Three of Square Enix’s most senior creators say the next generation of game developers will have to learn AI tools, and that no tool will replace the person who has the first idea. PC Gamer’s headline, syndicated on HeadTopics, summed it up: devs of the future will need to “master the use of AI,” but the human touch stays vital. The comments came from Yoshinori Kitase, Hiromichi Tanaka and Koichi Ishii, whose careers run through the Final Fantasy and Mana series.
The story is circulating widely today, October 28, 2026. iXBT.games ran its own version on October 27 under a title that captures the second half of the message: only humans should turn zeroes into ones. Two named outlets, one consistent framing, and three developers who rarely agree to talk in slogans.
Why does it matter? Because the people speaking built the franchises that define Square Enix. Kitase is a Final Fantasy veteran, and the series’ story choices keep drawing debate (we covered one such argument in our look at the FF7 Remake story discussion). When veterans of that rank say AI literacy is now a career skill, studios, students and hiring managers pay attention.
Don't miss new tech stories on Google
Add Tech Insider once in the Google app and our stories appear in your news suggestions.
What the Square Enix Veterans Actually Said
The confirmed content is short, and that is a strength. According to the reports, the three developers discussed the potential role of generative AI in game development and kept returning to one point: human creativity remains necessary. They did not announce a product, a policy or a budget. They offered a view of how the craft will change and what the next generation must bring to it.
Tanaka framed the skill question. Kitase framed the creative question. Ishii framed the emotional question. Read together, the three remarks form a simple model: AI handles volume, humans supply origin and feeling, and the people who thrive will be those who can direct the tools without handing over the idea. The table below lays out who said what, using only wording that appears in the reports.
| Speaker | Known for | Core point (as reported) | Where reported |
|---|---|---|---|
| Hiromichi Tanaka | Final Fantasy and Mana projects | The future comes down to how well people master AI | PC Gamer, Automaton |
| Yoshinori Kitase | Final Fantasy series | Humans must turn zero into one, and that will never change | PC Gamer, Automaton, pro-gamer.gg |
| Yoshinori Kitase | Final Fantasy series | AI and game engines can speed up one-into-a-hundred work, freeing time for the first creative stage | Automaton, GamesRadar |
| Koichi Ishii | Mana series | Fully AI-driven output may look polished but can lack human warmth | Automaton, GamesRadar |
| All three | Square Enix veterans | Generative AI has a role, but human creativity stays necessary | PC Gamer, iXBT.games |
Hiromichi Tanaka: Mastering AI Becomes a Core Skill
Tanaka made the most career-relevant statement of the three. As Automaton West reported, he said: âIt will definitely come down to how well people can master the use of AI.â
Read plainly, that sentence treats AI fluency the way earlier generations treated engine fluency. Nobody in 2010 asked whether a programmer should learn a commercial game engine. They asked which one. Tanaka’s remark suggests the same shift is underway for generative tools. Developers who can specify, review and correct AI output will move faster than those who cannot. Developers who refuse to touch the tools risk falling behind on throughput.
The word “master” matters. Tanaka did not say “use,” and he did not say “accept.” Mastery implies judgment, which means knowing when a generated draft is good enough, when it needs rework and when it should be thrown out. That is an editorial skill, and it explains why the veterans paired their AI comments with a defense of human taste.
Yoshinori Kitase: From Zero to One, Then One to a Hundred
Kitase supplied the most quotable lines. In the version carried by Automaton West, he said: âHumans need to be involved in the initial step of turning zero into one. That will never change. The next generation of creators absolutely needs the ability to turn zero into one.â
Another outlet, pro-gamer.gg, renders the same remark with slightly different English: âWhen it comes to creating something brand new and going from 0 to 1, you absolutely need a human being. That will never change: the next generation of developers must possess that ability to create something from nothing.â The wording differs because each outlet words the remark independently, and the exact original venue and full text are not confirmed. The meaning does not change.
The one-into-a-hundred half of the argument
Kitase did not stop at defending human origination. He also described where AI helps. Per the reports, he said AI and game-engine support could improve efficiency in the stage of turning a one into a hundred, the long production slog of building out, polishing and scaling an idea. If machines absorb part of that labor, he argued, people can spend more of their energy on the opening creative stage.
That is a notable position for a Final Fantasy veteran. Large-scale production is where modern blockbuster budgets balloon, and it is where schedules slip. Kitase effectively placed AI in the production-scale half of development and kept it out of the ideation half. Whether that boundary holds in practice is a different question, and one we return to below.
Koichi Ishii: The Human Warmth Problem
Ishii, associated with the Mana series, raised the softest but possibly most important concern. According to the reports, he warned that if AI does everything on its own, the result might be polished but could end up feeling like it is missing human warmth. Outlets translate his phrasing a little differently, so this article paraphrases rather than quoting a single English version.
Warmth is hard to measure, which is why it tends to get ignored in spreadsheet-driven planning. Yet players describe it constantly. They talk about the odd character who feels written by a person, the level that rewards curiosity, the soundtrack with a point of view. Ishii’s worry is that a pipeline optimized purely for output will sand those qualities away. A game can pass every technical check and still feel empty.
The Mana series gives his concern a concrete anchor, since Ishii is associated with it. A creator in that position naturally cares about the tone of the work, and that is the quality he fears a fully automated process could flatten.
What the Reports Do Not Confirm
Precision matters here, because AI stories attract embellishment. The available reports do not confirm any specific AI product, software version, budget, sales figure, game title, price, technical specification or release date. They also do not establish the exact original venue and date of the interview or its full wording beyond the quotations above.
That means several claims you may see online are unsupported by the sources behind this story. Nobody has said Square Enix is shipping a specific AI-built game because of these remarks. Nobody has said a Final Fantasy title uses a named generative tool. And the three developers are not announcing company policy. They are sharing views as veterans, and readers should treat the comments as perspective rather than corporate roadmap.
The Industry Data Behind the Debate
The veterans spoke into a market where AI adoption is already high. Several 2026 studies, reported by named outlets, show how far the tools have spread and how uneven their integration remains. GamesIndustry.biz, Video Games Chronicle and Hitmarker all covered a study from Vivrato, which interviewed 101 senior studio leaders between April and September 2026, mainly at studios in North America and Europe. The research used a structured 44-question format rather than an online poll.
In Japan, a 2026 preview from the Computer Entertainment Supplier’s Association (CESA), reported in coverage of Tokyo Game Show, found even heavier use. A separate 2026 Game Developers Conference survey of more than 2,300 industry professionals, as cited in press summaries, showed a far more skeptical workforce. The table collects the headline numbers.
| Study | Sample | Finding | Takeaway |
|---|---|---|---|
| Vivrato 2026 AI Adoption Report | 101 senior studio leaders | 97% used AI in some form during development | Near-universal experimentation |
| Vivrato 2026 AI Adoption Report | 101 senior studio leaders | 94% used AI in at least one of nine production stages | Use spans concept to live operations |
| Vivrato 2026 AI Adoption Report | 101 senior studio leaders | 52% rated pipeline integration 3 out of 10 or lower | Adoption is wide but shallow |
| CESA 2026 preview | Japanese game developers | 85.8% use generative AI at work | Japan leads on usage |
| CESA 2026 preview | Japanese game developers | 63% use it daily | Routine, not occasional |
| GDC 2026 survey | 2,300+ industry professionals | 36% use generative AI at work | Rank-and-file use trails leadership |
| GDC 2026 survey | 2,300+ industry professionals | 52% say AI hurts the industry | Sentiment is negative |
Compare the leadership numbers with the worker numbers. Studio heads report 97% usage, while the GDC sample shows 36% of professionals using generative AI at work and more than half calling its effect negative. Different methods explain part of the gap, since Vivrato interviewed decision-makers and GDC polled a broad cross-section. Still, the pattern is clear. Executives adopt faster than the people doing the day-to-day work, and that tension is the backdrop to the Square Enix veterans’ careful tone.
Where AI Fits in a Production Pipeline
Kitase’s zero-to-one and one-to-a-hundred language maps neatly onto a production pipeline. The table below applies that framing to five common stages. It is this site’s analysis of how the veterans’ logic plays out, not a statement from Square Enix, and it should be read as a working model rather than a company plan.
| Stage | Likely AI role | Human role | Main risk |
|---|---|---|---|
| Concept and ideation | Reference gathering, rough variations | Original idea, direction, taste | Generic, derivative concepts |
| Prototyping | Fast scaffolding of mechanics and scenes | Judging what feels fun | Prototypes that never get rethought |
| Asset production | First-pass variants, cleanup, scaling | Art direction, final approval | Style drift and lost warmth |
| Testing and QA | Repetitive checks, bug triage | Judging feel and edge cases | False confidence in coverage |
| Live operations | Support tooling, analytics, localization drafts | Community judgment, tone | Tone-deaf automated output |
The pattern mirrors the veterans’ argument. AI earns its place where work is repetitive and verifiable, and humans keep the stages that depend on taste. The CESA preview points the same way. According to the coverage, common governance practices among Japanese developers include human verification, correction and supervision, along with discouraging the use of AI output unchanged. In other words, the working industry already treats the human as the final gate.
The cost side deserves attention too. Our earlier reporting on AI tools creating extra review work shows why mastery matters. If a team generates ten drafts but a senior artist must review all ten, the saving shrinks fast. Tanaka’s emphasis on skill is partly about making that review loop efficient.
Square Enix’s Wider AI Context
The veterans’ comments do not exist in isolation. Nova Crystallis reported this month that Square Enix is re-launching its Collective co-publishing operation, with a featured game developed using generative AI. We have not verified the details beyond that headline, and no specific title, tool or figure is confirmed in the sources used here.
Even so, the juxtaposition is instructive. A publisher can host a project built with generative tools while its veteran creators insist that origination stays human. Those positions are compatible if the company treats AI as a production and experimentation layer and the core franchises as human-led. That reading fits the veterans’ remarks, though the sources do not confirm it as company strategy.
How Other Studio Voices Compare
The Square Enix message lands in a crowded conversation. Nintendo’s Shigeru Miyamoto has voiced a positive outlook on generative AI, which we covered in our breakdown of his key quotes. Masahiro Sakurai, meanwhile, has warned about a flood of generic games on Steam, a concern we examined in our report on his 19,000-titles warning. Together with the Square Enix trio, that makes at least three of Japan’s best-known creators on record, each with a slightly different emphasis.
Miyamoto leans optimistic, Sakurai leans cautious about quality and volume, and the Square Enix veterans land in the middle with a skills-plus-soul formula. What unites all of them is the refusal to say AI will do the creating. Even the enthusiasts frame it as a tool in a human’s hands. For Western studios, where public debate often swings between hype and backlash, that shared Japanese framing offers a steadier template.
Market and Labor Impact
Statements about future skills carry weight in a tight labor market. The games industry has absorbed waves of layoffs and closures, and workers read every executive and veteran comment on AI for signs of what comes next. Our coverage of EA closing BioWare and funding four bets shows how restructuring and AI investment now appear in the same news cycle, which is exactly why developers listen closely.
The veterans’ message offers a partial reassurance and a clear warning. The reassurance says original creative work stays human. The warning says that people who cannot work with AI may lose ground in production-heavy roles. Broader labor analysis points to the same tension. Our report on Goldman Sachs flagging AI hiring headwinds shows hiring concerns spreading across industries, and games are unlikely to be immune.
For students and juniors, the practical reading is this. The one-into-a-hundred tasks, the ones AI absorbs first, are often where newcomers learn the craft. If those tasks shrink, training paths have to change. Studios that want future directors will need to keep apprenticeships that teach origination, not just supervision of generated output.
Historical Context: New Tools, Same Argument
Veterans like Kitase, Tanaka and Ishii have watched the medium absorb several technology shifts. The move from 2D to 3D production, the rise of commercial game engines and the growth of large outsourcing networks each triggered fears that craft would suffer. Each time, the tools changed who did what without removing the need for a strong creative lead. Both Final Fantasy and Mana have outlived many of those shifts.
Generative AI differs in one respect. Earlier tools automated execution, such as rendering, physics or asset pipelines. Generative systems can produce candidate ideas, which blurs the line between execution and origination. That is precisely why Kitase drew a bright line at zero-to-one. He is defending the boundary that earlier tools never threatened.
Five Predictions
The following are this site’s own forecasts, built from the reported comments and the survey data above. They are analysis, not statements from any named person or company.
- AI fluency enters job listings. Within a couple of years, expect more production and design roles to name AI-assisted workflows as a requirement, echoing Tanaka’s point about mastery.
- Disclosure fights continue. The gap between leadership enthusiasm and worker skepticism in the GDC and Vivrato data suggests labeling and consent debates will intensify, a theme we explored in our coverage of the Steam AI disclosure dispute.
- Integration stays shallow for a while. With 52% of leaders in the Vivrato study rating integration at 3 out of 10 or lower, deep pipeline change will take longer than the headlines imply.
- Human-made becomes a selling point. Ishii’s warmth argument points to marketing that stresses authorship, as audiences learn to value visible human craft.
- Apprenticeship models evolve. Studios that want future directors will build training that teaches zero-to-one thinking alongside AI supervision.
What Developers and Studios Should Do Now
Turn the veterans’ advice into habits. Learn the tools you are likely to meet, and learn them well enough to judge their output. Keep a portfolio of work that shows original concepts, since that is the skill Kitase says will never be outsourced. Practice reviewing and correcting generated material, because verification is where human value concentrates.
For studios, the message is to set clear governance. The CESA coverage describes human checks and restrictions on unchanged AI output as common practice. Define which stages allow generation, who signs off and how credit is handled. Do that before a deadline forces ad hoc choices. For more on how these debates connect to the rest of the industry, browse our gaming coverage.
FAQ: Square Enix Veterans on AI
Who said what in the Square Enix AI comments?
Hiromichi Tanaka said the future will come down to how well people master AI. Yoshinori Kitase said humans must turn zero into one and that AI can help with the one-into-a-hundred stage. Koichi Ishii warned that fully AI-driven results might look polished yet lack human warmth.
Is Square Enix announcing an AI-built game?
No. The reports behind this story confirm no specific AI product, game title, budget or release date. The comments are the personal views of three veterans, not a company announcement.
What does turning zero into one mean?
It means creating something new from nothing, such as an original concept, world or character. Kitase argues that this first step must stay human and that the next generation of creators needs that ability.
How many studios already use AI?
In Vivrato’s 2026 report, based on 101 senior studio leaders, 97% had used AI in some form and 94% used it in at least one of nine production stages. However, 52% rated integration at 3 out of 10 or lower.
Do game developers support AI?
Views are split. A 2026 GDC survey of more than 2,300 professionals, as cited in press summaries, found 36% use generative AI at work and 52% think it has a negative effect on the industry. A CESA preview found 85.8% of surveyed Japanese developers use it.
Why do the quotes read differently across outlets?
Each outlet words the remarks independently, and the exact original venue and full text are not confirmed. Wording varies slightly, but outlets agree on the substance: humans originate ideas, and AI supports production.
What should aspiring developers learn?
Build original work to show zero-to-one ability, and learn AI tools well enough to review and correct their output. The veterans’ message ties both skills together.
