Bad taste and bad copy
The column Computing refused to print
Somewhere back I have ancestors from Wales — a place where, as everyone knows, you need half a pint of phlegm in your throat before you can pronounce place names.
Sadly I didn't inherit the corollary trait, a sing-song voice — vital not just for getting your larynx around multi-syllabic town names but to entertain while struggling to get a sentence out: "So, you want to get to Phwoarrtaffytottiebuyaleak, boyo? Just climb across Snowdon and head west. The mountain can be windy, like, but it'll blow all that spit out of your hair."
My voice, instead, is more on a musical par with a waste disposal unit. But I'm not complaining and one day you might even hear my grinding tones when text-to-speech systems become the norm. Indeed, it might happen sooner than you think, given that text-to-synthesised speech is already available as server software, while PC voice recognition systems are selling in their millions now.
In short, the industry is of age and even has its own show, Voice Euro taking place in London next month. However there will be no humans admitted — the computers did discuss it but decided people were far too primitive, prone to dangerous throat-clearing habits and unable to converse intelligently at 112kbps.
Stanley Kubrick, of course, envisaged talking computers in his film 2001: A Space Odyssey, though for once it looks like the prediction is coming true ahead of time. Hal and well met, you might say.
IT guru Nick Negroponte, of the Massachusetts Institute of Technology, has also long forecast voice recognition systems which, aligned with the trend towards miniaturisation and mobile comms, will soon have us barking phone messages into our cufflinks, Dick Tracy-style.
But I can already see a few pitfalls. What, say, if the command "Beam me up, Scottie" gets phonetically garbled to "Seam up me bottie?" You could easily get a few nasty volts up the jacksie, what with long-range satellite lasers and private surgery.
In fact, the more you think about it, the more worrying the technology becomes. I used to have an ageing neighbour, for instance, who loved a good natter with the television and was convinced newscasters addressed her personally. Yet once Web TV becomes widespread, offering digital interaction, what's to stop viewers having a chat with newsmen via the set's in-built speech recognition system?
"'And over now to Trevor McDonald with News at Eleven."
Trevor: "Good evening. Tonight's headline — Concorde crashes on Buckingham Palace.."
Rose of Ealing: "Aw my gawd Trev. Woz 'er Majesty in?"
Trevor: "No Rose, the Queen was in Balmoral at the time."
Fred of Hornsey: "I've got an aunt who lives in Balmoral. She reckons the Queen has taken to god big time, so much that she's having a fling with the local pastor who's now dumped his missus."
Rev Spooner, junior, calling from Oxford: "Well, Christianity is a different lay of wife."
Trevor McDonald: "Er, thank you for your contributions viewers, though Dr Spooner I think you need to upgrade your speech recognition software...meanwhile, back to the crash."
You can see where it's going, can't you? It will only be time before the gimmick industry also swings into action. If mobile phones no longer need keypads thanks to voice recognition, why not house them in plastic banana casings if demand is there? Prince Charles would doubtless be an early patron, given his fondness for talking to all things vegetable. Camilla Parker-Bowles might conversely prefer to whisper her sweet nothings to a turnip, again in keeping with instinctive habits.
Before long, it will be quite normal to talk to trees sporting the BT phone kiosk logo, or have some Lan-linked supermarket lettuce home in on your loyalty card buying patterns and then purr seductively from the rack: "You can do anything you want with me. I'm really fresh." Evans the Voice has spoken. Be warned. ®
Sponsored: Hyper-scale data management