Artificial intelligence (AI) is changing the way we do things, how we interact with others, but also what we know about ourselves. It's already a tired cliché. This fourth industrial revolution is expected to have a huge impact on industrial, social, and economic changes on humankind in the 21st century and beyond.
What is it exactly? Some see AI as a technology that allows computers and machines to function intelligently. Some see it as a machine that replaces workers and is, to boot, more effective, flexible, and speedier. Others see it as a system that interprets external data, simulates human learning from such data, solves problems, and uses the knowledge acquired to achieve certain goals and tasks.
Most people do not realise that AI is the fruit of some 70 years of development. It emerged in the 1950's with some pretty basic machines that exhibited human intelligence. By the 1980's it had evolved into machines learning from historical data. By the 2010's, we had the first machine learning models that mimicked human brain functions. A decade later, we started seeing the deep learning models of Generative AI that create 'original' content.
It's easy to get lost in all the categories, ranging from Reactive Machines (responding to input from pre-programmed rules and having no memory of past experiences, like in simple chess); Limited Memory AI (systems that store past experiences and use them to make decisions, like self-driving cars); Theory of Mind AI ̶ still a hypothetical type that can understand and model the thoughts, beliefs, and emotions of others; and Self-Aware AI ̶ another hypothetical conscious model that is self-aware and has its own desires and beliefs, like in science fiction.
The applications of AI are innumerable. The earliest forms were the robots that automated manufacturing and quality control, consumer assistants like Siri and Alexa, followed by recommendation systems for e-commerce (e-Bay, Amazon), then expanded to fraud detection (banks), transport (autonomous vehicles), chatbots, image and facial recognition, and medical diagnosis and treatment (the robots dispensing medicines at Mater Dei or assisting in precision surgeries).
The list is endless. AI is in education, entertainment, cybersecurity, agriculture, legal affairs, real estate, lifestyle, navigation, vision, gaming, social media, marketing, astronomy, and travel, among others. It is now in newspapers. We have already had several instances of news reports and opinions being written by AI. Serious journalists or newspapers would indicate whether this is the case or not (always look at the bottom of the piece).
Do we really need AI? Take me. Choice No1 is to spend several hours reading about the forms of AI, who has written about it and said what, who are the companies involved in the industry, where it is used, the extent of its use, what it costs, what energy is involved in machine-learning, and all sorts of other information. Choice No 2 is to ask ChatGPT or another AI application to write 1,400 words about all the matters I have just mentioned. I did this just now, and it took all of three seconds! If I were to request 2,400 words about it, it would still take three seconds.
Some people would say that it is obvious what I should do, particularly since I am not paid for writing this article. In my case, however, the downside of doing it by AI is that I get no joy out of it. Being semi-retired, I prefer occupying myself for a couple of hours rather than for a mere three seconds. Often, I get way-laid in my search by something which is not related at all to what I am writing about, although it is interesting to learn. AI would never do that.
Until recently, the problem with using AI to write this article would have been that it would not have been in the style that I have developed over the years, and which is instantly recognisable to people who follow me. But that is now no longer a problem. There are tools that will read all the articles I have ever had published, analyse my linguistic footprint, and then compile the 1,400 words I need in my own style.
Naturally, my choice No 1 is totally contrary to Malta's National AI Strategy, which would have me use AI to gain 'a competitive advantage in the nation's effort to be a leader in the AI field'. The government's plan is built on three pillars: investment, start-ups and innovation; public sector adoption; and private sector adoption. Malta Enterprise, MITA, the educational institutions, MFSA, and dozens of other entities are busy working on this, spending millions out of public, EU and private funds.
Modern-day Luddites would say that AI will have huge negative impacts. Millions will become unemployed, they claim; wealth inequality will increase as investors and professionals cash in, at the expense of uneducated and unskilled people; ethical issues will become more complicated; AI could even destroy humanity itself. No doubt, some of these ills will happen. As the liberal intellectual Henry Steele once said, "Change does not necessarily assure progress, but progress implacably requires change."
I was struck by something that the former Greek leader George Papandreou said at the United Nations AI for Good Summit in Geneva. Despite the tech industry's push to position AI as the foundation of all human knowledge, Papandreou thinks Plato would see AI "not as the sunlight outside the cave but as the shadows inside: intriguing and persuasive - but misleading, and far from the truth."
He was referring, of course, to Plato's cave, where humans watch silhouettes flicker on the wall, oblivious that fires just behind them are distorting reality into a shadow world. If someone managed to escape the cave, his eyes would be blinded by the sunlight, though he would eventually see the world for what it truly is. Unfortunately, the escapee would not manage to rescue the others because he would fail to convince them to leave the cave.
Plato wrote his cave allegory around 380 B.C., but he may as well have been writing today, according to Papandreou, who thinks that AI might be used by demagogues or powerful people to concentrate power and bully, if not dominate, society. They would do this by distorting the facts and convincing enough people to live in the shadows of their virtual mreality echo-chambers, rather than by interacting openly in genuine dialogue with others.
The danger is there, like it was when Plato's mentor Socrates drew attention to the flaws of democracy. Socrates asserted that people should debate respectfully and understand each other, rather than bully others to win arguments. Of course, we all know he was killed for it. Had he lived today, would Socrates see AI as something that would contribute to a just society, or would he see it as a threat at the hands of those who own it and control the algorithms?
Socrates's answer, Papandreou believes, would be that today AI is another one of the many paradoxes and threats to democracy. Democracy was invented by the Greek philosophers to make sure that power is not concentrated. Yet, today a few hundred individuals are exerting power and control beyond imagination. In the early days of AI, the fear was that humans would be replaced by robots. Now, we should be more afraid of humans being turned into robots.
N.B. This is not an AI-generated opinion.