By Nick Fisher, Portfolio Manager
AI is reshaping the world and the investment landscape is no exception, according to Dan Rasmussen and his team at Verdad. The following is a summary of their excellent work which cautions against hubris when investing in such a hot segment such as AI.
Over the past decade, betting against Silicon Valley innovation has proven to be a costly mistake for investors. Major technology firms have repeatedly defied skepticism, with companies like Tesla, Uber, and Netflix overcoming significant challenges to achieve substantial success. A similar scenario may unfold for artificial intelligence (AI), but in the short term, the fate of the U.S. equity market hinges on the success of the Magnificent 7 (Mag 7): Apple, Nvidia, Microsoft, Amazon, Alphabet (Google), Meta, and Tesla. These companies are investing heavily in AI, making high-stakes capital expenditures that could either redefine their industries or result in significant financial losses. My guess is the latter.
Market concentration has reached levels unseen since the dot-com bubble of 2000, with the Mag 7 allocating vast portions of their net income to AI-related capital expenditures. Goldman Sachs initially projected $1 trillion in AI investment, a figure that continues to rise. However, historical precedents from the shipping industry and the late-90s telecom sector suggest that simultaneous over investment by competitors often leads to declining returns and market inefficiencies.
Challenges to AI Profitability
The primary challenge facing AI investment is the uncertainty surrounding its economic viability. Many AI models, particularly large language models (LLMs), are capital-intensive, requiring massive computational power while generating limited immediate revenue. Unlike traditional software models, AI operations are probabilistic, consuming significantly more resources for simple tasks. The current AI expenditure model mirrors an industrial manufacturing approach, where high upfront costs may not yield expected long-term profitability.
The AI industry’s reliance on scaling laws suggests that bigger investments will lead to superior AI models. However, the recent emergence of China’s DeepSeek model—offering alleged comparable performance at a fraction of the cost—challenges the prevailing investment assumptions of the Mag 7. If lower-cost models continue to disrupt the market, U.S. tech giants could struggle to recover their massive capital expenditures. The historical fate of tech incumbents like IBM and Xerox serves as a cautionary tale—giant firms can be blindsided by smaller, more agile innovators.
Microsoft’s CTO, Kevin Scott, emphasizes that AI is a model, not a product. While AI applications like ChatGPT have demonstrated viral adoption, they have yet to generate sustainable revenue streams. AI’s long-term financial success will likely depend on its integration into broader digital infrastructure rather than standalone consumer products. Future winners in AI may not be today’s Mag 7 but rather specialized firms that build industry-specific solutions leveraging open-source AI and existing infrastructure.
While the Mag 7 have dominated the tech industry for years, their aggressive AI investments pose significant financial risks. AI-driven infrastructure may ultimately benefit newer, specialized companies rather than the incumbent giants. As history has shown, large-scale technological shifts often favor emerging innovators over established market leaders. Whether AI investment yields substantial ROI or becomes a cautionary tale of over investment remains to be seen.
In the meantime, we will continue to remain prudent in our investment principles, ignore the crowd, look for opportunities as they present themselves and focus on the business performance over a long period of time. This is and will always be the focus of our work at Pilot Wealth Management.