Interhuman AI wants to teach AI what humans say without words
AI has become good at understanding what we say. But human communication has never been just about words. We pick up on hesitation, confidence, uncertainty, engagement, and agreement, often without an...
Interhuman AI, a Copenhagen-based startup, is developing technology to help artificial intelligence interpret non-verbal human signals such as body language, tone of voice, and facial expressions. The company's latest model, Inter-2, can analyze social cues across text, audio, and video in real time. Interhuman's co-founders, CEO Paula Petcu and COO Frederik Sally, were inspired to create the company after noticing that large language models like ChatGPT were largely blind to these social signals.
Petcu, who previously worked in the pharmaceutical industry, saw an opportunity to combine large language models with video and audio analysis to better understand patients in clinical trials. The team built an initial prototype, an AI coach that responded not only to what someone said but also to how they said it. Interhuman's approach combines behavioral science with machine learning to translate human communication into structured signals that AI systems can process.
The company is designing its models to be transparent, allowing end users to understand which cues led to a particular conclusion. This is crucial, as human behavior is highly individual and social cues can have different meanings depending on the person or context. Interhuman is also working to collect diverse data and annotate it appropriately to account for cultural, linguistic, and individual differences in communication.
The company aims to build Artificial Social Intelligence, giving AI systems a more human-like ability to navigate interactions and making real-time signal detection more practical at scale.
Written by urgent.news from Tech.eu's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
