The AI Opportunity that goes beyond Models

a16z AI Apps partners argue software-eats-labor dwarfs software-eats-software, with portfolio cases showing 50% higher collections and zero-to-$100M ARR in under two years

• Salient (auto loan servicing AI) collects 50% more revenue than human call centers; first client had $50M/year call center with 40–70% annual employee churn.
• Eve (plaintiff legal AI) hit ~$2M→$30M ARR; 100% of attorney cases flow through it; proprietary outcome data trains intake model, compounding the moat over time.
• VLex quintupled revenue after adding AI to 26 years of digitized legal records — the shift from raw data subscription to finished-product pricing unlocked the multiple.
• Open Evidence holds exclusive licenses to New England Journal of Medicine and peers; two-thirds of U.S. doctors use it weekly — same UX as ChatGPT, but labs can’t replicate the data.
• Workday will charge $500/reference check via AI on existing customers — incumbents monetize their hostages with AI rather than ceding ground to startups.
• Zero-to-$100M ARR in one to two years is now normal for top AI app companies — unprecedented in PC, internet, cloud, or mobile cycles.
• AI roll-up playbook: buy a declining company with five blue-chip clients at 3× EBITDA, transform with AI, skip hiring a sales team entirely.
• Model aggregation (Kayak for AI) beats single-model apps in creative/coding — labs are definitionally blocked from using rivals’ models, so aggregators win.
• Differentiation ≠ defensibility: voice in 50 languages is differentiating but not a moat; owning end-to-end workflow plus proprietary outcome data is.
• Vibe coding drops software build cost to near-zero, making moats more critical than ever — ‘your margin is my opportunity’ now plays out in weeks, not five years.

Guests: Alex Rampell (a16z General Partner), David Haber (a16z Partner), Anish Acharya (a16z Partner), Jen Kha (a16z Partner) · 2026-01-19 · Watch on YouTube