Penelope Maclachlan

Artificial Intelligence (AI) is deployed by unconscious machines whose operating system is digital, not biological. Human Intelligence (HI) encompasses the ability to solve complex problems by deploying experience and intuition. 

Before we decide whether AI is better than HI, we may look at the pros and cons of both.  Some of the advantages of AI are that its agents (computers and robots, for example) work fast. They don’t need breaks and rest, as human do.  AI is more efficient at ordering  purely routine jobs, such as moving objects from one place to another, than HI.

Nevertheless, some people are apprehensive about AI, or suspicious of it.  They fear job displacement; if a robot cleans offices faster than the hardest-working human, why should a chief executive officer pay the human’s wages? Robots needs maintenance and updating, but in the long run the costs of operating one are less than employing a human. Besides, however well-meaning human cleaners are, they may fall ill or become injured, and unable temporarily or even permanently to work. Nick Ludd’s story may be legendary rather than factual, but most British people have heard of the story of his breaking machines for taking jobs, and so bread, away from human workers. Fear of technology is widespread.

AI has limitations. Its agents can’t think outside the box.  The follow algorithms (procedures or rules) and are unable to adapt them to changing circumstances. When you engage with a “chat” service on a computer, one of the signs is that your statements are turned around. Another is that the agent is always online and ready to respond. Replies never contain typos. The conversation always proceeds in the same direction. Every response comes with a disclaimer.

AI threatens data privacy. A manager who dislikes an employee can, by exploiting AI, pry into the employee’s private life and use unfairly gained information to block promotion or compile excuses for dismissal.

The salient fact to understand about AI is that it is built by humans. Humans make mistakes. Some have biases which, consciously or subconsciously, they embed in their programs.  Like AI, HI has limitations. Some humans are good and trustworthy. Others are evil. Murderers and torturers are evil. It is incorrect to ascribe virtues or crimes to AI; it is capable of neither. An AI agent cannot kill unless humans have programmed it to do so. Then the evil arises not from the agent, a machine, but the human programmer.

Humans can and do misuse AI. An example is in the assessing of replies to questions in academic examinations.  The assessor may suspect an answer is generated by AI because the language is simple and terse. If the examinee is writing in a language other than their own, though, they are likely to use language which may seem clipped and formulaic, and lacking the fluency and vocabulary available to a native speaker.

AI is a bad master but a good servant. It helps people with certain illnesses. One is ALS – amyotrophic  lateral sclerosis, a kind of MND – motor neuron disease. An effect of ALS is gradual loss of speech. If AI is introduced while the patient can still speak, it can record and store their words. Later, after the patient has lost the ability to speak, they can choose words from a personalised program onscreen to express their meaning.

Choice is a human ability which good leaders and teachers deploy with care. We can choose to utilise AI   for good purposes, such as helping people with disabilities. It is imperative, though, that the head of AI services has the training, qualifications and personal strength of character for so responsible a position. Reputable lawyers and scientists may be the most suitable to select such a person. The next step must be to promulgate laws to protect the vulnerable and innocent, and punish those who try to evade or flout the laws.    

---------------------------------------------------------------------------------------------- *Material should not be published in another periodical before at least one year has elapsed since publication in Whispering Dialogue. *أن لا يكون النص قد تم نشره في أي صحيفة أو موقع أليكتروني على الأقل (لمدة سنة) من تاريخ النشر. *All content © 2021 Whispering Dialogue or respective authors and publishers, and may not be used elsewhere without written permission. جميع الحقوق محفوظة للناشر الرسمي لدورية (هَمْس الحِوار) Whispering Dialogue ولا يجوز إعادة النشر في أيّة دورية أخرى دون أخذ الإذن من الناشر مع الشكر الجزيل

Leave a Reply