a16z's Justine Moore: AI creative app market keeps expanding with more specialized tools, not fewer

Summary

Justine Moore, partner at a16z, pushes back on the consolidation thesis for AI creative tools. Early expectations pointed toward market concentration, but the opposite is happening. The market keeps expanding, and specialization is accelerating as AI enables first-time creators who have no history with traditional creative software. That new-user influx is opening verticals that legacy tools like Photoshop or Premiere never reached.

Moore frames the likely structure as two parallel tracks. Horizontal tooling layers will serve developers and prosumers broadly, with companies like Fal cited as an example on the developer side. Vertical-specific applications will layer on top, mirroring what has already played out in the LLM market with purpose-built tools for accounting, law, and finance. Different industries and workflows, Moore argues, will not converge onto a single interface.

Mobile vs. Desktop

The platform split maps closely to use case intensity. Consumer-facing, low-effort products dominate on mobile. Moore points to Lensa and Sora's mobile feed as examples where users scroll and consume rather than edit. Deep creative workflows requiring precise controls favor larger screens, and Moore flags a notable trend toward iPad as a serious professional creative platform, driven by its combination of stylus input and screen real estate.

Monetization for AI-Native Creators

Monetization for creators building entirely on AI output is still nascent. The current dominant model mirrors traditional creator economics: ad revenue and platform view-share on TikTok, Instagram, and X. A significant secondary revenue stream has emerged around selling prompts and courses, driven by high demand from non-expert users. Moore sees the next frontier as AI-native intellectual property, characters or performers generating streaming revenue on platforms like Spotify or Netflix. She points to early signals in music, particularly in Asia, and an AI actress as early proof points, while noting the category remains at a very early stage.

Trend Velocity and the Tracking Problem

Moore abandoned her impressions-tracking spreadsheet roughly six months ago because the pace of launches made it unmanageable. Nano Banana, VO3, Sora, multiple Chinese model releases, and the World Labs launch were all hitting simultaneously. The fragmentation now runs not just across products but across audiences: practitioners deep in one corner of AI creativity are often unaware of major launches in adjacent areas.

On viral formats, Nano Banana generated multiple distinct waves, including AI Polaroid recreations, action figure renders, and a current TikTok trend of users generating and then hugging AI images of their younger selves.

Bubble Assessment

Moore is not concerned about an AI bubble under current conditions. The indicators she watches are retention and gross margin trajectory. Both are moving in the right direction. She draws the comparison to Amazon, Uber, and DoorDash, all of which ran at gross margin neutral or negative early before scaling into efficiency. The concern would arise if margins were universally negative and retention was weak, which she says is not what the data shows. User scale at companies like ChatGPT, approaching 1 billion users, and the ability to monetize directly through subscription infrastructure like Stripe at $20 per month distinguishes this cycle structurally from the late 1990s.