Nvidia CEO Jensen Huang predicted a “crazy good” fourth quarter following the company’s strong earnings report on November 19, 2025, a performance that helped calm fears of an AI bubble amid a broader market selloff as he told investors, “We excel at every phase of AI.” However, SoftBank’s $5.8 billion Nvidia stake sale on November 11, 2025, stirred fresh anxiety by signaling that one of the sector’s most aggressive backers might see overvaluation in AI investments. Even as AI leaders touted agents and a more practical phase for the industry on November 13, 2025, investors remain skeptical, with bubble concerns easing only slightly by November 20, 2025 as markets wait for AI to deliver on its promise.
SoftBank’s Stake Sale Sparks AI Bubble Worries
SoftBank’s decision to sell a $5.8 billion stake in Nvidia, announced on November 11, 2025, immediately raised questions about whether one of the most visible champions of artificial intelligence was quietly calling time on the sector’s froth. The transaction, detailed in reporting on SoftBank’s $5.8B Nvidia stake sale and its ties to AI investment and OpenAI, was interpreted by traders as more than routine portfolio management, because it involved a company that has been central to the AI boom and to SoftBank’s own narrative about the future of computing. By trimming exposure to Nvidia at a moment when AI valuations were already under scrutiny, SoftBank signaled that even insiders with long time horizons were willing to lock in gains rather than ride out the next phase of the cycle.
The timing of the sale, coming in the middle of intense AI hype and just ahead of Nvidia’s latest earnings, contrasted sharply with the upbeat tone from chipmakers and software firms that have been promising years of growth from generative AI and large language models. Investors who had been watching Nvidia’s market capitalization surge on expectations of unrelenting demand for data center GPUs saw the SoftBank move as a potential inflection point, a reminder that even strategic investors can decide valuations have run ahead of fundamentals. For pension funds, retail traders using platforms like Robinhood, and sovereign wealth funds that have piled into AI-linked exchange traded funds, the sale underscored the risk that early backers might be exiting while public markets are still pricing in perfection, reinforcing the sense that the AI trade could be vulnerable to any disappointment in earnings or adoption.
AI Leaders Push Back on Bubble Narrative
Just two days after SoftBank’s sale rattled sentiment, a group of AI industry leaders moved to counter the idea that the sector was in a speculative bubble, arguing that the technology was entering a more grounded, practical phase. In a detailed account of that shift, coverage of AI leaders touting agents and waving off bubble fears as the industry enters a practical phase described how executives and founders highlighted the rise of AI “agents” that can perform multi-step tasks, integrate with enterprise software, and automate workflows in areas such as customer support, code generation, and financial analysis. By emphasizing concrete use cases rather than abstract promises, these leaders sought to reassure investors that the current wave of spending on AI infrastructure is tied to real demand from businesses that are already deploying the technology at scale.
Those same leaders also pushed back explicitly on concerns about overvaluation, arguing that comparisons to the dot-com bubble overlook the revenue already flowing from AI services and the structural need for accelerated computing in cloud data centers. Their message, as relayed in the November 13, 2025 discussions, was that AI is moving from experimentation to deployment, with companies in sectors from automotive to healthcare embedding AI agents into products like driver-assistance systems in 2025 model-year vehicles and diagnostic tools inside hospital software. For institutional investors weighing whether to stay in Nvidia and other AI names after the SoftBank sale, that framing matters, because it suggests that current capital expenditures on GPUs and AI platforms are the foundation for multi-year productivity gains rather than a short-lived speculative mania.
Nvidia’s Earnings Boost CEO Confidence
Nvidia’s latest earnings report, released on November 19, 2025, gave Jensen Huang a powerful data point to support his argument that AI demand is not a passing fad. In an account of the results, coverage of Nvidia’s earnings report in which Huang said “We excel at every phase of AI” and quelled Wall Street fears of an AI bubble amid a market selloff described how the company’s performance reassured investors who had been bracing for signs of slowing growth. Huang used the earnings call to stress that Nvidia’s business spans training, inference, and deployment of AI models, telling analysts, “We excel at every phase of AI,” and positioning the company as an indispensable supplier to cloud providers, enterprise customers, and AI startups alike.
That confidence carried into the following day, when Huang predicted a “crazy good” fourth quarter, a phrase highlighted in reporting on how Nvidia’s CEO predicts a “crazy good” Q4 after strong earnings calm AI bubble fears, and framed Nvidia’s outlook as a direct rebuttal to bubble talk. By tying his optimism to concrete metrics from the just-reported quarter, including continued strength in data center revenue and robust orders from hyperscale cloud operators, Huang aimed to show that demand for AI accelerators is broad-based rather than dependent on a handful of speculative projects. For portfolio managers who had watched Nvidia’s share price swing sharply during the market selloff, the combination of strong results and a bullish forecast reduced immediate panic, but it also raised the bar for future quarters, since any stumble could revive questions about whether the AI buildout has peaked.
Investors Remain Cautious Amid Easing Fears
Even with Nvidia’s upbeat guidance and the industry’s pivot toward practical applications, many investors are not ready to declare the AI bubble debate settled. A detailed look at market sentiment in coverage of how bubble fears ease but investors are still waiting for AI to live up to its promise noted that by November 20, 2025, concerns had eased compared with the immediate reaction to SoftBank’s sale, yet skepticism persisted about whether AI can deliver earnings growth commensurate with current valuations. Fund managers and analysts cited in that reporting pointed out that while Nvidia’s numbers justify its recent rally in the short term, the broader ecosystem of AI software and services companies still needs to prove that pilot projects can scale into durable, high-margin businesses.
That lingering caution reflects a broader pattern in Wall Street’s response to transformative technologies, where early enthusiasm often gives way to a demand for hard evidence of returns on investment. Events like SoftBank’s $5.8 billion stake sale, the November 13 push by AI leaders to emphasize agents and practicality, and Huang’s “crazy good” fourth quarter prediction have become reference points in a debate over whether AI is a once-in-a-generation platform shift or a classic case of over-extrapolated expectations. For corporate technology buyers deciding whether to commit to multi-year AI infrastructure contracts, and for policymakers watching the concentration of power in a handful of chip and cloud providers, the outcome of that debate will shape not only stock prices but also the pace at which AI tools move from experimental pilots into the core of the global economy.
