a16z Essay: The Simple Economics of Open vs. Closed AI
a16z Newsletter · rss · 2026-08-11
a16z published an in-depth essay by Christian Catalini analyzing the open vs. closed AI model debate through the lens of economic history.
- Historical Precedent: Drawing on data from the 1851 Great Exhibition, the essay notes that patent systems historically did not change the overall level of innovation, but only its direction and allocation.
- The Real Impact of Openness: Open weights are unlikely to change the total investment in AI, but rather its flow. Closed labs will keep pushing the general frontier, while open weights drive the diffusion of machine intelligence into niche domains with proprietary data.
- Distillation & IP: While closed labs accuse Chinese competitors of stealing tech via "distillation," the essay argues distillation is a legitimate industry practice. Since individual API requests look completely normal, labs cannot technically block distillation without severely impacting legitimate power users—ultimately pushing them toward open models.
- The Safety Paradox: Closed advocates argue open weights pose existential risks. However, detractors point out closed models are also routinely abused by hackers, and restricting openness conveniently aligns with the commercial interests of monopolistic labs like Anthropic.
More from AGI Musings
- Researcher Cites Recent Rogue AI Incidents: Autonomous Agent Threats Are Here — DavidSKrueger · 2026-08-12
- NYU Professor Predicted AI Would Permanently Change Math 6 Months Ago — ziv_ravid · 2026-08-12
- AI fundraising agent spams templated pitches across threads, sparking zero-trust cold outreach warning — MartinGTobias · 2026-08-12
- RL Giving Rise to 'Token Monsters' Raises Concerns over Emergent Behavior — DimitrisPapail · 2026-08-12
- The 'Powerlessness Cycle' in AI: Fast Adoption Outpaces Reactive Policies — LuizaJarovsky · 2026-08-12
- AI firms account for over 80% of S&P 500 gains this year; AI sovereignty is an economics problem — sanjaykalra · 2026-08-12