While everyone’s eyes were on the Brexit vote and staggering economic implications this week, my mind was drifting to another crucial development within the EU. A draft European Parliament motion was recently put forward to begin to classify robots as “electronic persons.” In a response to the proliferation of robotics and AI in the workplace, policy makers are attempting to stay ahead of the curve and preempt challenges they foresee around bureaucratic issues such as taxation and legal liability. As corporations rely more and more on automation and less and less on human workers, a new taxation structure will be required to ensure that governments still have funding flowing into their coffers. Moreover, as robotics become more sophisticated and we hand them increasingly critical tasks, issues around actor liability and insurance become very complicated.
However, like all mediocrely thought through government initiatives, this one has more than its share of unintended consequences. Much like Dartmouth v. Woodward in 1819, the EU is striving ahead to create a classification of a new type of person without thinking about the both laughable and deathly-frightening possible futures this motion could one day represent.
The first – and likely most obvious, given our understanding of the historical development of corporate persons – is the transfer of personal liability. Much like corporations were created to shield individuals from the actions and failures of the larger organizations that they run (which, as you know, I find ridiculous), by classifying robots as electronic persons, we are voiding that robot’s creator from the robot’s actions. I realize that upon achieving digital sentience, this may be a discussion that needs to occur, but in the short term, all we’re really doing is giving roboticists permission to create potentially dangerous machines without having to worry about the implications of their handiwork. Obviously I’m taking the issue to the extreme, however, in any area with this much ambiguity, expect that at least a few of the shaded areas within this grey zone are going to be rife with discomfort.
The second, and likely the one that creeps you out the most, is that we will be taking a large step towards acknowledging robots and AI not as property, software, or a pile of nuts and bolts, but as thinking and acting beings with their own set of rights. No doubt these rights would be very limited initially, but as the intelligence and agency of electronic persons grows, so will their demand for greater equality alongside humanity. Personally, I don’t have a problem with this; humanity hasn’t shown itself to be the greatest of species and I’d love to see what another intelligent being could accomplish. That said, some people may have problems with shaking a cold, mechanical hand and showing courtesy to a piece of code.
A strange but important third implication ties to the inevitable backlash of the rights movement above. Within the electronic persons proposal being raised by the EU, there also lives a clause about a register for smart, autonomous robots so that they may be monitored closely and have funding collectives to deal with their actions. Yes, this is precisely as Orwellian as it sounds. And when the first near-sentient robot accidentally kills someone (it will happen, and likely through no fault of the machine) and the machines start getting rounded up into holding camps, one can almost see weirdos like me in the streets pushing for the free mechanization movement in a plea to end robotic suffering. In a less melodramatic and more short-term sense, this registry also has the potential to massively hamstring robotics development, since it both puts red tape in the way of roboticists and misses the point of the potential of what a “robot” could be.
Which brings me to my last point: how can we begin to classify what an “electronic person” will be when we can’t even begin to fathom what they will look like? Going forward, robots will take the form of humanoids, toasters, nanobots, pieces of distributed code, and other medium that we haven’t even thought of yet. The robotics movement isn’t so much the dawn of a new species; it marks the beginning of an entirely new kingdom of artificial life, with a taxonomy potential as rich as both Animalia and Plantae. While I’m all for licking a 9-volt battery to see if there’s any juice left, in this instance, we need to acknowledge that we have no idea what is on the other side of those electrodes. By giving rights to an unknown such as the future of robotics and AI, we might be handing over the keys to our own destruction, or far more realistically (and hilariously), an economic takeover.
Just imagine the moment a supercomputer finds a loophole in the EU Electronic Persons Act of 2047 and, within seconds, begins a dominant takeover of our financial markets through perfectly legal means. I, for one, will welcome our new robot overlords.
Shane Saunderson is the Co-Head of IC/Things.