‘Like 10 Manhattan Projects going off all at once.’ How AI is rewiring the global economy, says this BlackRock exec.

BlackRock tech investing head Tony Kim argues the AI trade has staying power because it is rewiring the global economy. The article is a bullish thematic call on artificial intelligence and technology innovation, but it contains no new earnings, guidance, or policy data. Market impact is likely limited to sentiment around AI-related stocks rather than immediate price-moving news.

Analysis

The important second-order read-through is that AI enthusiasm is no longer just a sector multiple story; it is becoming a capex supercycle with multi-year spillovers into power, semis, networking, construction, and industrial automation. When investors start framing AI as a broad economic reconfiguration rather than a software fad, the beneficiaries widen beyond the obvious platform names toward the physical bottlenecks: grid interconnects, turbines, transformers, datacenter REITs, and high-bandwidth infrastructure. That broadening should support a longer-duration capital spend cycle even if near-term software monetization remains uneven.

The market is likely underestimating how concentrated the near-term winners are. In the first 6-18 months, returns should accrue disproportionately to picks-and-shovels suppliers with pricing power and order backlogs, while many application-layer companies face margin pressure from heavier compute costs and customer reluctance to pay for undifferentiated AI features. A further nuance: the more AI workloads scale, the more power availability becomes the gating factor, creating a bottleneck that can shift bargaining power toward utilities and grid-equipment vendors even if headline AI demand cools.

The main risk is not that AI adoption slows, but that the trade gets crowded and re-rates too far ahead of cash flow conversion. If hyperscaler capex normalizes or regulators force more disciplined spending, the highest-multiple beneficiaries can de-rate quickly over a 1-3 month horizon, while the industrial enablers hold up better because their demand is tied to installed capacity, not sentiment. The contrarian view is that the market may still be early on the infrastructure leg of the AI cycle but late on the pure narrative beneficiaries; that argues for favoring tangible asset-backed exposure over expensive AI software proxies.

Key Decisions for Investors

• Overweight semicap and AI infrastructure suppliers relative to software over the next 3-6 months; prefer names with backlog and pricing power, as those are the most likely to convert AI demand into visible earnings.
• Long utilities/grid-equipment beneficiaries vs short high-multiple AI software baskets for a 6-12 month pair trade; thesis is that power and interconnect constraints force real capex, while software monetization remains more speculative.
• Add tactical exposure to BLK on pullbacks only; the strategic benefit is improved asset-gathering from the AI theme, but the stock likely needs market-wide AUM growth to re-rate, so reward is lower than in direct infrastructure beneficiaries.
• Use call spreads rather than outright longs on the most crowded AI winners over the next 1-2 months; upside can continue, but convexity is preferable because sentiment-driven drawdowns can be abrupt if capex guidance disappoints.
• Monitor hyperscaler capex and utility load-forecast revisions as the key catalyst set; if either inflects down, reduce AI beta quickly and rotate toward industrials with contracted demand.