Back to News
Market Impact: 0.25

Innovative ways the world used AI in 2025

GOOGLGOOGAMZN
Artificial IntelligenceTechnology & InnovationHealthcare & BiotechLegal & LitigationMedia & EntertainmentEmerging MarketsPrivate Markets & Venture

AI adoption accelerated across multiple sectors in 2025, with ChatGPT ranking as the fifth most-visited website and technology firms investing hundreds of millions to refine models. Real-world deployments cited include near-50% chatbot mental-health use among young people in parts of China, an AI assistant that quadrupled a Brazilian pharmacist’s prescription-clearing capacity, over 140 AI projects in Brazil’s judiciary handling 70m+ pending cases, and generative tools cutting film production costs in Indonesia; risks include flawed medical advice, LLM underperformance in low-resource languages and widening inequality. These trends point to broad addressable markets and efficiency gains across healthcare, legal and media verticals, suggesting selective investment opportunities in enterprise AI tools and regionally focused model developers while mindful of quality, ethical and regulatory execution risks.

Analysis

Market structure: Hyperscalers and cloud/infra owners (Alphabet, Amazon, MSFT, data‑center REITs, GPU suppliers) are the primary winners as enterprises outsource model hosting and inference; expect Alphabet/Google to capture ~5–15% incremental revenue mix from AI services within 12 months while unit economics improve for those owning the stack. Losers include small AI consultancies, legacy media/content producers and mid‑market SaaS without proprietary data — pricing power will concentrate at scale, compressing margins for ad‑heavy and content‑dependent incumbents. Risk assessment: Key tail risks are regulatory (EU AI Act, US liability frameworks) and catastrophic hallucination events that could trigger class actions; a credible regulatory shock could repriced multiples down 15–35% in 3–12 months. Near term (days–weeks) volatility will track earnings/AI product demos; medium term (3–12 months) depends on enterprise adoption and GPU supply; long term (1–3 years) is adoption-driven revenue growth offset by rising capex and talent costs. trade implications: Primary actionable trade is overweight large-cap cloud/Ai owners (GOOGL/GOOG, AMZN) and underweight pure-play small caps. Use option structures to size asymmetric exposure: buy 9–12 month call spreads on AMZN (20–30% OTM) and 12–18 month LEAP calls on GOOG as core exposure, while allocating 0.5–1% to protective puts to hedge regulatory shocks. Rotate 3–5% from small‑cap/VC AI names into infra, copper/energy exposure (data‑center power) over next 3–9 months. contrarian angles: The market underestimates localization and language gaps — niche local players in emerging markets will maintain pricing power; conversely, consensus overestimates near‑term monetization timelines (expect most large AI incremental revenue to materialize over 4–12 quarters, not instantly). Historical parallel: late‑90s tech adoption (real revenue followed hype), implying selective concentration rather than broad market gains; unintended consequence risk includes surge in litigation/legal services demand, increasing costs for health/legal deployments.