Gen X CEO uses AI versions of Steve Jobs and Warren Buffett as a ‘fantasy board of directors’ to help him prepare for meetings and performance reviews

Markup AI CEO Matt Blumberg built a 15-member “fantasy board” of AI-generated thought partners (including profiles of Warren Buffett, Oprah Winfrey and Steve Jobs) by using ChatGPT, Gemini and Claude to create ~5,000-word personas, a process he says took an hour or two and is now used to accelerate preparation for board and employee meetings. The experiment illustrates a broader corporate trend—JPMorgan has replaced external proxy advisors with an internal AI voting platform and Klarna has deployed AI avatars and an “AI CEO Hotline”—signaling growing use of AI to augment executive decision-making and governance. The developments suggest potential efficiency gains and shifts in corporate governance practice, though Blumberg and others caution about AI imperfections and the need for human oversight.

Analysis

Market structure: Direct winners are enterprise AI infrastructure and compute providers (NVIDIA, MSFT, GOOGL) and SaaS vendors that can productize decision‑support AI; direct losers include proxy advisors (ISS/Glass Lewis revenue pressure) and legacy consulting hours. Internalization of governance/CEO augmentation shifts pricing power toward platform owners with proprietary data and scale; expect 10–30% incremental demand for cloud/GPU hours by mid‑2025 for firms that roll out internal agents at scale.

Risk assessment: Tail risks include regulatory constraints on AI use in fiduciary decisions (proxy/SEC guidance) and operational liability from AI ‘‘hallucinations’’ leading to governance failures; both could trigger 10–25% downside in affected names within 6–12 months. Near term (days–weeks) risk is noise around earnings/proxy season; medium term (3–12 months) is adoption/cost‑savings realization; long term (2+ years) is structural margin reallocation across tech/finance.

Trade implications: Favor long positions in AI compute and cloud infra for 3–12 months (expect revenue upside and margin leverage), and selectively long JPM (JPM) for margin benefit from reduced external advisory fees; favor short or underweight legacy consulting services (e.g., ACN) and fee‑based proxy advisers. Use concentrated options (3–9 month calls on NVDA/MSFT) to express upside and buy protection or put spreads on exposed incumbents if regulatory headlines escalate.

Contrarian angles: The market underestimates governance/legal friction — adoption may be slower where fiduciary risk is high (asset managers, banks) creating a 6–18 month window where infrastructure names rally but end‑user productivity gains lag. Also, cost savings could be eaten by increased cloud/GPU spend (an ‘‘arms race’’) so net margin impact for adopters may be muted; asymmetric trades should size for 1–3% conviction with tight risk controls.

Key Decisions for Investors

• Establish a 2–3% long position in NVDA (or equivalent AI‑GPU exposure) with a 6–12 month horizon; complement with 6‑month ATM call options (allocate 0.5–1% of portfolio to calls). Target +35–45% upside, cut if position falls 25%.
• Initiate a 2% long in JPM (JPM) ahead of proxy season to capture cost savings and control benefits from Proxy IQ; use a 3–6 month call spread (buy 1x ATM, sell 1x 15% OTM) to cap capital at risk. Trim to 1% if price rises >10% within three months.
• Enter a relative value pair: long MSFT (1.5%) / short ACN (1.0%) over 3–9 months to capture cloud hosting gains vs. consulting hour compression. Set stop losses at 12% absolute for each leg and rebalance if spread narrows >50% of initial width.
• Avoid meaningful conviction on KLAR (KLAR) >1% until regulatory clarity on AI‑customer interaction and data usage; if taking exposure, size 0.5–1% with a 30% stop and monitor EU consumer/privacy guidance over the next 60 days.