Interview:Edge, August 1996

Edge, August 1996
Issue No.	35
Page(s)	22-26
Interviewee(s)	Anil Malhotra
Interview Subject	Millennium Interactive
URL available?	Yes

An interview with Anil Malhotra from August 1996 published in Edge magazine issue 35.

Transcript

Currently making headline news with its Cyberlife artificial life technology, Millennium Interactive has come a long way from the days of James Pond. Edge tracks down the company breeding Creatures.

An audience with... Millennium

In 1988, from the ashes of Logotron, a small purveyor of educational software, rose the phoenix of Millennium Interactive. Initially distributing and publishing games picked up on an ad hoc basis, usually from 'bedroom independents', co-founders Michael Hayward and Ian Saunters soon directed their burgeoning resources to in-house development. The early nineties saw a string of popular games for Sega and Nintendo, most notably the James Pond series, and in 1993 Millennium ditched its publishing arm to become pure developers. Since then it's appeared to the casual observer to be largely dormant, with only a handful of work-for-hire jobs (such as Raymond Briggs’ The Snowman) and the fairly well received Defcon 5 to its name. In reality, however, it's been busy creating a totally self-sufficient development facility complete with state-of-the-art video, audio, and digital video suites. In fact, in the last year alone, Millennium has doubled in size. So where are the games? Edge paid a visit to the Cambridge HQ to talk turkey with its dynamic and loquacious business development manger, Anil Malhotra.

Edge It's been a difficult 12 months. Do you think the industry has emerged healthy?

AM Yes and no. In terms of other consumer products - films, records, books - computer games sell in pathetically small quantities. A company like Nintendo can claim comparable profits but it is reaching much fewer people. The problem is, most games are still designed for the same kinds of people who like the same kinds of things. I think the PC in particular will break down those barriers. And if I were to back one horse in the platform race it would be Windows 95, undoubtedly.

Edge With a graphics accelerator? And if so, which one?

AM VideoLogic seems like a good company and I like its technology. But although the 3D boards will be important, the interactivity of Windows 95 will carry the day. The ability to just plug and play on the network - when you see Microsoft that determined you know they're going to be a player in the market. I guess a lot of the games will be a little different from the standard console fare. There'll be a skew towards intellectual fascination rather than fast reflex. That's going to be the mould for PC-style gameplay.

Edge Interactivity seems to mean different things to different people...

AM What I mean is a bit like playing cards or Battleships. where what's important is that you're playing against me and we can make things happen in realtime.

Edge The usual conception of 'interactive' is a two-way bridge between player and software. You seem to be envisaging a bridge between two people.

AM I'm interested in lifelike software rather than just software that's clever at solving problems or copying how life systems work. And if this advances much further over the next ten years, you will, fo a limited degree, be able to surrogate your own personality in a software agent. Imagine sending that off down the line while you were asleep or on vacation. You could make it how you want. You could be truthtul or you could emulate a woman. It's cheaper than buying a new wardrobe.

Edge And what about the consoles?

AM You have to bet on Nintendo. Largely because they just know how to make damn good games or get damn good games made for them. You hear all these spurious reports leaking through the Internet about delays due to 'hardware problems', 'overheating on the motherboard', etc, but the real issue is that they want this to be the best game system you've ever seen and the thing that will make it the best is the games. I think they're being shrewd. Just waiting until the software is perfect. However, unless Mario 64 and the other launch titles are the best videogames you've ever played, they've blown it.

Edge Who else is there room for?

AM Sega are very bullish at the moment but on the other hand the genuine reality of your prospects are often inversely proportional to how bullish you feel about them. The PlayStation's interesting. I've got mixed feelings about it. It's a good machine and Sony launched it well but a lot of people are disappointed by the average quality of its games.

Edge Perhaps due to the fact that it remains a slave to third parties. Twenty mediocrities don't make a single genius.

AM Right. As I understand it there were in the region of 500 to 600 PlayStation titles in development last Christmas. It's too many, I think Sony realises this. We've just had a couple of original concepts approved by SOA and it’s suddenly changed the rules of engagement. Now it's very particular about what it'll allow you to develop or publish and suddenly very interested in whether you're leading on the PlayStation.

Edge But is this a sign of positive discrimination or negative conservatism?

AM If I was Sony I would be a little anxious right now because I've not yet had a mind-blowingly good hit. The other reason I'd be worried is that there's probably a short fuse in the marketplace for whether a console's going to make it or not. With N64 just around the corner how much time has Sony got left to turn its machine into the must-have console - Christmas, probably.

Edge What games are you working on and for what platforms?

AM All our games will appear on Win 95, but we'll be leading on the PlayStation with Brains in Planes [see boxout], a 3D racing game featuring futuristic cartoon craft. The general idea is to make the thing as unfeasible as possible. Racing up the side of a skyscraper and through the window of the block opposite. Keeping the physics persuasive is obviously part of the trick. We want the speed and pizzazz of a Wipeout-style game but incorporating a greater feeling of freedom and openness. We also want to engender a feeling of close-quarter competition in 3D. So you can overtake above and below as well as left and right.

Edge You must have a pretty smoking engine, then.

AM In fact, we just dumped the last 3D engine - it wasn't quick enough. The one we've got running downstairs is just a week old. We're going to get the speed almost too fast to start with and then slow it down to the right speed as we add more and more detail.

Edge What about the Saturn?

AM We're working on Medievil. a 3D arcade-action adventure akin to the Ghouls 'n' Ghosts genre. You're looking at some platform elements and some puzzle-solving elements in a tightly focused set of levels. The premise is that you control a knight. Sir Daniel Fortesque, who has been brought back from the dead in skeletal form by accident by an evil wizard. Unlike most of the creatures in this world you remember that you were once a good guy and set about following the wizard's trail to his lair. The action is viewed from a thirdperson perspective but the viewpoint alters according to the style of level. We move the camera logically to provide the best angle as well as to heighten the drama.

Edge Both projects look extremely promising. But neither, perhaps, are likely to 'break down the barriers', as you put it.

AM What you have to do to break down the barriers, and what we're trying to do, is to launch a product almost based on its novelty. Which is risky because it flies in the face of the classical marketing model, ie give the people what they want.

Edge So what have you got up your sleeve?

AM We have Cyberlife. We're not trying to simulate the way life works using computer software (the AI solution), we're actually trying to model the way life works by copying biology. Which is the difference between AI and AL (artificial life). Biology is made up of a load of different things that make people tick and we want to copy those, put them together, and see if we can make software tick in the same way.

Edge How did Cyberlife come into being?

AM One of the things that’s always disappointed us about games. racing games. or a Doonrstyle game, is the terribly predictable characters you have to deal with. Imagine what it would be like to imbue your opponents with a measure of life-like response and the ability to learn how you play the game. Once the bad guys have seen some of their colleagues lain to waste they're going to think 'Hmm, perhaps it's not such a good idea to hang around here, I'm going to hide for a bit and consider my next move.' That's a much richer experience.

Edge How ts this distinct from clever AI?

AM I mentioned earlier what I wanted from interactive games. I wanted to play another human being. AI can't achieve this because from a programming point of view you have to prescribe everything that could happen at the outset. Even if you manage 50 permutations the player's still going to say after a while, 'Oh, it’s that one again'. Fortunately for us, one programmer in our company was a biologist by trade rather than a computer scientist. He understood that what was needed was software that captured the essence of human behaviour.

Edge Which is unpredictable but not erratic. The difference lies in the ability to learn.

AM Right. We started by mimicking a system that could have punishment and reward effects, ie. if an in-game character got shot coming round a corner he might try another approach next time. But trying to write motivational rules becomes absurdly complex. So we took a completely obverse view. What if we model the way motivation works. How an organism responds to a set of stimuli. Take hunger. When your glucose levels fall your biological system gets hungry. That stimulates your brain and if you've never eaten before you've got to figure out what to do. Stick some things in your mouth and see if that helps. Which it does, and the brain remembers. It's like a feedback loop system. Add fear, anger, boredom and the other drivers and you're getting towards a real person.

Edge Even taking things bottom-up rather than top-down must have required a prodigious amount of programming.

AM Over three years. But we can now create a creature for whom the property and nature of an event causes a series of things to happen in its system triggering a decision. From just trying to make in-game characters a bit more authentic we were actually beginning to push towards creating a proper little virtual life form. Which is how Creatures, the first game to use the Cyberlife technology, came to be developed.

Edge How would you describe Creatures?

AM It's a genuine life sim for the PC. It's the first attempt by anybody in the software industry to develop such a thing. You hatch a creature into a virtual world full of objects - fire, water, food - and modify the environment in ways that allow it to learn. Once Creatures is released we'll be applying the fruits of that research to more traditional game genres. People are going to be able to do things with computers that they have never been able to do before. I think you'll be quite surprised.

Boxout

Creatures

In Creatures, the first product to use the much vaunted Cyberlife technology, the player starts with a set of eggs which they can hatch onto their PC and into a 2D world of about 12 screens wide and three high. The cuddly infants environment is populated with about 40 objects (food, fire, enemies, etc} for it to interact with and learn about and, after about a week, an averagely precocious creature adolescent with a stirring in its loins will be in a position to breed. Its offspring will have a distinct genetic code and, ipso facto, features and traits of its own. In fact, scientists at Millennium claim they have no way of predicting what sort of lifeforms will emerge a few hundred generations down the line.

Millennium is clearly hoping people will adopt its 'creatures' as virtual pets and, moreover, will participate, via the Creatures Website, in 'the world's largest artificial life expenmment' in which whole communities of creatures will evolve on the Net. This all sounds most intriguing - not to say downright scary - but whether or not these 'creatures' have sufficient personality to engage the heart as well as the mind remains to be seen.

Brains in Planes

Millennium's first 3D racer is set in a 23rd century dystopia of peace, harmony and total boredom. where outlaw racers get their kicks from racing ultra-fast planes around the cities of the future. However, because the human body can't withstand the extreme G-force, the racers must remove their brains and plug them into the craft's circuitry.

The Brains team hope to have at least six circuits on offer, all boasting multiple routes, short cuts, and horizontal and vertical hazards (pedestrians, police vehicles, moving scenery) in true 3D space, Game modes will include multiplayer, tag, battle and time trial. The first track is still being tested on Alias but even so the team feels confident that Millennium's own 3D libraries will deliver an awesome visual experience. All the planes will be generated on SGI and ingame will be tracked by a selection of three thirdperson camera positions. Cleverly, anything that could obstruct your view is rendered in transparent polys so vision is unimpaired. The crudity of depth shading will be avoided by constructing 3D maps to limit the viewable distance from any point on the map. Graphical effects will include a particle system allowing for spectacular explosions, fountains, exhaust streams; multiple light sources and lens flare; and a dazzling chrome effect achieved through the use of environment mapping within the graphics API.

Brains in Planes certainly has the credentials to be an attractive and original title, and, if the team's commitment to gameplay holds true, it might even be fun.

Medievil

With the avowed intention of matching the rich, twisted art design of Tim Burton's Nightmare Before Christmas, the classic arcade gameplay of Capcom's Ghouls and Ghosts, and the dramatic 3D of the Alone in the Dark series, the Medievil team could hardly be accused of lacking ambition.

The quest of the hero, Sir Daniel Fortesque, encompasses a mixture of shoot 'em up and exploration set across 11 areas ranging from a graveyard, through an asylum, to a battleground, each offering a distinct playing environment. Sir Dan can jump, duck, and use both handheld (sword, battle axe) and projectile (lightning bolts and crossbow) weapons against the sorcerer Zarok's malevolent minions. Plus, for sheer variety, there are Clockwork Knight style minisections in which you get to control Dan's sidekick, Morten, a small worm.

Technically, things are shaping up nicely. All the game characters and backgrounds are true 3D objects (either texture mapped, Gouraud, or flat shaded), lit with parallel ambient and point light effects, with the realtime 3D display engine delivering an update of 30fps. If the game matches the technical expertise and the imagination and thoroughness of the design, Medievil could be a serious contender early next year.