Andrew Lee of a16z on the coming AI Pixar, why gaming companies are pivoting from B2B, and consumer entertainment's AI wave

Summary

Andrew Lee, partner at a16z across its games fund and Speedrun accelerator, expects a structural shift in entertainment built around an AI-native studio. Incumbents like Disney or Pixar face a real constraint: using AI modestly triggers PR backlash, as The Studio on Apple TV depicts. A new entrant launched as AI-native has no such friction. Consumer pull is already forming. Stormtroopers vlogging, Bible recreations, and viral four-video accounts reaching 120,000 followers in two days suggest the appetite exists.

Distribution remains the harder problem. Steam is the default PC gaming platform but increasingly costly to work with. Wishlisting schemes, rising customer acquisition costs, and thin organic demand all cut into margins. Lee compares this to DTC brands, where big tech platforms consumed most of the margin. The escape route is a product with enough pull to sidestep paid acquisition entirely, the way Midjourney grew organically inside Discord.

On infrastructure versus application, Lee follows the founders. Most are going into AI tooling first, which he sees as correct. The cycle works like this: B2B infrastructure enables powerful video or world models, apps are built on top, content creates new network effects, and that feeds back into demand for better infrastructure. Unreal Engine is the canonical example, a B2B platform that eventually birthed Fortnite.

a16z Speedrun backed TR Games, a VR studio built by teenagers who grew up natively in VR and won Apple awards. Their latest game, Yeps, is a social hangout experience targeting the under-16 VR-native cohort. Lee is direct about VR's constraint: headsets need to get cheaper or churn needs to drop sharply before the platform can support large games businesses. No VR game has yet delivered the 100-hour progression loop that Metal Gear Solid, Final Fantasy VII, or GTA IV did for their platforms.

Hedra, another a16z-backed company, caught a wave when Studio Ghibli-style AI content exploded. Users needed a tool to animate in that style, and Hedra was ready.

At YC's W25 demo day, Lee counts roughly five or six consumer companies in the batch, including founders who pivoted away from B2B because they found it boring. That signals something healthy. Funds chasing B2B AI infrastructure are indirectly funding the wave that will eventually produce a breakout consumer entertainment product. Whoever builds the 10x product that earns genuine network effects won't need to buy their audience.