A recurring issue that engineers have with brain-computer interface is that they do not know what it’s like to have brain implants. As a result, they can’t test the technology on themselves to find bugs in the device.
A team of Neuralink engineers hit the road to work alongside a study participant. According to the company, the team watched the participant use the system in his living room. When ideas didn’t work, they axed them on the spot. When they found a bug, they recoded the next day. Everything was happening on site during a 72 hour test.
Fixing Neuralink on the Spot

A lot of the improvements focused on daily annoyances. Before this setup, a person had to hit a physical button to start and stop speech recording. The team built an automatic detector so the app knows on its own when someone wants to talk.
Advertisement
They also worked on Bloom, a mobile app meant to make the Neuralink easier to use every day. The engineers added a way to use the system from an iPhone lock screen. Additionally, they set up an iOS keyboard so people can dictate directly into any app without jumping through hoops.
The system can still make mistakes, however. So, the team worked on quick editing tools so a user can fix typos right inside the keyboard. For computer tasks, they built a prototype where thinking the words “Hey Neuralink” triggers an assistant to handle computer jobs on its own.
Bringing Back Everyday Moments
The biggest win was hooking the app into regular iPhone calls. The system feeds speech straight into the phone call audio, so the person on the other end hears clear sound. It can even play a message like, “This call is from someone using an assistive communication device.”
For the participant, this update meant he could easily call his wife and joke around with her on the phone again, just like when they first started dating. Building alongside users gave the team clear focus, and it proved that the best tech comes from listening closely to the people who actually use it.