Mark RadarMARK RADAR
About
中 EN
Sign in
Event File AI AI Models Humanoid Robots

ACE Robotics Sees Embodied AI Reaching ‘ChatGPT Moment’ by 2027

1 reports · First detected 2026-08-23 · Last active 2026-08-23

Embodied AI aims to give physical machines, including humanoid robots, the ability to perceive their surroundings, reason about situations and carry out tasks. The field is emerging as a major frontier beyond generative AI, but a shortage of real-world training data remains a critical obstacle. ACE Robotics says advances in world models could improve how robots understand environments and translate decisions into reliable actions.

The chairman of Chinese startup ACE Robotics expects embodied intelligence to reach an explosive “ChatGPT moment” by the end of 2027 as world models and data-collection methods improve. The company is building a large training database and accelerating efforts to deploy humanoid robots commercially. ACE Robotics has not disclosed the investment involved, a firm launch timetable or shipment targets for the initiative.

All Coverage

1 original reports
DECRYPT.CO 2026-08-23
Robot Brains Could Have Their ‘ChatGPT Moment’ by 2027, ACE Robotics Chairman Says

The Backstory

The history behind this event
Unitree Founder Sees Embodied AI Breakthrough Within Three Years2026-07-23 · 1 reports · similarity 0.83

Embodied intelligence combines artificial intelligence models, sensors and robotic control systems to help machines understand instructions and act autonomously in the physical world. Unitree founder Wang Xingxing said rapid advances in humanoid robotics could eventually push the technology beyond structured settings such as factories. A breakthrough in general-purpose capabilities would be significant because robots could take on a broader range of work and deliver substantial productivity gains.

Wang said embodied intelligence could reach its “ChatGPT moment” within two to three years at the earliest. At that point, humanoid robots may be able to follow commands and complete tasks in most unfamiliar environments, reducing their reliance on preset programming and tightly controlled conditions. The report did not specify the date of Wang’s remarks and provided no figures for planned production, investment or commercial deployment.

Mark Radar｜MARK RADAR
Past Issues・by MarkReadFintech

If you search news on Google, you can set Mark Radar as a preferred source—our coverage will show up more often in your results. Set as preferred source on Google →

All times are in Taipei time (GMT+8)