from Ricardo Silva

I have not interacted with Sam as much as others here, but he still managed to make a strong impression on me.

The first time I talked to him was during the reception of ICML 2006 in Pittsburgh. As he met others during that event, he couldn't help but talk enthusiastically about some neat algebraic trick he had just learned. Right there I realized that taking pleasure in sharing was typical of him. After all, for years I had kept his notes on matrix identities in the desktop area of my computers!

We interacted most when I provided some help in the organization of ICML 2008. Just a few months before I was done helping with the 2007 conference, and I wasn't really expecting to do it all over again so soon. That with no effort he convinced me to participate again speaks volumes on how easily he could motivate people.

One of the last times I remember seeing him was in the Newton Institute in Cambridge. He gave a talk was about searching for patterns in astronomical data [http://www.newton.ac.uk/programmes/SCH/seminars/042911001.html]. One thing we all know about astronomical data is that it can be truly astronomical in scope, and searching through it is a daunting problem. But with Sam presenting it, it just sounded like a lot of fun.