It would be a mistake to think that the interface technologies we're talking about are necessarily new. Speech recognition and voice control, for example, have been around in one form or another since the 1960s, with the first commercial applications appearing in the early 1990s. Ten years ago this was still an emerging technology, with the consumer market dominated by IBM's ViaVoice, Lernout and Hauspie's Voice Express and Dragon NaturallySpeaking. All three applications were largely effective, but required a period of training and reasonably quiet office conditions to work. While it was entirely possible to control Windows, dictate documents and browse the Web, there were too many barriers for speech recognition and voice command to really hit the mainstream.
Today, that's changing. For one thing, speech recognition and voice commands are now an integral part of Windows, having been made native to the operating system with Windows Vista and enhanced for Windows 7. Its key use is as an accessibility tool for Windows users with mobility impairments (though you'll also appreciate it if you ever suffer from bad RSI). Sure, it's a little odd to be literally telling your PC what to do, but it takes less time to adjust than you might think, and while most of us will be faster typing documents than dictating them - as a writer, speaking rather than typing seems to throw up a weird kind of barrier - the system is surprisingly accurate after a minimum amount of training. Meanwhile, the latest version of Dragon Naturally Speaking claims to be 99 per cent accurate, and as, on average, most people speak at around 140 words per minute using it should be faster than typing too.
Even ignoring speech recognition, voice command is due a resurgence. In the past, people have been put off by clunky voice-enabled Web browsers, mobile phones and early PDAs, but as we start interacting with more devices or networked devices within the home or office, where you might not necessarily be sat in front of a mouse and keyboard, voice command begins to make more sense. As is so often the case, games are leading the way. For example, for years console gamers have suffered with second-rate RTS games that were designed for a mouse rather than an analogue stick. Last year's Tom Clancy's EndWar, while not totally successful, changed that by mapping the command of troops to voice commands. Shouting "Squad X attack Target Y" works well within the context of a strategy game, and the technology - if not always the gameplay - stood up.
You might also remember SWAT: Global Strike Team and Tom Clancy's Rainbow Six 3, Vegas and Vegas 2, where you could issue simple vocal commands to your squad. So far it's been a gimmick - and not always an effective one - but as games start to use player to AI dialogue in more and more sophisticated ways, speech recognition and voice control will be the natural way to do it. Expect Peter Molyneux's Natal showcase, Milo and Kate, to be a benchmark, as much of the gameplay involves spoken interaction between you and the titular characters.
Voice command is also becoming increasingly popular for in-car mobile phone, audio and navigation systems, where - for obvious reasons - hands free operation is a must. BMW and Mercedes are already doing great work here, and you can expect the tech to trickle down to cheaper makes and models over time. We might even have it operating our TVs. At this year's IFA conference in Berlin, a pan-European research project, Dicit, was demonstrating voice control of a DVB set-top-box, though the masses of hardware required and the stilted nature of the language used showed that this was still a technology in development, not one for immediate commercial release.