Sam was one of those people that made you feel proud to be doing machine learning, just because he was doing it too.
Memories of Sam:
Giving a talk at Gatsby when I was a PhD student ... so many questions from Sam and Zoubin (I think in those days I did almost think of them as brothers) that I was stuck on one slide for what felt like 15 minutes. Then flicking through the next 10 slides because they'd covered all the material we'd just discussed!
Breakfast with Sam, Ralf and Alice at the end of the NIPS workshops in 2002.
Sam entering the Nonparametric Bayes workshop at NIPS for the discussion to tell all assembled that Bayesians didn't have all the answers. Typical Sam --- I'll miss this!
Being at the table at UAI in Banff when Sam was talking about early ideas behind Neighborhood Component Analysis. Watching him pull the whole table into the idea.
For my ICML talk in Pittsburg, feeling proud that Sam was in the audience. Feeling prouder when he told me he liked it.
Seeing Sam three months ago at NYU. Having a great conversation with him and Yann in Yann's office. Later, seeing Sam on his way home (in a floppy flat cap), but him going out of his way to make sure I got safely to my next meeting.
Sam was always smiling, always joking, and always flowing ideas.
No comments:
Post a Comment