Bold claim: the AI gold rush isn’t a fickle fad, yet the market is buzzing with questions about whether the fever will cool. Tech leaders insist this is no bubble, even as colossal sums flood in, startup valuations run high, and many AI projects stall at pilots rather than delivering production-scale results.
HPE is riding the wave of demand for high-performance hardware that powers AI, and in Barcelona at the Discover event, Rami Rahim, president and general manager of HPE Networking, said the current trajectory shows no sign of slowing. The conversation turned to whether AI projects ever reach full-scale production. Rahim acknowledged plenty of pilots exist but argued there’s real value emerging in production products as well. Within HPE Networking, he noted, developers are increasingly efficient by using copilots to write and verify software.
This perspective intersects with concerns about code quality produced by AI assistants. Rahim observed an inflection point: trust and capability have grown as experience accumulates. He emphasizes that rapid change is rare in this industry, but the trend toward tangible value is clear.
Some observers compare today’s AI landscape to the dot-com era’s bust, but Rahim cautions against using past patterns as a crystal ball for the future. Technologies differ, and today’s appetite for AI products and GPU cycles is exceptionally robust. Predicting a slowdown a year or two out remains difficult.
AMD CEO Lisa Su shares a parallel viewpoint, asserting that AI is not a bubble. Speaking at UBS’s Global Technology and AI Conference 2025, she described a multi-year, “ten-year super cycle” in which computing enables steadily increasing capabilities and intelligence.
The conversation shifts to the evolution of OpenAI’s expansion goals. Su notes a shift from model training to inference, with no one-size-fits-all model. This variety sustains demand for infrastructure, as customers seek more compute to accelerate outcomes.
When pressed about OpenAI’s sky-high valuation and the potential need for substantial capital to cover losses and datacenter investments, Su argued that rising capex forecasts reflect confidence in future capabilities. She argues that, despite questions about a bubble, well-capitalized firms are leveraging this unique moment in AI progress.
Even as OpenAI’s Sam Altman acknowledged earlier in the year that the industry might be in a bubble, Su reports tangible ROI from internal AI pilots that have translated into productivity gains. She asserts a measurable return on AI investments for enterprises that adopt these tools.
Recent headlines illustrate the tension in the market: HSBC identifies a potential $207 billion gap in OpenAI’s expansion plans, and sector commentators weigh whether AI stocks are overheated or poised for corrections. The trend lines include billion-dollar rounds, a flurry of collaborations, and ongoing debates about the pace and profitability of AI initiatives.
Industry voices from broader markets, including SK Group’s chairman Chey Tae-won, caution that stock prices may adjust after rapid rallies, while analysts from research firms like Forrester note that some large organizations plan to defer significant AI spending into 2027 due to gaps between vendor promises and real-world outcomes. Bank of England officials similarly warn of the risk of a sudden market correction tied to AI equities.
In summary, leaders across tech and finance portray AI as a transformative, long-term cycle rather than a speculative bubble. Yet the path forward remains contested, with questions about ROI, the speed of adoption, and the right balance of investment and risk. What’s your take: is this AI surge sustainable, or are we heading toward a correction? Share your perspective in the comments.
