Warren Buffett's Successor, Greg Abel, Is Wagering Heavily on an AI-Driven Future

Berkshire Hathaway's new CEO Greg Abel more than tripled the company's Alphabet stake in Q1 and added another $17 billion in Q2, making Alphabet Berkshire's third-largest holding. Abel cited Alphabet's AI capabilities and Berkshire's internal experience with AI as key investment drivers, while Google Cloud revenue rose 82% year over year to an annualized run rate above $99 billion. Abel also sees AI data-center expansion supporting steady electricity-demand and operating-cash-flow growth at Berkshire Hathaway Energy, although he emphasized that infrastructure investments should not raise costs for existing customers.

Analysis

The investable implication is less Berkshire's portfolio shift than external validation of Alphabet as a durable AI monetizer rather than a model-training spender. A large, long-duration holder can reduce perceived governance and capital-allocation risk around Alphabet's AI capex, supporting multiple resilience through the next two earnings cycles; however, the cited cloud growth figures require confirmation in Alphabet's filings because an unusually high growth print may reflect revenue-recognition, comparison-base, or segment-definition effects rather than a sustainable run rate.

Berkshire's regulated-utility exposure creates a cleaner read-through to the AI power bottleneck than BRK.A itself. Data-center load growth benefits generation owners with contracted capacity and transmission/construction vendors—CEG, VST, PWR, EME and ETN—but the economics split sharply by jurisdiction: utilities facing retail-rate caps or mandated customer protections can absorb the grid-upgrade cost before earning allowed returns. Over 6-18 months, the scarce asset is interconnection and dispatchable power, not simply electricity demand.

Near term, the likely market response is incremental institutional demand for GOOG and AI-power proxies rather than a fundamental step-change. The contrarian risk is that consensus extrapolates data-center demand into unbounded utility earnings: hyperscalers are increasingly pursuing self-generation, power-purchase agreements and geographic load shifting, which can cap merchant-price upside after 2027. The thesis is falsified if Google Cloud growth decelerates without corresponding operating-margin expansion, or if utility regulatory dockets disallow timely recovery of incremental transmission and generation investment.

Key Decisions for Investors

• Accumulate GOOG on post-earnings volatility only if cloud growth and segment operating margin both meet or exceed prior-quarter levels; target a 6-12 month long versus equal-dollar short QQQ to isolate company-specific AI monetization. Exit on two consecutive quarters of cloud-growth deceleration without margin leverage, which would reframe AI capex as dilution rather than a premium multiple catalyst.
• Establish a 6-12 month pair: long CEG / short a diversified regulated-utility ETF such as XLU. CEG has more direct exposure to incremental high-load demand and contracted clean-power scarcity, while XLU contains rate-base utilities vulnerable to political resistance on customer bills; size modestly because power-price and nuclear-outage risk can dominate the spread.
• Prefer PWR and ETN over broad utility exposure for a 12-18 month grid-build thesis. Their revenue is tied to transmission, electrification and power-quality capex irrespective of whether merchant generators or regulated utilities capture the ultimate electricity margin; reassess if hyperscaler capex guidance or utility transmission-award pipelines weaken.
• Do not initiate a standalone BRK.A long solely on this signal. Use Berkshire's next disclosed portfolio filing and Berkshire Hathaway Energy capital-plan/rate-case updates as watch items; a meaningful uplift requires evidence that incremental load earns acceptable regulated returns without customer-subsidy constraints.