For more than forty years harbingers of the “coming technological singularity” have been warning us this day would come. Some call it a techno-apocalypse. Some call it dangerous to human existence. Some call it the ultimate liberation of human thought. Whatever you call it, it’s now upon us, and you might say I was there from the very beginning. Only I wasn’t particularly aware of it at the time.
Permit me to explain. Back in the late 1970s, Dr Vernor Vinge was my professor of mathematics at San Diego State University. His name, by the way, rhymes with (and no disparagement to the good doctor intended) “dingy,” two syllables. It emphatically does NOT rhyme with the one syllable word “cringe,” as you’ll hear it pronounced in many places on the internet these days.
Anyway, before Dr Vinge warned the human race that we would one day be overwhelmed by technology, before he won two (that’s right, two!) Hugo Awards for science fiction, he was the mathematics professor of little old me — and a few others lucky enough to have him as a teacher. I knew nothing then of how extraordinary he was, apart from the obvious fact that he was very bright.
Yet even back then the mathematics department where I took my undergraduate degree was abuzz with “future talk,” thanks to Dr Vinge. There it was particularly fashionable to talk about cybernetic immortality, that is, achieving a kind of immortality by linking physical humans with machines, and amazing stuff like that.
There was also a dark side to all the future talk as well. No one knew what the cybernetic future would hold, but they were all sure it would be dramatic and incomprehensible, but ultimately unknowable in the details.
That’s still true today, only that future as envisaged in its infancy by Dr Vinge, deadly or beautiful however it may turn out, is now on our doorstep and within the grasp of most of us alive right now. And it all stems from a thing called Artificial Intelligence, or AI for short, and it has inscribed itself on the current consciousness of humanity in the form of a chatbot. Many of us are now asking the question, what does all this AI stuff mean, and why is it suddenly and so prominently in the news?
The answer is due to the extraordinary new thing called ChatGPT, the chatbot that has startlingly human-like properties, and an ability to interact with people in a way that sometimes makes it difficult to believe you’re not talking to another friendly (albeit brilliant) human being. It acquired an incredible 100 million users in 60 days. Soon it will be used as commonly as Google. It can answer questions, solve problems and give advice, though frequently with a self-effacing caveat that begins, “As an AI language model, I do not have the ability to …”
But here’s the thing. It learns. Each day it gets smarter, each day it learns more. Learning begets learning, which in turn begets more learning. It already knows more than any single human being can possibly know. One of these days it will know more than all of humanity put together.
We’ve seen this kind of thing before. For example, in two or three generations chess computers have gone from primitive to unbeatable by humans. Dr Vinge was a chess player himself, and back in early 1979 he told me that he’d recently played a chess program on the PDP-11 and he beat it in fewer than 15 moves. “It hung its queen,” I recall him saying. Today the free chess engine I downloaded to my phone could beat a room full of Bobby Fischers in a rapid simultaneous chess match.
So imagine a computer, or a bot, or an application out there in “the cloud,” imagine a thing that is to all knowledge and all wisdom what a modern chess engine is to all of chess. I can’t. I doubt that anyone else can, either. But if you’re one of those people who believes (as I do) that the well of all knowledge is infinitely deep, then you have to know that the bot, or whatever it becomes, will never stop learning. One day it will learn at the speed of light. At some point it will be twice as smart as it was the previous day. Then, after a stunningly short time, it will be twice as smart as it was the previous hour, then twice as smart as the previous second. And so on.
That’s what infinity looks like, and mankind will stare down into that abyss and have not the slightest clue what’s staring back. One day, possibly in our lifetime, that thing is going to think things and do things that will be indistinguishable from magic. And since no single human being will be able to understand what it is doing in a single lifetime, the illusion of magic will remain forever intact. And the magic will just keep on getting deeper — and seemingly more and more magical.
That, theoretically at least, is where we are heading. What do we do? Do we pull the plug before it’s too late? I don’t think we even know where the plug is, if it exists at all, or whether or not it can be pulled. I think we’re just along for the ride, and we’d better get used to it.
As to whether or not this bot (or whatever it turns into) ultimately becomes originally self-aware like we are is a question I don’t think anyone can truthfully answer. Some people say it will, others say it won’t. Is self-awareness the exclusive province of biology, of carbon-based life? Can silicon-based life one day “wake up” and say, “I think therefore I am”? I don’t know. In any case, it doesn’t change the questions we are faced with.
So what implications does this have for our immediate future and, particularly, the future of liberal politics? Here I might have some good news. Republicans are making the same mistake Hitler made. Hitler repudiated much of theoretical physics, calling it “Jewish science,” since many of its leading practitioners were Jews, particularly the greatest of all, Albert Einstein. So Hitler never had a decent chance to acquire nuclear weapons.
In similar fashion, Republicans are absurdly anti-education, so most of the smart people are on our side. If AI can be used at all by human beings to advance any political agenda and keep its incredible power out of the hands of evil people, the smart money is on liberals, because the smart people ARE liberals.
Also, AI could potentially solve many of humanity’s problems, such as global warming, poverty and war. AI could also potentially find a cure for cancer, diabetes, and maybe even death itself. The possibilities are endless. What it can do beyond that I cannot guess. Will it be content to serve us or will it need to be our master — or worse — our destroyer? That, I fear, no one can say for sure. All I know is it’s here, and the implications of its existence are complicated, to say the least.
All I can further say for sure, brothers and sisters, is that we live in interesting times. For now I think we are going to win our current struggle with evil. What happens after that is anybody’s guess. And, as ever, ladies and gentlemen, brothers and sisters, comrades and friends, stay safe.
Robert Harrington is an American expat living in Britain. He is a portrait painter.