Database Reference
In-Depth Information
that have been built this way—even, to an extent, companies like Cloudera,
where they had already built a lot of the hard technology when Yahoo! spun
out Hadoop. Someone has to be footing that bill, andVCs do not have the risk
appetite or patience to let you try something for three years. They'll let you
try it for a year, and they'll probably still keep funding you over the next two
years, but what's going to happen is you'll bleed through your cap table, and
you're going to wake up and maybe finally have a success and realize you've
sold 90 percent of the company. Oops.
Gutierrez:
What do you look for when you're hiring people?
Jonas:
It depends a lot on the role. The first thing to find out is if they are an
asshole. Life's too short to work with assholes. At MIT we used to talk about
how it would take freshman a while to de-frosh. They would come in thinking
they were the smartest person ever because they grew up being the smartest
person they knew. Then they get out into the real world and realize that no,
they're not. They have to have that arrogance beaten out of them. There are
some people who never lose that. There are some people who very much
think that being smart is an excuse to not have interpersonal skills. And the
world is just too collaborative for that to work anymore.
My cofounder Beau Cronin made the comment the other day when I was
talking about an academic who I was working with who was a little bit
difficult, and Beau said, “The nice thing about doing a startup is you get to say,
'Nope! No! No! Do not talk!'” In academia, because of the way the incentive
structures are often set up, that's not as much the case, so you might end up
working with difficult people.
At P(K), we evaluated a lot of really smart people that just weren't a good fit.
Startups spend too much time talking about culture these days, and often
culture is a euphemism for “not exactly like me.” Which is a terrible way
to look at culture. What really does matter, and how we looked for fit was
by asking ourselves the following questions about them: Are they excited
about the same technical problems as we are? Are they excited about being
collaborative? Do they like sharing their successes and failures? Do they
possess some degree of appropriate humility and understand why it's impor-
tant? Finding the right person with the right fit is hard, especially in the machine
learning and data space. But that's the most important thing—making sure
they are a good fit.
Obviously, a strong math background is necessary. If I have to explain probability
to someone, it's going to be a really hard slog for everyone involved. I would
rather take someone in the top 20 percent of quantitative skills who also is
a great software engineer over someone in the top 5 percent who doesn't
know how to code. The quantitative finance model really popularized the
notion that raw cognitive talent is all that matters. This is the D. E. Shaw and
Renaissance Technologies model of “We're going to take people who have
