Guideposts: MIT AI Report Misreads the Market: Early Failures Don't Mean an AI Bubble | | By Dr. Robert Castellano
08/25/2025 | | | SPONSORED CONTENT How to Get A Double in 3 to 10 Days Hello. My name is Jim Fink.
A unique trade just triggered in my system that you need to see to believe.
This "Friday Phenomenon" could generate returns of up to 100% or more… in the next 3 to10 days… and continue to pay out week after week.
However, I can only send the details on this trade to a limited number of investors. And I'd like you to be one of them.
Due to the nature of this trade, I have to keep the numbers small to make sure everyone I share this information with has an equal opportunity to get in on the action.
Get in on the action by clicking this link. | | | A recent MIT report claiming that 95% of companies see zero returns from their generative AI investments is a prime example of the "publish or perish" plague where research is presented in ways to make headlines rather than provide balanced insight.
I've seen it for decades from my time at Stanford University, and from my role as Editor-in-Chief of the scholarly Journal of Active and Passive Electronic Devices. But the MIT report is an egregious example.
The study made headlines, triggered a sudden market sell-off, and raised questions about whether AI is already in bubble territory. In reality, the study is just one more example of reports that got it wrong by measuring transformative technologies against short-term profitability.
Technology Diffusion and Early Low Returns
Every major technological revolution began with an extended period of disappointment before it transformed industries. When electricity was first introduced into factories, productivity gains were negligible because plant layouts remained tied to the architecture of steam power. Only when managers redesigned production lines did electricity's potential begin to be fulfilled—a process that took nearly two decades. Computers, installed in offices during the 1960s and 1970s, were initially treated as glorified calculators that did little to change how work was done. It took a full 15 years before their integration into enterprise workflows began to pay measurable dividends.
Remember the dot-com boom of the 1990s: thousands of websites with no business models went nowhere. Yet within a decade, online commerce, advertising, and communications had become indispensable. Cloud computing went through a similar phase in the 2000s—dismissed as an expensive experiment with no return on investment (ROI) until scale, standardization, and enterprise trust unlocked its true value.
Table 1 shows how generative AI fits squarely into this historical sequence. Today's lack of immediate ROI is simply evidence that AI is in the same stage that electricity, the internet, and the cloud once occupied: early, awkward, and misunderstood. Just as Amazon and Google grew from internet "failures," and AWS grew from the cloud's early skepticism, companies like Nvidia (NASDAQ: NVDA), Microsoft (NASDAQ: MSFT), Meta (NASDAQ: META), and Google (NASDAQ: GOOG) are laying the groundwork for AI's inevitable maturation.
Refuting MIT's Framing
The MIT report repeats the same mistake by framing the 95% of early projects that showed no ROI as failures. In reality, generative AI is encountering the same barriers electricity once did—companies are layering it onto existing workflows instead of redesigning processes around it. Businesses today need new operating models that integrate AI at the core. The MIT report mistakes transitional friction for structural weakness. Nvidia, Microsoft, and Google are not selling short-term productivity boosts; they are building the platforms that will anchor global enterprise for decades. | | $11.5 Trillion Material Sparks New Tech Revolution This once-in-a-lifetime chance lets you profit from the $11.5 trillion revolution a tiny company is sparking. Claim your stake in the next Amazon-sized surge through this secret investment -- revealed for free. Top scientists are producing this steel-beating miracle material that will shape the future of tech. Tap into massive wealth BEFORE this tech revolution reshapes our world. | | | The 5% Success Rate as the Real Signal
The same report that emphasizes failures also acknowledges that 5% of projects achieved extraordinary success, with startups leaping from zero to $20 million in revenue in a year. These successes are easily dismissed as anomalies, but they are the very indicators investors should focus on. In every prior technological cycle, the early winners looked like outliers before they became industry-defining. Amazon (NASDAQ: AMZN) in e-commerce, Google in search, and Netflix (NASDAQ: NFLX) in streaming were once small exceptions in a sea of failures. Today, they represent trillions of dollars in market value.
Table 2 shows that the dichotomy between the 95% that fail and the 5% that succeed is not a verdict on AI's futility, but a reflection of how disruptive technologies scale. The 5% are the seeds of future giants, and their rapid revenue growth confirms that generative AI already has commercially viable pathways. Nvidia's GPU dominance, Microsoft's integration of OpenAI, and Meta's open-source push show that the commercial winners are already visible, even if most companies remain stuck in pilots.
Buying vs. Building: A Familiar Pattern
Another headline figure in the MIT report shows that of companies attempting to build proprietary systems internally, only about one-third achieved any measurable success. By contrast, two-thirds of those that bought tools from external vendors reported successful deployments.
Table 3 shows this is another familiar pattern in enterprise technology. The same divide played out in cloud computing, CRM, and office software. Companies that tried to build their own equivalents fell behind, while those that adopted standardized platforms—AWS, Salesforce, Microsoft Office—thrived. The message for investors is clear: AI platform providers will capture the lion's share of profits. Nvidia, Microsoft, and Google are already benefiting from this buy-not-build reality. Meta's open-source approach is positioning it as the low-cost platform alternative. The report's numbers do not undermine the AI investment case—they validate it.
Valuations, Rotation, and Market Context
The Nasdaq 100 trades at nearly 27 times forward earnings, roughly a third higher than its long-term average. Such valuations make technology stocks hypersensitive to negative headlines. The sell-off triggered by the MIT report was not evidence that investors no longer believe in AI—it was evidence that valuations were stretched and ripe for a pullback. Defensive rotations into staples and utilities are cyclical and have occurred in every market era, from the dot-com boom to cloud computing's early years. Nvidia's 3.5% dip on the day of the MIT headlines, followed by rebounds in subsequent sessions, reflects this cyclical volatility rather than any structural weakness in AI's trajectory. | | The 1 Sector You Want To Be In This Week Every market rotation tells a story. This week's story is about one sector that's dominating while others struggle.
VantagePoint's neural networks identified this opportunity by analyzing patterns that most traditional analysis overlooks. The A.I. processes global financial interconnections to predict which sectors will benefit from institutional rotation before the movements become obvious.
[Discover the sector A.I. is targeting →] | | | MIT's Numbers Reinterpreted
The clearest way to understand the gap between the MIT report and investor reality is to see how the same numbers can be spun in opposite directions. The statistics themselves are not in dispute—95% of projects show no immediate ROI, 5% do, and buying tools from vendors outperform building in-house. MIT chose to emphasize the negatives, portraying the 95% as failures and the short-term lack of ROI as proof of hype. But seasoned investors view the same figures differently: early high failure rates are the hallmark of every new technology, and the small group of early successes are exactly the companies that become industry leaders. The "buy vs. build" data is proof that vendors like Nvidia, Microsoft, Google, and Meta are already in the strongest position to profit.
Table 4 shows the MIT report and the investor lens diverge not on facts but on framing. MIT highlights the failures; investors recognize the early signals of success.
The same statistics that MIT calls alarming are the signs seasoned investors recognize as evidence of the early stage of technological transformation.
Vendors Already Prove the Case
While corporate users may struggle to monetize AI quickly, the vendors enabling AI adoption are already reporting massive revenue growth. Nvidia's data center revenue has surged more than 100% year over year. Microsoft has guided to a steady rise in AI-driven Azure demand. Google Cloud continues to expand, while Meta has raised its CapEx guidance explicitly to fund AI infrastructure.
According to Table 5, the vendors themselves are monetizing AI now—disproving the idea that generative AI is too immature to generate real money. The companies selling the picks and shovels of the AI gold rush are already cashing in.
Investor Takeaway
According to history, the 95% of early failures are the compost from which the 5% of winners grow. Investors should look past short-term disappointment and focus on the platform providers and standout adopters already proving that AI can generate revenue. Just as electricity, computers, the internet, and cloud computing reshaped entire industries, generative AI will follow the same trajectory. The question is not whether, but when. Nvidia, Microsoft, Meta, and Google are already ensuring that the "when" is closer than skeptics imagine.
Sincerely,
George Gilder, Richard Vigilante, Steve Waite, John Schroeter, and Robert Castellano
Editors, Gilder's Guideposts, Technology Report, Technology Report Pro, Moonshots, and Private Reserve | | About George Gilder:
 George Gilder is the most knowledgeable man in America when it comes to the future of technology and its impact on our lives. He’s an established investor, bestselling author, and economist with an uncanny ability to foresee how new breakthroughs will play out, years in advance. George and his team are the editors of Gilder Technology Report, Gilder Technology Report Pro, Moonshots and Private Reserve. | | | | | |
Visit our YouTube Channel - Eagle Investing Network
No comments:
Post a Comment