What Speech Does for You

Its interesting to see the ideas people have regarding speech recognition. Alot of people will have definite ideas about whatever application they used. For example, there is a cinema application - a 1-800 app. You dial in, and you're supposed to say the name of a film and then it will locate the cinema you'd want to find..

I was talking with someone today about this application, and they pulled their phone out and showed it to me. The dialog begins with an advertisement.

Suppose I were to talk to you, one day - and the first thing I said was a message that my sponsors required me to say? Thats how we took it. And then, the application wasn't location aware. It would return a list of locations. And your eyes are glazing over . Places you have no interest whatsoever to go to, start rattling off the list one after another.

When we speak with people, the assumption we make is likely that the person we're talking to is aware of where they are. Thats an important assumption. Directions are given relative to that person's position. If you spoke with someone and asked them for directions, and they started you off in another city you weren't even standing in - it wouldn't make alot of sense to you.

Speech does alot for you, it can route your call and locate companies in a huge directory for you - it can whisper turn by turn directions to you in your car. And in your life - human speech is a huge part of living. You remember your child's first word, the vows of your marriage, or a memorable speech. You even remember the quiet phrases and sentences that go along with experiencing natural beauty - things that a friend might say to you, by really not saying anything - two legs hanging over the edge of Half Dome.

But Speech can't do much for you if it isn't able to interact with you, tolerate error, and be able to figure out your position. Why, then, would someone write an application that exists to tell you an ad? And then you would use it once, and never again - when it doesn't work.

And whats interesting too, is how vicious a person can be about these types of applications. Its almost as if we invest a bit of ourselves in every conversation, and we're upset if we are dissed or the conversation becomes strange. Every sentence relies on a set of hidden, mutually shared but often unspoken assumptions regarding how that sentence will fit into the conversation or how one will acceptably respond.

Failed Speech Applications become a symbol of frustration , a way for a person to vent their feelings about an increasingly technological and impersonal world. It doesn't help that banks, and such designers as the film hotline - are intent on simply not letting that person speak to a human being.

But one day Speech will be natural. There's a neat film out there called 'Code 46' and speech is a part of the film - its set in a kind of unsettling view of the future where humanity is constrained to live in the cities, and weather a harsh sunlight of day by operating mostly at night. Each person has a "palabra" - a codeword that they can speak that grants them access to their applications - all of which seem to follow them everywhere - they open doors, start cars, purchase items using their credit card - all by this "palabra"

And , coincidentally - its a future in which genetic engineers have designed viruses that can alter human thought.

And in which English has become part Spanish, part other languages. Maybe there will come a point in which these applications will be able to do things for us that are truly useful. I wonder what the first real application will be, that can do that... is it banking?


M@ said…
It's just a matter of adding more complexity. Interesting that scientists are building sensors to be placed in the ocean that operate on their own and "scavenge" for energy by consuming microbial life. It's almost a new species....