Story Threads
When multiple sources cover the same story, Current automatically groups them into threads with AI-synthesized headlines.
How Story Threads Work
When you subscribe to multiple news sources covering similar topics, you'll often see the same story from different angles. Instead of showing each article separately, Current detects related coverage and groups it into a single thread.
Detection
Current uses entity extraction to identify when articles share the same subjects:
- People mentioned in the articles
- Places and locations
- Organizations and companies
- Topics and concepts
- Events and timeframes
When enough entities overlap between articles from different sources, they're grouped into a thread.
Synthesis
For each thread, Current generates:
- Synthesized headline: A neutral summary of the story
- Synthesized summary: Brief overview of the coverage
- Perspectives count: How many sources are covering it
Thread Display
In your river, threads appear as a special card showing:
- The synthesized headline
- A "2 perspectives" (or more) badge
- Time span of coverage ("over 3 hours")
- Thumbnails from multiple sources
Expanding Threads
Tap a thread to see all related articles:
- Each article from each source
- Their individual headlines and summaries
- Publication times
- Tap any individual article to read it
Verification
Current tracks the confidence of its thread groupings:
Confidence Score
Each thread has a confidence score indicating how certain Current is that these articles are about the same story.
Alignment Levels
Current also evaluates how aligned the coverage is:
- Aligned: Sources agree on the key facts
- Mixed: Some disagreement or different angles
- Divergent: Sources present contradictory information
Benefits of Threads
Reduce Duplicate Coverage
Instead of seeing the same story five times from five sources, you see one thread. This dramatically reduces clutter when following major news.
See Multiple Perspectives
Tap into a thread to see how different sources cover the same story. Compare angles, identify bias, and get a fuller picture.
Better Context
The synthesized headline provides a neutral summary, helping you understand the story before diving into any particular source's take.
Time Awareness
See how a story develops over time. A thread that spans "over 2 days" is evolving; one that spans "over 3 hours" is breaking news.
When Threads Don't Form
Not all related articles become threads:
- Unique takes: If an article has a significantly different angle, it may stand alone
- Insufficient overlap: Articles need enough shared entities to be grouped
- Time gap: Articles too far apart in time stay separate
- Single source: Threads require at least two different sources
Privacy
Thread detection and synthesis happen on-device. Your reading data isn't sent to external servers for analysis.
Tips for Getting More Threads
- Follow multiple news sources: Threads form when the same story appears in different feeds
- Include major publications: Big stories are more likely to be covered by multiple sources
- Mix perspectives: Different viewpoints create richer threads