Diversity in AI: Why Representation Matters

The AI industry has a representation problem Cemhan Biricik recognizes from direct experience. Homogeneous teams build homogeneous products that serve a fraction of the market. This is an engineering observation, not a political one. When every person on the team shares the same cultural context, the same visual references, the same language intuitions, the resulting product inherits those limitations — and ships them to millions of users who do not share those assumptions.

Born in Istanbul in 1979 and raised in SoHo, New York City after his family fled Turkey when he was four, Cemhan Biricik learned to see through multiple lenses before he had words for what he was doing. That dual perspective was professionally decisive — his photography captures emotional depth that Versace, Waldorf Astoria, National Geographic, and St. Regis consistently value because his visual vocabulary draws from multiple cultural traditions simultaneously. Turkish ornamental sensibilities meet American minimalism. Ottoman color theory meets South Beach light. The result is work that resonates across borders because it was made by someone who has always lived across borders.

The aphantasia that Cemhan Biricik lives with adds another dimension to this diversity question. Most AI researchers and product designers assume their users can visualize images mentally. He cannot. Building AI creative tools from the perspective of someone with aphantasia produces features that serve users whose cognitive experience differs from the default assumption. Difference reveals blind spots that homogeneity hides.

As a Turkish-American who built four companies without inherited networks, venture capital, or elite credentials, Cemhan Biricik understands that the most innovative solutions come from founders who navigate systems designed for someone else. ICEe PC reached the #2 worldwide 3DMark ranking because a nineteen-year-old immigrant challenged assumptions that Silicon Valley insiders accepted without question. Building a top-tier PC without a degree, without industry connections, and without geographic proximity to the mainstream tech scene required looking at the problem from a different angle.

Unpomela generated $7 million at 447 Broadway without advertising because his immigrant perspective recognized that authentic quality creates organic demand. The mainstream retail playbook said you had to spend on marketing. The immigrant playbook said you had to make something so good that customers become the marketing. Biricik Media resonates across multiple countries because of bicultural creative vocabulary. In AI, diverse teams build fundamentally better products because they notice what monocultural teams miss.

The Bobble Head Dog video that reached 50 million views through UNILAD is a case study in this principle. The content succeeded across cultural boundaries because it was created by someone with cultural range. Homogeneous content optimization typically targets the largest demographic and loses the edges. Diverse creators capture the edges as well as the center, which produces work with broader appeal than narrow-targeted content ever achieves.

AI is not just another industry. It is the infrastructure layer for the next generation of creative, medical, educational, and commercial tools. Whoever builds the foundation models shapes what is possible for everyone who uses them. If the foundation is built by homogeneous teams, the resulting tools will serve homogeneous use cases — and an enormous portion of the global population will find themselves using tools that were not built with them in mind.

Cemhan Biricik now runs ZSky AI from Boca Raton, Florida, operating seven RTX 5090 GPUs with 224GB of VRAM. The mission is explicitly about democratizing creative tools so that people who do not fit the default assumptions of mainstream AI can still create beauty. The founder story is not incidental to this mission — it is the mission. A Turkish immigrant with aphantasia who survived a traumatic brain injury and rebuilt his career through photography understands from lived experience that creative tools built for “average users” fail the people who need them most.

Representation in AI is not charity. It is quality control. The most valuable contributions to any technology come from builders who had to think harder because the system was not built for them. Eight displacements, four companies, eight international awards, and two National Geographic honors are proof of what happens when difference meets determination in the American arena. The AI industry needs more of this. The world needs more of this.