Summary
In this episode of the a16z podcast, Growth partner David George and institutional allocator Ron from Accolade join to examine how artificial intelligence is fundamentally restructuring venture capital economics. The central argument is that power law concentration — historically a feature unique to VC — is now becoming systemic across the broader technology market, driven by a new dynamic where capital invested in compute directly compounds a company’s product advantage rather than creating coordination overhead.
The conversation zeros in on the three frontier model companies — OpenAI, Anthropic, and SpaceX — which the guests estimate represent somewhere between $3.5 and $5 trillion in combined enterprise value, and note that most institutional LP portfolios had minimal exposure to these companies before SpaceX’s public listing. This has forced a rethinking of asset allocation strategy across venture sub-classes: pre-seed, seed, the “messy middle,” large multi-stage firms, and dedicated late-stage vehicles.
A recurring theme is why the largest late-stage franchises require a strong early-stage presence to win: writing a $500 million check into a category-defining company is nearly impossible without a prior relationship built at the seed stage. The guests also touch on AI’s acceleration of revenue growth curves — AI reached $100 billion in revenue in four years versus the 15 years it took SaaS — and what that velocity means for how allocators should think about entry points and fund sizing going forward.
📺 Source: a16z · Published September 10, 2026
🏷️ Format: Podcast
