Garick Elordi
Masahiro Sakurai, the legendary director behind the Super Smash Bros. and Kirby franchises, has voiced significant concerns regarding the integration of generative artificial intelligence in video game development, cautioning that while the technology will inevitably lead to a massive influx of titles, it risks smothering the industry in unoriginal, low-quality content. In an extensive interview with the Japanese publication Denfaminicogamer, Sakurai argued that while AI can replicate existing systems with alarming efficiency, it remains fundamentally incapable of mirroring the pioneering human creativity exemplified by designers such as Fumito Ueda, the visionary behind Ico and Shadow of the Colossus.
The discussion arose following Sakurai’s recent retirement from his long-standing role as a judge for the Game Designers Awards. Reflecting on the evolution of the industry during his tenure, Sakurai highlighted a staggering shift in the volume of game releases. In 2010, the Steam platform saw approximately 275 new games published. By 2024, that number had skyrocketed to over 19,000 annual releases. Sakurai predicts this trajectory will only steepen as generative AI tools become more accessible to developers, though he warns that this quantitative explosion will likely lead to a qualitative crisis.
The Exponential Growth of Digital Marketplaces
The statistics cited by Sakurai underscore a transformative era for digital distribution. According to data from SteamDB, the number of games released on Steam has seen a near-vertical climb over the last decade. In 2018, the platform saw roughly 8,000 releases; by 2023, that number surpassed 14,000, and current projections for 2025 suggest the 20,000-title threshold will be comfortably breached.
Sakurai posits that a significant portion of this growth is fueled by the ease with which developers can now generate assets, code, and even entire gameplay loops using AI. However, he remains skeptical of the commercial viability of these projects. "I’m sure the number will increase even more from now on because AI will increase the number," Sakurai stated. "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."
The director’s critique centers on the idea of "reproduction." He suggests that AI-generated games often lack a unique identity, instead functioning as amalgamations of systems already present in the market. This phenomenon, often referred to in developer circles as "AI slop," refers to content that is technically functional but lacks the cohesive vision or "soul" that typically defines successful interactive media.
The Limitations of Algorithmic Creativity
A core component of Sakurai’s argument is the distinction between "referencing the past" and "forging the future." Generative AI models are trained on vast datasets of existing human work. Consequently, their output is inherently derivative—a synthesis of what has already been achieved.
Sakurai illustrated this point by referencing the work of Fumito Ueda. Games like Ico and Shadow of the Colossus are celebrated not just for their mechanics, but for their specific atmospheric and emotional resonance—qualities that were unprecedented at the time of their release. Sakurai argued that while an AI might eventually be able to mimic the visual style of Shadow of the Colossus, it cannot predict or create the "next" revolutionary idea that a creator like Ueda might conceive.
"Basically, AI refers to things that have been created so far, but it can’t refer to things that haven’t been created before," Sakurai elaborated. He emphasized that as the market becomes saturated with mass-produced goods, the value of a "unique flavor" will only increase. Designers who act as pioneers—those who understand their own creative voice and can offer something that does not yet exist in a dataset—will become the most critical assets in the industry.
The Rise of the "Clone" Economy and Developer Anxiety
The industry is already seeing the practical fallout of AI-assisted development. Sakurai pointed to simple but compelling puzzle games, such as Librarian: Tidy Up the Arcane Library! and the viral hit Needle in a Haystack Simulator, as examples of concepts that are highly susceptible to AI replication.
Needle in a Haystack Simulator, developed by Studio Bitdot, features a deceptively simple premise: finding a single needle among millions of pieces of hay. Its success triggered a wave of near-identical copies on digital storefronts, many of which appeared to be rapidly assembled using automated tools. To protect its brand, Studio Bitdot was forced to add a prominent "Original" tag to its Steam store artwork to distinguish it from the influx of clones.
This environment has created a climate of fear among independent developers. Recent reports indicate that some indie studios have opted to cancel public playtests or delay demos for fear that their unique mechanics will be scraped and replicated by AI-driven "asset flip" developers before the original game can even reach the market. This year’s Steam Next Fest, for instance, saw over double the number of demos compared to previous years, a surge that many analysts attribute to the lower barrier of entry provided by generative tools.
Industry Response and Policy Shifts
The proliferation of AI content has forced major industry players to reconsider their terms of service. In early 2024, Valve, the operator of Steam, updated its policies to require developers to disclose the use of AI in their games. This includes "Live Generative AI," where the game uses AI to create content on the fly while the player is engaged. Valve also introduced a new system allowing players to report illegal AI content, specifically focusing on copyright infringement and "deepfake" material.
Despite these safeguards, the sheer volume of submissions makes manual curation nearly impossible. Sakurai’s comments reflect a growing sentiment among veteran developers that the "democratization" of game development via AI may be a double-edged sword. While it allows those without traditional coding or artistic skills to create, it also threatens to drown out high-effort, human-centric projects in a sea of algorithmic noise.
Chronology of AI Integration in Game Development
To understand the current tension, it is helpful to look at the timeline of how automation has entered the creative space:
- 2010–2016 (The Era of Procedural Generation): Games like No Man’s Sky used algorithms to generate vast universes. This was math-based and rule-bound, not "generative" in the modern AI sense, but it set the stage for large-scale automated content.
- 2018–2021 (Machine Learning for Optimization): Technologies like NVIDIA’s DLSS began using AI to upscale graphics, focusing on performance rather than content creation.
- 2022–2023 (The Generative Explosion): The release of Large Language Models (LLMs) and image generators like Midjourney allowed developers to generate dialogue trees and concept art in seconds.
- 2024–Present (The Market Saturation): The influx of AI-assisted titles on Steam reaches record highs, leading to the "AI slop" discourse and calls for stricter platform regulation.
Broader Implications for the Future of the Medium
Sakurai’s perspective suggests that the industry may be heading toward a "great sifting." As consumers become more adept at identifying the hollow nature of mass-produced AI content, the market may see a flight to quality. This mirrors trends in other creative industries, such as film and literature, where the novelty of AI-generated work is beginning to give way to a renewed appreciation for human craftsmanship.
However, the economic reality remains challenging. For small indie developers, the temptation to use AI to cut costs is high. The "unprofitable titles" Sakurai mentions are often the result of developers trying to compete in a market where the cost of human labor is rising, but the expected price of games remains low.
The director’s retirement from the Game Designers Awards marks the end of an era, but his final warnings serve as a roadmap for the next generation of creators. His message is clear: survival in the age of AI requires more than just technical proficiency; it requires the "human flavor" that no algorithm can yet extract from the history of art. As the industry moves forward, the distinction between a game that is "built" and a game that is "designed" will likely become the primary battleground for the future of interactive entertainment.
