ISTANBUL

Artificial intelligence (AI) tools like ChatGPT and Google Gemini have become indispensable in government and market operations worldwide, as countries and large tech firms are investing heavily in the technology.

The billions of dollars invested in data centers, as well as the rapid rise in market valuations of AI firms, fuel concerns about an AI bubble.

Gloria Shkurti Ozdemir, an AI researcher and faculty member at Azerbaijan's Hazar University, told Anadolu that previously, governments and academia led AI research and development (R&D) efforts, but now, the private sector is at the helm.

Ozdemir stated that the growing military and geopolitical importance of AI has created a self-sustaining cycle of technology that involves governments and businesses. “Companies are competing for commercial and technological superiority, while countries are prompted to fund and adopt these technologies to keep up in competition and security, triggering new waves of R&D with public funding,” she said.

She noted that previous projects in the tech industry where expectations exceeded reality demonstrated that bubbles can burst—the metaverse, for example, was built on a yet-to-be mature. "Interest quickly faded, but the concepts and goals did not disappear entirely; they were only postponed."

She emphasized that an AI bubble is still on the table, but that a potential burst is unlikely to cause a systemic collapse and may instead result in a temporary slowdown rather than a complete halt.

“Even if the AI bubble bursts, economic competition, geopolitical imperatives, and the foundational role of AI will prevent the development and funding of it from stopping entirely,” she added.

Agah Tugrul Korucu, an AI expert and an associate professor of IT, told Anadolu that the AI bubble debate is fueled by data quality concerns, a cost-benefit balance, hallucinations, reliability issues, and a lack of mature engineering support.

“The collected data of companies is often scattered, incomplete, or inconsistent, which can lead to AI responding with incorrect information while using definitive language—this requires caution, especially in health care, finance, and law,” he said. “The cost of model development sometimes does not offset the short-term gains, dampening expectations—but this doesn’t mean it’s an AI bubble; there are other key factors to tell the technology’s maturation.”

“I can say this for certain: AI will become as natural as electricity in our lives in the long term,” he added.

Korucu stated that the sustainability issues also fuel AI bubble discussions, and the industry is responding with smaller but specialized models to reduce costs and energy requirements.

“With AI, we’re seeing a rise in labor productivity—an industrial project analysis that would otherwise take an expert eight hours may be done in minutes by AI, which gives experts time to use for more creative or strategic tasks instead of routine work,” he said. “AI-powered systems in robotic production lines can monitor micron-level errors, which in turn improve quality and reduce costs.”

“Repetitive, heavy, risky, and delicate tasks can now be done by AI-powered robots, ensuring job safety and uninterrupted operations, while data-driven decision-making mechanisms enable managers to act with evidence rather than based on intuition,” he added.