The AI Opportunity that goes beyond Models
TL;DR
Alex Rampel from a16z outlines how AI represents the fifth major technology product cycle, enabling unprecedented revenue velocity for application-layer companies. He identifies three core investment themes—AI-native traditional software, software that replaces labor, and walled-garden data models—emphasizing that enduring value comes from owning end-to-end workflows and proprietary data rather than just AI features.
📈 The AI Product Cycle & Adoption Velocity 3 insights
AI builds on 50 years of infrastructure layers
Unlike previous cycles (PC, Internet, Cloud, Mobile), AI leverages existing smartphones and cloud infrastructure to achieve instant global scale, with 15% of adults worldwide now using ChatGPT weekly as part of daily routines.
Enterprise adoption inflected in January 2025
Expense data from Ramp shows a dramatic spike in AI software purchases beginning January 2025, as forward-thinking enterprises moved from experimental 'magic tricks' to mission-critical deployments that save time and money.
Unprecedented revenue velocity
AI application companies are achieving $0 to $100 million in revenue within 1-2 years, a velocity rarely seen in traditional software, driven by the universal human desire to be 'richer and lazier'—achieving more economic value with less work.
🎯 Three AI Investment Themes 3 insights
AI-native traditional software targets greenfield opportunities
Rather than displacing incumbent systems (brownfield), winning companies target new company formation or inflection points—such as Real (AI-native ERP) capturing companies outgrowing QuickBooks—becoming systems of record that create 'hostages, not customers.'
Software is eating labor in new categories
The largest opportunity lies not in competing with existing software budgets but replacing human labor entirely, such as EVE automating plaintiff law firms—enabling attorneys to take 5x more cases by handling intake, evidence gathering, and drafting demand letters.
Walled garden data models create compounding advantages
Companies that generate proprietary outcome data (like EVE's case resolution values) create flywheels where accumulated intelligence improves decision-making over time, creating moats that foundation models cannot replicate since the data is non-public.
🛡️ Building Defensible AI Businesses 2 insights
Differentiation without defensibility is insufficient
While AI capabilities like multilingual voice agents provide differentiation, true defensibility requires owning the end-to-end workflow and becoming the system of record, making the product contextual to all customer operations rather than a replaceable widget.
Align with business model incentives
The most compelling targets operate on contingency or outcome-based models (like plaintiff attorneys who only get paid if they win), where AI productivity gains directly increase revenue rather than eroding billable hours like in traditional hourly billing.
Bottom Line
Build AI applications that own the end-to-end workflow in greenfield markets or high-value labor categories, ensuring you accumulate proprietary data that compounds into an irreplaceable system of record with high switching costs.
More from a16z Podcast
View all
Why Every Satellite Needs Earth | Northwood CEO on a16z
Northwood CEO Bridget explains how vertical integration is solving the satellite industry's critical bottleneck—ground infrastructure—reducing deployment timelines from three years to three months and enabling the next wave of space economy growth.
Inside Palantir: Building Software That Matters | Shyam Sankar on a16z
Palantir's Shyam Sankar argues that America's defense industrial base has become isolated and uncompetitive after post-Cold War consolidation, and now faces a 'late-1930s' geopolitical moment requiring urgent whole-country mobilization led by founders and institutional 'heretics' to rebuild deterrence.
Inside the New Media Team with Marc Andreessen & Ben Horowitz
Marc Andreessen and Ben Horowitz detail the shift from defensive, leak-fearing 'old media'—where narrow channels and corporate blandness reigned—to an offensive, infinite-channel 'new media' paradigm where flooding the zone with authentic, long-form content and embracing controversy as 'interesting' is the only viable strategy.
Why Scale Will Not Solve AGI | Vishal Misra - The a16z Show
Vishal Misra argues that large language models operate as compressed Bayesian inference engines—updating probability distributions through in-context learning—but remain fundamentally incapable of consciousness or novel discovery, meaning scale alone cannot achieve AGI.