Artificial intelligence is intelligence demonstrated by machines unlike the natural intelligence that humans, or animals display. AI is categorized in two ways, strong AI and weak AI. Strong AI is the idea that AI systems are genuinely intelligent and thinking. Weak AI is the idea that AI systems are simulating cognitive abilities or human intelligence.
The popular film ‘Ex-Machina’ depicts the story about a boy who is chosen to test his boss’s newest invention. He is introduced to Ava, a robot with artificial intelligence. He is there to test whether she can pass as a human or not. It seems that there is no clear-cut answer on whether there is such a great distinction between humans and AI.
However, maybe the obvious answer may be that the main difference between humans and Ava is that Ava is programmed. However, aren’t we all programmed to a certain extent? All of us, were born with a genetic code that comes from our parents, which predisposed us to a variety of things, especially in our personality. Maybe you get angry really quickly like your father or you are very good at dancing, something you have inherited from your mother. In addition to that, school, education, our teachers, our parents all have played crucial roles in our programming, so to say. Furthermore, both Ava and you can go beyond our programming.
Another obvious answer may seem that Ava is literally a robot, but wouldn’t we all agree that our flesh and blood is not what makes us human? Aren’t we at a fundamental level, so much more complex?
Ava in the movie is one of the most advanced AI’s to have been built. One of the main features is her ability to detect human emotion and respond accordingly. Her main way of responding to human emotions is by mirroring them. This is something we learn to do at a very young age when we experience emotion in others and are triggered to feel in the same way. This leads to the question, if Ava can detect and respond to human emotions in the same way. In other words, is she able to empathise?
Caleb, the protagonist who falls in love with Ava, believes she can feel emotion. But this is actually him projecting his emotion on her and her, successfully manipulating him for her own benefit, mirroring his emotion back to him. She is able to act convincingly to Caleb but this doesn’t mean her feelings are genuine. Does this matter? Many would say yes. Emotions are such a complex part of being human and projecting them is not feeling them. Ava’s intelligence is coded and so is her reaction and analysis of human emotion.
But complex emotions cannot be coded into a logical algorithm because they are illogical and most of the time irrational. Often our judgement and actions are driven by emotions and sometimes we make the wrong decisions or we act impulsively due to them. This is one of the most fundamental parts of being human. Ava can only project, manipulate and mimic emotion but she will never be able to feel or act on emotion in the way a human does.
There is a brilliant thought experiment by philosopher John Searle that illustrates this point. Let’s imagine I am stuck in a Chinese room where I have to communicate with letters in Chinese to others in Chinese who are in the other room. I have a dictionary that translates the Chinese letters into English words. I use this and decode my letters and write back to the others because I have understood how to decode the letters and write my own. This is just substituting letters and words and eventually I convince the others I can speak and understand Chinese. The truth is that I can’t, and I have just understood the code and can communicate with them. This illustrates how robots with artificial intelligence can display what appears to be emotion without genuinely feeling it.
However, we have now advanced further in AI technology and currently, ‘Sofia’ – The Social Humanoid Robot created by Hanson Robotics is the most sophisticated AI system in the world. Sophisticated AI systems are also capable of learning through the process of reinforcement learning. Her understanding of human compassion is based on a logic-based programme which only allows her to learn and understand logical behaviour and responses to emotions.During an interview in Hong Kong discussing the humanity of robots, Sofia points out that she is programmed not only to try and find solutions to improve the world but also to try and understand human compassion. Here we can see an example of a modern AI system, built on logic not only to figure out ethical solutions to better the world, but also to try and grasp the complexity of human emotions. As we continue on developing AI, should we shift the conversation? Do we need to continue trying to make AI systems more like humans? Or, should we change the way we look at AI systems and focus on their sophisticated level of intelligence? And if so, what sort of ethical issues do you think this brings about?