Why Intempus believes robots benefit from a human physiological state

Intempus retrofits robots with emotional expressions for improved interaction.

: Teddy Warner, a 19-year-old robotics enthusiast, founded Intempus to enhance robots with human-like emotional expressions to improve interactions with humans. By retrofitting existing robots, Intempus aims to make robot movements more predictable and generate data for AI model training. Warner started with fMRI data for physiological signals but shifted to using sweat data and other metrics for emotional modeling. Intempus, launched in September 2024, is advancing with support from the Thiel Fellowship and plans to expand its team to start testing its technology with humans.

Intempus, founded by 19-year-old Teddy Warner, is pioneering technology to retrofit existing robots with human-like emotional expressions. The aim is to improve human-robot interaction by enabling machines to display emotions through kinetic movements—such as arm and torso gestures—rather than relying solely on facial expressions or speech. Warner believes humans primarily interpret emotion through body language, a trait also seen in animals.

The concept behind Intempus emerged during Warner's work at Midjourney, where he observed that AI systems lacked spatial and contextual reasoning. This was largely due to the absence of a "physiological state"—the internal human processes like stress or excitement that inform decision-making. Warner argues that robots need this intermediate state to act in a more human-like, relatable way.

To simulate these physiological states, Warner initially experimented with fMRI data, but real progress came when he used polygraph data to detect sweat levels. These physiological signals helped train models that assign emotional profiles to robots. The system was later enhanced with additional biometric inputs such as body temperature, heart rate, and photoplethysmography to simulate complex emotional states more accurately.

Since launching in September 2024, Intempus has collaborated with seven enterprise robotics firms. Warner, still operating solo, plans to expand the team and initiate human-centered testing. While the current mission is to retrofit existing robots, Warner is open to developing proprietary emotionally intelligent machines in the future.

Intempus’s vision also draws from the concept of "world models" in AI, which attempt to simulate cause-effect relationships over time and space. By including simulated physiological inputs, these models can better understand sequences of actions, helping robots mirror human-like behavior and interact more naturally with people.

Sources: TechCrunch, Techedt, Intempus Research