From Arc to Dia: Lessons learned building AI Browsers
summary
Arc becomes Dia, a ground-up AI-native browser that moves prompts, tools, models and context inside the product so anyone at the company can prototype and iterate, claiming a 10x speedup in ideation. Engineering builds a built‑in prompt editor and a hill‑climbing system called Jeba that mutates and selects prompts instead of fine‑tuning weights. Memory and knowledge‑graph tooling underpin personalized assistants, and features like autofill require an autofill confirmation step to mitigate prompt‑injection exfiltration.
AI-written summary · about
chapters (7)
- 0:00AI browsers introduction
- 1:19Arc to Dia evolution
- 3:54Faster AI iteration tools
- 6:44Jeba prompt evolution
- 9:54Model behavior craft
- 13:05AI security prevention
- 16:39AI shift transformation
related talks
17mAI-Driven Frontend Development: Beyond Traditional
25mThe Ai-Assisted Developer Workflow: Build Faster and Smarter Today
15mThe future of AI is now: Real-life case studies for on client-side AI adoption in web apps
37mBuilding AI-powered user interfaces
23mWebAI: Building the next user experience in the browser
53mAI Didn’t Kill the Web, It Moved in!
24mThe truth about AI development
33m@Cypressio x AI: Automating Browsers Using Natural Language