>>11294692I'd ask him a question and he'd go on for like fifteen minutes in a really roundabout way. I was at his house for almost two hours but only asked like five questions. It never got boring. He also has this really distinct voice that sounds really nice and almost poetic on the ears. He gave me an "arthritic handshake" as he called it where we shook pinkies.
And in case anyone is really interested he basically told me about how he got into his field and then showed me ADALINE.
His story is basically:
>Kid from smallish Connecticut town>Gets accepted to MIT>Takes digital signal processing class, loves it>Becomes professor at MIT>Develops adaptive filtering at some point>On a whim attends some "artificial intelligence" (new terminology at that point) conference>Becomes enthralled with that idea>Spends months in his free time thinking about a machine that can think>Realizes how to do it>Realizes how to do it cannot be done on current hardware>Realizes that although you can't make a machine that thinks at that point you can make a machine that learns>Uses techniques from adaptive filtering to begin working on machine learning (he wasn't the first but he was definitely one of the first)>While working on this problem another professor passes off one of his graduate students who seems uninterested in his work to Widrow>This man is Ted Hoff>After explaining his work to Hoff, the two work together on the blackboard until they figure it out>What they discovered was the LMS algorithm>With this Widrow develops ADELINE and MADELINE>At some point he becomes a professor at Stanford>Then he retires>Then he turns 90 
