Neura Pod (All Things Neuralink) The Man Who Uses a Brain Chip to Control A Robotic Arm
Jan 15, 2026
In this discussion, Alex Conley, a participant in Neuralink's Convoy Study, shares his journey using a Neuralink brain implant to control a robotic arm. He dives into his personal experience with spinal cord injury and rehabilitation, and the innovative technology that enables him to manage everyday tasks. Alex talks about the surgery, calibration process, and the impressive low latency of the system. He also offers insights into the emotional value of regaining independence and the development of assistive tech like QuadStick for gaming and control.
AI Snips
Chapters
Transcript
Episode notes
Three-Plane Control Mapping
- The robotic arm uses three 'mouse models' (table, mirror, wall) tied together to map imagined motions into 3D space.
- Combining those plane calibrations yields the user's first closed-loop control model.
Refine Don't Restart
- When a model degrades, run a brief closed-loop refine session to quickly produce a better model.
- Use existing movement and click data in five-minute recalibrations to outperform scratch restarts.
Lack Of Sensory Feedback Limits Grasping
- During early grasp tasks Alex lacked tactile feedback and visual reference, making precise grabs like a pretzel difficult.
- The team plans cameras and tactile sensors on the claw to improve feedback for users.