[kfajdsl]

Sure, but have any made as much progress as Neuralink? Not a rhetorical, genuinely asking as someone who doesn't know much about this field. Though, even if others have, isn't this kind of technological achievement something to be lauded regardless of who owns the company that did it?

[Veserv]

Yes: https://www.youtube.com/@BCIcanDoBetter

Shaking President Obama's hand with "touch feedback" in 2016: https://www.youtube.com/watch?v=itkgmMLi7l4

Eating a taco in 2018: https://www.youtube.com/watch?v=fUjfA78FuZM

Robot arm in 2018: https://www.youtube.com/watch?v=MjFr0rnbT24

Playing Final Fantasy 14 with a BCI in 2019: https://www.youtube.com/watch?v=WjNHkRH0Dus

Non-invasive robot arm control in 2011: https://www.youtube.com/watch?v=8eOSlzDdOpg

Non-invasive robot arm control in 2020: https://www.youtube.com/watch?v=asDwupMbE2I

Speech/voice generation in 2024: https://www.youtube.com/watch?v=v8frSsvwPp4

The technology to do these sorts of things as proof-of-concepts is fairly old. You do not see widespread deployment because brain surgery betas are not a very good idea. There is insufficient evidence the technology is mature or safe enough to support full-scale deployment. A common class of problem being brain scarring on the invasive insertions that reduce efficacy of the implant requiring further damaging brain surgery to remove the implant in a few years.

When you have insufficiently mature technology for deployment you optimize for research. For that, you only need enough to saturate your researchers with data and well-designed tests which is usually achieved with only a small number of units. This is similar to the reason why you only need a few prototype cars even when you are going to make millions of them. If you are not deploying, then you do not need a lot to saturate your design/development process and making a bunch of each half-baked version prior to the final release candidate is a waste of time.

When the technology is minimally adequate, then you scale up. In contrast, deploying middling quantities of proof-of-concept versions as if that "tests" anything is a recipe for a slow-burning disaster. Nobody else is "trying to compete" on who can deploy more because competing on who can deploy more half-baked brain implants would be unethical.

[dpnmn]

very interesting links, thanks! I was not aware that the technology had progressed that rapidly (outside of Neuralink, which captures a lot of the attention)

[mrguyorama]

BCIs have been doing what Neuralink is showing off since at least the 90s. It is emphatically not a difficult concept to understand that you can put wires in the brain and someone can learn to influence signals on those electrodes. Hell, Deep Brain Stimulation has been an FDA approved use of putting electrodes in your brain since 1997.

The hard parts of BCI are: Electrode sensing, but that's a much less difficult problem nowadays. Implant longevity, probably an unsolvable problem without massive advancements in understanding the body. Brain surgery, which will never not be a huge deal because piercing the barriers that protect the brain is just inherently a huge deal and risky to do.

I'm pretty sure Neuralink is the only one mass killing monkeys though.

Note that Elon has also helped push for the killing of US science funding, like funding used to further study BCIs. How convenient for him that all his competition is suddenly going to struggle.