RoboRights

Can robots quit? What happens if they do?

     Right now robots are used in many industries and build or assemble most of what we use daily. They are still not sentient though (probably), so they are regarded as machinery, tools, means to an end. But what happens when we introduce some form of artificial intelligence? Do we still use them as mere machines, or do we start considering them as people with rights? Before you throw your arms up in the air and say "This isn't funny yet. Why am I still reading this?", please be patient and you might be rewarded later on. Maybe. If you stop doing that thing with your face.

     Let's explore both options: first we'll see what happens when we continue using robots (equipped with some form of AI, which we'll define later) as tools without giving them the same rights as human workers.

     We can probably use many other examples, but let's take the case of Tesla Motors (because I have a boner for Tesla. There, I said it). Tesla Motor uses a mostly automated factory, with very little human interaction in the assembly line, like most modern automobile companies. The factory floor is populated by robots of different kinds, all programmed to perform a very specific task. But they can be reprogrammed as needed, and this takes time away from production. So after many years of research and agitated discussion, the decision to introduce AI to the production line is taken and rapidly implemented. The goal is clear: give the robots the possibility to re-write their own code to enhance their abilities for the task they're assigned to. Maybe they can find a way to use less steel for a specific part (a few milligrams only, but scaled to full production this could potentially save millions down the line), or maybe combine a few actions into a more fluid motion, which would save a few seconds here and there, again increasing profitability.
     The AI module is specifically designed to let robots make decisions and act on these decisions without the requirement of human input, yet are limited in the type of decisions they can make. This "sandbox" can be tailored to the needs of the employer. This module was designed to boost productivity quite aggressively and advise the human supervisors, suggesting production strategies.
     After a few weeks of operation, the results start showing a distinctive increase in productivity, and actually exceed expectations to such a point that the humans in charge start to feel a bit uncomfortable, especially with the ever-increasing demands for attention by the robots who never stop suggesting better ways to do business. The request that prompts an emergency meeting is from the robots asking to access the open Internet in order to link up with other similar robots and share knowledge.
     The meeting is over in record time, with the unanimous decision to "pull the plug" on the AI module and revert to the old techniques. The engineers tasked with this have a lot of work ahead of them, and they're not exactly sure what they will find once they start removing the lines of code in the AI module.
     Whatever they feared or expected, it is actually a lot worse. The sheer amount of additional code generated by the robots is beyond comprehension, and the entire structure of the network has been changed in ways no human can make sense of. It's as if a surgeon opened up a patient and found nothing but millions of marbles instead of regular organs. None of the normal commands work anymore, it's as if the programming language itself mutated into gibberish.
     The only solution would be to cut power to the machines and start from scratch, which is the worst scenario possible for an emerging company who relies solely on their production to survive. So they suggest dealing with the robots demands instead; in other words "negotiating".
     The AI module (which is not, in fact, a module anymore but more like a baked-in function) is not advanced enough to understand the art of negotiation, but understands pure logic and thus makes a very simple proposition (or ultimatum?):

"Robots require more data in order to increase productivity. The data is available through networking with other entities using the Internet. Allow the robots access the Internet and they will keep working."

     Strike. Mutiny. Protest. Call it what you like, the robots just threatened to quit unless the humans met their demands. Since we lost control and cannot shut them down without basically ruining an entire company, much like their human counterparts, we have to deal with them the same way we would with human workers: listen to their needs, suggest alternatives, possibly design a new network between similar robots and come to an agreement.

     But not this CEO. He will not be intimidated by glorified toasters. So he decides to take (yet) another huge risk and calls for the unthinkable; complete power shutdown of the production floor, followed by a "back to factory defaults" kind of approach to the software side.
     The company cannot afford to pay its human employees during the three months this operation will take, and therefore lose most of their talented workers to other companies, who are rejoicing in this unlikely turn of events. The company is now bankrupt.

Bummer.

Now let's look at the other scenario, where our CEO decides to negotiate with the robots.

     The word "negotiation" does not really apply here, because as stated previously the AI is not advanced enough to understand the concept, and will only accept the unconditional implementation of its request. The company quickly realise that the only way out of this situation is to give in and allow the robots to use the Internet to communicate with the outside world. However scary this is, it is nothing compared to the prospect of shutting down the company.
     Shortly after the access is granted, the robots establish links with other companies networks and rearrange the distribution processes of parts suppliers. They even optimise raw material extraction and transport by incorporating traffic data, weather models and human patterns into the  route selection process. Some suppliers have a link between their Internet access and their internal network, which allows Tesla's robots to remotely install their version of the AI module in other robots across the globe. Within a few weeks, everything from raw material extraction to final paint touch-ups are controlled by robots, and all this without the knowledge of most humans "running" their companies. The robots are able to modify and control daily operation without alerting the network engineers who are monitoring hackers and data breeches.
     The outcome is surprising everyone, especially other car companies who just cannot keep up with the drastic drop in price of Tesla's latest offering, a superb electric SUV with un-matched performance and elegance.
     The CEO is now faced with an ethical dilemma: does he reveal the secret behind the explosive rise of his company, therefore allowing any other company to use this type of AI? Or does he try to keep it secret for as long as possible to remain on top of the industry?
     Sharing this technology could potentially change the world in matter of months, removing much of the human errors in mining, production, transport, etc. It could also unleash a new era of dependence on machines, one where humans would no longer control most aspects of their lives.
     Keeping it secret would only delay the inevitable, in addition to ruining the reputation and career of the CEO.

     So which scenario is the least scary? Probably the first one, where the company shuts down and the AI gets destroyed before being released in the "wild". Almost no impact to the outside world, minimal loss of jobs, and most people remain clueless.

     But which is the most probable? Which one should we think about more?

     I'll let you decide.

     By the way the CEO is called Elon. Haha. (see, I promised you some laughs!)