Listen To Woz, And Perhaps Type Madly

Shawn King of The Mac Show Live talked a few days ago with Apple co-founder and knowledge-omnivore Steve (The Woz) Wozniak. Shawn graciously agreed to post the interview, formerly Quicktime only (downloadable or streaming), as an MP3 file -- so now most anyone can listen. This is an interview worth listening to: Woz talks about his lifelong motivations, his years with Apple (up to the present), OS X, the Newton, and what the future holds for him. He also talks about building TV jammers and the only prank he got caught for in high school, one which might not fly so well right now. (The interview starts about 55 minutes into the show, and lasts for nearly an hour.) What's this got to do with typing madly? Well, since Shawn's program is all-audio (no pictures, and only the barest explanitory text), it's a lot less useful to those on text-only or just-plain-slow links than it could be. Read on below for your chance to change that with just a few minutes of your time. Update: 10/20 20:43 GMT by T : Thanks to everyone who's volunteered to transcribe, and to the several alternates who are already in line! No need for more voluneers right now :)

Transcribing an hour of text takes a long time. But if you (yes, you!) are willing to transcribe a 3-minute (well. 3:15) chunk of this interview, I will spend my putative day off gluing chunks of interview together. Shoot me an email with "WozScript" in the subject if you'd like to participate, and I'll give the first volunteers (it shouldn't take that many) a randomly-drawn three-minute segment to type up, as well as more instructions on how to format it. No compensation except your name in lights, and the knowledge that lynx users everywhere appreciate your efforts. I'll update this story if and when the transcription is complete. (And if anyone can suggest a good Quicktime audio --> .ogg converter, Shawn and I would both appreciate it.)

Comments

How about algorithmic voice transcription?

[Knobby]

I thought we were suppsed to be geeks? Come on guys. Transcribing an hour of audio into text should take one line to fire up a voice recognition code, and no more time than the wall time required to listen to the interview..

There's a huge group of people hear who would love to see a free variant of *NIX that can compete with windows for the desktop market. I think that before this happens you're going to need to sit down, spend some time in your local technical library researching voice, image, pattern recognition algorithms.. I'd love to be able to type:

voice2text -mp3 woz.mp3 woz_interview.text

and get a transcribed version of a speech, or lecture notes.. How about combining this with an answering machine app to record and transcribe messages then send those messages to the IMAP server or atleast place them in a searchable database for future reference..

This is way off-topic but it's something I started thinking about when rumors bagan floating around concerning Apple's iPhoto app.. I thought it would be pretty incredible if Apple could piece together an app to project photos onto an empirical basis set and then use the coefficients from that projection to sort images.. Think of it like a generalized face recognition routine only more useful..

I guess the point I'm trying to make is that gnome and kde are nice, but to take over the desktop market you we really need to crawl out of the box, and burn it to the ground!