�԰���ɫ��

Game industry veteran Stéphane Assadourian: “Experience can’t be taught in class”

In more than 20 years in the gaming industry, Stéphane Assadourian has worked on a range of successful titles like the original Outcast and the first three installments in the Assassin’s Creed franchise, where in the end he was the Lead AI and Gameplay Programmer. In our interview, he reveals what sparked his passion for games and articial intelligence - and shares which one of his games he is proudest of:

In the late 1990’s, you were working on your Ph.D. thesis at UPMC University Pierre and Marie Curie when you decided to instead become a game developer. What triggered that decision? What was your original motivation to become a game developer?

Actually, I’ve always wanted to become a game developer. I’ve been playing video games since I was 8 years old, playing Pac Man and Space Invaders in 1980. There was a strong passion for video games burning inside of me since that age, and to this day it is still with me. I was fascinated precisely by the fact that such a lifeless piece of metal could bring me so many deep emotions, so many various stories to bring with me to school or to bed, and this for me was the birth of my passion for Artificial Intelligence. Later I ended up starting a Ph.D. in AI but could not resist the call of learning how to become a video game developer. So I decided to not finish my Ph.D. and instead learn the craft I always wanted to learn.

You joined Belgian game studio Appeal to work on Outcast with little prior experience in game programming. How did that work out?

It was such a unique opportunity! After 2 years working on my Ph.D. thesis, I had to do the military service (still mandatory in 1997-1998 in France), but toward the end, I did like all my younger colleagues who had just finished their engineering schools: I sent resumes to the various places I wanted to work at. And Appeal wanted to see me, so I went.

I vividly remember to this day the emotions of entering this small 20-people studio in the suburb of Namur, in Belgium, and I remember being naturally excited by the work environment. I had the chance to enter the video games industry, and it was now. I had not prepared anything for the interview, really, I just could only tell them that I had not done any programming for quite some time (my Ph.D. was approaching Artificial Intelligence from the non-classical logics standpoint and it was rather heavy on mathematics, more than anything else). So basically, I told them I was absolutely not up-to-date and did not feel ready to code anything!! They told me they were interested by my academic profile because they wanted to build an AI pipeline for their next series of games. Not being able to program anything on day 1 was not a problem for them, what was important was the ability to understand Academic AI. Then I got the job, moved to Namur in a split second and started my career in October 1998, 20 years ago.

Outcast (1999), Appeal

Do you have any plans of returning to university and finishing your doctorate thesis at some point?

No. During these two years I realized the work environment was very political, and research was not at all what I had in mind. For me, research was about discovery, exploration, new findings. For me, research was Gödel, Turing, Einstein, Kantor, Tesla... Most of the other theses I was seeing were about the state of the art of one or the other well-known algorithm or approach and were not actually bringing anything new to the table. It felt more like documenting than researching and I wanted to push the limits.

I remember being disappointed: I had idealized scientific research a bit too much. Then I understood the grant systems, the politics behind it and felt you really needed to have a vocation for research to turn it into your profession, despite the environment in which you would actually perform this research.

Looking back at the previous two decades making games, which of the titles you worked on do you feel proudest of, and why?

Assassin’s Creed Brotherhood, without any doubt. I was Lead AI and Gameplay Programmer and was supervising a team of 40 people. It’s essentially a producer job, really: you must drive in the right direction this pool of talent who can achieve anything but not everything, you must constantly assess and challenge a seemingly infinite list of features to implement as the development is going on, and we did probably as much content as in AC2 in half the time. This was state-of-the-art production, with friends more than colleagues and with so many good memories and some tough moments. But the ride was well worth it.

I knew I wanted to change after producing this game because the franchise we had built from the ground up was already seven years of my life, and making this game at this position was clearly the peak of this experience. Ubisoft wanted me to continue on the franchise but I declined, as my thirst for more knowledge was stronger and I wanted to learn more and different things, new skills. So I went on to learn different topics inside the company after that.

After AC Brotherhood, I’d say in second place it would be Assassin’s Creed, the first one. And in third, Assassin’s Creed 2. But those three games are really tight in that ranking, through the specific filter you’re asking.

Assassin’s Creed Brotherhood (2010), Ubisoft

Do you feel that your own university studies prepared you well for your career?

Not at all. No school in the 90’s could prepare you for a career in video games development. It was too new, it was too fragile, not well understood and too specific - yet encompassing so many different crafts. Today we have free, very capable game engines, free tutorials, free courses, affordable online courses.

What school taught me though, was to remain focused for a long time on topics that were not naturally interesting to me. School taught me to listen and pay attention, I guess. But I would say school is not for everybody, it only teaches well to one kind of students, and if you cannot switch your behaviour to being that student, school will be difficult, seem useless and potentially even be useless. Today there are other ways to learn and they all have their pros and cons. Education is vital, but it’s also something that needs to be tailored to each individual because we don’t all learn in the same way, using the same process.

AAA game productions with large budgets are said to often suffer from intense workloads, especially leading up to release dates. How did you cope with that?

I had suffered from it on my previous project at a different company, before moving to Canada. During my interview with Ubisoft, I knew exactly what I did want and what I did not want. I simply asked them and they gave me a good enough answer: there is overtime, but we care for people and we watch out for them. I could not really expect more than that, but was rather happy that they would give this answer, at the very least.

As an expert for Artificial Intelligence, what do you think about recent advances in the field? Should we be worried about a Skynet-type scenario and stock up on supplies and weapons?

I think we have become very good at recognizing patterns. And that’s about it. The worry does not come from the fact that we need to defend against AI (we are already killing each other well enough to be threatened by anything else than ourselves), but more about big corporations selling our data and trespassing our intimacy. The next AI revolution should be about representing the world and what it is to be human. This, for me, is the missing part that will connect everything tomorrow and eventually lead us to the kind of AI people are talking about today.

Did you spend any part of your university studies abroad? If not: Where would you like to have studied abroad?

No I did not, but I had always wanted to. I would have gone anywhere really, just exploring was exciting enough for me.

What have you learnt after graduation that couldn’t have been taught in a classroom?

What is difficult to teach in a classroom is experience, which can only be acquired through work and failure cycles. Also, learning that failing is part of experimenting is usually something that you learn in a real game dev environment over the years, and not even in a video game school (no matter how often students are being told this). Also, how to react to journalists or reviewers criticizing your work publicly is a skill that you will usually acquire way after you graduated.

You left Ubisoft in 2016 and have since then been worked as a freelance advisor to game studios and publishers. What direction do you see the industry headed in that you are most excited about?

I think the Nintendo Switch is a great console and will see more and more interesting games. I also believe there could be a new AI-based genre breaking through within the next 5 years, creating a trend as successful as the one we have these days with Battle Royale.

Photo: Stéphane Assadourian at TUMO Conference 2018, Yerevan, Armenia