Vapor World: Over the Mind had every reason to expect a good launch week. The Korean studio behind it, Alive Inc, a team of roughly five people, built a Soulslike Metroidvania with a painterly art style drawing comparisons to Lies of P and Playdead's narrative platformers. A demo had been public since January 2025, and it pulled in 100,000 wishlists ahead of early access. Then it launched on August 19, and within days it was sitting on 64 "mostly negative" Steam reviews, almost none of them about the actual game.

What Actually Went Wrong

The specific complaint driving the backlash was narrow but real: the game's opening cutscene, generated using AI, couldn't keep the main character's model visually consistent from shot to shot. The top-voted Steam review called it out directly, noting the gameplay looked genuinely fun but that the studio should have waited until it could deliver that art style with human-made work instead. That's not an unreasonable thing for a player to notice or care about. A jarring, inconsistent character model in the very first thing you see is a legitimate first impression problem, full stop, regardless of how it was made.

The Director's Response Was Better Than Most

Studio director Young Kim didn't get defensive, and it's worth actually reading what he said rather than assuming the usual corporate non-apology. He explained the studio had originally tried building its cutscenes entirely by hand, in-engine, and found the results consistently fell short of what the story needed. Turning to AI was meant to solve that specific problem, not to cut costs or skip effort. In his own words, he expected some disappointment over the decision. He didn't expect it to become this big.

He also didn't hide behind the tool. He directly acknowledged that unless AI reaches a point where the difference is genuinely undetectable, what he'd hoped for was closer to an ideal than a reality, and that the fault sits with his team's judgment, not with generative AI as a concept. The cutscenes are already being replaced with the original in-engine version, delivered within days of the backlash rather than weeks of stalling.

Why This Deserves More Nuance Than It's Getting

Here's what's genuinely frustrating about how this played out: a five-person team spent real effort building something with a distinctive visual identity and combat system worth talking about, and the entire conversation around its launch got swallowed by 30 seconds of rough cutscene footage. That's a real cost, and it's worth naming plainly. A small studio experimenting with a tool to solve a genuine production problem, being transparent about why, and correcting course within days when it didn't land, is about as good-faith a version of this situation as you're going to get. Piling 64 negative reviews onto a game's entire Steam page over one cutscene, rather than reviewing the actual 20-plus hours of handcrafted gameplay surrounding it, is a genuinely disproportionate response to the actual scope of the problem.

At the same time, it's worth being honest rather than reflexively defensive: the specific AI output here wasn't good, by the director's own admission and by outside assessment. That's a separate question from whether AI belongs in game development at all. A studio trying a tool, getting the result wrong, and owning it openly is exactly the kind of good-faith experimentation this industry needs more room to do without it turning into a referendum on the entire technology. Punishing honest attempts this harshly doesn't produce more caution, it produces more silence, studios quietly avoiding disclosure rather than risking exactly what just happened to Alive Inc.

The Actual Lesson Here

Vapor World joins a growing list of small studios who've walked back AI use after backlash, alongside bigger names like Larian and Sandfall who've faced similar pressure. The pattern worth watching isn't whether AI gets used, it's whether honest, quick correction like Kim's actually gets rewarded with anything close to proportionate goodwill, or whether the review bomb happens regardless of how the studio responds. If it's the latter, every future developer weighing this same tradeoff learns the wrong lesson: not "be careful and transparent," but "never admit it in the first place."

Sourcing note: this piece is built on PC Gamer's original reporting, including director Young Kim's full public statement and independently gathered Steam review and wishlist data.

Does the studio's quick, honest response change how you'd judge this, or does the initial choice still outweigh it for you? Let us know your thoughts in the comments. And if you've got a game you'd like us to review, send it over to editor@gamevibe.io.