Imagine a person treated for cerebral palsy by instrumenting their brain so that computers could compensate for the injury. Similarly, a blind person could be outfitted with a system of electrodes to allow them to see. Anyone like this could also have electrodes in other parts of their brain, from the frontal lobes to the speech areas.
These people will be augmented to make them “normal”. However, there is no reason why that augmentation could not include designing interfaces to improve their memory, ability to calculate, imaging, or anything else. Our only limitation now is that in the developed world it would be impossible to justify such augmentation without treating some disability or injury. I think this will mean that the first people who will be fully augmented will be elderly people with Alzheimer's.
An Alzheimer's patient has a progressive disability that could be, in theory, compensated for by instrumenting that person's brain. This could allow a slow takeover of functions by external software. This is no small task to be sure. The first several people treated this way will have mistakes made, and their augmentation will start very crude and slowly evolve. But such a project will allow us to learn so much.
At some point there will be people without any disability who want to be augmented. Mark Zuckerberg has talked about something on the order of telepathy, but I think that we will go beyond that. Just as our cell phones evolved from simple voice devices to text, apps, and vide, with videophone arriving as an almost unnoticed afterthought, what is possible with direct brain interfaces has yet to be known. Feeling and experiencing what others do should be possible, and sharing on a level unknown could allow intimacy beyond imagining. I can foresee a time when it could be considered a strange or traumatic experience to be cut off, alone in one's head the way everyone is today.
It is then that the truly radical new future will begin. Today, the law recognizes that in some ways a prothesis is part of that person, it's not just a thing. I think that will extend into mind protheses. How soon could it be difficult or impossible to perform work without prosthetic extensions of self? And it gets more complicated than that.
Now, we can say that 'this person' stops here, and 'that person' starts there. Nobody questions that idea, because it is self-evident, or seems so to most although some physicists or philosophers might argue. However, in the near future that won't be true anymore – at least not for everyone. And this is where I part company with Mark's vision.
Where is the boundary of the mind of someone linked to a group of others, all in constant contact? In whose mind did a thought originate? We have our corpus calossum that links the two halves of our brain. I think it is obvious that the longer a person is linked brain to brain, the more those who are linked will be non-separate. One could ask where is the self of those people? Today, we can ask where is the self in our body and not be able to decide. Linkage will extend that. And when a person in such a network dies, will that have the same significance as it does now? How will we adjust to that?
It is this new reality that is the singularity. It will arrive when we have become selves linked together that develop collective self-hood. It won't be some foreign AI taking over the world. It will be us, seamlessly integrated together. There will be pieces that are AI, pieces that are memory, that link together and integrate large numbers of people. Childhood's End had this kind of concept, although it proposed no specific technology for achieving it.
What we should worry about is whether it is equitable. We should worry about leaving people out. And we should worry about how to prevent incarnations not compatible with free will. There are issues. However, I think that in the end, this singularity of ourselves will be far to complex to be controllable.