Artificial Intelligence is just an “intelligence simulator”
ChatGPT from Microsoft and now a similar chat bot from Google are all over the news. We also saw recent hype about “AI Art” and “AI writing”. Some people worry that soon robots and computers will take over most jobs, even creative ones. They panic that machines will one day reach “singularity”, rebel against their creators and completely dominate society.
I don’t. In fact, I don’t really believe “artificial intelligence” exists or can exist, as real intelligence is derived from consciousness, which cannot be created. With consciousness comes conscience, which allows us to distinguish right from wrong. This process cannot be generated artificially, except as a crude simulacrum — and indeed, we have now bots programmed with all kinds of politically correct filters which are, I suppose, the morality of the age. But they don’t really “think” for themselves.
All these “AI bots” can do therefore is to provide a clever simulation of intelligence, creating an appearance of something that is not really there.
That’s not to say that AI doesn’t have some uses. Unfortunately, most of such uses are negative — mas surveillance, for instance. Facial recognition, movement detection and other forms of AI based on pattern recognition can — and will — be used for pretty nefarious ends, in order to create a totalitarian state — we already felt its first birth pangs during Covid.
Then you have the planned merging of people with machines, sci-fi things like artificial organs or chips in the brain. Ted Kaczynski, aka the Unabomber, warned about this in his manifesto. They sell this as turning people into super-human cyborgs, but the real purpose is to turn people into slaves, with no control over their own body or mind.
Other uses of AI are more benign. Machine-translation, for instance, has greatly improved in the last years, providing very readable texts. However, for good, literary-quality translation, you still need a human brain.
AI automation, as in self-driving cars, planes or trains, can be useful but has some drawbacks, such as hacking, maintenance costs, and the risk of serious hardware or software malfunction at critical points. I am not sure totally autonomous self-driving cars are a good idea.
Chatbots as customer service are awful. They are really no better than pre-recorded messages on a phone. Sometimes, you really need to talk to a human, and preferably not someone from Pakistan with an incomprehensible accent.
I am not impressed by AI poems or AI art. Almost everything I saw is pretty bad or bland, with no there there. Even if AI manages one day to create something of aesthetic value, it will be in the same manner of the proverbial hundreds of monkeys typing a Shakespeare play.
That said, I am sure we will soon see humanoid robots (androids) walking around and being used as servants, companions or even sexual partners. Why? Because even if machines are incapable of real intelligence, conscience and feelings, a simulacrum is enough for most of us. Didn’t people in the 1980s get attached to “tamagotchis”, which were basically very primitive digital pets? Don’t children get attached to dolls that play just one or two pre-recorded messages? A robot or a chat bot that simulates feelings and intelligence is almost as good as the real thing. We don’t mind being fooled. Aren’t most human relationships also based on deception or self-deception anyway? We pretend all the time to be what we are not. Machines can do the same, only better.
There are many films and novels about robots becoming sentient. The last one I read was “Machines like me”, by Ian McEwan. It was pretty lame. McEwan has written some good novels over the years — unfortunately, this was not one of them. I think the problem is the chosen subject. AI is not real, and robots cannot develop self-consciousness or real intelligence. Still, in a fictional context, it could work. But McEwan is not really a science fiction writer. I think the problem is that he tries too hard to be serious about a fundamentally unserious problem, and ends up creating a novel where all characters are bad caricatures that feel very artificial. And I don’t mean just the androids.
Even Asimov cheap robot novels felt better and more realistic. At least, in my teen years, I enjoyed reading “The Caves of Steel”, “The Naked Sun” and “The Robots of Dawn”, the trilogy about a detective and a robot in a weird planet where people feel afraid of diseases and all human contact and live isolated from each other (sort of like the Earth during Covid).
AI will certainly play a pivotal role in the near future and, sure enough, it will take over many jobs. Lots of people will become unemployed, or will receive a UBI or some other type of welfare in exchange of getting a chip in the brain or a genetically-modifying injection, or both. Our society will become increasingly automated, and mass surveillance will be a constant feature.
But sentient robots will not rebel against humans and become our new overlords.
4 comments on “Why I'm not worried about AI”
So much we agree about AI. Its function as a translator or transcriber on YT videos is what I value most. It augments my understanding of what’s being said. My hearing not being what it was, it’s appreciated and I notice its lack on other platforms that can’t afford it. Same thing in movies and tv w/ or w/o subtitles.
My only disagreement in your assessment is about conscience being the foundation of intelligence. When you first wrote it, I thought you meant consciousness. But then you repeated it; so apparently it’s your philosophical stance. I think consciousness or awareness precedes conscience. I also believe conscience is more learned than innate and would point to the epidemic of virtue signalling among liberals as an evident outgrowth of a false, conditioned conscience.
The impact of AI on aesthetics in problematic. When we spot it, we feel it and deplore it. But I had cases recently where I didn’t spot it at first. When it was pointed out I felt tricked and reacted according to my conscience leading to moral actions on my part: warning others and supporting the source of my enlightenment. Conscience is important, but reading reality comes first. I think.
I actually meant both. I updated the text to make it more clear. English is not my native language and sometimes I make mistakes… Consciousness and conscience are actually the same word in Spanish (conciencia), ergo the confusion.
Regarding AI in Art, isn’t it in the end a result of the programmer and the algorithms and the database put into it? So it is a human process in some way, even if the result is somewhat random because the choice is not direct. I’ve seen the AI Rembrandt, and I’m not impressed. Technically well done, I suppose, but it’s not a Rembrandt and it’s not even art, as I understand it.
All art used to be pure: words on a blank page, paint on a blank canvas, form on a formless block of material, even images on an untreated piece of film, as long as the renderings were not mediated by computerized or digital processes. But now most art involves technology as does life in general. Your point that a human’s intention is behind even completely AI generated art is true and that’s what the one photographer said in his defense when I confronted him w/ his deception – an image of a beautiful woman posted on a page devoted to pre-Raphaelite art.
What you were saying about self deception as being intrinsically human… so worth pondering. We like it when we do it to ourselves and don’t consider it delusion. But we hate it when it’s done to us.
I guess transparency is important. Knowing the process involved in creating the images would help eliminate accusations of duplicity, but it would also probably destroy the power of the illusion the artist is attempting to create.
One of the things that intrigue me about AI (and I forgot to add this on the text), is those stupid “CONFIRM YOU’RE NOT A ROBOTt” messages that now and then appear when you try to access some webpage. And then, to prove that “you’re not a robot”, you’re supposed to do a stupid, very robotic captcha test or quiz. Are bots incapable of identifying, say, cars or planes or mountains, or of typing letters and numbers? It seems very silly to me, and it shows that AI is not as advanced as they hype it, if bots can’t even pass those stupid tests, and if webpages cannot in a more automatic and less intrusive manner determine that “you’re not a robot”.
Comments are closed.