Yai.news
Project Pilot: Can AI models fly drones?
How to Win the Largest Market in AI
a16z’s investment thesis treats AI inference as a stable, recurring workload where tokens per watt matter more than flexibility. Its sharper claim is that production—not chip design alone—is the moat: memory, power, cooling, interconnects, supply chains, and reliable rack-scale delivery determine unit economics.
Investing in Neo
AI agents break a core endpoint-security assumption: actions can no longer be attributed to the person at the keyboard rather than software acting with that person’s authority. a16z’s Neo thesis reframes control around real-time identity, agent guardrails, and attribution at execution; it remains an investment claim.
