Arrogant man applies for job, uses machine learning, gets kicked out of office
(This will make more sense in a minute)
Back in 2005 I was looking for a job and I saw one at a startup in Austin that was working on a tech to improve images from cell phones. Well that should be an easy one. I happen to know a thing or two about THAT, I said, patting myself on the back super duper hard.
They were looking for an "image specialist" to work under some big shot guy who had written some papers on it. I do not remember the details, also I don't care.
But this was a fascination of mine already, and I'd built this little engine called iStill that would pull images from a webcam and improve them based on how other images from that same web cam had been improved.
I had the machine try to improve images, and then I'd pick the ones that were the best, and it would store that info for later. That's a high level view of how it worked.
Today we'd call that human-guided machine learning but at the time I called it whatever the hell I'd cobbled together in SuperCard and C. I set this up on my web server so anyone could submit a photo and it would automatically enhance it and send it back. These are anonymously submitted images. Actually, the truck photo is mine, from a $20 webcam.
So then I'd apply that on top of something else I made up, which was kinda *like* image stacking, but again it was greatly improved by info on the camera I had collected from having the camera calibrate on thousands of images before. So the last photo in this set is from the whole package.
Anyhoo, I applied for the job and they wrote an angry email back saying I was lying, I was obviously working for their competition, and no way in hell I would get an interview where I could steal their secrets. Then they told me to never contact them again.
These photos still make me wonder what would've happened if I'd been living in Silicon Valley, or maybe even Austin at the time. Who knows. And it taught me a valuable lesson about job interviews. Don't try so hard.
Huh. It feels like I learned the wrong lesson.
Comments
Post a Comment