Saed News: According to an engineer and entrepreneur in the field of artificial intelligence, concerns about the complete replacement of human labor with AI may be overly optimistic or exaggerated, because the real cost of using AI models at large scale has now become a serious challenge for major technology companies.
According to SAEDNEWS, Microsoft recently canceled internal licenses for the Claude Code tool; a decision apparently driven by the high costs of token-based pricing models. This becomes even more notable considering that Microsoft is one of the world’s largest cloud infrastructure players, and a significant portion of Anthropic’s processing is hosted on Azure servers.
It appears that even a company of Microsoft’s scale has concluded that the cost of widespread use of AI coding tools may exceed their economic value.
In another part of the report, Uber’s CTO reportedly warned in an internal memo that the company’s teams had consumed the entire 2026 AI budget within just four months, highlighting the growing financial pressure of AI development and usage.
At the same time, prices for AI services from American companies have increased by 20 to 37 percent, and GitHub is gradually shifting toward a usage-based pricing model.
These changes indicate that the era of heavy subsidization of AI services is likely coming to an end. In recent years, many companies hid the real cost of running models to attract users, but with rising demand and increasing computational load, the true infrastructure costs are now becoming more visible.
The mentioned entrepreneur says that token-based pricing has forced companies to face the real cost of running large-scale language models. According to him, the figures companies see today are much higher than what fixed subscription plans previously suggested.
In such conditions, it seems that the future of artificial intelligence depends not only on the technical capabilities of models but also on their economic viability—a factor that could slow down the pace of full replacement of human labor by AI compared to initial predictions.
