The Frontier Ideas a16z Wants to Fund | Marc Andreessen, Ben Horowitz & Chris Dixon

What It Covers

a16z announces the expansion of its crypto fund into "a16z Frontier," broadening its mandate to cover emerging technologies beyond crypto. Marc Andreessen, Ben Horowitz, and Chris Dixon explain why AI is creating the largest capability gap between early adopters and incumbents they have ever observed, and why founders should now ask "is this ambitious enough?" rather than "is this too ambitious?"

Key Questions Answered

• •Fund Architecture Shift: a16z restructures its crypto fund into "a16z Frontier" to capture "outer planet" investments — companies that fall outside existing verticals like bio, fintech, or consumer. The rationale: core vertical teams are fully occupied with high-quality deal flow, creating a coverage gap for genuinely novel, category-defining companies that may take years to develop and don't fit existing classification systems.
• •Ambition Threshold Inversion: The firm has reversed its internal screening question from "is this too ambitious?" to "is this ambitious enough?" The reasoning: AI capabilities are advancing so rapidly that moderate ideas lack differentiation. Founders building incrementally risk being overtaken by the AI capability wave itself, so the defensible position is to build further ahead, not closer to current benchmarks.
• •AI Adoption Gap as Investment Signal: The delta between what AI-native early-stage founders believe is buildable versus what incumbent companies understand is currently the widest Andreessen has observed in his career. This gap is widening week-by-week as models improve. Investors and founders should treat this asymmetry as a primary signal for identifying industries structurally vulnerable to displacement by AI-native entrants.
• •Software-to-Hardware Unlock: AI tools like the open-source "REA" (Reverse Engineer Anything) now allow a 30-person team to build full-stack hardware companies that previously required 1,000 people. Any hardware device can be reverse-engineered, reprogrammed, and given a new interface via AI-generated code. Home automation, consumer electronics, and industrial hardware are cited as sectors where this capability removes historical barriers to entry.
• •Internet Business Model Disruption: The foundational "covenant" of the web — where publishers gave Google indexable content in exchange for traffic — is breaking down as AI answers replace click-throughs. Sites like Stack Overflow have already seen precipitous traffic drops. Entrepreneurs should rebuild internet economics from first principles: rethinking how attention flows, how transactions occur via agents, and how content producers monetize in an agent-mediated web.
• •Open Source as Structural Safeguard: Open-weight AI models are growing faster than closed models by token consumption on platforms like OpenRouter. Open source prevents AI from centralizing to one or two companies or governments. Practically, open source enables local model deployment — wilderness explorers already run offline LLMs on phones for survival scenarios — and allows the global research community to address unsolved safety problems like reward hacking that proprietary labs have deprioritized.