AI Era, Is Google Ultimately the Winner? Dan Niles, who predicted Intel's revival, talks about Anthropic, OpenAI's decisive limitations | Dan Niles [Global Money Talk]
Speakers
Dan Niles
— Founder & Portfolio Manager, Niles Investment Management
Summary
Dan Niles discusses the AI capex cycle, highlighting explosive revenue growth at Anthropic and OpenAI driven by enterprise agentic AI adoption. He argues that Google is best positioned to dominate AI due to its full-stack advantage, while Intel benefits from the shift toward CPUs in agentic workloads. He warns that software sector budgets are being cannibalized by AI spending, making it an unattractive area for investment.
- Enterprise agentic AI is driving token demand growth of 120% in two months, supporting continued hyperscaler capex.
- Google's in-house AI chip, Android distribution, and cash flow give it a structural edge over rivals.
- Intel's CPU importance is rising relative to GPU as the compute model shifts for agentic AI.
- Intel's foundry is attracting major customers, but execution risk remains.
- Software sector faces existential risk as enterprise AI spending crowds out traditional software budgets.
- Dan Niles avoids software entirely, comparing the current environment to the dot-com bust.
- S&P 500 valuation is elevated but can stay high due to strong AI infrastructure demand.
Trade Ideas
Google's full stack makes it AI winner.
Google has the full stack — its own AI chip (TPU) avoiding the NVIDIA tax, massive advertising cash flow to fund AI capex, Android distribution covering ~80% of global smartphones, and a leading AI model (Gemini). These advantages position Google to become the dominant winner in the AI era, unlike OpenAI or Anthropic which lack the same infrastructure and monetization breadth.
Intel benefits from CPU shift in AI.
Intel stands to benefit from the shift from GPU-heavy to CPU-heavy workloads as agentic AI gains traction, with the GPU:CPU ratio moving from 8:1 toward 1:1. Additionally, Intel's foundry business is winning interest from major customers like Tesla, Nvidia, Google, and possibly Apple, offering significant upside potential if execution succeeds. However, the stock could still drop 15% in a macro shock, so it's a longer-term watch.
Avoid software sector due to AI cannibalization.
Enterprise AI spending by OpenAI and Anthropic (run-rate revenue exploding from $70B to ~$700B) is cannibalizing budgets from traditional software vendors. CFOs will cut other software spending to fund AI, leading to a wave of failures in the software sector, similar to the dot-com bust where even strong companies like Amazon lost 95% of their value. Therefore, he avoids the entire software sector.
Up Next
This 3PRO TV (삼프로TV) video, published May 17, 2026, features Dan Niles discussing GOOG, INTC, IGV. 3 trade ideas extracted by AI with direction and confidence scoring.