Apple’s Siri – Is the Era of Speech Finally Upon Us?

Even if the technology behind Apple’s new Siri works as advertised (check out Apple’s video demo), there are still a number of big hurdles to Siri becoming an effective enterprise tool.

I guess I was less disappointed by the iPhone 4S announcement than many. No matter what you call it, there's a lot to like, and, and I previously noted, sales are indeed brisk. Dual-core A5. Better camera. iOS5. iCloud. And, very interesting all by itself, the new Siri voice capability. Voice has been the holy grail of mobility for many in mobile over the years. Speech is natural. It would be hard to imagine simpler user interface. And yet, throughout its history, speech recognition technology has been nothing if not problematic. I've used products like Dragon NaturallySpeaking and all manner of command-oriented systems. And I've found that while one usually needs a large vocabulary, high performance (minimal time to recognition), speaker independence, and high accuracy, getting all four simultaneously has never been achieved, at least not in a consumer product. Will Siri, a combination of speech recognition and an inference engine, be any different?

I hope so, but I'm skeptical. Speech recognition is obviously tough. There are numerous environmental factors (including microphone quality and positioning, as well as ambient noise, a/k/a interference), fundamental limitations in the technology itself (accents can be tough, as is supporting the very large number of languages in use today, and what happens if Siri screws up and sends the "wrong" message to the "wrong" person?), and the big one that everyone seems to forget: the "other people" factor.

By this I mean that we really owe it to those near us to hold the noise down lest we cause a disturbance. This is, of course, the essence of good manners, which, OK, has taken a bit hit in the past couple of decades, but which still remain, IMHO, one of the hallmarks of civilization. In other words, just because you can talk to your iPhone anytime/anywhere doesn't mean you should. And, of course, there's the issue of security - some things just shouldn't be spoken out loud, especially in an arbitrary location while surrounded by strangers. One can often, for example, gather tremendous amounts of business intelligence while sitting in any Red Carpet Club. Without conscious consideration of the consequences, Siri just moves the leaks to any random location.

Paranoid? Me? As I noted in my talk at Interop last week, information is the core element in creating value in the enterprise today. Mobility is at its core about moving information where it needs to be and thus gaining competitive and market advantage by compressing time. But if that information is compromised, well, then, what's the point? Some things just shouldn't be spoken out loud.

So, will Siri work? It's not really just a question, then, of cool technology alone.

Join the Network World communities on Facebook and LinkedIn to comment on topics that are top of mind.

Copyright © 2011 IDG Communications, Inc.

IT Salary Survey: The results are in