Amazon's new Help Me Decide button picks what you were going to buy anyway – just faster (www.techradar.com)

Amazon has rolled out a "Help Me Decide" button in the U.S. on its mobile app and browser that uses AI to pick a single, personalized product for you after you browse several similar items. The model ingests your browsing history, shopping patterns, hover times, clicks and past purchases plus customer reviews to choose—and explain—why that item suits your needs, and it also surfaces a cheaper alternative and a premium upgrade. The button appears at the top of a category or search page and acts like a compact, assertive version of Amazon’s prior tools (Lens Live, Interests, Shopping Guides) or conversational assistants such as Rufus, but with stronger intent interpretation rather than relying solely on collaborative filtering. For AI/ML practitioners and product teams, the feature is a clear example of context-aware recommendation systems moving from ranking multiple candidates to making a single, authored decision and providing natural-language rationales. That increases utility and conversion potential but also amplifies persuasive power and privacy concerns: the system depends on fine-grained behavioral signals and could shift user agency by shaping choices rather than just surfacing options. Technically, it demonstrates combining personalized signal fusion, intent inference and review-grounded explanation generation—an architecture other platforms (Google Gemini, ChatGPT shop assistants) are also racing to refine.