RoboBoom
Humanoids Summit· February 5, 2025· 15:41

Jerome Monceaux (Enchanted Tools) | Transforming Robotics with AI & Human Connection

Jerome Monceaux, co-founder and CEO of Enchanted Tools, discusses designing social, character-driven robots with Jeff at the Humanoids Summit 2024 in Mountain View.

In this talk

  1. 1:33

    Design robots as characters

    After producing 40,000 Pepper robots, Monceaux wanted interactions to offer more than basic answers and movements. He came to see humanoids as characters with stories that people can encounter in everyday life. Sharing a character’s mission and personality beforehand can help people know how to approach it.

  2. 3:37

    Make interaction feel reciprocal

    Monceaux found that a robot seemed more alive when it responded to people’s presence. Sound localization led one robot to turn toward him when he made a sound, making him feel noticed. He says this kind of feedback helps children feel that the robot recognizes them.

  3. 5:14

    Set expectations through design

    Human-shaped robots can lead people to expect human-level perception and behavior, which current robots may not deliver. Enchanted Tools instead designed a creature with features such as long ears, giving it an identity that is neither human nor familiar animal. This invites curiosity about both the robot and the person it meets.

  4. 7:18

    Build for social settings and real needs

    Monceaux argues that people should understand what a robot can do well from its own perspective, such as balancing, sensing objects, and detecting nearby people. In shared places like hospitals, homes, museums, and factories, robots need social skills for groups as well as useful movement and manipulation. For children receiving radiotherapy, a digital character on an iPad can help build familiarity before they meet the robot, with pediatric teams providing appropriate knowledge.

  5. 10:32

    Use AI to handle uncertainty

    Monceaux says rule-based systems built around fixed “if this, then that” instructions struggle with the variability of real situations. Deep neural networks can work across many probabilities, opening the way for interfaces and behaviors that adapt to what is happening. He frames the broader aim as using technology to bring joy and practical and psychological support.

Briefs summarise talks and interviews; timestamps open the video at that moment. Claims in a talk are the speaker's own.

More talks