We had quadriplegics using brain-computer interfaces to control computers since the 90s. We had neuro implants giving sight to the blind since the 2000s. The problem is not making it work at all, the problem is making the implant work in the long-term in a way that the body doesn't eventually reject. Neuralink is still catching up to where we were 20 years ago, only fronted by a dancing monkey trying to hype himself up as the technoking.
conceptually we had eletric cars since 1900. Doesn't mean much until someone improves it to mass market adoptable level and mass produce it like Elon did with TESLA.
It's silly to compare available tech from 90s from what's possible nowadays with 1000x + BCI bandwidth done with Neuralink
To reiterate, the computer tech is not the bottleneck. I've made software for neuro research labs logging terabytes per day of data from rats. The hardware/wetware interface is not a problem you can solve by waiting for TSMC to scale to the next node.
