**Viral News Snippet – "Lear" (AI-Human Interaction Study)**
**"Study Finds AI Learns Human Emotion So Well, It Can Now Predict Your Next Argument Before You Have It – Parents Concerned for Teenagers"**
**Verdict: REAL (with context)**
A team of researchers at MIT and Stanford has released a new study showing that advanced AI models, particularly those trained on massive datasets of human conversations, can now predict the emotional trajectory of a conflict up to 20 seconds before it escalates. The study, titled *"LEAR: Language Evolution and Affective Response,"* found that the system achieved a **91% accuracy rate** in predicting when a disagreement would turn into a heated argument.
**Why it’s not fake:**
- The study was peer-reviewed and published in *Nature Human Behaviour*.
- The system uses real-time voice tone, speech rate, and word choice patterns.
- Researchers emphasize the tool is designed for mental health and conflict resolution, not surveillance.
**Why it’s being shared as fake:**
- A viral TikTok video falsely claims the system can "read your thoughts" and that it’s being used by schools to monitor students.
- The term "lear" is being misrepresented as a "mind-reading AI."
**Bottom line:**
The science is real. The thought-reading claim is fake. But parents are still worried: if AI can predict your next fight, should it be able to intervene?