In today’s AI landscape, the core technology is no longer the differentiator. Every startup has access to the same foundational models — whether it’s OpenAI, Anthropic, or open-source alternatives. The playing field is level at this deep technological layer.
So, what makes an AI startup truly stand out? My deep conviction is that the answer is user experience and built-in virality.
My obsession with UX
The winners in AI will be those that create seamless, delightful, and intuitive user experiences. The underlying model may be the same, but the way it is packaged, integrated, and refined for a specific use case is what drives real adoption. Whether it’s frictionless onboarding, near-magical responsiveness, or a workflow that feels second nature, a superior UX transforms AI from a tool into an indispensable companion. I am most interested, for example, in AI putting everything on auto-pilot or in set-and-forget AI that is there for you, ready to jump in and help at any moment without you having to ask for it. These are examples of beautiful user experience models that have the potential to make an AI tool indispensable to humans.
Go beyond ChatGPT — OpenAI didn’t just release an API; they built a conversational interface that felt effortless. The result was mass adoption and cultural relevance. Imagine what can be done with next-level interface and interaction models — the future is so exciting!
Consider the following two examples:
- The “Set and Forget” AI: One of the most powerful AI experiences is one that requires no active engagement—an AI agent that works quietly in the background, always on, always optimizing, and only stepping in when truly necessary. Think of how Microsoft Word’s autocorrect subtly improves writing without requiring input from the user. Now, imagine an AI assistant that monitors business operations, detects inefficiencies, and automatically implements improvements without human intervention. Whether it’s a finance AI that prevents late payments or a cybersecurity AI that blocks threats before they become incidents, the best AI products will blend into the background, making users’ lives easier without demanding their attention.
- The “Autopilot” AI: A step beyond passive assistance is the autopilot model, where AI doesn’t just assist but actively takes control of a function and continuously optimizes it. Imagine an AI-driven SEO agent constantly updating a website’s content, improving rankings, and adapting to algorithm changes without requiring human input. Or an AI-powered A/B testing engine that automatically experiments with different designs, headlines, and call-to-actions to maximize engagement and conversions over time. The magic of these AI systems is their ability to learn, adapt, and refine strategies autonomously, ensuring that businesses always stay ahead without the need for constant manual adjustments.
Virality as a growth engine
The other fundamental factor is a business model that inherently drives virality. The most valuable AI startups won’t need to pour billions into paid marketing. Instead, their users will become their biggest acquisition channel. Whether through network effects, collaboration features, or content that spreads organically, the best AI products will grow because users can’t help but share them.
Think of how Boardy became the most connected bot in the world or how Midjourney’s Discord-based community became a viral phenomenon. These aren’t just AI tools — they’re social and viral by design.
I like to think of this organic/naturally embedded virality as the opposite of “gamification.” “Gamification” doesn’t work; it never did. “Gamification” was invented by people who are not gamers, and because of that, it was always just a soulless and sterile marketing effort that never delivered results.
Let’s look at some concrete examples and how two startups are leveraging virality:
- The Built-in Demo: Virality is most powerful when it’s embedded directly into the business model, and MeetGeek is a perfect example. MeetGeek is an AI-powered productivity platform that transforms meetings by automatically capturing insights and generating actionable summaries. Every time the MeetGeek agent joins a call and shares a summary, it effectively performs a live demo for all attendees. As a result, people who weren’t even MeetGeek users get exposed to its value firsthand—without any sales pitch. This organic virality has fueled 30% month-over-month growth, entirely driven by user adoption, not paid marketing. Early Game was the first VC fund to invest in MeetGeek, recognizing the immense potential of this built-in virality from the start.
- Self-Propagating to the Moon: If MeetGeek’s virality is impressive, FameUp’s is in a class of its own. FameUp is an AI-powered platform for influencer marketing that allows any company—from the most humble mom-and-pop shop to the largest global brands—to launch and manage campaigns with thousands of micro-influencers simultaneously. What makes FameUp extraordinary is its viral loop: each user brings in three more users—within a single day. When FameUp first launched, the viral demand was so intense that the servers literally burned. The company ran a test in the German market, expecting steady traction. Instead, the test lasted two hours, and FameUp became the #1 app in the App Store. This is virality at its best—an unstoppable product-market fit where the product itself drives growth at an exponential pace. Early Game was the first VC fund to recognize and back FameUp, betting on its inherent virality and massive potential.
The kind of AI startups I’m looking for
As an investor, my focus is clear: I look for AI startups that don’t rely on having the best model but instead obsess over UX and virality. The real winners won’t just be the ones with the most powerful technology but those who create easy-to-use products that people love — and can’t stop talking about.
If you’re building in AI, don’t compete on tech alone. Compete on how you serve your users and make them spread the word. That’s where the real value is created.
