A wake-up call: NVIDIA's reassessment of OpenAI, why should it most concern Microsoft?

The AI collaboration chain formed by NVIDIA, OpenAI, and Microsoft is showing signs of loosening due to differences in "execution power" and "operational discipline."

On Wednesday, Bloomberg technology writer Parmy Olson stated in her latest column that Jensen Huang's hesitation regarding OpenAI is not just a shift in the direction of a single investment, but rather a reminder to Microsoft that exclusive models and intellectual property do not automatically translate into product advantages. Market competition is shifting from "whose model is stronger" to "who can implement it better."

NVIDIA CEO Jensen Huang recently told industry insiders that his previously announced $100 billion investment in OpenAI is "non-binding" and privately criticized OpenAI for lacking business discipline. Huang denied any dissatisfaction with OpenAI and told reporters last Saturday, "We will invest a significant amount of money," but he also chose to leave room for commitment.

Parmy Olson believes that this investment, originally tied to infrastructure expansion, is now more likely to appear in the current financing process of OpenAI in the form of "hundreds of billions of dollars," and is happening before its potential IPO. Reports also indicate that OpenAI is in discussions with NVIDIA, Microsoft, and Amazon regarding approximately $100 billion in financing, which is independent of the previously proposed infrastructure deal.

The more direct market impact falls on Microsoft. Despite Microsoft gaining exclusive access to OpenAI's intellectual property and models until 2032 through a restructuring agreement announced last October, and integrating Copilot with OpenAI's latest GPT-5 model by August 2025, the product experience of Copilot is still criticized by users, and its functional form lags behind competitors, highlighting the contradiction of "holding top models but struggling to realize advantages."

Huang's "Buyer's Remorse"

According to Parmy Olson, the AI industry has long been "numb" to massive bets, making Huang's caution regarding commitments particularly noteworthy. The Wall Street Journal mentioned that Huang not only emphasized that the investment is non-binding but also privately questioned OpenAI's management and execution discipline, casting variables on OpenAI's financing narrative and partner confidence.

For investors, this statement signals two layers: first, the capital and computing power planning around OpenAI may still be rapidly adjusting; second, even one of the core industry partners is reassessing risk exposure and return paths.

Parmy Olson pointed out that Sam Altman's management style continues to cause unease among outsiders, including his dramatic firing at the end of 2023 and subsequent return, as well as a series of complex and astonishing transaction arrangements. She wrote that these arrangements have burdened OpenAI with commitments of up to $14 trillion in computing resources, approximately 100 times OpenAI's projected revenue for 2025.

On the product side, OpenAI's progress also appears rushed. Parmy Olson cited that OpenAI attempted to establish a developer market through the GPT Store and custom GPTs, but due to a lack of clear strategy, it "fizzled out." Between "model capabilities continuously leading" and "commercialization and organizational execution fluctuations," OpenAI's partners need to price for uncertainty.

Microsoft's Execution Dilemma

Microsoft's capital return on the books is extremely impressive. Parmy Olson wrote that Microsoft's early investment of $13 billion in OpenAI corresponds to about 27% equity, with a valuation of approximately $135 billion, more than ten times the original investment; at the same time, the restructuring agreement allows Microsoft to enjoy a cleaner exclusive intellectual property and model access arrangement before 2032.

However, at the product level, Microsoft has not fully realized this advantage. Parmy Olson points out the key issue in the article: Since OpenAI can still continuously launch one of the world's strongest AI models, why does Microsoft's flagship product Copilot, based on OpenAI technology, still lag behind the competition? She wrote that user feedback on Copilot focuses on "confusing, limited, and difficult to use," highlighting the gap between model quality and product implementation.

The pressure of competition has concentrated on the application side. Parmy Olson wrote that last month, Anthropic released Claude Cowork, an application that reportedly completed its self-developed AI coding tool in 10 days and can operate personal computers, organize files, generate PowerPoint and Excel based on documents, and respond to LinkedIn messages after obtaining authorization.

In contrast, she pointed out that despite Microsoft having Windows, Office, and LinkedIn, Copilot cannot achieve the aforementioned capabilities. This gap makes it easier for the market to attribute the problem to Microsoft's internal research and productization mechanisms rather than solely to the model itself.

Parmy Olson cites David Rainville, head of Sycomore Sustainable Tech, who states that some industry observers are paying attention to the direction of Microsoft's AI project leadership. He indicated that if Microsoft cannot release a product equivalent to Claude Cowork in the next six months, "someone will have to step down," and noted a clear disconnect between "model quality" and "execution capability" at Microsoft