Perplexity CEO Aravind Srinivas’s launch of the Comet web browser is a critical case study in product strategy for the current AI landscape. He launched it knowing the underlying models weren’t ready for his full vision of an “operating system for the AI era.” This wasn’t a mistake; it was the entire point.
The New Go-to-Market: Build for the Future Model
The core insight here is a fundamental shift in product development. As Srinivas states, “You’ve got to position your product and your technology with the assumption that the models are eventually going to be great and also going to be affordable.”
Waiting for perfection in the underlying tech is a losing strategy. By the time the models are flawless and cheap, the market will already be saturated by fast-movers who built for the trajectory, not the current state. OpenAI’s Chief Product Officer, Kevin Weil, echoed this, saying it makes you “much more open to building products that only kind of work,” because they will be great in a few months.
This is the tightrope walk for modern founders: launch a product that is too early and it fails, or launch one that is perfect but too late.
A New Product Category: The Agentic Browser
Comet isn’t just a search engine or another Chrome competitor. It’s an attempt at a new category: an agentic browser. The vision is to automate tedious online tasks—from finding job candidates on LinkedIn to filling out forms—directly within the browser interface.
Srinivas offers a powerful analogy for this new paradigm, framing it as a merge of search and self-driving: “The browser feels like a car. Agents feel like autopilot.” This clearly defines their direction and helps users grasp the ambition beyond simple search.
The Founder’s Mindset
Srinivas’s journey from a PhD to a founder, guided by Ilya Sutskever’s advice on the “AI flywheel,” is telling. His focus is clear: business isn’t about “charisma” or “Steve Jobs marketing,” but simply “problem solving.”
This directness is key. However, he also understands the game, learning from Marc Andreessen that on platforms like X, a statement that is “50% correct and 50% wrong” generates the highest engagement. It’s a calculated approach to building a public persona while staying grounded in execution.
Inevitable Friction and High Stakes
Of course, the product has limitations. Comet struggles with complex, multi-step functions, and the web isn’t built for AI navigation yet. There is also a brewing debate over whether Perplexity’s tool is a sophisticated AI agent or simply a “crawler” that ignores website access rules.
This kind of friction is expected when introducing a disruptive technology. With a recent $500 million fundraising round and reported acquisition interest from Apple and Meta, Perplexity has validated its approach. But with OpenAI planning its own browser, the stakes are incredibly high. Perplexity’s strategy of shipping an imperfect but visionary product might be the only way to stay ahead.