Fireside Chat, Interview
Marc Andreessen's 2026 Outlook: AI Timelines, US vs. China, and The Price of AI
- Revenue growth rates for leading AI companies are expected to surpass previous technological waves, with consumer applications projected to reach 5–6 billion mobile broadband users at "light speed" and monetize at high price points such as $200–$300 monthly tiers.
- Per-unit AI costs are predicted to decline faster than Moore's Law, driving hyper-deflation where "tokens by the drink" become extremely cheap, while GPU shelf life extends to seven years and supply chains transition from shortages to gluts within five years.
- The industry is expected to develop a "pyramid" structure over the next decade featuring a handful of "God models" in giant data centers alongside a volume market of smaller models on embedded systems, with small models catching up to big model capabilities within six to 12 months.
- Competition will intensify globally as China accelerates in robotics and native chip ecosystems, while US companies like Google, Meta, Amazon, and Microsoft face emerging rivals like xAI and Anthropic, with new incumbents catching up to OpenAI and Anthropic capabilities in under 12 months.
- Pricing strategies will shift from standard SaaS models to usage-based or value-based pricing, with venture capital firms aggressively funding contradictory strategies to capture plausible opportunities across proprietary, open-source, big model, and small model verticals.
- Policy dynamics in the US are forecasted to shift from a chaotic state of over 1,200 disparate state bills to federal intervention, potentially negating restrictive measures like California SB 1047, while global competition is expected to prevent US self-sabotage.
- Technological revolutions beyond AI, including impacts on biotech, healthcare, crypto, and energy sectors, are anticipated, alongside the acceptance of AI as essential by the general public within one to five years.
- The venture firm expects its public stance to remain a competitive hiring advantage, maintains a focus on maintaining humility regarding investment accuracy, and anticipates AI transforming its own business operations and downstream demand for energy and materials.
- Skepticism regarding current product forms persists, with expectations that the industry will experience fits and starts, overpromising, and occasional operational failures, even as underlying capabilities continue to advance rapidly with daily research breakthroughs.