[KronisLV]

> with sufficient direct brain I/O we could interact with these patients in a metaverse setting

I feel like the complexity behind this idea is often lost in optimism about our current capabilities.

While Neuralink and other brain/computer interfaces should be developed (ideally in a non-invasive mode of operation) and have a wide range of possible uses in the future, i think that we're still decades if not centuries away from something like a "metaverse" of any kind.

Instead, currently instead of a simulated room (a la VRChat), it would probably be more like getting direct access to the spindles of a HDD and not really knowing much about the file system or the data on the drive, being able to write/read some arbitrary data and hope for the best.

If it were at all viable at the moment, you could probably communicate in a basic manner (e.g. morse code or yes/no) with someone who cannot utilize their muscles for whatever reason. This is still immensely useful, for example, in the cases as described in the article, but is still far off from immersive environments.

Or maybe my knowledge is out of date and there is promising research out there? Has anyone directly streamed video/3D environments or even pictures to someone's brain, maybe through the optical nerve? What about creating body sensations or interacting with one's sense of balance, or even smell/hearing?

[wswope]

Video has been done, and been around for a while, but IIRC the main barrier to long-term success has been the tendency of scar tissue to form around implanted electrodes (this is not my wheelhouse at all though, someone else can probably offer much more detail).

https://en.wikipedia.org/wiki/Visual_prosthesis

[KronisLV]

That's a pretty impressive technology, but isn't it primarily used on blind people, instead of someone who'd need to experience images/video by alternate means temporarily?

It's probably still a viable approach, it's just that care should be taken to not disrupt the functionality of one's eyes whilst adding the necessary implant, which may or may not be doable, depending on the method.

Of course, the point about scar tissue is also a valid one.

[rcxdude]

It's not exactly photorealistic. most of those in the linked article are <2000 (monochrome) pixels total. Better than totally blind, but still legally blind.

[niemandhier]

use transient gene therapy to mutate a light sensitive channel into the relevant neurons, guide an optical fiber to the area and trigger via light coupled into the fiber.That is as precise as we can make the targeting of the mutation, and does not cause scaring. Even if some of the vector goes where it is not supppsed to there is litte light inside a skull...

The biggest problem is neural rewiring as response to the unnatural stimulus and ethics comitees.