Why AI’s greatest challenge is leadership, not innovation

At the Fortune CEO Forum, executives highlighted social platforms like TikTok as real-time discovery engines for shifting consumer tastes while emphasizing that companies must track cross-region resonance to identify scalable trends. Leaders expressed cautious optimism about AI’s productivity gains, citing Gartner’s projection that AI infrastructure software spending will rise to nearly $230 billion this year from roughly $60 billion in 2024, but stressed boards now demand demonstrable ROI, governance, reskilling and leadership changes to convert experimentation into measurable business value.

Analysis

Market structure: AI infrastructure re-rating (Gartner $230bn vs $60bn) shifts pricing power to software, consulting and cloud providers; winners are services-led incumbents that convert pilots into managed contracts (IBM, large system integrators) while commodity-dependent incumbents face margin pressure. Short-form discovery (TikTok) accelerates brand-level demand volatility, favoring firms with fast digital go-to-market and agile supply chains; expect higher realized volatility for retail/consumer names and sticky revenue for platform/consulting deals over 6–24 months.

Risk assessment: Tail risks include rapid regulatory action (EU AI Act enforcement, data/privacy fines) and a macro slowdown that compresses IT budgets; both could wipe 20–40% off discretionary AI project spend in 6–12 months. Hidden dependency: productivity gains require org-level investment (training, governance) — absence of this will yield low ROI and stranded capex; catalyst set: large multi-year contracts and FY results showing >15% incremental margin lift will re-rate winners.

Trade implications: Favor high-conviction, asymmetric exposures to services/consulting and industrial tech enabling AI (IBM, TT) while trimming energy capex/leverage to transition names (SHEL). Use 6–18 month calibrated option structures to cap downside: allocate 1–3% NAV per idea, target 20–40% upside within 12 months, cut if position falls >12% or if no material contract news in 9 months.

Contrarian angle: Consensus underestimates implementation lag and governance spend — this favors consulting/managed-service margins over pure-play hardware where growth is priced in. Historical parallel: early cloud adoption where services captured initial value; unintended consequence: faster AI rollout raises compute energy demand, creating rotational risk back into energy and grid/utility capex in 12–36 months.

Key Decisions for Investors

• Establish a 2–3% NAV long in IBM (IBM) via a 12-month call spread (buy Jan 2026 call, sell a higher strike) sized to cap premium ~1–1.5% NAV; target 20–30% upside in 12 months, stop-loss if position down 12% or if no >$500m consulting/AI contract wins in next 9 months.
• Trim/hedge Shell (SHEL) exposure by 40–60% of current position or add a 1% NAV short via 3-month puts if oil/energy equities rally >8% in 30 days; rationale: allocate capital from slower-transforming energy capex to faster AI-driven ROI opportunities.
• Initiate a long Trane Technologies (TT) 9–12 month call spread sized 1% NAV to play ESG+productivity tailwinds (industrial efficiency); target 25%+ upside if HVAC/order book shows >5% sequential margin improvement, cut if macro softens and orders fall >10% q/q.
• Execute a pair trade: long IBM (1.5% NAV) vs short SHEL (1.5% NAV) to express AI services outperformance versus energy transition execution risk; rebalance in 3–6 months or earlier if pair diverges by >15% absolute performance.