Rockstar co-founder Dan Houser is using AI, but cautions "it's not as useful as some companies would have you believe yet"

Dan Houser, Rockstar Games co-founder and founder of Absurd Ventures, says his team is 'dabbling' in AI for forthcoming IP but cautions the technology is overhyped and not yet capable of solving many development problems. Absurd is developing new live‑action and video game universes that will require several years of work, a timeline investors should note for capital needs and monetization horizons; Houser's comments reinforce a cautious industry view on near‑term AI productivity and commercialization, coming shortly after Ubisoft disclosed an AI‑generated image slipped into a final build.

Analysis

Market structure will bifurcate: large-cap compute and cloud suppliers (NVDA, MSFT, AMZN, GOOGL) pick up secular demand for GPUs and hosting while legacy IP owners (TTWO, ATVI, EA) gain relative pricing power because exclusive, monetizable franchises become more valuable when AI tooling underdelivers. Small-cap AI content/tool vendors and boutique studios face compressed pricing power and higher churn as customers delay adoption; expect higher implied volatility and funding yield premia in that cohort over 3–12 months.

Key risks center on regulation and operational failures: a legal or regulatory limit on training data within 6–12 months could rapidly revalue startups reliant on scraped content (potential downside 30–60%). Short-term (days/weeks) the biggest risk is PR-driven volatility from AI mishaps (5–15% drawdowns in affected stocks); long-term (2–4 years) the primary risk is capital intensity — studios will need meaningful equity/debt over multiple funding rounds to monetize new live‑service universes.

Trade implications favor convexity into compute and durable IP: favor option-led exposure to NVDA and selectively add long exposure to IP-rich publishers while hedging development timelines with protective puts. Rotate away from small-cap AI content/tool names into large-cap cloud/semis and top-tier publishers over the next 2–6 weeks, scaling positions over 3 months as catalysts (earnings, regulatory guidance, major AI product demos) resolve.

Contrarian view: the market underprices the monetary value of licensed IP and long-dated live‑service cash flows — if AI adoption remains slow, top publishers could rerate +20–40% over 12–24 months while hardware/cloud vendors compound demand. Conversely, consensus is underestimating regulatory tail risk; a binding restriction on training data would disproportionately hurt small AI startups and re-rate valuations faster than revenue can be redeployed.

Key Decisions for Investors

• Establish a 2–3% portfolio long in NVDA via a 12‑month 10–15% OTM call spread (buy 10% OTM, sell 25% OTM) to capture sustained GPU/cloud demand while capping premium; exit or roll if NVDA reports QoQ revenue guide miss >5% or shares appreciate >40%.
• Allocate 1.5%–3% longs split between TTWO and ATVI (1–1.5% each) for IP exposure, hedge 25% of notional with 3–6 month 5–10% OTM puts; hold 12–24 months and trim half if either stock outperforms +25% or raise hedges if new-game release slips >3 months.
• Implement pair trade: long TTWO (1.5%) / short Unity Software (U) (1.5%) via short stock or buy 3–6 month 20% OTM puts on U; unwind if Unity posts >5% beat on bookings and showcases measurable AI-driven revenue within 6 months.
• Within 30 days, reduce exposure to public and private sub-$1bn market‑cap AI content/tool companies by 50% and reallocate proceeds to large-cap cloud/semis (MSFT, AMZN, GOOGL) and top publishers; if EU/US regulatory text limiting training data appears within 6 months, increase shorts in affected small caps and buy protection on creative IP owners (3–6 month puts).