Investing in AI for More Jobs, Less Pay

AI investment is accelerating to historic levels following Nvidia’s reported $500B partnership with asset managers to finance AI infrastructure, signaling sustained capital deployment into AI. Former IBM CEO Sam Palmisano argues this wave is driven by strategic “goals and milestones,” reflecting a structured buildup rather than ad hoc spending. Apollo’s Torsten Slok suggests AI could increase job creation but compress pay, implying broader labor-market distributional effects alongside continued investment momentum.

Analysis

The market’s real edge here is not the abstract AI story; it is the widening gap between firms that sell scarce compute and those that merely talk about transformation. NVDA remains the clearest beneficiary because broader financing lowers the hurdle rate for buyers and extends the capex runway, which should support order visibility and multiple durability over the next 1-3 months.

The second-order effect is margin pressure elsewhere in the enterprise stack: as buyers demand faster payback, software and services vendors will face tougher ROI scrutiny, and that tends to favor vendors with direct productivity capture over open-ended consulting narratives. IBM can still benefit from integration demand, but it is a slower, lower-conviction way to play the cycle unless the company can prove that AI is pulling through recurring services revenue rather than just displacing labor.

Contrarianly, the consensus may be underestimating how long “more jobs, lower pay” can coexist with heavy AI spend; that is supportive for infrastructure demand but not automatically bullish for end-market monetization. If enterprise budgets start to reallocate from pilots to production, the winners will be constrained by power, packaging, and supply chain capacity, while financing-sensitive players could see a fast unwind if capex discipline returns. For ARI, the only relevant angle is whether AI data-center financing becomes a crowded private-credit theme; without disclosed exposure, it is more a watch item than a trade.

Key Decisions for Investors

• Go long NVDA on pullbacks over the next 1-3 months; this is the cleanest way to express a prolonged AI capex cycle. Risk/reward improves if hyperscaler spending commentary stays firm; falsify if major buyers guide capex flat/down.
• Use a 6-9 month NVDA call spread rather than outright stock if you want upside with defined downside; the thesis is momentum plus broader financing, not a guarantee of linear upside.
• Treat IBM as a selective, lower-beta beneficiary only if upcoming results show AI-driven bookings or mix improvement; otherwise fade rallies as a relative laggard in the compute-led spend cycle.
• Avoid forcing an ARI position until there is verified disclosure of data-center or AI-infrastructure lending exposure; if that exposure is material, it becomes a credit-beta trade, not an AI-equity trade.
• If NVDA rallies hard on sentiment alone without follow-through in customer capex commentary, trim into strength; the falsifier is any sign that AI investment is being front-loaded without durable utilization.