the facts. Inside two years of development we have seen Watson go from one API and a limited set of application-specific deep Q&A capabilities to more than 25 APIs powered by over 50 technologies.

“Since introducing the Watson development platform, thousands of people have used these technologies in new and inventive ways, and many have done so without extensive experience as a coder or data scientist,” said Mike Rhodin, senior vice president, IBM Watson. “We believe that by opening Watson to all, and continuously expanding what it can do. [IBM is introducing advances in services that enable cognitive applications to understand the ambiguities of natural language in text].”

The IBM Watson Natural Language Classifier enables developers to build products and applications that understand “intent and meaning”, finding answers for users even when questions are asked in varying ways. The computer will even hold a conversation tailored to the individual style a person uses — whatever their personal jive.

“IBM Watson Visual Insights for the first time allows developers to build apps that reveal insights from social media images and video. Available initially as an experimental service, this API applies reasoning to the content of images to deliver deeper insights, assess trends and patterns and get a more comprehensive view of what users are communicating to get the big picture,” said IBM, in a press statement.

There are also speech-to-text and text-to-speech capabilities here. Watson speaks multiple languages including Japanese, Mandarin, Spanish and Brazilian Portuguese, with additional languages being continually introduced.

IBM’s big play here is a suggestion to software application developers to embed Watson APIs in any form factor from mobile devices, cloud services and connected systems. IBM is also previewing IBM Watson Knowledge Studio where the company will open up its machine learning and text analytics capabilities in a single tool.

Is IBM Watson amazing? Mostly, yes. Will IBM go on and on and on about Watson at its forthcoming Insight symposium? Mostly, yes. Is IBM Watson starting to get out of its more experimental adolescent years to become a tangible ‘cloud service’ that any businessperson might actually touch? Mostly, yes. Is the Watson playing Jeopardy! YouTube video still online to watch? Mostly, yes.

Can human intelligence still beat Watson at the end of the day? Mostly, for now, yes.

Source: Forbes.com