Your algorithm wraps you in more of the same. Lens keeps what’s relevant — and pulls in the angles it filtered out, each with a reason and a source.
Waitlist opens soon
Try the live demo ↓{{modeNote}}
{{card.summary}}
{{card.why}}
Demo: four curated public sources and a simulated reading profile. Ranking is a transparent rule — score = (1 − c) × familiarity + c × gap — not a trained model.
People lean on passive feeds instead of searching for what they need. Even search is moving to WeChat, Douyin and Xiaohongshu, where it runs on discovery and social proof, not precise lookup.
say “big data + algorithm” recommendations have trapped them in an information cocoon.
China Youth Daily & wenjuan.com survey · 1,501 respondents · July 2023
searches a day on Xiaohongshu by late 2024, about double the mid-2023 level.
Zhou Tian Finance estimate, via Tencent News · December 2024
of Douyin users start a search within 30 seconds of opening the app. Over 80% of monthly users search.
Ocean Engine (ByteDance), Douyin Search Behaviour Report · 2023 data
search most often with AI tools, and 53.3% with social platforms. Only 21.7% name traditional search engines.
2024 survey cited in YuanXun GEO, China AI Search White Paper · August 2026 · original survey not named
Switching the algorithm off isn’t the answer — a generic feed is broad, noisy and irrelevant. You need a feed that stays relevant, shows its work, and lets you choose how far to stretch.
Keep author, date, publication and original URL for every permitted source.
Cluster related items and keep reported facts, author opinion and Lens labels apart.
Match the reader’s question, topic and goals — not a generic “opposing view”.
Bring credible missing disciplines or stakeholders forward and cut duplication.
Show why it appears, link the source, and let readers report errors or switch feeds.
A bilingual AI-governance pilot for 100 early readers. We never sell ads — we’re paid by subscribers, not by your screen time.
Waitlist opens soon
We’re not collecting sign-ups yet.