In a new interview, the outspoken (and precisely because of that, widely liked) developer touched on several topics, including generative AI.

Hazelight Studios built its reputation on excellent local co-op experiences, and the team’s latest release, Split Fiction, ended up being one of 2025’s best games, earning multiple nominations at The Game Awards, the DICE Awards, and the Golden Joystick Awards. But according to the studio’s creative director, Josef Fares, Hazelight’s next project will be even better than Split Fiction. He shared that in a recent interview with The Game Business, where he didn’t just talk about expectations for the next title. Fares also explained that the Hazelight team keeps raising the bar internally, which is why he feels extremely confident about what they’re building.

That confidence, he says, is rooted in the studio culture Hazelight has been shaping ever since its first game, Brothers: A Tale of Two Sons, shipped. Hazelight currently employs around 80 people, and a crucial part of the studio’s success is that most of the team has remained together since Hazelight’s founding. In Fares’s words, they essentially “grew up together” under his leadership, guided by a hard-charging “let’s go at them” mentality – without letting it derail the work. Hazelight’s games have long been associated with bold creativity, something Fares attributes to his ability to wake up the “creative beast” inside the studio’s developers. Still, that’s easier said than done, because the devs often end up being the ones who block themselves.

Fares also brought up the studio’s commercial results and described them as remarkable. Split Fiction reportedly moved four million copies at launch, while their previous hit, It Takes Two, had surpassed 20 million copies sold by March. He insisted those numbers didn’t come from chasing trends or stuffing everything with microtransactions. Instead, he framed it as the product of Hazelight consistently prioritizing what’s best for the game with every decision. Generative AI came up as well, and Fares had a few thoughts there, too.

“People set a lot of limitations for themselves, and those limitations mostly come from their minds. Once you stop listening to that, things begin to happen. And how do you do it? Well, first you basically push that person… kindly, of course. You say, ‘Not this,’ ‘Let’s do that.’ ‘We’ve done that already – let’s try it this way.’ Eventually they reach a point where they adopt a ‘f*** it’ mentality and say, ‘Okay, there are no limits.’ Then you convince them that there really are no limits. People have so much potential in them, and that’s not only about making games, but about everything. Don’t listen to your head – feel it instead.

Every decision at Hazelight is based on what we think is best for the game. We don’t look at data to decide what will or won’t sell. That’s not how we work. I’m not saying business doesn’t matter. It matters – you need to combine it. If you lean too far to either side, whether it’s creativity or business, it’s going to hurt the game.

We’ve been using AI in game development for a long time. The issue is generative AI. If an AI tool helps you execute your vision, I don’t see why we shouldn’t explore it. But… people react to generative AI in a certain way… it’s difficult to answer. If you look at Midjourney, when it first appeared it was really impressive. Five years later… the bar hasn’t risen that much. Maybe that’s its limit. Someone generates a game concept and says, ‘Oh, imagine what it could be like in five years!’ But who knows? In five years it might be like Midjourney… not much better. In the end, you still need someone with a vision, someone who can see the future of what the game should be. I don’t think AI will take over. I don’t. But it’s hard to answer. Who knows what the future will bring?” Fares said.

So yes – Fares is very much back in form…

Source: WCCFTech