Wednesday, December 21, 2005

Thoughts on Text-to-Speech and More

If you're like me, your thoughts come faster than you will ever be able to type.

Why is it that we still don't have truly GREAT voice recognition software? This is an advancement that the net is ready for. We have so much processing power. So many great advances in software development.

I find myself with an almost insatiable need to record all my thoughts, but my hands are never fast enough to record them -- and I have yet to figure out how to attach a ram chip and SATA interface to my gray matter.

I also have this strange desire to capture many visual moments of my life -- and make them searchable. I am really looking forward to what the folks at MSR are doing, and what their My Life Bits project will bring about.

So much data already exists that we can passively capture, index, and search. And still I wish for a way to capture every word that I've ever spoken, and have a private record of every place I've ever been (not just destinations, but places), all arranged in an elegant visual interactive timeline overlayed on a visual Google Map like interface.

And then, what if we could search one another's voiceprints? What could we know from one another--talk about the speed of innovation!

If there were a convenient way to do it, I would record virtually all of my waking hours with an always on digital visual capture device. Today's digital cameras allow me to do most of this, but not all of it. And certainly not in an always-on mode.

Is really good text-to-speech, always on, geolocational data capture MEETS always-on audio capture and always on photo/video image capture really that far away?

I don't think so. In fact, I think we have a lot of the above in the "mostly-on" (vs. always-on) devices/capabilities available today. So if the data is already there, whose going to give us the tech to capture and create intelligent value from it?

What's really interesting to me is all the passive data we can capture RIGHT NOW. Time stamps of phone calls, emails, photos. Geodata of photos, voice calls, cell phone location. Social networked maps of IM, friendster accounts, and email interactions. This technology has come so far already, I can't wait to see what people will build for us, and also participate in those innovations, in the future.

3 comments:

  1. Check out what Steve Mann has been doing over @ UofT with his eyetap devices. He has a permanant visual record of everything he's done for the last decade and even does stuff like erase billboards from his vision before he sees it (the camera sees it first, and his wearable computer does the processing before projecting it on the inside of his glasses/viewfinder).

    Documentary about his work was done a couple of years ago as well, 'Cyberman' I believe was the title.

    I think there is a point of too much information... tough to figure out where that point is though.

    ReplyDelete
  2. Chris, thanks for the comments... I've been following the works of Steve Mann loosely over the past few years now. Its pretty interesting. I'm also seeing more and more wearable computer stuff which is absoultely amazing. I still find the iPod nano and the credit card thin sony cameras to be astonishingly cool. I can't wait to see what's coming down the pipes this year.

    I don't think the issue is "too much information" but rather, how you go about mineing that information that's key. I'm looking forward to working on solving some of these problems in the next 20-100 years. =)

    ReplyDelete