Do you mean EEG stuff,

For the most part, yes. If we're just needing enough input to control something like a mouse, then there's no real reason to go with an invasive implant. You can pull the same data from eeg and ocular tracking.

Do you mean counteracting potentially deadly seizures, or causing them?

It would be counteracting seizures.

BCI, or your opinion? Maybe like, challenges, what you see as being the most promising stuff, that sort of thing?

The problem with BCI is that there's just not a lot of uses for them. The quadriplegic community is already small, and their range of cognitive ability runs the gamut. So creating a cbi that is useful to the entire patient population is going to be tough. The largest obstacle would be patient education, and training care takers.

This is part of the reason I discount Musks interest in BCI as medical device, there's just no money in it. I think his only real motivation is to sell it to gullible wealthy people.

Another inherent problem with BCI is that it's not seamless. It takes a lot more concentration to operate a mouse with your mind than it does with your body. People don't really understand how much of their movement is handled by their spinal chord instead of the brain.

People have a hard time utilizing interactive spaces when we separate them from physical input. Which is why a lot of people struggle with VR,. When your physical senses like proprioception don't reflect the interactions the same as our visual senses we can become physically ill.