Demis Hassabis, part 1
12th Jul 2006 | 12:39
Demis Hassabis, ex-Bullfrog and Elixir gaming genius and the AI guru behind titles like Theme Park, Republic: The Revolution and Evil Genius, has been a bit quiet recently. The man with a brain rumoured to be larger than a planet has dipped below the videogame horizon to embark on new AI research.
However, when CVG recently caught up with him, Hassabis revealed during an extensive interview that he's turned his mind to a new game that's already underway, and that some of his work on artificial intelligence will feed back into what is currently a top secret project. But what else did the man have to say for himself and his colourful videogame career? Well, read on to find out. Part one of the two-part interview is published below. We'll be bringing you the second and concluding part soon...
How did you get into the games industry?
Demis Hassabis: Professionally, I got into the industry when I was 15. My foot in the door was a competition run by Amiga Power where you had to write a Space Invaders clone. So during my school summer holidays I thought I'd give it a go, and my game actually came in second to a guy called Mike Diskette who went on to do a lot of things in the industry.
The first prize was a job at developer Bullfrog, an amazing prize, so as I came second I decided to phone up anyway and they found out how young I was. I think Peter Molyneux was intrigued to know "who is this kid!" So I went there for two summer holidays, play-tested Populous and designed a few levels for Syndicate. I then took a year off before going to Cambridge University, co-wrote Theme Park with Peter and it kind of blossomed from there.
With Theme Park, who came up with the idea and how did you develop the ideas into a full game?
Demis Hassabis: Peter's idea was already there when I got there after finishing my A-levels and I was a wide-eyed 16-year-old. There was all this cool stuff going on at Bullfrog, which at the time, only had about 25 people working there - before the big EA expansion of it. We were in a little, fairly run-down office in a Surrey research park, and it was a really great time.
Development of Theme Park had just kicked off - they were about two months into it and remember seeing Populous graphics to show off what the game idea might be, and it seemed like a really cool project to get involved with. The original idea was Peter's, but as with any project, there are thousands of ideas along the way that everybody contributes until it's finished.
What are you most proud of with Theme Park?
Demis Hassabis: My responsibility in TP was the simulation bit, and having never really written anything like that before, or anything of that size or complexity, I was pleased that when I was thrown in at the deep end I managed to cope and do a reasonably good job. It turned out really well. I wrote all the computer players and how the technology progression should go plus other game mechanics that I can't remember now! But I had a brilliant time helping Peter with the design.
When the game was released, it had a great reaction. How did that go down with yourself and the Bullfrog team?
Demis Hassabis: It got amazing reviews and sold really well - especially in Germany and Japan. But at that time I'd already left to go to college. I don't regret college, as they were some of the best days of my life, but I kinda missed out on the excitement of seeing the game's success as I was away from it by then. Of course I kept in touch with everyone, and it was exciting to see it on the shelves while I was at college - it was a bit surreal actually!
What did your friends at university think of your success?
Demis Hassabis: At first they thought I was kidding, but I managed to get hold of a copy and they saw my name on there - although they did suspect I'd hacked it or something! Eventually they believed my story (laughs).
After graduating, did you make up your mind to start making videogames as a career?
Demis Hassabis: Yeah, I mean ever since I can remember I've always loved games and particularly computer games and programming. I've always played games from a very young age - I was four years old when I started playing chess, and played for England and stuff. So videogames seemed like the perfect marriage of games and programming. I never really thought about any other career, even though I could have done all sorts of things.
So after graduating, did you always have an idea of setting up your own company?
Demis Hassabis: Oh yeah, I always had a dream of doing that, but it was working out the logistics of setting it up and so on. I was going to do it straight after university, but at that point Peter Molyneux had sold Bullfrog to EA and was looking to start something new. So it seemed like the timing was perfect to go back with him and help set up something new.
Lionhead had just started with five or six people, and basically it was a chance to see how a company was set up and work with a bunch of guys I'd had a lot of fun with at Bullfrog. I guess also it was less daunting than setting up a new games company straight away! Also, Black & White - although still an embryonic concept - already sounded pretty cool in terms of the AI needed and it was something that I was interested in working on.
What was the philosophy behind Lionhead? Did you and Peter have similar ideas about creating original games?
Demis Hassabis: Peter and I have very similar ambitions in terms of trying to do things different from what other people are doing and to always push the envelope a bit. I guess we have slightly different tastes on some things, which often works well when you have two people with different takes on things.
Peter quite likes the fantastical element, whereas I like a more strategic element I suppose. B&W could have been taken in very different ways, and what turned out was very cool, but it could've been a more strategic game too. We worked together as a really good team at Bullfrog and Lionhead, but it was Peter and his vision that was driving the company.
You've never been scared of taking risks?
Demis Hassabis: No, I don't know why that is, but it's something I've never worried about. I'm actually more worried about not taking risks and playing safe. Not pushing myself enough. It's a bit perverse I suppose, and asking for trouble! I've always been prepared to jump in at the deep end and see if I can swim or not.
After Lionhead, you set up Elixir...
Demis Hassabis: Yeah, I basically wanted to take a lot of the great things I'd seen at Bullfrog and Lionhead, the culture, and nurture that more and take it to the next level. Also there was a lot of talented friends of mine from university who were looking at setting up a company, so it was a good window of opportunity.
Plus I had some investment from some friends in the city - a small amount, but enough to set up a company, and an idea for the game Republic, which I really wanted to do. Lionhead were busy with B&W, so it would've been a few years before I could've got started on it if I'd have stayed there.
Republic: The Revolution was an ambitious game - why did it take so long to develop?
Demis Hassabis: I can look at Republic now with hindsight and I think it's clear that we bit off more than we could chew at the time. We tried to build this incredibly ambitious graphics engine, plus create some fairly insane AI technology, for living breathing cities, as well as - the thing I underestimated the most - boiling down a topic as massive as politics into a game.
If I had to do it again, I'd still make Republic, but concentrate on two of those three areas, perhaps even one, and that would have been enough, rather than spread ourselves a bit thin. Because the company was quite small I only had one or two people on each of those areas - when you think of today's development team sizes, that was asking too much.
I'm proud of what we achieved, given the trials and tribulations along the way, producing an original and entertaining game. I was hoping it would be even more ambitious than it actually turned out to be, as we actually had to tone down some of that ambition towards the end. The idea of the game was to make you feel like you were a revolutionary such as Che Guevara, to make you feel you were taking part in a revolution, and I guess we didn't get far enough down the line to convey that idea fully.
Did you ever consider doing a sequel to Republic?
Demis Hassabis: I would've loved to - there's a lot of mileage in the topics of the game. I still believe that games can be more than they usually are. While there's definitely a place for games where you shoot people, if we want games to be taken seriously like novels or films, as a cultural medium, then we need games that are about more mature subjects. However, they still obviously have to work as games, and that's the challenge - how can you make something that's deep and meaningful and also fun to play?
Check back soon for the second and concluding part of our interview with Demis Hassabis.