Scouting the Web: What Happens When the Entire Web Is the Agent’s Context? with Dhruv Batra

Watch the entire AI Superstream: https://learning.oreilly.com/videos/ai-superstream-context/0642572273927/0642572273927-video403832/?utm_medium=social&utm_source=youtube&utm_campaign=free+trial&utm_content=ai+superstream What happens when you need an agent to continuously track the entire live web? In his answer from the AI Superstream, Yutori cofounder Dhruv Batra takes you back to the very beginnings of web search, traces the 20-year evolution from keyword-based Google queries to chatbot QA, RAG-based search, and ultimately to "scouting": an always-on AI agent that monitors the web like a cron job and alerts you the moment something you care about happens. Watch to discover why fitting the entire web into an agent's context is a dead end, and what you should do instead: orchestrating networks of subagents, where some prioritize recall and others precision; automatically enriching vague user queries before dispatching them; and training a vision-based browser model like Yutori's Navigator that operates websites like a human to unlock the long tail of the web that has no API. Follow O'Reilly on: LinkedIn: https://www.linkedin.com/company/oreilly/ Facebook: http://facebook.com/OReilly Instagram: https://www.instagram.com/oreillymedia BlueSky: https://bsky.app/profile/oreilly.bsky.social