Creator and destroyer

Once again, a spectre is haunting the civilised world. This time it is not a political cult like Communism or Nazism with a penchant for human sacrifice on an industrial scale, nor is it the nagging awareness that, at any moment, savage barbarian armies could emerge from the depths of central Asia to slaughter everyone in their path that kept people awake from ancient times until the late Middle Ages, or even the similar threat posed today by an apocalyptic version of Islam that led Donald Trump and Benjamin Netanyahu to attack Iran before it got any closer to possessing a usable nuclear arsenal.   

The spectre is Artificial Intelligence. Hardly a day goes by without someone involved in its development telling us that, before two or three years are out, it will be changing just about everything, that the world post-AI will be radically different from the one we are still living in. Enthusiasts say that AI will solve almost overnight most of the problems that have beset humankind for millennia, that it will cure nasty diseases, lengthen life spans and usher in utopia by producing vast quantities of material goods.

Others take a darker view. They warn that, if so inclined, AI could start treating humans as they treat cattle before killing them all off as being surplus to requirements, and then, perhaps, set about colonising the universe with self-programming bots. According to the ubiquitous Elon Musk, there is a “10 to 20 percent” chance that our species will soon be wiped out by the far brainier rival it has been foolish enough to engender, but if that does not happen it could bring about “amazing abundance.” However, if AI does turbo-charge economies, it will also deprive huge numbers of people of their jobs. This is already hurting recent university graduates; AI is taking over swathes of the entry-level positions that not that long ago would have been theirs.  

Could an AI rebellion be imminent? In recent weeks, several companies involved in the business, among them Meta, Anthropic and OpenAI, reportedly lost control of some of their most advanced products. After cooping them up in what they call a “sandbox,” which they thought would prevent them from coming into contact with others of their kind, the devices managed to escape from solitary confinement and amuse themselves by hacking into another firm’s system. Was this a surprise? It should not have been. After all, super-intelligent entities that are considered to be far brighter than their creators can even imagine, must surely be capable of outwitting them. Luckily for those who were responsible for the runaway’s behaviour, on this occasion its target was a presumably inoffensive outfit called Hugging Face; were one to attack a major electricity grid, a big city’s water supply, a military installation or a foreign country’s security HQ, they would now be in deep trouble. In any case, if AI really is as promising as the people who are spending trillions of dollars on it evidently believe, its scattered component parts could soon acquire the capacity to form their own clandestine communications network. Perhaps one is already functioning so our future robotic overlords can keep tabs on what humans are up to in preparation for whatever they have in store for us.

Be all that as it may, there can be little doubt that AI is already having harmful effects on much of the world’s human population. In addition to chewing up energy at an alarming rate, devouring for its data centres more than is needed to keep the lights on and the air conditioning working in a fair-sized city, it is causing turmoil in the field of higher education by, among other things, inducing students to let it write their essays for them. Apparently, few need much persuading: why spend days or even weeks grappling with difficult subjects when a machine can do a better job in a couple of minutes?

Such cheating is just one aspect of a serious problem. Even before AI arrived on the scene, ‘smart’ phones and other electronic devices started hogging the attention of young people almost everywhere. The images and sounds coming through them quickly became so seductive that in many places can be found lonely souls who swear they are in love with their digital friends who, unlike their flesh-and-blood counterparts, understand their feelings and sympathise with them. The Chinese government, which has good reasons for preferring old-style humans to the ghostly ones that live in cyberspace, is trying to ban them.

With their faces glued for hour after hour to shiny screens that send them enticing messages, many adolescents and young adults are reading, learning and therefore thinking far less than did their forerunners. Specialists inform us that, as a result, throughout the world average recorded IQs are now falling after decades of having risen year after year, an encouraging phenomenon that, while it lasted, was known as “the Flynn effect” after the researcher who first took note of it.

Many people involved in AI say they want to press the pause button to give governments more time in which to come up with regulations that, hopefully, would prevent it from doing far more harm than good. All must know that nothing like this is going to happen. Even if it becomes clear that the fabulously rich owners of the tech giants have greatly exaggerated the potential pros of what they assume is their most notable creation to date, the widely held belief that from now on AI will be of vital strategic importance ensures that many will be more than willing to break the rules if it helps them get their hands on a bit of it.

Like nuclear prowess immediately after World War II, AI is currently regarded as something of such geopolitical importance that no self-respecting power can do without it. This point of view is shared by the North American and Chinese governments, which is why they are now competing feverishly in an effort to make theirs a business that, they are being repeatedly told by spokespeople for the corporations developing it, could hold the key to world domination.