WhoBuilt

Inside the $550M AI Browser That's Taking on Google Chrome

Jun 9, 2024· 44 min· 2.7K views·Watch on YouTube

Speaker

Josh Miller, co-founder and CEO of The Browser Company, talks with Peter Yang about building Arc and Arc Search, the company’s product culture, and the challenges of growing an AI-powered browser.

Why reinvent the browser

Miller says people spend much of their day in browsers but rarely care about the experience. The Browser Company wants to make the browser as meaningful to use as the iPhone became compared with earlier cell phones.

Hiring and building with users

Miller describes valuing people who bring care, originality, and craft to their work, including designers with very different résumés. The company’s Membership team brings customer support and research into product decisions, using channels ranging from surveys to direct conversations.

Arc Search focuses on one job

Arc Search began as a side project to find the fastest way to look something up on a phone. Its Browse for Me feature opens multiple pages, reads them, and creates a synthesized answer; early external testers helped shape the product before launch.

A short window to reach scale

Miller believes AI could reshape computer use and says the company has roughly 12 to 24 months to earn a place in that shift. He sees reaching browser scale as the central business risk, with monetization potentially coming from the attention and distribution a widely used browser provides.

Craft, publishing, and personal motivation

Miller argues that high craft and early testing can coexist: Arc Search was tested with people outside the company and revised using their feedback. He acknowledges that AI summaries challenge publishers’ economics, says Arc Search has added prominent source links, and advises listeners to follow their own sense of meaningful work.

Takeaways

  • Choose a specific user need and make the experience exceptionally good at solving it.
  • Bring customer research into product decisions from the start.
  • Look for originality and evidence of care, not only résumé credentials.
  • Treat reaching distribution as a core business risk when scale drives the model.
  • Test ambitious products with users early while maintaining a clear bar for craft.

Mentioned in this talk

Search inside this talk

What was claimed 13

Show all 13 facts

Related talks