Michael Lewis reveals he’s got a deal to write the Sam Altman book—when ChatGPT is ready to write a rival draft

Author Michael Lewis disclosed a verbal agreement with OpenAI CEO Sam Altman to write Altman’s biography only after ChatGPT can produce a competitive draft, a framing device for a broader debate held on SoFi’s podcast about AI’s economic impact. Fundstrat’s Tom Lee argued AI is already driving productivity gains, labor displacement and a software-sector repricing with 2026 as a potential inflection point, while Lewis emphasized limitations of current models and political risks from rapid job disruption. The discussion underscores sectoral winners and losers and distributional risks but contains no immediate earnings or macro data likely to move markets materially.

Analysis

Market structure: AI’s next phase concentrates value in cloud/compute, data, and infrastructure owners (Google/GOOGL, cloud peers, chipmakers) while compressing margins for labor‑heavy, legacy enterprise software. Expect a bifurcation: top‑5 infra players gain pricing power (20–40% incremental gross margin over 2–3 years), mid/small SaaS face 10–30% revenue/margin downside as agentic products replace point solutions.

Risk assessment: Tail risks include regulatory clampdowns (EU/US legislation in 6–18 months), catastrophic AI misuse triggering liability suits, and compute supply shocks (TSMC/NVDA capacity) — each could wipe 30–60% off richly priced AI plays. Short term (days–months) expect elevated IV and episodic drawdowns; medium/long term (2026+) true productivity gains materialize if adoption curves match Tom Lee’s thesis.

Trade implications: Favor durable infra over application software: overweight GOOG (AI stack exposure) and selective financials (GS) that can monetise M&A/venture exits; underweight mid‑cap SaaS (IGV/individual names) and software multiples. Use directional LEAPs on infra and short-dated puts/call spreads to express conviction while limiting capital at risk; scale in over 3 months and add on 10% pullbacks.

Contrarian angles: Consensus underestimates distributional and regulatory backlash—history (dotcom, flash‑frozen adoption) shows multi‑year diffusion and many winners vanish. The market may be underpricing durable moat capture by mega‑infra players and overpricing rapid monetisation by endpoint software vendors; a regulatory shock could abruptly re‑rate both sides.

Key Decisions for Investors

• Establish a 2–3% portfolio overweight in Alphabet (GOOGL/GOOG): allocate 1% to outright stock and 1–2% to Jan 2027 LEAP calls (~ATM) to capture 12–24 month AI infra upside; scale in over 3 months and add on any pullback ≥10%; trim at +25–35% or if quarterly AI revenue misses by >15%.
• Initiate a 1.5% short exposure to legacy enterprise software via IGV: buy 3–6 month puts 5–10% OTM sized to 1.5% notional (or short ETF outright) anticipating 10–25% downside in 6–12 months as AI reduces spend on traditional software licences.
• Add 1–2% long position in Goldman Sachs (GS) to capture M&A/venture exit activity tied to AI (hold 6–12 months); set a stop‑loss at 8% and take profits at +20% to lock gains from deal flow and trading revenues.
• Implement portfolio tail hedge: buy a 3‑month XLK 10% OTM put position sized to offset a 2–3% equity drawdown OR purchase a 30‑60 day VIX call spread (2x multiplier) before major AI/regulatory headlines; monitor EU/US AI regulatory announcements on a 30–90 day cadence and reduce tech gross exposure if binding rules pass.