"Big Tech That Nurtured AI Now Faces Discomfort as It Risks Losing Everything" [Tech Talk]

Microsoft: "AI's Monopoly on Knowledge Must Be Prevented"
Cracks Emerging in the Alliance Between Frontier AI and Big Tech
Big Tech Faces Dilemma as AI Encroaches on Their Core Businesses

Editor's NoteFrom AI and semiconductors to telecommunications and biotech, we break down the essential yet often unfamiliar technologies that shape our daily lives in an easy-to-understand way.

"We cannot allow artificial intelligence (AI) to consume the global economy."

Satya Nadella, Chief Executive Officer (CEO) of Microsoft. Photo by Reuters and Yonhap News

This was the statement made by Satya Nadella, Chief Executive Officer (CEO) of Microsoft (MS), during an interview with the U.S. financial media outlet The Wall Street Journal (WSJ) on the 21st (local time). He explained that it is impossible to continue making massive AI infrastructure investments while high-performance AI models are causing large-scale job losses.

As models from so-called "frontier AI companies" such as OpenAI and Anthropic surge ahead, a sense of discomfort is emerging among big tech firms. This is not solely due to concerns over mass layoffs. AI is now encroaching on the business domains traditionally dominated by big tech, yet these very tech giants are exhausting their cash reserves by building data centers for AI companies.

Cracks in the Big Tech–AI Alliance

At the 2023 OpenAI Developer Conference, Sam Altman, CEO of OpenAI (left), and CEO Nadella. OpenAI

In the interview, CEO Nadella stressed, "We cannot pour all our resources into building data centers while all office jobs are disappearing because of AI," and added, "It is unacceptable for a handful of companies and AI models to monopolize all knowledge."

Instead, Microsoft plans to focus on much more affordable and straightforward AI models. The company is targeting AI agents for ordinary office workers, with token pricing that is accessible to the average user. Microsoft has already launched a product called "Copilot Cowork," and is considering integrating Deepseek, a low-cost, open-source model from China.

This marks a departure from Microsoft's previous AI strategy. In the past, Microsoft forged a de facto alliance with OpenAI through large-scale strategic investments, entered into partnerships with Anthropic, and provided them with investment funds totaling trillions of won. In effect, today's frontier AI companies have grown thanks to the support of big tech. However, big tech is now trying to rein in its AI-related expenditures.

Big Tech Spends, AI Reaps the Rewards?

As frontier AI companies continue to grow, big tech firms are facing a dilemma. In the past, programming assistants and office software were core revenue sources for companies like Microsoft and other tech giants. But as AI automates more of this work, big tech is losing these cash cows.

Moreover, big tech is engaging in "hyperscale" businesses by building and leasing data centers for AI companies. In order to develop ever more powerful AI models, big tech firms have poured vast amounts of cash into infrastructure investments, draining their cash reserves. According to Wall Street forecasts last month, the free cash flow (the cash companies can use for dividends, debt repayment, etc.) of the four largest U.S. tech firms this year is expected to hit its lowest level since 2014.

Microsoft's hyperscale data center. Microsoft

Weakened financial statements are also affecting investor sentiment. While hardware companies have rallied this year, Microsoft and Google, among others, have continued to underperform. In other words, AI companies are developing more advanced models using the data centers built by big tech, penetrating the business domains of these giants, while big tech finds itself in the paradoxical position of spending cash solely for the benefit of AI firms.

Big Tech Voices Frustration: "AI Must Not Consume the Economy"

On the 15th, CEO Nadella posted a lengthy message on his official X (formerly Twitter) account, clearly explaining the reasons behind Microsoft's revised AI strategy. He argued, "The real opportunity in the AI era is not in building the best AI model, but in creating a cycle of iterative learning that brings together models and humans," adding, "To achieve this, every company must be able to develop its own AI agent system."

He emphasized, "We do not want a world where every company and every industry has their value absorbed by AI models. Neither society nor policymakers will allow such a reality." He further stressed, "Rather than a world where a handful of AI companies swallow up the economy, we want a world where new value flows to all companies, industries, and nations."

This content was produced with the assistance of AI translation services.