Ghost in the Shell

 Rian Benoit

HON 255-2

November 3, 2022

 

If we are part human/part machine, do Asimov’s rules still apply?

 

Isaac Asimov developed a set of rules designed to hold robots and humans to high ethical standards. Created in 1942, this take on responsibility for artificial intelligence appeared in most of Asimov’s works and is still commonly referenced in pop culture and related literature. While Asimov’s rules were instrumental in further developing the popularity of both science fiction novels and films, his laws are impractical and fail to reflect what most science fiction stories entail. The laws of protecting and not harming humans, obeying orders from humans that won’t cause harm, and protecting oneself as a robot unless this would violate the preceding standards. Many are only familiar with Asimov’s 3 laws. However, he later added another one, called the zeroth law, that took priority over the others. This fourth law stipulated that “a robot may not harm humanity, or, by inaction, allow humanity to come to harm.” With these laws established, many have debated and altered them to fit a new narrative or express their own opinions on what the future of artificial intelligence could entail.

            A hybrid between a machine and a human would be entirely unable to adhere to Asimov’s rules. This is because humans act and think emotionally, even though we might like to think we are more logical in our decisions and actions. And robots follow algorithms and employ fact-based analysis in their decision-making. These concepts clash with one another and don’t leave room for following specific, very logical laws. In short, the human part of a hybrid model would be unable to refrain from bringing harm to other humans because we often cause emotional damage to others without knowing it. Additionally, humans struggle with taking orders unless specifically predisposed or trained. Therefore, an inability to follow directions would deter any part of full human creation from carrying out orders the way a robot is programmed to do. And lastly, humans have always placed their own safety above all else. While humans may be capable of putting their own protection first, a robot is told to do the same, creating this never-ending push and pull that ultimately would destroy any being composed of a natural and mechanical part.

            Alternatively, this combination of human and robot lifeforms would perform even better than just a robot or a human. This is because the concept of two minds being better than one might allow the natural and mechanical components to protect and instruct each other in accomplishing tasks. For example, acting without considering all the options is the human body’s way of protecting its brain from mental harm in stressful situations and its body from physical damage. By employing fight or flight instincts. Whereas robots do the complete opposite. They follow rational thinking patterns, are programmed not to act rashly, and don’t have emotions and mental struggles barring them from accomplishing their goals. With these two options being so far from one another on the spectrum of Artificial Intelligence, it would make sense that 50% human and 50% robotic thinking would create this incredibly talented being entirely capable of following each of Asimov’s 4 laws without failure.

While both opinions deserve a space in the pursuit of Artificial Intelligence, the same can be said of any theory. There is so much yet to be discovered. So much R&D to be invested in, disallowing us from being able to determine what a cross between a human and a robot might be capable of. For all we know, they could match their strength and weaknesses to create an incredibly intelligent and understanding being that works to follow Asimov’s overarching fourth law. But even Asimov admitted it. Humans can and will always injure other humans in the pursuit of power and are rarely able to think 100% rationally, so we may never reach our full Artificial Intelligence capabilities unless something changes within our species first.