62 AI MAGAZINE
Recipe: Success in AI à la carte
Serves: The entire field
Your time and commitment in helping the field of AI
and the larger scientific community in a number of
ways. The higher the quality of these ingredients, the
better. The exact amounts are up to you — and you
should make sure you are able to balance these with
the rest of your career — but the more, the better.
1. Do Good Work.
This one pretty much goes without saying, but it’s a
necessary step. If you want to be recognized in the
field and if you want to be able to help others achieve
the same, you need to do good work. Good, but not
necessarily great. There are so many incredible
researchers in our field that killing yourself trying to
compete is not necessarily the best way to achieve
success. Strive to do all you can, but don’t burn out
trying to get yet another paper out the door. Find balance in your professional life between the many
things you can find that are rewarding to do and the
many things that can help other AI researchers as
well. In particular, this work can include a number of
service-related activities that will help both the field
and your own career, as I’ll outline.
2. Pay It Forward.
As you move through your career, especially as a
researcher, there are many things you’ll need to do.
You’ll want — and need — to attend, and publish in,
workshops and symposia and conferences. These
events don’t organize themselves! Committees are
needed to help organize and run the meetings, program committee members must be found to review
papers, and so on. If you are attending a successful
meeting, you owe it to all the people who have volunteered to do these things. As the recipient of the
gift of their time, the best way you can repay that gift
is to donate your time to future events. When you’re
asked to be on a program or organizing committee,
please accept. It’s a great way to help the field and to
help others to be able to attend the kind of events
you are going to now.
3. Say “Yes” to Review Requests.
Although we are living in a time where publishing is
going through some changes (many for the better),
in one way or another the process still requires peo-
ple to read what is being submitted and to make
comments — whether as formal reviews in the peer-
reviewed literature or in the various other forms
being experimented with in open forums and com-
ment-based models. If you’re a researcher, your career
depends on the papers you get into journals, top con-
ferences, and professional publications like AI Maga-
zine. In most places, papers need at least three
reviews, some more, and thus for every paper you’ve
had accepted, three people (not including editors
and others involved in the production of the publi-
cation) have spent their time. While this step is a
more specific instance of paying it forward, I call it
out particularly because it’s becoming harder and
harder to find reviewers these days, since AI publica-
tion is booming. We’ve reached the point where
many conferences are starting to limit submissions in
some ways, and several journals I’m involved with
are looking at instituting some kind of (nonbinding)
agreement that if you publish a paper with them,
you’re expected to review three papers over the next
one to two years.
Senior colleagues, I know that there are more and
more demands on your time. I know it gets harder and
harder to find the time to do the reviewing. However,
our students and other junior members of our field
cannot succeed unless we do it. I admit I cannot say
“yes” to everything I’m asked to review, but I try to do
what I can. Unless and until new ways of publishing
emerge, we need to support the current model.
As well as agreeing when asked to do things, sometimes it pays to be the one who offers to do things.
AAAI and most other professional societies have a lot
of committees and other volunteer opportunities,
and they need people to fill those spots for the societies to continue. One of the first positions I ever
held for AAAI was with the Spring Symposium Series.
I had attended one of the meetings and really
enjoyed it. I thought this (then new in AI) model of
a set of AI meetings held in one place, combining the
intimacy of a workshop with the social scale of a conference, was a really great idea. I talked to the chair
of the committee organizing the series and asked,
“How can I help?” A of couple years later, I was the
chair. While this step has a bit of “be careful what
you wish for,” as it was indeed a lot of work, it turned
out to be professionally rewarding for me, as well as
of value to the field. The people I met, both the senior people creating the meetings and my more junior
peers who were attending, ended up being an important social network that brought me into contact
with many people from all parts of AI, not just the
particular area I was pursuing at the time.
5. Remember, We’re All in This Together.
That last point, about meeting people from all over
the field, is crucial. When one is immersed in a particular research area, it’s really hard to find the time
to track what else is happening in the field. However, when you’re involved in the various kinds of service activities I’ve been describing, you meet people
from other parts of AI as a matter of course. Chatting
with someone over dinner is one of the easiest ways
I’ve ever found to learn about what was hot and who
was doing it. This natural networking also meant
that when I needed to find someone from another