Quite a bit of this feels like "if it were possible, someone would have figured it out by now".
But this is an attitude that doesn't believe in new tech breakthroughs in general.
I believe drexler style nanobots and atomically precise manufacturing are possible.
I don't know exactly what compute looks like when you have atomically precise manufacturing, but probably quite a lot better than anything we are currently using.
And there are some really out there possibilities, like micro black hole based computing.
Well, I think the fact that something hasn't happened is at least a useful heuristic. But it shouldn't be our only one!
In many of the cases of unconventional computing, we have several examples of companies with smart people and funding going bust. This is also a useful signal. Some forms of computing look daunting because of the physical limitations.
I gave nanocomputing short shrift though. My skepticism here comes from a few data points.
First, that top-down nanotech (semiconductors) already approach atomic scale. Electron beams can achieve features below 10 nm with a recent paper getting some features close to 2nm. That's like 9 silicon atoms wide!
Second, the few people who have taken up the charge of building nanocomputers seem to be proposing logic components that are quite large. Making the size/speed advantage with e-beam fab smaller. But maybe designs will get better!
Third, I believe Drexler, while still working on APM, has become more excited about protein nanotech rather than computing per se, but I could be wrong about that.
> In many of the cases of unconventional computing, we have several examples of companies with smart people and funding going bust.
This is the sort of thing that often happens with new technologies.
Because to not go bust, the company has to do enough R&D to get their unconventional computers to be comparable to conventional computers. Starting with whatever tools and tech is available in their day. And these companies will have a lot less total R&D funding than conventional computing.
> already approach atomic scale.
Fair. Going from 10 atom features to 1 atom features is a 1000x improvement in volume. But also, a major limiting factor in chips is heat/power. And also price of components.
If nanotech can make components that aren't any smaller, but are orders of magnitude cheaper and more efficient, there is still a lot of room for in practice gains.
You’re right, spaceborne data centers seemingly don’t make economic sense. But the oligarchy apparently wants to build them anyway. So given that, what alternative motives might the oligarchy have for doing so? Answer, safety for the datacenters, space is out of angry mob range.
Quite a bit of this feels like "if it were possible, someone would have figured it out by now".
But this is an attitude that doesn't believe in new tech breakthroughs in general.
I believe drexler style nanobots and atomically precise manufacturing are possible.
I don't know exactly what compute looks like when you have atomically precise manufacturing, but probably quite a lot better than anything we are currently using.
And there are some really out there possibilities, like micro black hole based computing.
Well, I think the fact that something hasn't happened is at least a useful heuristic. But it shouldn't be our only one!
In many of the cases of unconventional computing, we have several examples of companies with smart people and funding going bust. This is also a useful signal. Some forms of computing look daunting because of the physical limitations.
I gave nanocomputing short shrift though. My skepticism here comes from a few data points.
First, that top-down nanotech (semiconductors) already approach atomic scale. Electron beams can achieve features below 10 nm with a recent paper getting some features close to 2nm. That's like 9 silicon atoms wide!
https://pubs.acs.org/doi/10.1021/nl304715p
Second, the few people who have taken up the charge of building nanocomputers seem to be proposing logic components that are quite large. Making the size/speed advantage with e-beam fab smaller. But maybe designs will get better!
https://github.com/philipturner
Third, I believe Drexler, while still working on APM, has become more excited about protein nanotech rather than computing per se, but I could be wrong about that.
Regardless, let's hope I'm wrong!
> In many of the cases of unconventional computing, we have several examples of companies with smart people and funding going bust.
This is the sort of thing that often happens with new technologies.
Because to not go bust, the company has to do enough R&D to get their unconventional computers to be comparable to conventional computers. Starting with whatever tools and tech is available in their day. And these companies will have a lot less total R&D funding than conventional computing.
> already approach atomic scale.
Fair. Going from 10 atom features to 1 atom features is a 1000x improvement in volume. But also, a major limiting factor in chips is heat/power. And also price of components.
If nanotech can make components that aren't any smaller, but are orders of magnitude cheaper and more efficient, there is still a lot of room for in practice gains.
You’re right, spaceborne data centers seemingly don’t make economic sense. But the oligarchy apparently wants to build them anyway. So given that, what alternative motives might the oligarchy have for doing so? Answer, safety for the datacenters, space is out of angry mob range.