What if human employability doesn't end?

Okay so we know that the robots are coming to take all our jobs, and it's very likely that those jobs will not be replaced by new jobs staffed by humans, because right now we're building machines that know how to build machines that can handle different tasks.

But, for the sake of argument, let's assume (falsely) that the capitalist model of individual worth isn't doomed to catastrophic failure in the near future. Let's assume (falsely) that there are enough jobs to go around now, and there will continue to be enough jobs to go around into the future, because new jobs will replace those that we know are pretty much about to go away.

Pro-capitalists like to argue that this is a reason that the government shouldn't guarantee personal security to all residents. The argument is that the market can supply, and will supply, sufficient opportunity to everyone that large scale welfare support is unnecessary.

But no matter how perfectly the market generates new kinds of work to make up for the work that is, right now, going away, there's still the logistical problem of the gaps between an individual losing their job, and finding and becoming qualified for one of the magical new jobs. 

That gap is not going to be an occasional experience. It's going to be a constant fact of everyone's lives, because we can be sure that right now virtually every major form of employment is replaceable and there's not any particularly good reason to believe that the magical new jobs won't be. 

So everyone can expect, more than once in their lives, to find themselves suddenly unemployable, pending their self-education in a currently-incomprehensible field. 

The cost of that gap has to be paid for -- and it can be paid for in one of two ways. 

One is unconditional welfare. I say unconditional because the cost of verification that a person really needs help will likely massively exceed the saving it generates.

The other way is human suffering: let people slowly burn through their resources as they struggle to keep their heads above water while trying to wrap their minds around whole new invented worlds of knowledge. People will lose their homes, start to fail to feed themselves or their children, people will die, or develop medical complications that hasten their death. People will develop stress-based mental illnesses that inhibit them from fully pursuing new means of employment. 

To be totally clear: That option is evil. Picking it is morally unacceptable.

Again, this is the most charitable fantasy-world defense of the free market I can think of that accounts for the reality that technology isn't gonna suddenly stop being a thing by 2017.