Best of the Pod: Would You Shut Down Your Most Successful Product? The Arc to Dia Story
Timestamps are as accurate as they can be but may be slightly off. We encourage you to listen to the full context.
Podcast Summary
In this in-depth conversation, Josh Miller and Hursh Agrawal, co-founders of The Browser Company, share their remarkable journey from Arc to their new AI-powered browser, Dia. (09:00) The episode explores their difficult decision to pivot away from Arc—a beloved product with millions of users—to build something entirely new. (07:00) Despite facing intense public backlash and a year of internal struggles, the founders remained convinced that AI would fundamentally change how we interact with computers. (10:00) The conversation reveals the emotional toll of leadership during uncertain times, their prototype-driven approach to innovation, and how Dia represents their vision of a "personal intelligence layer" that gets better with every tab you open.
- Core themes: The episode centers on strategic pivoting, the intersection of AI and browser technology, building consumer products for mass adoption, and the personal cost of betting on an uncertain technological future.
Speakers
Josh Miller
Josh Miller is the co-founder and CEO of The Browser Company, creators of Arc and now Dia. He previously co-founded Branch, which was acquired by Facebook when he was 22, where he worked on social media and video strategy. Miller is known for his focus on consumer product design and his ability to anticipate major shifts in how people interact with technology.
Hursh Agrawal
Hursh Agrawal is the co-founder and CTO of The Browser Company. Like Miller, he co-founded Branch before its acquisition by Facebook. Agrawal has been instrumental in the technical architecture and AI integration strategy for both Arc and Dia, demonstrating particular foresight in recognizing the potential of AI to transform browser technology and personal computing.
Key Takeaways
Trust Your Conviction When Everyone Says You're Wrong
Miller and Agrawal made the difficult decision to pivot from Arc—a product with millions of users and strong market validation—to build Dia because they believed AI would fundamentally change computing. (10:00) Despite facing intense public criticism and internal doubts, they trusted their conviction that browsers needed to be rebuilt from the ground up for an AI-first world. This decision required them to ignore external noise and focus on what they believed was the inevitable future of human-computer interaction.
Prototype Your Way to Understanding the Future
The Browser Company's breakthrough came through ArcSearch, a mobile experiment that wasn't meant to be a major product launch. (13:15) This prototype helped them realize that AI could transform the browser's core function—from routing people to existing web pages to creating personalized content on demand. Their "assume you don't know" culture enabled rapid experimentation without the pressure of perfect execution, allowing them to discover insights that informed Dia's development.
Build for Mass Market, Not Just Early Adopters
Arc's challenge was what they called the "novelty tax"—too many new features that attracted tech enthusiasts but created friction for mainstream users. (05:21) With Dia, they prioritized simplicity and performance, rebuilding their architecture in Swift to ensure the browser was fast enough for anyone to adopt as their default. The lesson: revolutionary products still need to feel familiar and performant to achieve mass adoption.
Personal Intelligence Requires Deep Context
Dia's core innovation is memory—the ability to get better with every tab you open, learning from your browsing patterns and preferences. (46:00) After nine months of failed attempts, the team cracked memory functionality just six weeks before launch using newer AI models. This breakthrough enabled Dia to become truly personal, moving beyond generic AI assistance to contextual intelligence that understands your specific needs and workflows.
Enable User Creativity Through Flexible Platforms
The most surprising early adoption pattern for Dia has been users creating custom "skills"—AI-powered workflows for specific tasks like extracting guitar tabs from YouTube videos. (50:24) This demonstrates that the future of AI software may not be about building specific features for large audiences, but creating flexible platforms that let users craft personalized tools for their unique needs. The browser becomes a canvas for individual creativity rather than a one-size-fits-all solution.
Statistics & Facts
- The Browser Company had 70 employees and millions of monthly users when they decided to pivot from Arc to building Dia. (08:00) This statistic highlights the magnitude of their decision to abandon a successful product with substantial scale.
- Chrome has 4 billion monthly active users, making it one of the most widely used pieces of software globally. (27:27) Miller shared this to contextualize the massive opportunity in building a better browser for the mass market.
- Arc's mobile companion app generated more explosive growth than anything The Browser Company had built previously when Miller casually tweeted about it on a Sunday. (13:38) This unexpected success helped validate their hypothesis about AI-powered browsing experiences.
