This is a YouTube video titled "The VC Lens on What Qualifies as a True AI Startup" containing a discussion between two speakers about the criteria venture capitalists use to define and evaluate AI companies, focusing on the distinction between native AI startups and those that merely incorporate AI as a feature. The content was created in May 2025 and explores differing investor perspectives on what constitutes a defensible AI business model.
When you talk about venture funding to AI companies, sort of inherent in that is what is an AI company. And the problem here is that there are a lot of different kinds of AI companies and everyone will have a different definition about what is defensible and what is not defensible. So some VCs might say, if you're not building a foundational model, I don't want to talk to you. Some VCs might say, if you're not fine tuning a foundational model with your own proprietary data, I don't want to talk to you. All investors treat this differently. I have not found one unifying definition that everyone agrees to.
For us, the easy question is if we took the AI out of your business, does your value prop hold firm or does it fall apart? If it holds firm, then you're not an AI company, right? That means you're really a SaaS. But if we take the AI out of your startup and your whole startup falls apart, then that means it's native AI. It's the AI that's delivering the magic, the value. So that's how we look at AI startups and defining native AI.
