Apple Tries to Make Voice Recognition a “Sirious” Feature

The “Best Feature” of the iPhone 4S Must Move beyond a Party Trick


Since the iPhone 4S was announced last month it’s been interesting to hear people’s thoughts on the phone’s “voice assistant”, called Siri. Some have used it. Others have not. But most people already have an opinion about Siri. And in most cases it tends to be rather cynical. This isn’t surprising, considering that voice recognition technology has never really lived up to its own hype. Even now there are numerous voice applications available to download on Android, BlackBerry or Symbian. Yet none of them have really altered the way we use mobile phones. So what makes Siri any different?

I would argue that Siri’s main difference is that the application is built into the handset. This means it takes full advantage of the iPhone’s hardware specifications. And Apple has “opened” the device to Siri so that inbuilt apps can respond to voice commands. While this has raised some security concerns about access to the device when it is supposed to be locked, it works better than downloadable voice software. Take, for example, Google Voice which must support a range of Android devices from entry-level phones through to the high end.

I was impressed with the Siri demo during the unveiling of the iPhone 4S. Voice recognition demos are notoriously risky to do and a lot can go wrong. Microsoft’s embarrassing show of Windows Speech Recognition during a demonstration of Windows Vista at a Microsoft financial analyst meeting in 2006 is well known. Nevertheless, Apple was confident enough to show off Siri as the world was watching. What’s more, it had the courage to build up the announcement and describe Siri as the phone’s “best feature”.

Yet I was surprised at the prominence given to the new voice recognition software. Siri is still in beta testing, so by definition isn’t complete. Testing will prove particularly difficult for voice technology, given the infinite richness of language. And the software will have to content with accents and background noises. I’ve already heard of a few customers disappointed with Siri’s inability to process certain accents, such as Scottish.

At the moment, Siri seems to be gaining ground as a party trick, responding to deliberately stupid questions. Interestingly, a recent poll by PYB Gadget of iPhone 4S buyers at Apple’s Regent Street Store in London indicated that a third thought Siri was a gimmick. The danger is that if this perception continues, Siri might never be taken seriously. This would be a shame, not just for Apple’s product, but for voice technology in general.

Of course, I wouldn’t underestimate Apple. The company has turned touch screens into the standard interface for mass-market mobile phones. It may have found a new, fun way for people to interact with their devices. There’s no doubt there’s a niche market for voice technology, especially in Apple’s hands. But for me it seems unlikely to replace the touch screen or keypad. Its role may be more as a supplementary interface in very specific environments.