IBM exec asks job candidates whether they think we are in an AI bubble as a make-or-break interview test

Prominent economists and investors are divided over whether an AI-driven market bubble is forming: Nouriel Roubini expresses skepticism, a dotcom-bubble-calling analyst warns of a painful end, and Michael Burry has reportedly positioned roughly $1 billion betting on an AI bust. Beyond market views, corporate leaders such as IBM’s EMEA managing partner Dave McCann are using candidates’ perspectives on an AI bubble as a decisive interview question, signalling that employer and talent sentiment around AI could shape hiring and strategic positioning in the tech sector.

Analysis

Market structure: The AI wave is concentrating economic value into infrastructure and hyperscalers—NVIDIA (NVDA), Microsoft (MSFT), Alphabet (GOOGL) and AWS/AMZN capture disproportionate gross margins from GPUs, cloud and model-delivery. Small-cap pure-play AI vendors and consultancies face margin compression and funding risk as customers prefer bundled cloud+model offerings; expect 60–70% of gross AI profit pool to sit with top 5 providers over next 24 months. Cross-asset: equity volatility and option IV for chip/cloud names will remain elevated; speculative equity issuance and high-yield paper tied to AI startups are vulnerable to repricing if sentiment flips.

Risk assessment: Tail risks include regulatory clampdown (EU AI Act enforcement, US liability rules) or a sharp chill in GPU demand if NVIDIA supply normalizes or China export rules tighten—either could knock 20–40% off speculative valuations. Near term (days–weeks) headline-driven swings dominate; medium term (3–12 months) earnings/capex cycles and model commercialization milestones matter; long term (2+ years) is structural adoption. Hidden dependency: industry’s leverage to NVDA GPU cadence and energy costs; catalysts: NVDA earnings, major LLM launches, AI hardware guidance and EU/US regulatory moves within next 30–180 days.

Trade implications: Favor concentrated long exposure to NVDA (hardware), MSFT/GOOGL (cloud + model monetization) and defensive IBM (IBM) at 1–3% each depending on risk budget, with stop-losses of 20%. Hedge with small short positions in speculative baskets (ARKK or small-cap AI ETFs) sized 0.5–1% or use put spreads; buy 3–6 month call spreads on NVDA to capture upside while selling premium in crowded small-cap AI names. Rotate away from unprofitable AI services/consulting stocks into software/cloud infrastructure over next 3–12 months.

Contrarian angles: Consensus fears of a total AI collapse ignore durable enterprise switching costs and multi-year software revenue uplift—if adoption is real, enterprise ARR could sustain 15–25% incremental margins for hyperscalers. Conversely, optimism may underprice regulatory and energy constraints; the mispricing is in mid-cap and micro-cap AI issuers, not the infrastructure leaders. Historical parallel: 1999–2001 tech bubble concentrated returns in a few survivors (MSFT/GOOGL equivalents); expect similar winner-take-most dynamics but with earlier regulatory friction.

Key Decisions for Investors

• Establish a 2% long position in NVDA and 2% long in MSFT (total 4% portfolio) over the next 30 days; target holding horizon 3–12 months, trim if NVDA rises >30% or MSFT rises >20% from entry.
• Initiate a 1% short position in ARK Innovation ETF (ARKK) or equivalent small-cap AI basket as a hedge; size to 0.5–1% of portfolio and add 10% of that notional as 3-month 10% OTM put spreads to limit downside cost.
• Deploy options: buy 3–6 month NVDA call spreads ~5–7% OTM sized 0.5–1% of portfolio to capture hardware upside; concurrently buy 3-month put spread on ARKK ~10% OTM sized 0.5% as volatility hedge and skew trade.
• Cut exposure to unprofitable mid-/small-cap AI equities and convertible/high-yield debt tied to AI startups by 50% within 30 days; reallocate proceeds into cloud/hardware leaders and maintain cash buffer for regulatory or sentiment-driven dislocations expected over next 3–6 months.