Claude Introduces 'Reflect' Dashboard
新智元 · wechat · 2026-07-11
Anthropic launched a new Reflect panel for Claude, designed to help users review their past AI usage and consider "when to use it and when to turn it off."
What Reflect Does
- Tracks usage themes, time distribution, and peak hours over the past 1/3/6/12 months
- Reviews tasks delegated to Claude, such as drafting strategy docs, managing emails, or finding meal inspiration
- Prompts you on tasks better done yourself and guides you to create Projects to reduce repetitive context-setting
Design & Privacy Boundaries
- Offers Do Not Disturb hours and break reminders, easily toggled off
- Does not read incognito chats
- Does not display raw files from connected tools directly, retaining only high-level summaries
- Health-related conversations are excluded entirely, and other sensitive content is only summarized
The Underlying Issue
The article mentions a study by OpenAI and the MIT Media Lab based on massive real-world conversations and randomized controlled trials, suggesting that heavy chatbot use correlates with higher loneliness, emotional dependence, and problematic usage, while potentially reducing real-life social time.
The core takeaway isn't just the feature itself, but that AI products are now building "using less AI" as a capability. However, the better they can review user behavior, the easier it becomes to lock users into their platforms.
Related event: Claude Launches Reflect Usage Review Feature(10 posts)→
More from AGI Musings
- The Evolution of LLM Business Models: Selling Outcomes Over Tokens — yacineMTB · 2026-07-22
- Bindu Reddy says GPT-6 is coming soon, with Alibaba, DeepSeek and Kimi close behind — bindureddy · 2026-07-22
- Bindu Reddy says the industry still lacks a way to train 20T models and scale post-training RL — bindureddy · 2026-07-22
- Advanced AI Models Are Becoming Impossible to Plug and Play — emollick · 2026-07-22
- AI suggested a better composition, and that made one user uneasy — Sydde · 2026-07-22
- The Thimble and the Waterfall: AI's Data Bottleneck and Feedback Loops — dyamins · 2026-07-22