OK, so how many million computers can you make each year before the billions of people out there have bought them all up?
I don't think you understand what he means, he is saying we are coming to the Apex of Technology with computers, which in turn will put many people out of business..
He makes extremely valid points he backs up with good reasoning, Fiber Optic, once it's out there in full force, will be the end of the data transfer bottleneck... He is saying in essence, we are reaching the maximum possible tech we can achieve, he uses the process of printing transistors on the wafer as an example...
If you can't make something better, now what? Computers will either:
1) Lock into a set price..
2) Continue to get cheaper.. (Doubtful)
3) Stop being such a huge demand...
4) Continue to be a huge demand...
No matter what happens, it's not speculation to say many people will eventually lose their job, because in truth, it's costly to pay a team of engineers to develop tech...
Once the MFGs have all the tech they need, they will just start mass producing stuff, have a massive sale off, and then lay everyone off..
I've seen it many times, a company will mass produce something then shut the product line down permanently... Once we reach the apex, that's going to start happening...
Companies will be mass producing computers, because they won't get any better, those with the most, will announce the closing of their plant and sit back and collect dividends to rising prices / inflation... When the MFGs cut off the supply chain there will be this fictitious world wide panic, "Oh no, they aren't making any more computers"!
Nobody can really predict the future, truly, though we can take a good educated guess, and that I believe this writer has done well..
Yes, the specific market of personal computing would likely scale back in production, but people always have been and always will be innovative. New devices will be thought of, that while implementing current technology, would still require engineering to reconfigure new combinations of circuitry on different sized PCBs.
Software and firmware will always have room for efficiency improvements as well. New instruction sets, more optimized code, and more concentration on bringing along the best possible performance on the software side of things would ensure that new devices and software are still sold.
All of this only matters if you believe that technology improvements will grind to a halt, which I do not. People said that 10, 20, 30 years ago, but look where we are today? Just because we haven't discovered the next breakthrough in technology yet does not mean it doesn't exist. Or just because we do not yet know how to harness the technology (quantum computing) doesn't mean it will remain unharnessable.
Going on the example of fiber optics: It's not done yet. Modern fiber-optic based backbones can carry upwards of 14tbps. That's certainly not the maximum of the medium the data is traveling on, but I'm sure the switched and endpoint technology can be further upgraded to push more data through the lines.
My point is, I don't believe technology, overall, will ever decrease in demand. At least not for a long time. Engineers are always looking for ways to integrate technology more and more into our daily lives, and that means more R&D, more production, more consumers purchasing technology, etc. If you want to talk about a particular component, such as a personal computer, then sure, demand may go down over time for such things. But the same could be said of land-line telephones. Those people who designed land-line telephones have been able to move on and design other things.
Basically, what's the big deal?