An era of mass production and low-quality goods is, paradoxically, the moment when human craft becomes most valuable. That is the argument Masahiro Sakurai has laid out while dismissing the wave of machine-generated titles flooding Steam as “unprofitable” releases that nobody can keep track of.
The Smash Bros and Kirby creator made his case in a candid interview with Japanese outlet Denfaminico Gamer, and it lands at a moment when artificial intelligence is reshaping how the industry talks about itself.
From 275 Releases a Year to 19,000
The conversation began with the Game Designers Award, a prize Sakurai launched in 2010 and stepped away from this year after failing to find a successor to take over judging duties.
When the award began, roughly 275 titles arrived on Valve’s storefront each year. Today the figure sits closer to 19,000 — and the veteran designer has no doubt the climb will continue.
“19,000 is amazing, but I think that number will increase even more from now on,” he said. “AI will increase the number. However, I think most of them will become unprofitable titles. There are too many to keep track of, and it’s impossible to understand them all.”
Key point: Steam releases have grown from around 275 a year in 2010 to nearly 19,000 — and most of the next wave, he expects, will be commercially unviable.
The numbers frame everything that follows. A storefront that once felt curated now absorbs more releases in a single week than it did in an entire year of the award’s early life. Mass-produced projects, he argued, simply reproduce systems players have already seen elsewhere — and “there are already a lot of those now”.
The Clone Economy: Needles, Haystacks and Supermarkets
To illustrate, he pointed to the puzzle genre and the success of Librarian: Tidy Up the Arcane Library, a charming little game about shelving books in a magical library. Imitators arrived almost immediately: tidy the supermarket, put the screws back, find a needle in a haystack.
That last concept reached its logical extreme with Needle in a Haystack Simulator, which tasks players with locating a single needle among five million drinking straws. After it went viral, copycats multiplied so aggressively that developer Studio Bitdot stamped “Original” across its Steam artwork in self-defence.
None of this is new in principle — clone culture has existed since Pong — but the cost of imitation has collapsed towards zero. Concepts this simple, the designer argued, are exactly what automated tools reproduce most easily.
“AI refers to things that have been created so far. It can’t refer to things that haven’t been created before.”
Why the Next Shadow of the Colossus Cannot Be Generated
The most striking passage of the interview concerns Fumito Ueda, the director behind Ico and Shadow of the Colossus. Sakurai conceded that algorithms may one day study the atmosphere of Ueda’s back catalogue and produce something comparable.
What the director creates next, though, is beyond the reach of any model. “What kind of artistic style will his future works possess? That’s something only he can create. I think that’s where human creativity lies.”
Creators who forge their own path, he concluded, will become “increasingly important from now on” — especially in an era saturated with mass-produced, low-quality content. It is a defence of authorship, and a bleak forecast for the discoverability of everything else.
What It Means for Players and Indie Studios
The warning arrives against a fraught backdrop. Generative mods have become one of the loudest trends in games, with players porting classics to web browsers and mashing unrelated franchises into single experiences. Epic has spoken openly about an “AI-first” production model, while layoffs continue to hit studios of every size.
At this year’s Steam Next Fest, demo submissions roughly doubled year-on-year — and with them came a surge of what players now bluntly call “AI slop”. Several independent teams withdrew their demos entirely, worried plagiarists would lift their ideas before launch. For players, the practical effect is a discovery problem: genuinely good work risks burial under sheer volume.
The counter-argument Sakurai offers is not that the technology should be banned, but that originality is about to become the rarest commodity in games. As shelves groan under cheap imitation, the designers who venture into uncharted territory become the ones worth following — and, in his telling, the only ones building anything that lasts.
