🤖 AI Summary
            Microsoft’s Edge Copilot Mode is a new, agent-style AI layer for the browser that can interpret what’s on your screen, synthesize multiple tabs, and perform web tasks — including populating shopping carts — with minimal user input. ZDNet’s hands-on test had Copilot compare open product pages for space heaters, generate a comparison chart (power, coverage, efficiency, noise, price), and recommend models like the Dr. Infrared (higher heat via dual system) and the Dreo Smart Wall Heater (smart scheduling, Eco Mode). Key features include an intent-aware search/chat bar with selectable reasoning modes (Quick, Smart/GPT‑5, Think Deeper), @tab scoping to limit analysis to specific tabs, Copilot Vision that reads page transcripts and comment sentiment, Journeys that save and surface past browsing sessions, and Actions (opt-in) which can add items to carts and prefill checkout but won’t finalize payments without user confirmation.
For the AI/ML community this matters because Edge is shipping a tightly integrated, multi-modal browser agent that combines real‑time web retrieval, on-screen visual/text understanding, stateful session memory, and constrained automation — all accessible in everyday workflows like product research and shopping. Technically notable are the hybrid search/chat intent detection, on-page transcript parsing, and agentic task orchestration tied to browser UI and permissions. The experience is faster than some competitors and useful for productivity, but it currently requires explicit opt-ins, has deliberate safety limits (can’t complete purchases), and raises questions about transparency, control, and privacy as browser agents become more capable.
        
            Loading comments...
        
        
        
        
        
            login to comment
        
        
        
        
        
        
        
        loading comments...
        no comments yet