Database Reference
In-Depth Information
I read the paper?” If the answer is yes, then I'm more likely to have believed
it. “Am I surprised? What is the entropy on the result?” The more surprising
results, the more thorough I want to be when evaluating the work.
On the industry side, I think that the ability to do software engineering is
something that is very important, but isn't really taught. You don't actually
learn it as a computer science undergraduate, and you certainly don't learn it
as a graduate student. So for me it's very important that someone has learned
it somehow—either by themselves or from someone else. I basically can't hire
people who don't know Git.
There's a universe where companies at a certain scale can afford to hire people
who have tool deficits. For almost anything I've been associated with, that's
not really the case, which is unfortunate because it means that you leave out
some very smart people. You want people to be productive on day 10, not on
day 100. I also really think that the Bayesian approach to machine learning has
incredible legs. I think that it encourages a certain kind of precise thinking. If
you're not doing it, it's very easy to confuse yourself and lead yourself astray.
So I generally look for people who are at least familiar with that part of the
universe.
Gutierrez: What have you've changed your mind about with regard to using
data?
Jonas: I started dating my best friend in undergrad around 2001. We were
together for nine years and then broke up. So I found myself single at 29, and
realizing that I was going to have to learn how to date. I decided that if I'm
going to start dating, I should keep data on it. So I create a dating spread-
sheet, and then went on something like 100 first dates and ended up with this
massive chunk of data. Coming out of MIT, you think you're stupid because
everyone else was smarter than you and you think all these sorts of negative
things about yourself. However, after looking at the data, it turned out that I'm
actually kind of a catch—which was a great thing to change my mind about.
I found that if I could get to the second date, then generally I could get to
the fifth date with someone. There's this initial evaluation process, but I was
generally pretty good at that. So I was actually in a much better position dat-
ing-wise than I had ever thought. You wake up and 30s are around the corner,
and your friends are marrying and you're single, you're like, “Oh, my God—I
missed the boat!” With the dating spreadsheet it was easy to see that no,
actually the data says the opposite. This was this phenomenal. Psychologically
it's so easy to fall into these kinds of anchoring effects where you always
remember the last date that went poorly. You don't think of the previous 10
out of 15 that went well. If I hadn't been keeping track of that data, I think it
my confidence would have taken a big hit, which interplays with all these other
things in my life. So I'm dating a great person now and I don't think that would
have been possible had I not had this epiphany that the data shows that some
people do in fact like me.
 
Search WWH ::




Custom Search