Vanessa Larco thinks 2026 is when consumer AI finally hits its stride.
On TechCrunch’s Equity podcast, the Premise partner and former NEA investor lays out why she believes consumer and prosumer products are about to change how people spend time online — and why OpenAI won’t crush every startup in its path.
From generic apps to ‘concierge-like’ AI
Larco has backed consumer and prosumer startups for years. Her core thesis: we’re shifting from static apps to AI-powered, ‘concierge-like’ services that actually do work on a user’s behalf.
That raises an uncomfortable question for incumbents: do standalone products like WebMD and TripAdvisor survive as separate destinations, or do they get swallowed by general-purpose agents like ChatGPT and Meta AI?
Larco’s answer isn’t to run from OpenAI and Meta, but to build in the gaps they’re unlikely to fill.
The AI products OpenAI won’t want to kill
One of those gaps, she argues, is labor-heavy marketplace businesses.
According to Larco, OpenAI is unlikely to build or run marketplace models that require managing real humans — the kind of operations that come with support, compliance, onboarding, and messy edge cases. Those aren’t the clean, infinitely scalable software problems foundation model companies tend to chase.
That leaves room for startups to build vertical AI layers on top of human services instead of competing head-on with pure software assistants.
AI apps as ‘disposable software’
Larco also challenges how founders think about product durability.
She describes a new category of ‘disposable software’ — AI-native apps that should be spun up, used, and discarded without ceremony. In her view, many AI apps ‘should be treated like Word docs’: lightweight, easy to clone, and not precious.
For developers, that means shorter product cycles and fewer illusions about moats. For users, it means expecting a world where spinning up a new AI tool is as trivial as opening a document template.
Voice first, screens optional
Hardware is quietly reinforcing that shift.
Larco points to Meta’s Ray-Ban smart glasses as the device that turned her into a serious believer in voice interfaces. After using them, she now sees screens as optional for most everyday tasks.
If that vision plays out, the winners in consumer AI won’t be the prettiest apps on a phone screen. They’ll be the services that feel like ever-present concierges, reachable through voice, earbuds, or glasses — whatever happens to be closest.
2026: M&A and stablecoin experiments
Larco expects 2026 to be another big year for M&A as incumbents race to buy AI capabilities rather than build everything in-house. That could mean more exits for small, focused AI teams that slot cleanly into larger platforms.
She’s also watching what stablecoins make possible. The episode teases new business models that could emerge when you combine AI agents with programmable, price-stable digital money — though the details are still early.
Why this matters for founders
For consumer and prosumer founders, Larco’s framework is blunt:
- Don’t assume your app will remain a destination; assume an AI concierge will sit in front of it.
- Don’t bet your company on becoming a horizontal assistant; look for messy, human-in-the-loop markets OpenAI won’t want to run.
- Design products as if software is disposable and distribution lives in other people’s interfaces.
The full Equity conversation, hosted by TechCrunch senior reporter Rebecca Bellan and produced by Theresa Loconsolo, digs deeper into where Larco is putting money to work — and which consumer AI products she thinks the giants won’t bother to kill.
