Haven wants to clarify the hazy future of machine learning
Yesterday we posted about a panel where Haven Studios CEO Jade Raymond and CTO Leon O’Reilly spoke with PlayStation 4 and 5 hardware architect Mark Cerny about the studio’s experiences in artificial intelligence and machine learning, and how these technologies could shape the industry in the future.
After the panel, we have the chance to speak with O’Reilly and Raymond about the Montreal studio’s hopes for the new technology. Both have been in the gaming industry for a while, and they’ve seen a variety of high-profile new technologies arrive – digital distribution, stereoscopic 3D, AR, VR, cloud streaming – and succeed or fail to very varying degrees. With that in mind, we ask if their attitude towards every “next big thing” has changed over the years.
“I’m still very optimistic,” O’Reilly said. “We’re very focused on developing games for a lot of these technologies, and we’re very clear about what problems we’re trying to solve and how these will solve them most effectively. When you have consumer-facing technologies, you don’t You don’t always know whether or not there will be a market for your technology, and that’s often the reason why they fail.
“I think we want to position Haven as a place where we do pioneering technology. And we’ll try things, and not everything will work. And that’s okay, as long as the things that don’t work, we fail. quickly, and then we iterate, we try something, we pivot, we change, and we keep going. But I think fundamentally these technologies that we’re looking at have a very promising future.
Raymond adds: “The key here is that we’re trying to solve real development problems. The real development problems have to do with our desire to make bigger and bigger games, bigger and bigger teams and types of games that gamers really want to change. We’ve seen people want more social games – even more so since the pandemic – where people expect a certain amount of content updates. Obviously the graphics have become And to deliver the kinds of experiences that players really want, we need to rethink things within the development team so that we can create those player-centric and community-centric experiences.”
Dozens of Haven’s founding employees came from Google Stadia, where Raymond had led the development of proprietary software for the cloud streaming effort. And while the consumer side of the streaming tech has yet to pan out (Google itself announced the shutdown of Stadia late last month), Raymond says the cloud tech was a perfect solution for developers. seeking to solve the very specific problem of how to work remotely during a pandemic.
“We see the promise of a lot of this technology, but we’re really looking at very concrete things that we’re trying to do to address specific issues as a development team trying to deliver new kinds of experiences that players want,” she said.
During the panel, Mark Cerny spoke about the potential of machine learning to allow “moderately sized” teams to make bigger games. We wonder if, in the end, this means fewer jobs for artists.
“To be honest, we don’t think that will be the case,” Raymond said. “Are there fewer people making a blockbuster movie than making a game? There are also directors, art directors, costume designers…we’ve replicated a lot of those roles in digital. Can Maybe at some point we’re going to find new roles that don’t even exist. And that’s been the case with game development ever since we started making games.
“Back then, people were looking at this programmer who did everything, right? And then we made these amazing games and now we created needs for voice directors and composers who didn’t have role in the games originally, so I think the development teams will continue to evolve, but I think it will be more based on what our imaginations allow and where we want to take the games ourselves.”
“Maybe at some point we’ll find new roles that don’t even exist. And that’s been the case with game development ever since we started making games”
O’Reilly stresses that they don’t want to downsize the development team in any way.
“But if we can, with that same development team, create more interesting content, try more ideas, iterate more, iterate faster…that’s what the tools allow us to do,” he says. . “It’s about building those muscles for the creative team, improving them and empowering them.”
While some of the most spectacular use cases of machine learning in development are still not ready for prime time, others are already implemented and impacting the pipeline, such as a system that O’ Reilly describes it as “almost a texture compression system” that takes a gray scale texture with four colors and interpolates between them to avoid an artist creating a color version from scratch.
“It’s by no means a particularly creative endeavor, so instead of spending their time doing this, we had a machine learning algorithm figure out the best way to group the data,” says O’Reilly. “And it actually worked better than artists would by hand. We’re seeing significant savings in memory and disk size, as well as time savings for artists. This is just one example of where we can put these things in the production pipeline for artists, save mundane work and allow them to be more creative with their time.”
And given the team’s current focus on using these technologies for development tools rather than user-facing parts of the game, they say there’s less concern about the “black box” qualities of machine learning, where the model produces unexpected or unwanted outputs.
“The kinds of problems we’re trying to solve are pretty analogous in a way. When you’re trying to improve your creativity, you’re not looking for a specific outcome”
“The kinds of problems we’re trying to solve are pretty analogous in a way,” says O’Reilly. “When you’re trying to improve your creativity, you’re not looking for a specific outcome. You see what comes out and you tweak or iterate on it. It doesn’t matter what happens internally. If we were to do, I don’t know, a traffic light system that needed to work accurately that would be a problem but if you’re using it to enhance your artists creativity and concepts it’s actually I think an advantage to be a bit fuzzy .”
During the roundtable, Raymond said the North Haven star was “creating a new IP that could belong to the fans, where the fans are really leading this.” Given the subject of new technologies and the nature of many recent start-ups that have invoked the concept of ownership in games, we ask them for their opinion on blockchain technology and the metaverse.
On the latter, at least, Raymond describes herself as “a huge fan.”
“I don’t know what this means for our industry, but everyone seems to care now,” Raymond says. “As a geek who loves sci-fi books, I love those little things that happen and the people who try to make them happen.”
“As a geek who loves science fiction books, I love those little things that happen and the people who try to make them happen”
O’Reilly adds: “It’s great to be able to dream. Whether it comes to fruition or not, it’s exciting. I think we live in very exciting times.”
Having not heard them discuss blockchain, we specifically press the subject.
“Video games have always had digital elements,” says Raymond. “We’ve always had NFT-like things in video games. It’s like, you do a certain thing and you get a collectible or a trophy, something like that.
“One thing that we’re really excited about is the concept of scarcity and items in games – not related to blockchain per se – but I find it interesting to see how with blockchain and related to crypto the things we’ve had in games take on a new form. [Items and rarity] is more something interesting that we talk about and think about.”
O’Reilly offers a more definitive statement.
“Blockchain as a technology is not something we engage in at all,” he says. “At all.”