Very smart people who know how the new AI works are issuing warnings about its dangers. These well-informed folks are worried about what the AI might do as it becomes bigger and better in the future. Like, for example, maybe it will get smarter than us – and then dominate us. Or maybe it will take work away or even eliminate jobs of people who work with their minds. Surely it will make it possible for bad people to do more and bigger bad things.
The new AI is indeed a game-changer, a transformational new technology destined to change civilization in major and unforeseen ways. Like, for example, the printing press was a game-changer in 1440. Imagine what might have been said back then, and probably was said, about that new technology: scribes will be put out of work; bad people will print bad things, misleading things, even wrong things; people, ordinary people, will have information and learning they should not have, that they can’t handle; people who control the presses will control the world; and so on. Printing was as scary and dangerous a development then as the new AI is today.
What mainly happened back then, though, was that the masses of people learned how to read! Books, good and bad, were written, published, and read. Political pamphlets and advertising, both honest and dishonest, were printed and read. Religion and the literate elite no longer controlled all thought. Changed and richer, life went on.
Were there people who said “Bah humbug” and refused to learn to read, or who wouldn’t read the books printed by evil mechanical machines? No doubt, but history makes no account of them – just as history will make no account of those today and tomorrow who will not adapt and make use of the new AI.
The inventions of the Industrial Revolution (around 1760 to 1840) were also earth-shaking. The Industrial Revolution was a transition from making things by hand to making them by machine. Inventions had to do, not with thought and language, but with manufacturing, transportation., and energy (the steam engine, the internal combustion engine, electric motors). The nature of work changed fundamentally. People had to learn new skills and new ways of doing things. Those who did had jobs; those who didn’t, well, they were left behind.
The term Luddites originates from those days. It refers to an organization of English textile workers who destroyed textile machinery. The group is believed to have taken its name from Ned Ludd, a legendary weaver. Today, the term Luddite refers to anyone opposed for whatever reason to new technology. The new AI will certainly have its share of modern-day Luddites.
We don’t know yet what changes the new AI will mean for intellectual work and perhaps also physical work. However, we’d be awfully obtuse if we didn’t expect education to change, some mental jobs to disappear., and many others to be remodeled. There are those, including yours truly, who thought the computer would by itself create such changes, and it has. But one cannot avoid the anxious feeling that bigger, more fundamental, and more transformational changes are now ahead.
The reaction of Luddite-leaning people to technological change is interesting to observe. I remember when hand calculators became available to engineering and science students. There were faculty who prohibited students using the calculators because, those faculty argued, the students would lose their ability to do computation. The fear of computers was similar: some faculty agued that computers would lead to students who couldn’t solve problems. Such consequences never happened, of course; the students just used the new tools, worked at a higher level of abstraction, and got more capable. What happened to the Luddite faculty? They never learned or made use of computers, and so faded from view.
How will our society and our government react to the new AI? We really don’t know yet, but certainly slower than the developers of the AI will proceed. There is so much hype, so much competition (world-wide), and so much money at stake, the new AI is surely unstoppable.
However, the new AI has some problems. I am personally not worried about the changes the technology may make to jobs, to the possible discrimination and bias that some fear will happen, or to the evil things some evil-doers wiil try. These seem to me to be inevitable, like auto deaths that come with automobile technology. We’ll learn to cope with such things, and will come to believe that the benefits of AI outweigh the risks and costs.
Another problem is that training the AIs requires enormous amounts of energy. No-one is sure the sun and wind are up to the task.
But the worst problem is that AI makes mistakes, and, worse yet, it hallucinates; that is, it makes things up. When it doesn’t know what to say next, instead of saying it doesn’t know, it makes something up, it fantasizes. Apparently, the way this works is not understood.
Whoa!! Stop the presses! This is like a human neuropsychiatric disorder; it must be understood in the AIs and fixed. We can’t have AIs with mental illness. If AI is going to be used in serious ways, such fundamental flaws must be fixed. As Henry Kissinger has pointed out, It certainly can’t be used in National security. What if it made up a fake missile strike on us and we ‘retaliated’ erroneously? Or vice-versa?
What do we little people do with such a huge change-making new technology? We do like the people did with the printing press: they learned to read. So we learn to use this new thing. (It’s easy and free to get ChatGPT for your computer.) However, there’s a caveat with the AI. Since it makes mistakes and hallucinates, we can’t trust it. We take what it tells us with a grain or two of salt until we can check it out.
This new AI is here to stay. It won’t be stopped or even paused as some Luddites want. We’ll have to adapt, make adjustments, fix the problems, and live with imperfection just like we’ve done with every other new technology. Teachers may have the toughest problems with it, but also the biggest opportunities. Teachers who adapt and learn to use AI to the advantage of students will be teaching forever, and will be contributing to the changing world. Those who don’t will be doing hall duty.
The striking TV writers also have a very real challenge. Their union wants “no AI” used, none at all. Really? Well, no matter what results from the strike, some writers will anyway learn to use AI so they can write even better and faster than Taylor Sheridan (Yellowstone, 1923, 1883). They’ll be working and writing; those who don’t will still be carrying ‘No AI’ posters on the picket line.
We humans are tool-designers, tool-makers, and tool-users. The new AI is a powerful new tool. So we use it, bearing in mind that it still makes mistakes and hallucinates. So long as it does, we remain skeptical of its results. It is, after all, only a computer program.
Just Sayin’.
John R. Dixon 