Buffett's AI Decision: Why Sell Apple and Heavy on Google?
Berkshire Hathaway-B (BRK.B), led by Warren Buffett, has released its Q3 13F report. Berkshire Hathaway newly acquired 17.85 million shares of Alphabet Class A stock in Q3, spending approximately $4.3 billion, making it the tenth-largest holding. Meanwhile, Berkshire continued to reduce its Apple holdings. Amid the global AI wave, this transaction has been interpreted by the market as the "Oracle of Omaha" giving a clear endorsement to Google's AI strategy.
Coincidentally, from late this month to early December, Google is set to release Gemini 3.0, dubbed the "GPT-5 killer" by the industry. The question arises: Should ordinary investors jump on this bandwagon now? Decoder will break it down for you.
Why did Buffett suddenly go heavy on Google in 2025?
First, Apple's ceiling is gradually apparent. A decline in the first three quarters + a Q3 rebound, its premium pricing strategy is under pressure from local brands and government subsidies, but mitigated by strong iPhone 17 sales and rare discounts in China. Its forward P/E ratio has exceeded 32x, far above historical averages. Apple holdings have dropped from 280 million shares to 238.2 million, with the current investment size reduced to $60.7 billion. Although still its largest holding, it clearly reflects Buffett's cautious stance on Apple's future growth potential.
Second, Alphabet's valuation has returned to the "value investing comfort zone." Its current Forward P/E Ratio is the lowest among the Magnificent 7; free cash flow yield is around 5.2-5.5%, well above the U.S. stock average; net cash reserves stand at $98.4 billion, with buybacks + dividends increasing steadily.
What makes Gemini 3.0 so strong?
| Capability | Gemini 3.0 (Coming Soon) | Current Top Competitors (GPT-5 / Claude 4.5) |
| Context Window | 1M-2M tokens | 128k-200k |
| Code Generation | Full project-level (generates 2000+ lines of runnable code at once) | Requires multiple iterations |
| Native Multimodal Support | Real-time video understanding + generation, 3D environment modeling | Partially requires plugins |
| Reasoning Architecture | Built-in Deep Think mode (automatically plans multi-step tasks) | Still needs external Agent frameworks |
| Enterprise Deployment Cost | ~40-60% cheaper than GPT-5 (Google Cloud internal estimates) | High |
More importantly, Gemini 3.0 will be deeply integrated into: Google Search (AI Overviews already has 1.5B+ monthly active users); Gmail/Docs (high enterprise conversion rates); Android 16 (1.4B global shipments expected in 2026).
This means Google has finally turned "technological leadership" into a "business moat."
Let’s look at key financials. Q3 total revenue: $102.3B (+16% YoY); Google Cloud revenue: $15.2B; net profit $34.98B; diluted EPS $2.87; free cash flow $24.5B; CapEx (quarterly) $24B.
Decoder believes that for investment horizons exceeding 2 years, Alphabet is currently a "Buffett-style" ideal pick. Its moat is strengthening: AI + Search + Android triad; valuation is undervalued: 22x P/E + 5.8% FCF yield; growth is accelerating: Cloud + AI dual engines.
But don’t just focus on the positives. Short-term volatility risks exist. If Gemini 3.0 is delayed or underperforms at launch, a pullback may occur; it’s still in a high-CapEx phase: margins may face short-term pressure in 2026.
Buffett cast a $4.3B vote of confidence; Gemini 3.0 will soon deliver the technical answer.
$Berkshire Hathaway B(BRK.B.US) $Alphabet(GOOGL.US) $Apple(AAPL.US)
The copyright of this article belongs to the original author/organization.
The views expressed herein are solely those of the author and do not reflect the stance of the platform. The content is intended for investment reference purposes only and shall not be considered as investment advice. Please contact us if you have any questions or suggestions regarding the content services provided by the platform.
