IBM has announced an investment of more than $1 billion to establish a new business unit for Watson, its supercomputer system that attracted widespread media attention by winning the television quiz show, Jeopardy.
Watson's feat was to out-perform human champions in a TV game show in which contestants guess questions from answers rather than the other way round. This requires natural language understanding and cognitive skills that come naturally to humans but require vast amounts of processing power when tackled by a computer.
Winning Jeopardy back in February 2011 may have brought IBM Watson fame but translating that to fortune has proved more difficult.
This promo video was posted on YouTube at the time of the Jeopardy success but hasn't propelled the flood of paying customers that IBM might have anticipated. Instead Watson's total revenue from services is said to be under $100 million.
To try to achieve $10 billion in annual revenue within the next 10 years IBM is now setting up a new business unit for Watson. Based in New York City it will be headed by Michael Rhodin, previously senior vice president of IBM's software solutions group, and will have about 2,000 employees.
The investment includes $100 million for the Watson Developer Cloud, which IBM opened up to external application developers last November.
According to its website, IBM has opened up Watson as a development platform in the Cloud:
to spur innovation and fuel a new ecosystem of entrepreneurial software app providers who will bring forward a new generation of applications infused with Watson's cognitive computing intelligence.
It explains further that:
The Watson Ecosystem program is designed for ISVs whose business or application can benefit from Watson’s cognitive capabilities, specifically those looking to dramatically disrupt and transform an existing market. Partners can leverage the Watson Developer Cloud to develop their Powered by Watson prototype application.
IBM is gradually expanding the program and is inviting those with ideas for Powered by Watson applications to complete a form for the opportunity to join its community.
What is surprising about Watson and the whole enterprise is that it isn't clear what the breakthrough is that is being marketed. Watson seems to be a big data statistics machine and works because of the scale of the data and the speed of inference. If there is a novel AI technique at work here it isn't being made much of.
A C# SDK for the LUIS, Microsoft's Language Understanding Intelligent Service API is now available on GitHub, taking us another step forward towards being able to build bots that respond to natur [ ... ]