Google, Nokia and others have similar capabilities, but have not achieved the same level of profile as Siri, which was the only major highlight of last year's iPhone 4S launch. Some operators are having a go too, notably AT&T with its Watson product.
Now IBM could shift the goalposts completely by squeezing its own intelligent voice activated system, confusingly also called Watson, into mobile devices.
IBM’s Watson, famed for its ability at answering questions from the US quiz show Jeopardy, initially ran on many servers, running huge numbers of calculations to show off IBM's prowess in artificial intelligence. But now the firm says power consumption and resource requirements are “dropping like a stone”, its vice president of innovation, Bernie Meyerson, told Bloomberg. It could now be included in high performance mobile devices, potentially providing a sophisticated version of Siri for business users.
That could take Watson away from the rarefied market of data crunching for financial or medical giants, and into more mainstream mobile enterprise applications, once Watson 2.0 makes its debut next year. As Bloomberg points out, that could require some quick decisions by speech recognition specialist Nuance, whose technology underpins Siri but relies on some key IBM patents.
“One day, you will have ready access to an incredible engine with a world knowledge base,” said Meyerson, and mobilizing Watson would contribute to its goal of generating $15 billion (€11.9 billion) from business analytics in future.
Currently the system runs on ten racks of IBM Power750 servers in its Yorktown Heights, New York base and most computations occur on the server side – but the client-end application is still too power hungry for smartphones. That will need to change for release 2.0, and IBM will need to add better voice and image recognition so the service can respond to real world input and be more friendly to non-specialist users. Katharine Frase, vice president of industry research at IBM, explained: “In 2.0, we hope to give Watson more senses. A guy could say into his phone, 'Here's where I am and here's what I see,' lifting it up to take in images of the environment.”
The other Watson, at AT&T, is closer to biting at Siri's heels in the mobile world. The carrier has over 600 patents in this area and is opening up APIs so Watson can be leveraged in third party applications. The operators says its technology has been around for more than 20 years, and it is offering APIs for web search, local business search, question and answer, voicemail to text, SMS, U-verse electronic programming guide and dictation.
Gaming and social media APIs will come down the road. AT&T already makes Watson available via a free iOS and Android app called Translator, which converts speech into another language on the fly, and uses it internally for mobile voice directory search.