The use and development of Artificial Intelligence is only going to grow. I have absolutely no idea when the next big breakthrough will come, or what it will look like. Some scientists remain sceptical that we will ever arrive at the science-fiction version of AI that we have been promised for decades. But that doesn’t mean it isn’t possible.
As we continue to explore these technologies, the big question is how we control them — and how we stop them falling into the hands of people who might use them for the wrong reasons. Will AI eventually become our final invention?
For now, all we can really do is stay informed and keep talking about it.
But perhaps the more interesting question isn’t whether AI will replace human invention. It’s whether AI will eventually create breakthroughs that humans simply aren’t capable of imagining.
Every major leap in history has come from a mind capable of thinking beyond the accepted boundaries of its time. But what happens if AI begins to generate new scientific principles, new forms of intelligence, or entirely new ways of manipulating the world that aren’t based on human thinking at all? Would we even recognise them as inventions?
And if AI starts innovating beyond our ability to understand, what happens to the role of human creativity?
Perhaps that’s the point at which we finally discover whether we invented AI…or whether, rather worryingly, AI invented us. Which frankly, would be rather inconvenient.
AI: the Final Invention?
I am mostly positive about AI; I believe this invention would, in general, make more use than harm. Yes, it is affecting – and will continue to – a whole range of jobs, including mine, in both good and bad ways. However, there are ups and downs so you can’t really say that AI is a 100% good or bad thing.
Again, I have been positive about it, ‘till I recalled a story by Robert Sheckley, a profound American sci-fi author, actually, one of the greats of the genre’s Golden Age.
This is a short story, first published in Galaxy Science Fiction magazine in February 1953. In it, flying robot “watchbirds” are created to prevent murders by detecting intent through biochemical and electrical signals. Thanks to their learning circuits, they keep expanding the definition of “murder” — to include surgical operations, slaughtering livestock, even protecting insects and themselves — and eventually start killing or incapacitating people in the name of “protecting life.”
Well, I am not going to draw direct parallels between those watchbirds and the modern AI. Yet, if Artificial Intelligence goes out of control – and I am not qualified enough on the issue to claim it – then we can have a nice and scaled remake of that Sheckley classic. I only know that AI has powerful machine learning capabilities, constantly in progress. In our case, it would go offline, like in the story, and online, too.
Plus, I recall luddites who have been destroying new labor-saving machinery like power looms and stocking frames, in 1811-1816, at the dawn of Britain’s Industrial Revolution. In fact, they were right, from their point of view.
Sometimes I feel like many of us might be a modern version of luddites.
AI… The final invention?
Let’s just start at the top: Will AI kill us all? Not a chance in hell. Humans live under the sea, in caves, in orbit, on mountain tops, and on comfy sofas. I can’t imagine getting them all. Now that’s not to belittle the chance a big bunch of us could be killed but ending humanity doesn’t seem at all possible. Of course, that’s not good enough by half. Even a small percentage would equal hundreds of millions of people so taking chances is insane.
Enter, stage left, the people taking chances: Trump and the Tech Bros. I wonder where these guys minds are? Seriously, who do they think will be the first to be put on a pike if AI goes rogue? Pretend for a second: Facebook’s AI starts killing people, I assume Zuk would be more worried about his stock prices than about some enraged father of two who has sworn revenge for his babies. He’s right of course, a father is not coming for him… a half-million are.
Here’s what I honestly do believe will happen: AI will destroy the Internet. Completely. Let’s say we end up with two dozen AIs (targets) and let’s say there are one-million hackers & pirates out there. Only ONE, just one hacker, needs to succeed. Of course, they will sell their stolen software. Of course, several thousand will buy it. Within hours, every last phone, TV, printer, car, and computer, could be infected. Why, because these guys would hack everything by the millions… using AI!
Of course, everybody will unplug their house from the Internet. Like NOW! Think about it: a jet airliner cost like two hundred million, US. Will Delta pay $1,000,000 to get it back? Obviously. Vast sums of money will change hands that first week.
Millions of devices will become bricked over the coming days because the owners don’t know how to use crypto. We will go back to the 1980’s; a period just before the Internet. Of course, the Tech Bros will pretend they are victims. Correct that: The Tech Bros will be absolutely sure they are the biggest victims.
The fix will not be that easy. The Internet works on something called HTTP and I’ll not bore you to tears by explaining it. But I would think a new protocol/Internet would be designed and only computers with secure keys could get on. Apple, for example, uses a separate, secure chip that stores the owner’s fingerprint. Computer manufactures would also need secure facilities. Each computer would need to search it’s drive for AIs every time it’s signing on. This is likely to get cumbersome for anyone used to just surfing the web. If that doesn’t work, then no more general public internet. I think that’s the most likely outcome: small, private nets.
There would be some good things: No smart TVs, no connected cars, no phones tracking your location. If fact, did you know that your phone actually has a phone built right into it? I know! Crazy right?