You did pretty well at avoiding squirm-induction, I think - I'm glad you tried to put yourself in my/his shoes.
For the record, one of the best ways to avoid inducing squirms and induce a strong warm fuzzy feeling is to talk about what the one has inspired you to do - i.e., "PG, it's as a result of reading your work that I learned Lisp as my first programming language and started hacking on open source", even if you never founded your own startup, is probably going to induce a stronger warm fuzzy feeling than any amount of praise directed at particular essays you liked.
I find that sort of causality hard to draw, though. My choices are either to make it a hard correlation with a minor accomplishment (I tried programming in lisp, and came up with these few programs I found interesting: http://akkartik.name/lisp.html) or to make it a more fuzzy correlation with something bigger ("Everything I have done and everything I will do, would have been less good if you hadn't written what you did." http://akkartik.name/about. Or "When I write I try to imagine how you would react to each sentence." http://akkartik.name/?f=Me)
Neither seems satisfactory, especially since other people influenced these accomplishments as well. Let me attempt a middle ground -- using examples.
Eliezer, I still maintain an interest in blackboard systems to this day. For example, here's a program I wrote a couple of weeks ago: http://github.com/akkartik/brooks-ruby-warrior/tree/master. Thanks to reading Hofstadter and you and Brooks, I knew when I approached this problem that finding the heuristics to solve a particular AI problem is relatively easy. What is hard is finding an open-ended architecture to integrate lots of heuristics together. I ended up with a faint likeness of an idea others have explored in far harder domains.
PG, as I hinted above, inspired me to come to silicon valley 2 years ago. I do what I love. That was not something I could say a few years ago. Agh, this is more squirm-inducing. Here's a program that came out of conversations when I spoke to PG at a YC dinner in Boston last July: http://akkartik.name/newsflash. PG induced me to leave the problem of generating recommendations behind, at least for some domains.
I don't have more significant contributions. Hopefully that will change.
For the record, one of the best ways to avoid inducing squirms and induce a strong warm fuzzy feeling is to talk about what the one has inspired you to do - i.e., "PG, it's as a result of reading your work that I learned Lisp as my first programming language and started hacking on open source", even if you never founded your own startup, is probably going to induce a stronger warm fuzzy feeling than any amount of praise directed at particular essays you liked.