The Solo Developer Has Arrived. Nobody Is Coming After Them.

Ender Tınkır shipped Zen Mahjong in two months. Two employees, one of them himself. When he started, the benchmark for a mobile game at that scope was a team of ten and a runway of a year. AI did not just help Tınkır move faster — it restructured what a game studio is. A two-person company with his level of experience now does what a mid-sized studio did five years ago.

Turkey understood this shift before most countries did, partly because Turkey had more to lose. In 2021, according to Rest of World, more than half of the top-ten most-downloaded mobile games on Apple’s App Store were made in Istanbul. The country had positioned itself as the world’s mobile gaming capital through a combination of aggressive cost structures, imported design philosophy, and a tightly networked developer ecosystem that traced its roots to a single acquisition: Peak Games, sold to Zynga in 2020 for $1.8 billion. That deal produced what Naavik calls the “Peak mafia” — roughly 65 startups founded by Peak alumni, each one seeding the next generation. Dream Games. Spyke Games. Loop Games. Grand Games.

At the ecosystem’s height, Turkey was minting 200 new game studios a year. The figure comes from Turkey’s investment agency, as cited by Silicon Canals. By 2023, that number had dropped to 84. By 2024, it was 30.

The conventional explanation is market saturation, post-pandemic correction, and the broader gaming industry contraction that saw Microsoft cut 3,200 Xbox jobs and over 25,000 game workers lose their positions globally between 2023 and 2024. Those factors are real. But they do not fully explain the shape of the collapse — why the studio count fell so sharply while Steam submissions from Turkish developers actually jumped, from 67 applications in 2023 to 254 in 2024. More games, fewer studios. The math resolves once you see what AI actually changed.

What the Tools Actually Changed

Emirhan Gül is 24. He left a Turkish gaming startup after burning out on the company’s production philosophy — “the idea of churning out a game a week,” as he put it to Rest of World, “didn’t really appeal to me much.” He founded his own studio. The team he needed shrank from the dozen his previous employer maintained to two full-time employees. “With Claude Code, I can do all that easily.”

Claude Code, Copilot, Codex — the trio that now shows up in every experienced solo developer’s toolkit — are fundamentally amplifiers of existing competence. That is the part that makes the productivity numbers real and the labor numbers alarming at once. When Gül says Claude Code lets him do the work of a twelve-person team, he means that his years of experience building games can now execute across twelve domains simultaneously. The AI does not think about game design. It implements what a skilled designer directs, generates the boilerplate a senior programmer would otherwise write, and runs the QA loop a junior tester would otherwise be paid to run. Faster, cheaper, and without the friction of communication.

The workflow this produces is specific. A solo developer using Claude Code starts a feature by describing what they want in natural language — the game mechanic, the edge cases, the expected behavior. Claude Code writes the implementation. The developer tests, breaks it, describes the failure, gets a revised implementation. Meanwhile, Copilot completes boilerplate as they type. Codex handles the more complex architectural reasoning when the problem warrants it. What this eliminates, concretely, is the three-person loop that small studios used to run: junior programmer writes the first pass, senior programmer reviews and revises, QA engineer finds the bugs. The solo developer running AI tools now owns all three steps. The throughput is genuinely higher. The jobs are genuinely gone.

For the experienced developer, this is a genuine structural upgrade. Tınkır, who worked on hit titles at Peak and Dream Games before going independent, attracted investment for a two-person company that now ships games he could not have shipped alone two years ago. Naavik notes that Turkish publishers already demonstrated remarkable capital efficiency — roughly $1 million in revenue per employee, well above Finnish and Israeli competitors. AI tools press that ratio further. The solo developer with fifteen years of experience and Claude Code is a commercial entity. That is genuinely new.

The developer without fifteen years of experience is the problem the productivity story does not solve.

The Tier That’s Gone

Begüm Ceren Onar worked as a game designer until December. Her studio eliminated 17 of its 20 employees that month, not because the business was failing but because it had found it could replace most of its production capacity with AI-generated concepts. Onar, 34, told Rest of World what the math looks like from the outside: “More people are developing games on their own, but most are not making a living off it.”

She is describing two parallel realities that the industry has quietly allowed to coexist. The experienced developer with the right tools can now ship and earn. The volume of games being produced — the 254 Steam applications, the rising output from solo studios — does not translate into sustainable employment for the people who would previously have been in their first or second job. Nazım Adaklı, who teaches AI in games at Bahçeşehir University, put it plainly: instead of hiring five junior programmers, “my developer friend does all the work himself with AI.”

That sentence is where the logic collapses. The junior programmer Adaklı’s friend used to hire was not just a cost center. They were the next Emirhan Gül in training. The concept artist who spent three years at a twenty-person studio building their eye for game aesthetics is the art director at a studio in five years. The writer who got hired to do localization and ended up rewriting half the dialogue is the narrative lead at the next startup. These are not jobs that could be abstracted away and rebuilt later. They are the mechanism through which the industry reproduced its expertise.

The GDC 2026 State of the Industry report, which surveyed more than 2,300 game development professionals, found that 52% believe generative AI is having a negative impact on the industry — up from 30% in 2025 and 18% in 2024. The strongest opposition came from visual and technical artists (64%), game designers and narrative developers (63%), and programmers (59%). These are not people who dislike productivity tools. These are the practitioners who have watched the entry-level positions in their disciplines disappear and have done the arithmetic on what that means for who fills their roles in ten years.

28% of respondents had experienced layoffs in the past two years. 17% had been laid off in the past twelve months. 48% of those laid off had not found new employment.

The Udonis layoff tracker puts global game industry job losses at more than 14,600 in 2024 alone, with 2023 adding another 10,500. The bulk of those cuts came from larger publishers, but the tracker’s analysis of recent trends notes something more troubling: “Smaller studios and indie teams are now in the mix, not just the big names.” The contraction has reached every layer of the industry.

Güven Çatak, who founded Bahçeşehir University’s Game Lab, described the traditional career pathway as something that “has become a thing of the past.” His students are advised to launch their own ventures rather than seek employment — which is sound guidance for the well-positioned 22-year-old with strong fundamentals, and a brutal non-option for most people entering the workforce without them.

The Steam submission jump — 67 to 254 — is real, and it does reflect genuine creative energy. But the GDC 2026 data makes the commercial reality plain: the industry is generating more content without generating more employment. Royal Match alone accounts for 44% of Turkey’s entire $3.33 billion gaming revenue. Dream Games, with under 400 employees, generates $1.47 billion in annual in-app purchase revenue. The ecosystem’s value has concentrated at the top while the bottom has been reorganized around tools that do not need anyone in the seat.

The Math Comes Due Later

Emirhan Gül and Ender Tınkır are real success stories. The tools they are using are genuinely capable. The argument for the solo developer model is not manufactured. But both of them accumulated their skills inside the studio system that is now shrinking — working alongside designers and artists and engineers whose judgment they absorbed over years of shared production. AI amplified expertise they built through structures that no longer exist in the same form.

The solo developer who wins with AI is standing on a foundation the industry built before AI arrived. When Tınkır says he could not have imagined making a game by himself two years ago, the operative phrase is two years ago. His decade-plus before that was spent in rooms full of people whose positions are now eliminated.

What Turkey is showing the rest of the industry is not a Turkish story. It is the fastest version of a story playing out across studios in Europe, the United States, and Asia. The studio count is a lagging indicator; the junior hiring freeze is the leading one. The pipeline that produced the experienced developers who now thrive as solo studios is the thing being dismantled, quietly, by the same efficiency logic making individual developers more powerful than they have ever been.

The ceiling went up. The floor came down. They are not separate developments.

AI-generated editorial illustration · TemperatureZero · July 24, 2026