David Atkins seems a little oddly pessimistic in this post. On the one hand he is very correct about the progress of technology and what it will be able to do in the future. On the other he seems to ignore that a future in which all jobs are performed by machines is one in which either people must either all be enjoying immense prosperity or we are hiding from the robot death machines that are keeping insurrection in check (whether they are controlled by an AI or the future's 1% is kind of irrelevant).
While there are times I feel like the doomsday scenario is where we are headed, mostly I don't. There are a few reasons for that, some positive some negative. Let's start with the negative...
I don't know that we will really develop machines/software that can truly replace human creativity. It could be our limit, or that we realize it is a bad idea (more likely the former). I mean AI's will either enslave or kill us. It is the sensible thing to do. Whether our history is actually one of violence and destruction interspersed with periods of peace and prosperity, or the other way around doesn't matter much because our current trend is not good. Particularly with respect to...
...Global warming, which will vastly change how we live on this planet over the coming decades. Peoples will move, starve, thirst, die, and wars will start. Or we will resolve all these issues via a combination of technological development and a vast, global political force for good that sees that that development is distributed to maximize the aid to humanity. (Stop laughing!)
On a more positive note there is the fact that if you replace all people in the workforce with machines: who buys the products the machines make? In that world either governments/corporations would need to ensure that somehow the bulk of the population had the wherewithal to consume, or the wealthy could move to enclaves where the rabble couldn't kill them.
In reality I suspect that we will not approach any of these things (except global warming which we can't stop at this point and so will have to deal with). If machines displace enough middle class the economy will suffer so there will be less impetus for further development. There could be an asymptotic approach, but I think we would find there are local minima/maxima that would really require some heft to get past, and in our current political state we're nowhere close to having it.
I think in the US we will mostly be not as well off as we would like--with the wealthy being increasingly distant from the others--but still sufficiently comfortable that revolution doesn't really happen.
I'm not sure that is really optimism, but it isn't the downward spiral or killer robots of doom so it's something.