Reproduce “Orca: Browsing at Scale Through User-Driven and AI-Facilitated Orchestration Across Malleable Webpages” on OpenResearch

Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.

Universal app for Apple silicon and Intel. Requires macOS 11 or later.

About this paper

Peiling Jiang, Haijun Xia

Orca is a novel browser interface from UC San Diego that transforms traditional web browsing into a malleable, spatial workspace, augmenting user interaction with AI-driven orchestration to manage large-scale information tasks while preserving user agency. It demonstrates how a canvas-based design combined with contextualized AI can enable efficient parallel exploration, extraction, and synthesis of web content.

Key insights

  • Treating webpages as malleable objects within a spatial canvas fosters efficient parallel exploration and organization, allowing users to intuitively manage diverse activities and maintain task context.
  • Contextualized and atomic AI facilitation, operating under user direction, can effectively scale browsing capabilities while adapting to individual preferences and preserving user agency.
  • Direct user engagement with web content, even with AI assistance, builds greater trust in information sources and outcomes, leading to more personalized retrieval and deeper understanding.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.