Tags: human‑robot interaction

iRobot co‑founder Brooks recounts a near‑miss with an Agility Robotics Digit humanoid that fell when he got too close, leading him to keep a three‑meter distance from moving bipedal robots. He argues that current walking mechanisms make safety certification for shared human‑robot zones virtually impossible, limiting the deployment of humanoids in health‑care and factories. Looking ahead, Brooks predicts that within 15 years “humanoids” will look nothing like today’s bipedal machines, favoring wheels, varied arm counts, and novel sensor placements. He notes that billions spent on vision‑only, rigid humanoids may be misplaced, while academic work such as MIT’s tactile glove shows promise but remains far from true human‑like dexterity. Lire la suite

AI‑generated images and videos are increasingly populating social media feeds, often looking realistic at first glance but leaving viewers with a subtle sense of unease. Psychologists and human‑robot interaction researchers attribute this discomfort to the uncanny valley effect, where near‑human creations trigger disquiet when they are not perfectly lifelike. Experts such as Dr. Steph Lay and Prof. Christoph Bartneck explain that our brains are wired to spot small irregularities, a skill that once helped detect danger. While the technology improves, the feeling of something being “off” persists, prompting users to stay vigilant and occasionally step away from the screen. Lire la suite