EEG scanning, combined with models trained to recognize patterns in the signal, could allow humans (disabled or not) to do a set of things they otherwise would not be able to do. Such actions include moving prosthetics to gain some movement they previously lost, communicating with their loved ones if unable to speak through spellers and emotions detectors, enhance their memory, remotely control devices and much more. If we don't see all of this happening yet it's mainly due to the lack of data to train the models on, we're trying to face this problem and find a way to make it work with the data available.
Using AI to create synthetic EEG neural data would be a great way of solving the data scarcity. This solution would actually allow to build datasets that scientists and developers around the world can use to study the brain and create AI models that will then need only a short personalized fine-tuning period before becoming effective. More than that this technology could reduce the scanning period to a very short time, then augmenting the collected data through synthetic data generation. This means more personalized data in shorter time, more accuracy and better performance.