Will AI step outside the realm of science fiction entirely and begin to change existence as we know it?
Translated by Paola Pérez Castro and Ignacio Rauda Salazar of the original by Cassie Kozyrkov
Now that everyone is talking about ChatGPT, I’ve more often had to dodge some variation of the following question:
“Will AI step outside the realm of science fiction entirely and begin to change existence as we know it?”
The reason why I usually refuse to answer this question is that there may be some clever plan that will cause any intelligent discussion around the subject to come to nothing. And the reason for this is that the use of language can be a source of misunderstanding and there is no law against two people using the same word to explain different terms in different contexts, but let me answer the question for this one time.
There is no law that prohibits people from using the same word to explain different terms in different contexts. That’s what happens when people talk about AI these days.
Let’s look at some of the ways people use the term AI when talking about it. When we take a closer look at them, the answer to our trick question will come out on its own.
The AI that we find on the pages of science fiction novels is an invention! Most of the time, he plays the same role in the narrative as all the other suitably creepy, almost human-but-not-quite entities — demons, clones, demigods, aliens, golems, spirits, talking animals, animated puppets — that forces us to to confront what it means to be human. I hope no one is surprised that sentient, demi-human, antagonistic, evil version of AI doesn’t exist.
Sorry to disappoint you, but a HAL 9000 in assassin mode won’t be running a spaceship any time soon.
Science fiction is simply called “fiction” when it’s not running ahead of technology, so in an impressive display of just how tautological tautologies can be, let’s remove this version of the question: Are all sci-fi versions of AI in science fiction will totally come out of science fiction? No. Science fiction will continue to write about new tomorrows.
The sentient, antagonistic, evil, almost human version of AI does not exist.
I’m one of that special kind of curmudgeon who thinks it’s a crime against intelligence to talk about ”intelligence”, “sensitivity” y “awareness” on machines, as well as “the singularity”, without first clearly and unambiguously defining what these words mean to each person. It’s not nice to send yourself (and your audience) into a frenzy by saying a bunch of nonsense, but unfortunately that’s what the vast majority of discussions on these topics turn out to be.
As for creating artificial humans with our desires, feelings, and weaknesses…why would anyone want to do that?
As for making artificial humans with all our desires, feelings, and weaknesses…why would anyone want to do that? If we stop to think about it, it smacks much more of a plot twist in a novel than a serious goal. Regardless of whether you are very sentimental or heartless, if you wanted machines that could be more useful to you and/or humanity as a whole, it would be a better idea to build them without selfishness and other less pleasant aspects of human nature. While if you simply need more humans, buy me a beer so I can explain exactly how they’re made.
Definitely exist Great reasons to worry about the AI, but the it-reminds-me-of-the-uncanny-valley feel is just a distraction. Since a separate blog post would be needed to explain this, well… I’ve written a separate blog post for that here.
In technology, we use the term “AI” to refer to a specific way of converting data into computer code. When we see someone automate a task using patterns in data sources without directly looking for the answer, they are probably using machine learning to solve a problem. If it’s a particular type of machine learning algorithm, then it’s correct to call it AI, although good manners never stopped certain “AI” startups from jumping on this trend with their three IF statements up their sleeves. (You know who I mean.)
So,this version of AI already left the realm of science fiction? Absolutely, since it didn’t even go into it. Even before all the AI apps we use every day were released, the people who seriously worked on them were calling them AI, fully aware that they weren’t creating anything that would put them on good terms with sci-fi fans or neuroscience professors. It’s just a flashy name for a bit of math. Aren’t homonyms funny?
But it turns out that math is incredibly useful! They enable all sorts of cool apps and you interact with them every day. They have been “changing everything” for the past decade, from our smartphones and laptops to our cars and homes. This type of AI is quiet and unassuming, making our lives more comfortable without attracting attention.
Every time I see someone “fail” a Turing test, it tells me a lot more about the person than it does about the machine.
And that same math is useful for processing audiovisual data, so it’s a solid ingredient for all those magic tricks you’d need to fool the gullible into thinking you’ve built a self-aware self. But every time I see someone “fail” a Turing test, it tells me much more about the person than it does about the machine.
Finally, I would be remiss if I didn’t mention that there are a couple of additional homonyms nested within the one we looked at in this section: the AI that is more associated with academia and the AI that is more associated with industry. If you are curious about the difference between these two namesakes, you can refer to my explanation in “Why do companies fail doing Machine Learning?”
“Could AI break out of the realm of science fiction entirely and begin to change existence as we know it?” The answer is tautological either way. The answer will be a “no” simply by remembering any reasonable definition of science fiction, and it will also be a “no” because applied AI hasn’t had much to do with the science fiction version, at least for this millennium.
There is no more controversy here than two counterparts speaking to disparate purposes without defining their terms. So let me try to extract some controversial questions from this self-titled mix:
- Will the popularity of practical AI kill off science fiction AI? I could find an argument for the term becoming too mundane to be convincing… maybe one day an AI antagonist will be as hard to take seriously as an angry toaster. Will science fiction writers need to look for other entities to fill their uncanny valleys? I guess not, if creepy dolls are good enough for the fictional community, apparently anything goes, but what do you think?
- Did the creators of practical AI solutions do the right thing by letting their marketing teams capitalize on the sci-fi hype to secure funding and gain attention? How different would technological progress have been if we had stuck to terms like “layered weighted sums of nonlinear data transformations” instead of catchy phrases like “artificial intelligence” and “neural networks”?
- Should we just let nature take its course and let the public continue to confuse practical AI with sci-fi AI, or do we need to actively educate people about the differences between the two? If you are interested in my thoughts on why it is dangerous for society to combine the two, you can check out my blog post titled “Forget the Robots, Here’s How AI Will Get You.”
I’ll be very excited to hear your ideas as you digest these questions — join the discussion here.
Enjoy more data science translations in Spanish here.
If you had fun reading and are looking for an Applied AI course designed to be entertaining for beginners and experts alike, here’s one to keep you entertained: