AI Snake Oil—Part 1: The Skeptics
AI stocks continue to make new highs, not on fundamentals but on a simple narrative that the public is thoroughly convinced they understand. CEOs of the largest AI companies have been hard-selling their products for years and the mainstream media is happy to oblige without any real skepticism or investigative reporting.
Dario Amodei said a few months ago that Anthropic models may have a conscience. Elon Musk said AI will make every human wealthy and there will be data centers in space. The CEO of Google DeepMind said AI will cure all diseases. The CEO of Microsoft AI said all white-color work will be automated by August of 2027. Meta CEO Mark Zuckerberg said that they are working (in their “Superintelligence Lab”) on models that are showing signs of self-improvement. A new AI start-up says their AI can read your mind.
While the mainstream media treats these pronouncements as foregone conclusions, a growing number of skeptics including a large number of computer scientists, independent tech reporters and industry defectors are exposing the hype and snake oil sales tactics.
The volume and depth of material produced by AI skeptics is overwhelming but has been entirely ignored by mainstream media who—with very few exceptions—are committing journalistic negligence on a mass scale. Like the Whitehouse Press Corps, their access is bought with complicity. When the AI bubble bursts, their complicity will be well-documented.
I’m no AI expert, but I’ve spent a significant amount of time and effort learning from some. I’m deeply interested in trying to understand what is actually happening with AI so that I can form an intelligent opinion about what the future may hold for this transformative technology. If we can gain a better understanding of what’s to come, we may be able to position our portfolios ahead of other investors in the markets.
In Part 1 of AI Snake Oil, I want to focus on three topics. First, I’ll briefly outline my opinion about the current state of AI. Next, I want to introduce you to the skeptics I’ve learned from and share some resources so you can dig further, if you wish. Finally, I want to share my adopted vision for the future of AI. Specifically, what the AI landscape may look like after the over-investment bubble bursts.
The Current State of AI… in my humble opinion
AI is a revolutionary technology—the most important technological development since the internet. I agree with Nvidia CEO Jensen Huang that AI will soon be used by every major industry in the global economy.
AI companies and the mainstream media have been intellectually dishonest about the technology and are more interested in their own success than successfully delivering value to humans. The AI bubble is not a “technology bubble;” it’s the largest industrial over-build in American history. It’s an over-investment bubble that will become obvious to the market soon.
The biggest AI winners in the public markets will not be Meta, Google or Microsoft or massive private companies like OpenAI, Anthropic or xAI. They will be the small companies that emerge after the unsustainable investments implode. The winners will be the firms that have a deep understanding of the technology and how to deploy it to users. They are waiting for their opportunity to step in and deliver real value in the form of efficient AI solutions that solve real world challenges.
To summarize my view in a single sentence, I’m bullish on the technology and bearish on the hyper-scalers and their debt-engorged clients selling their LLMs as the all-powerful be-all and end-all.
The Skeptics
In Part 2 of AI Snake Oil, I will go deeper into the evidence of the AI investment bubble. Today I want to introduce some of the skeptics who have uncovered or publicized evidence of the AI mirage. If you’re watching this on YouTube, my client letter has 25 links to all of the resources mentioned. If you’d like a copy, email me: brian@alpharockinvestments.com.
In no particular order:
Julian Whatley spent 35 years in Hollywood and Madison Avenue learning how to manufacture perception. Whatley has produced an excellent 3-part series on the gap between what the AI industry is selling and what the physics actually allow. If you only check out one source from my list, watch his series: Episode 1: The Money Furnace, Episode 2: The Physics of the Collapse, and Episode 3: What Triggers the Collapse.
Arvind Narayanan and Sayash Kapoor are the authors of AI Snake Oil: What Artificial Intelligence Can Do, What It Can't, and How to Tell the Difference. Narayanan is a professor of computer science at Princeton University and director of their Center for Information Technology Policy. Kapoor is a PhD candidate in computer science at Princeton University. Their book was published in 2024 and focuses on separating AI hype from reality, discussing the differences between generative and predictive AI, and identifying "snake oil" products that fail to deliver on their promises.
Cal Newport is a Professor of Computer Science at Georgetown University, where his academic research focuses on distributed algorithms. He has a PhD in computer science from MIT and has posted much of his research on AI on his blog and YouTube channel, Deep Questions. Cal’s vision for how companies will use AI in the future is one that I’ve adopted and explain in the next section.
Ed Zitron is a technology reporter with confidential sources inside many AI firms. He has been published in The Atlantic, Business Insider, Wall Street Journal, USAToday and TechCrunch. Much of his work is posted on his blog and a number of YouTube channels, including his own, Better Offline Podcast.
Gary Marcus is a psychologist, cognitive scientist, and author, known for his research on the intersection of cognitive psychology, neuroscience, and artificial intelligence. His Substack, Marcus on AI includes most of his work on AI and his recent interview with Steve Eisman, on the massive problems facing AI and LLM scaling is excellent.
Michael Burry is a former physician and former hedge fund manager, most famous for profiting from the 2008 subprime mortgage crisis. His story was immortalized in Michael Lewis's book The Big Short and the subsequent Oscar-winning film, where he was portrayed by Christian Bale. Burry publishes his market commentary on his Substack, Cassandra Unchained.
Patrick Boyle is a hedge fund manager, a university professor and a former investment banker. His YouTube channel covers a range of current and historical events in finance and asset markets. His analysis on circular financing in AI is exceptional and witty.
The Future of AI
Setting aside the AI hype, major consulting and tech analytics firms have reported on what’s actually happening on the ground. Firms like PricewaterhouseCoopers, Deloitte, Gartner, and McKinsey all show a consistent pattern: the majority of businesses are not seeing any meaningful benefit from AI.
According to PwC, just 12% of CEOs say AI has delivered both cost and revenue benefits and 56% say they have seen no significant financial benefit from AI. Less than one-third believe it will increase revenue in the near future.
While e-commerce and “digital-first" companies see more apparent gains through coding models, AI-driven advertising and product suggestions, these are largely niche cases. For traditional incumbents like McDonald’s, GE, or Sony, AI-driven growth remains more of an aspiration than a reality.
Twenty-five percent of companies reported cost savings due to AI, however, 22% reported AI actually increased their costs. McKinsey found McKinsey & Company—the global leader in management consulting—found that AI hasn’t infiltrated any single business function by more than 10%, and less than 2% of companies have fully scaled an AI solution within a single department.
For the vast majority of respondents—between 70% and 90%—no steps have been taken toward advanced implementations like Agentic HR. Currently, AI use is characterized by localized trials and pilot projects rather than large-scale restructuring.
As a large-scale business transformation tool, AI is still in its infancy. There are few commercially viable, scaled solutions for giants like Boeing or GSK to adopt. Consequently, corporate strategies are being driven by a core belief in AI’s eventual importance rather than clearly defined business cases. This creates a gap where the jump from high-level strategy to daily operations is weak. At the operational level, priorities feel scattered because companies struggle to identify and prove real savings.
CEOs may dream of cutting employee costs by implementing AI, but so far it’s not happening. AI implementation in the real economy is sparse and intermittent. When will the “AI revolution” arrive?
I don’t know. But I’ll leave you today with a vision of the future of AI that I’ve adopted from Cal Newport, who outlined his vision in this interview with Rob Montz (46:30):
The future of AI will include a major shift from centralized, monolithic "frontier" models toward a fragmented, highly specialized landscape Cal calls “Distributed AGI.”
The current "arms race" among tech giants—OpenAI, Anthropic, Google, and Meta—is built on the flawed dream of a singular, all-powerful interface. Investors hope for a massive LLM so dominant that every tool must pay to access it, but this frontier LLM pure-play is economically unsustainable. Massive universal oracle models are too expensive to run and lack the versatility for specific tasks. The economic utility of a trillion-parameter model does not justify the trillion-dollar valuations of the companies maintaining them—the bubble will burst.
The true technological revolution will not be driven by massive LLMs, but by hybrid systems that may include LLMs and models that understand specific physical or logical environments like policy networks to help with decision-making, bespoke architectures designed for specific use cases (e.g., specialized AI accounting, engineering and law).
Just as we don't need a high-performance F1 car to run a daily errand, we don't need a frontier model like Claude Sonnet to power audio software or monitor the weather. The future belongs to Honda CRV-like models: smaller, cheaper, and more efficient tools tailored to their specific functions.
Newport predicts we will achieve AGI (artificial general intelligence), but it won't look like a single chat interface. Instead, it will manifest as 10,000 different tools integrated into our daily workflows. Companies selling the F1 oracle model will struggle because their business model is cost-prohibitive.
As models become smaller and more specialized, the barrier to entry drops. This will trigger a wave of startups focused on finding product-market fit within niche industries—experimentation that giant tech firms cannot manage at scale.
We will know this future has arrived when professionals across various industries report that their existing tools have become exponentially more powerful. This unlocking of value will come from specialized systems addressing specific problems until they find the perfect fit, rather than waiting for a central AI to solve everything at once.
The frontier model bubble will burst, giving way to an era of specialized, cost-effective, and highly integrated AI systems that prioritize utility over scale.
THE ALPHA ROCK DIFFERENCE - Our investment strategies are a compelling alternative to traditional buy-and-hold investing. By focusing on liquid alternative investments and active risk-management, we target absolute returns, not benchmarks. To see how we’re invested and why, download our most recent Quarterly Investment Outlook, The End of Easy Money.
REVIEW YOUR INVESTMENTS - When market volatility increases, it’s a good time to review your investments, especially if you’re using a passive, buy-and-hold strategy. Please schedule a call or meeting)
We’d love to hear from you!
Your thoughts are important to me, so don’t hesitate to share them. They give me great motivation and encouragement.
IMPORTANT DISCLOSURE INFORMATION
This commentary reflects the personal opinions, viewpoints and analyses of the Alpha Rock Investments, LLC employees providing such comments, and should not be regarded as a description of advisory services provided by Alpha Rock Investments, LLC or performance returns of any Alpha Rock Investments, LLC client. The views reflected in the commentary are subject to change at any time without notice. Nothing in this commentary constitutes investment advice, performance data or any recommendation that any particular security, portfolio of securities, transaction or investment strategy is suitable for any specific person. Any mention of a particular security and related performance data is not a recommendation to buy or sell that security. Alpha Rock Investments, LLC manages its clients’ accounts using a variety of investment techniques and strategies, which are not necessarily discussed in the commentary. Investments in securities involve the risk of loss. Past performance is no guarantee of future results.
The S&P 500 Index or the Standard & Poor's 500 Index is a market-capitalization-weighted index of the 500 largest U.S. publicly traded companies. The S&P 500 is a float-weighted index, meaning company market capitalizations are adjusted by the number of shares available for public trading. Note: Investors cannot invest directly in an index. These unmanaged indices do not reflect management fees and transaction costs that are associated with most investments.
Alpha Rock Investments, LLC provides links for your convenience to websites produced by other providers of industry related material. Accessing websites through links directs you away from our website. Alpha Rock Investments, LLC is not responsible for errors or omissions in the material on third party websites, and does not necessarily approve of or endorse the information provided. Users who gain access to third party websites may be subject to the copyright and other restrictions on use imposed by those providers and assume responsibility and risk from use of those websites.
