ranzware
In a recent episode of the Quad Damage podcast, Arkane founder Raphaël Colantonio spoke about his ideas for a new Dishonored game. He, Harvey Smith, and Dinga Bakaba all had their own ideas for a sequel, he said, although the odds of any of them being developed are low as all three devs now work on different projects.
Elsewhere on the podcast, the subject turned to Colantonio’s thoughts on AI and the future of immersive sims. In short, he sounds optimistic. “If AI could really remember everything, and take meaningful human decisions like a Dungeons and Dragons Gamemaster, then that would actually be very valuable. Things we could not do. There would not be a team that is big enough to do this kind of thing.”
“One of the things about immersive sims is that we are always trying, either to plan for everything that we think the player might do – and that’s authoring, that’s painful,” Colantonio said. “Or, we’re doing a lot of failsafes with systems that somehow fall back on their feet if the player does something we did not plan. There’s always some stuff the players do that we haven’t planned, so that’s why you have the net under. But that net’s a little janky. Sometimes it works, sometimes not.”
“Everybody hates AI … and for frankly good reasons. But there is space for AI to improve the tools,” he continued. “Not to do our job, but improve the tools. If you look at the way we made games in the nineties, it’s pretty similar to how we make games today. The tools have barely evolved. I mean, they have, but it’s still the same principles, the same ideas. If you need to make your level a little bigger, you still have to pull the props apart and add the middle prop that’s missing there, make sure the textures match. It’s an absolute pain.”
He thinks that level designers spend a lot of time doing “grunt work”, and that this could be improved if “there was an equivalent of a chatbot GPT embedded in some tools. And I’m sure it’s going to happen tomorrow. You’ll look at your building, and say: make it bigger. Make it wider. Change the style. And then it’s done very fast. And then you become the director of that level.”
The tools for making games Colantonio began his career with have evolved, he goes on, but not that drastically (as ever, the difficulty with genAI discussions is distinguishing between the current wave of new product announcements and older tools which make less showy use of automation and generation.) Meanwhile, team sizes and budgets have gone up dramatically. “We’re throwing people at the problems. Making games with 1000 people doesn’t make sense. It was not meant to be. There’s nothing admirable about that. It’s just a lot of money and a lot of people that are probably not having a fun job. I would rather go back to the way things were when teams were between 20 and 50 people.”
WolfEye, Colantonio’s current studio, indeed started with around 19 people, and now sits at “50-60”. They’re currently working on an unnamed first-person RPG that harkens back to the immersive sims their alumni are famous for. “The ingenuity and gadgetry of Dishonored and Prey with a ‘real RPG‘ experience redolent of Skyrim and modern-day Fallout.”
I’ve obviously made my own thoughts about GenAI known, but I’ve still got some learning to do, and there’s evidently a distinction between aiding artists and trying to replace them. And, hey, call me biased, but I’m more inclined to listen to someone like Colantonio than I am some LinkedIn bozo screaming about the future without giving me a reason to care in the present. That said, I’m concerned that he’s still speaking mostly speculatively here, about “space for” and “tomorrow”. And the sort of people gaining wealth and power in this space between “might” and “does” aren’t necessarily the ones that share Colantonio’s values. If WolfEye ever decide to make a new protagonist specialise in the ol’ six-finger discount, I’m going to live with the rats.
Source link
Kurt Indovina
Add comment