Apple cofounder Steve Wozniak admits he’s ‘disappointed a lot’ by AI and hardly uses it: ‘They just sound too dry and too perfect’

Apple spent $12.7B in fiscal 2025 capex—far below the roughly $300B collectively spent by AI hyperscalers—while opting to power Siri with Google’s Gemini rather than build a large in‑house AI effort. Cofounder Steve Wozniak and several tech leaders express skepticism about heavy AI use and screen time, and a jury recently found YouTube and Meta liable for youth-harm design, underscoring regulatory and reputational risk in consumer-facing tech. For portfolios, this signals Apple’s more conservative AI posture and modest capex commitment relative to peers, with limited near-term market disruption but potential long-term positioning implications.

Analysis

Big-cloud incumbents retain the most direct path to capturing AI's recurring revenue streams because they control both model hosting and enterprise distribution — this creates a multi-year revenue flywheel where every large enterprise deal locks in cloud spend, professional services, and model inference fees. That dynamic amplifies scale advantages: marginal cost of additional inference is low, so 12–24 month penetration into enterprise workloads can convert into outsized incremental operating leverage versus one-time device hardware sales.

Regulatory and legal pressure on attention-driven social platforms is a material asymmetric risk to ad-based monetization: forced product redesigns and higher moderation costs compress ad yield and increase churn among higher-ROI advertisers. The net effect should reallocate a portion of ad dollars toward search/commerce funnels and closed ecosystems that can better guarantee safety and ROI, creating a secular tailwind for search/commerce platforms but a headwind for raw attention marketplaces.

For hardware-first franchises, prioritizing on-device inference and privacy-first features preserves margins but delays high-margin cloud revenue — that tradeoff will show up in differing free cash flow trajectories and valuation multiples over the next 6–24 months. Investors should treat current positioning as a two-speed market: hyperscalers for growth and monetization, hardware/OS owners for margin insulation and optionality; capital allocation announcements, major enterprise AI deals, or adverse court rulings are the near-term catalysts to re-rate sector positioning.

Key Decisions for Investors

• Overweight MSFT (6–12 months): buy equity or buy Jan‑2027 calls sized 1–2% NAV. Thesis: Azure + enterprise AI monetization accelerates recurring revenue and operating leverage. Risk/reward: target 30–50% upside if model-hosting adoption continues; stop-loss at 12% drawdown.
• Long AMZN / Short META pair (6–9 months): equal notional long AMZN (AWS/commerce exposure) and short META (ad-risk exposure) sized 1–1.5% NAV each. Thesis: ad dollars shift toward commerce/search while social platforms face litigation/regulatory drag. Risk/reward: asymmetric — expect 2–3x payoff if reallocation occurs; close the pair on signs of ad yield stabilization or a major regulatory settlement.
• Selective AAPL exposure (12 months): add modest long position (1–2% NAV) or buy Jan‑2027 calls as contrarian hedge against ‘missed AI’ narrative. Thesis: on-device ML and premium hardware economics preserve margins and optionality to monetize AI without large capex. Risk/reward: 20–40% upside if device monetization accelerates; principal risk is slower software monetization — cap losses at 15%.
• Tactical shorts/puts on SNAP or META (3–6 months): allocate <1% NAV to near-term puts to hedge regulatory/legal outcomes. Thesis: courtroom liabilities and UX changes can compress CPMs quickly. Risk/reward: buys limited downside (premium) with potential 3–5x payoff if ad revenues and guidance decline; keep exposure small given headline-driven volatility.