Despite massive investments in artificial intelligence, the anticipated economic revolution has not materialized as expected. Companies had forecasted a 15% increase in productivity due to AI, but recent reports suggest the actual contribution to the U.S. GDP is a mere 1%. In the past year alone, Big Tech invested $400 billion into AI technologies, with projections for this year reaching $700 billion. This staggering amount could have eradicated world hunger for a decade, raising questions about the effectiveness of these investments.
While some economists initially claimed that these investments were driving the U.S. economy, Goldman Sachs’ chief economist, Jan Hatzius, reveals a starkly different narrative. He stated that AI investments contributed "virtually nothing" to GDP growth in 2025, debunking the notion that they are significantly stimulating the economy. Hatzius emphasized that the narrative surrounding these investments is often distorted and not reflective of their true impact.
The discrepancy in opinions among economists can be attributed to where the funds are being allocated. Much of the investment is directed towards importing computer chips, which ultimately benefits economies in Korea and Taiwan rather than the U.S. This situation effectively drains capital that could be used to bolster the American economy.
Moreover, despite the theoretical potential for AI to enhance productivity through automation, current implementations have failed to achieve this on a global scale. Analysts have long been aware of these shortcomings, and Goldman Sachs itself estimated that AI might only boost U.S. productivity by 15%, a figure that could be overly optimistic. In a more sobering assessment, ING predicted a mere 1% increase, highlighting a stark contrast to the 45% productivity growth attributed to the computer and internet revolution since 1980.
Recent studies underscore the challenges AI faces in meeting productivity expectations. Research from Carnegie Mellon University indicates that even advanced AI systems fail basic tasks 70% of the time. Additionally, a report revealed that leading AI models struggled with 97.5% of real freelance tasks, while some generative AI tools actually hindered developers by slowing their output due to the need for extensive error correction.
Corporate experiments with AI are often unsuccessful because businesses tend to over-rely on AI for tasks that would benefit from human oversight. The transition to effectively using AI requires a fundamental shift in mindset, where the focus is on how AI can enhance workflows rather than simply replace human labor.
This trend is reflected in broader corporate behavior; many companies are scaling back AI initiatives. Reports suggest that the usage of AI by major American corporations is declining rather than increasing. In fact, the cancellation rate of corporate AI programs surged from 17% in 2024 to 42% in 2025.
Despite the significant investments in AI, the reality remains that there have been no substantial job losses attributed to AI automation. The data indicates that the anticipated transformative power of AI has yet to be realized, with no clear path to improving its reliability or effectiveness.
As the market reassesses the potential of AI, competitors may find opportunities to innovate and adapt their strategies in ways that could lead to more fruitful applications of technology, ultimately reshaping the economic landscape.
Informational material. 18+.
