Martin, many thanks for tackling the issue!
Robotics and IT will challenge capitalism in that all productivity could be ‘dehumanized’, so the value creating potential of human beings would become, asymptotically, zero.
Computing replaces humans. “One employee can do this where before we could do this with x”. Banks have been almost depopulated compared to pre-computer levels of employment: computers replaced bank clerics, websites and apps replace branches.
Is any line of work robot-safe? Japanese research shows robots outperform pets (parorobots.com). There’s no reason to believe that a humanoid robot programmed to anticipate our desires would not be just as good a companion as a human. If not better. We can build robots that can render all of human history, poetry, prose and accomplishment in subjective perfection while through sensors and routine monitor and anticipate our emotional reaction. They’d never tire of us, never hesitate, would always care. Robots can become English language teachers who can create their own hamlet rendition based on endless human examples available, all the while performing while monitoring the attention level of the class, captivating its audience. A robotic surgeon could control its cuts all the way to the cellular level, never cutting too deep or too much never leaving anything behind while operating 24/7 at a fraction of the cost of a human competitor. An automated truck could drive around the world 24/7 delivering goods. Many of us will have a luddite reaction to such visions disbelieving them instantly and assigning them to the realm of scifi. But my generation (X) will have less trouble having grown up with computers, future generations will grow up with whatever reality we create as the norm.
For many of these devices to come into being, moore’s law no longer is a hindrance at all. It is our engineering and programming abilities that need to achieve momentum. That process is well underway, programming interfaces available today can abstract complexities such as facial and language recognition. Companies such as Google have made them widely available to programmers using simple interfaces over the web, see http://www.fastcolabs.com/3023656/a-google-hangouts-plugin-that-makes-you-a-nicer-person for a fun example. With a bit of programming experience you can use these tools yourself.
While we don’t yet really have a robot that is mass produced (except for Roombas), the convergence will be on devices that can be automated today. Cars, trucks, hospital beds, refrigerators, planes, drones. Daring companies will push the envelope, many already are well under way especially in Japan and silicon valley.
IT is responsible for very much of what created the disappearing middle classes today: outsourcing to China would not be possible without the almost fully automated just in time logistical systems that have been perfected for decades. With Chinese labour becomes more expensive, companies are rushing to bring labour costs down again, a process that may set off a robotics revolution leaving many unemployed and capitalism unable to find people jobs, income, meaning and the means to survive.
Without guidance, dystopia seems assured; the future we’ll create is one of inclusivity or exclusivity. Machine intelligence has the ability to augment us to where we’ll live in paradise, our every survival need anticipated far before it can reach our awareness. How many will be excluded (other than the Luddites who don’t want to be pampered or those who seek evolution for its own sake)? If we automate our entire economy and let computers generate demand, human beings won’t be necessary economically anymore. But of course the economy ought to be wholly subject to humanity.
I’d like to end this with two references: one is to Japan which has been grappling with robotics in its culture for a long time now, it prefers automation over immigration and its economic revival may be contingent upon its first mover advantage in various robotics endeavours. It may also have the social cohesion to share the spoils more evenly. The second reference Martin is to assign this to the realm of scifi by encouraging people to read the 1957 book ‘the robots of dawn’ by Isaac Asimov. Its reviews gloss over the fact that the book (and those in the series) are anthropological thoughts experiments of great relevance today. The novel contains a very surprising robotic character (not the obvious one) whose behaviour is motivated by a very interesting quandary.