The only human quality AI would never have…
What will happen when AI starts questioning our ways of the world?
I was discussing the challenges in realising general AI (AGI) with some of my acquaintances when I recalled a story from my school textbook somewhere around 1993–94. My English language book had this story named “Anukul”, written by one of the talented Indian polymaths, Satyajit Ray. He wrote the story in the late 1960s.
While it demonstrates fore-sightedness of some of the talented people of the past century, you may also find this story to be borderline spooky if not outright disturbing.
The story is about a humanoid-robot, an example of a general AI manifested in the form of a personal assistant. Another film-maker recently made a short film based on this story. The story also hints about what may happen when AI starts questioning our ways of the world.
What will happen when AI starts questioning our ways of the world?
The story
The main character (Nikunj Babu) of the story is a Hindi (language) teacher, who bought a humanoid personal assistant from a local robot supplier. When purchasing, the salesperson warns not to hit the robot named Anukul, under any circumstances.
A few days later, Nikunj Babu’s cousin brother (Ratan) comes to visit him. When Ratan sees the robot, first he seems surprised, but secretly he is not happy with his brother’s decision. Ratan had been recently fired from his workplace due to human-robot complications (this is why I recommend having a HAIR, i.e. Human & AI Resources department). Ratan then proceeds to hit Anukul with a clothing iron when Nikunj Babu wasn’t watching. This hit by Ratan results in Anukul’s shut-down. Local robot supplier is called for the repair.
In the second visit by Ratan, he tells Nikunj Babu that he had become a millionaire by inheriting wealth from a recently deceased uncle. While Nikunj Babu is away, Ratan again tries to hit Anukul. However, this time, Anukul gives him a high voltage electric shock resulting in Ratan’s immediate death.
Later, a lawyer visits Nikunj Babu to inform him that now since Ratan has passed away, he has inherited all the wealth of Ratan, circa $1.5 million.
I will leave it with you to make sense of the plot, the ending of the story, and conclusions you want to draw from it.
The role of conscience
The conscience is the leading part of our spirit. It is a moral sense of right and wrong that is viewed to be acting as a guide to one’s behaviour. As Wikipedia elaborates, conscience is a cognitive process that elicits emotion and rational associations based on an individual’s moral philosophy or value system.
Our ways of the world are still very much a grey area — we cannot explain many things in a straight forward manner with a repeatable logic. This ambiguity makes human behaviour least replicable. However, with such a complex (human) operation, conscience plays an essential role in binding us towards a meaningful goal all the time. It is what makes us rational in our actions and decision making.
However, if AI has to mimic humans in some way, whether it is via narrow AI or relatively broader AI or a general AI, certain human qualities are a must-have. Without which it will not be a good idea to let the AI roam around at free will. The human must always be kept in a loop, lest things go awry.
There are scenarios, where conscience may have no role. Classic repetitive actions are an excellent example of one such situation. However, if there is even the slightest impact on end-users, customers, employees, or society — human control is necessary. An AI may take a decision but should not be allowed to act on it without confirmation from its responsible master.
Important learning
The story is just a story until someone is so inspired by it that they make it a reality. Know what you want from AI and why — use that purpose as a guiding star, all the time. The purpose of using technology will help you as a human-AI owner to maintain control.
Know what you want from AI and why — use that purpose as a guiding star, all the time.
It is never a good idea to believe that the computer is always right. Do not think so AI blindly — when in doubt, prefer human decisions. Yes, we humans are flawed, yet, there are several human qualities that AI will never have or be able to replicate. Qualities and characteristics such as — conscience, empathy, emotional intelligence, leadership, love, hate, and several others. These qualities make us humans the way we are and differentiate us from the rest of the species.
It is never a good idea to believe that the computer is always right.
AI makes decisions based on how it has been taught to do so. Who taught it, how, with what inputs and when all these matters. Seek transparency and interpretability of AI decisions. If you cannot get hold of them, be very sceptical, you might be entering into a rabbit hole.
Seek transparency and interpretability of AI decisions.
No matter how intelligent and great an AI, the solution is, always seek for a kill switch and access to full control when necessary. You may not need it or use it regularly. However, when required, it should be accessible easily, just as quickly as the power button on your smartphone.
Remember, you know your business than anyone else, AI included — trust yourself and trust humans!
Whenever you are using or planning to implement AI in your business or life, think about the consequences. No matter how well-designed AI system is, it inherently carries certain risks at various levels. Estimate these risks objectively and then plan for implications, let me know if you need any help.
You will not become rich if your AI makes a mistake, it happens only in movies! 😉
About the Author: I am many things packed inside one person: a serial entrepreneur, an award-winning published author, a prolific keynote speaker, a savvy business advisor, and an intense spiritual seeker. I write boldly, talk deeply, and mentor startups, empathetically.
If you liked this article, subscribe to my newsletter for more such articles and connect with me on LinkedIn.
