Who owns the robots?

It is very difficult to stop people from having ideas, or to control them while they’re trying to see if they can make those ideas work. All the more so because most ideas don’t work out in the end – they swallow up a gazillion research dollars and then come to nothing. This unpredictability and lack of control makes lawyers very nervous, because the main purpose of law is to control everything and make sure nothing goes wrong.

​For regulation to work the initiator of an idea must be identified, and preferably licensed so that everybody knows what he's up to. This is not as easy as you'd think, in a world where so many creative experiments amount to nothing. For instance the idea for the driverless car can be traced back a really long way; there were many interesting but failed experiments long before companies like Google and Uber came along.

This chaotic and random pattern of innovation is a challenge for regulators who want to control innovation:

After the fact, a technological advance may look self-evident. But in the middle of the cycle it is chaotic in the extreme and therefore not easily subject to direct regulation. Indeed, the initial steps usually take place outside any regulator’s ken, when the initial rounds of financing come from the three Fs: family, friends and fools.

Richard Epstein, Innovation and Inequality: the Separability Thesis
​

Regulation would easily work if we could simply identify the companies who are developing driverless cars, and tell them to stop because those cars are going to put millions of people out of their jobs. Identify the decision-maker and get them to make the right decisions, namely the decisions that are best for Society.

But the invention of robot technology, and technological progress more generally, did not come about by chance: they reflected conscious decisions to make such an investment. We need to start, therefore, by asking how such decisions are made. This leads naturally to the question, Who makes the decisions? as well as to issues of countervailing power.

Atkinson, Inequality, p. 115.

The difficulty with asking ‘who makes the decisions’ or ‘who owns the robots’ is that markets exhibit something more like a fragmented network of decision-makers, all trying to incrementally build on what is already known. Look for the power broker behind innovation, and there isn’t one.

Sometimes it’s just a spotty teenager building robots in the basement of his parents’ house; multiplied by a thousand other spotty teenagers embarked on similar schemes and pastimes.

One way to regulate innovation is to wait until the product is developed and hits the market, and then pounce. This is quite straightforward for something like a new type of car, because road regulations can easily be changed to keep driverless cars out – so anybody can invent a workable driverless car if they want to, and drive it round their private yard. But if the car is not allowed on public roads then that’s the end of that. They'll never be able to sell it and recoup their investment in research and design.

The approach of banning innovation never works in the long-run, because there are too many tempting features of any successful innovation. Some country somewhere is always going to buck the trend and support the new invention, to gain a competitive advantage over other countries. This proves to be very inconvenient for luddite countries. Who wants to be the country that just bans innovation while everyone else is pressing ahead? And so innovation staggers on.
​