A.I. Isn't Real
Clockwork Sex Machines
The Old Turks
In 1770 a man named Wolfgang von Kempelen built a robot.
What’s that you say!? A robot!? In 1770!? Impossible!
Not true.
Robots, previously known as “automatons” have a long history. Going back at least to the 1400s and probably a good deal earlier if the clues and indications in ancient Egyptian sources are to be believed. They were traditionally made of clockwork… like with gears and springs and wind-up cranks… and they were, admittedly, of little actual use. More novelties for rich people than anything. A guy named Vaucanson invented a robotic duck in 1764 for example. A machine that could bend its neck to drink water and eat kernels of corn. Earlier, in the 1500s, King Phillip II of Spain commissioned a robotic friar to bow its head and pray. In 1768, the Swiss inventor Jaquet-Droz created a robot which could pick up a quill, dip it in ink, and then write, by “hand”, a series of pre-programmed sentences.
Like I say, mostly these were novelties. Without batteries or electric power… wind-up toys can only get so far. That’s not to say that such robots never served any practical purpose though. On the contrary, some would pour your tea for example, or hold a knife and carve the ham at dinner, stuff like that. There are even more than a few accounts of wealthy noblewomen spending a pretty penny to create large, mechanically thrusting clockwork men for…
Well for…
You know.
Robots have been around for a while.
And we’ve always wanted them to do the same sorts of things.
Ever since slavery started to become taboo (and it was starting to become taboo as early as the late medieval period) we have dreamed of creating slaves that we won’t feel bad about having. Robo-slaves. Slaves without souls or feelings that we can use and abuse without feeling guilty about it. Slavery, after all, is the original labor-saving device, and still unbeaten. Washing machines and vacuum cleaners are nice and all, yes. But have you tried just hitting women with a stick until your house is tidy? It’s much more efficient. All the cons accepted… slavery gets shit done.
Enter robots.
So okay. Okay. Maybe hitting a flesh and blood woman with a stick is bad. Fine. But… what if we just made a woman? From scratch? A mechanical woman who existed only for the purpose of mindless labor?
That would be okay.
Right?
I mean, especially if she didn’t have nerve endings or emotions.
And we quickly discovered that we could really do that! We could really make such mechanical women!
You know…
Sorta.
Our greatest engineers, our geniuses, they really could make automatons to do things and have been able to, like I say, probably as far back as Egypt. A clockwork man to turn the page of your book for you maybe, or to put out the light of a candle. But all these wonderful creations aside, no matter how hard they tried it quickly became apparent that in order for these automatons to be actually useful outside of a few gimmicks… we were going to have to make them intelligent. Intelligent but also mindless and emotionless. You know, so we could beat them.
That was the dream.
The perfect slave.
Unfortunately, for hundreds of years that seemed impossible.
Until 1770.
In 1770, Wolfgang von Kempelen built a robot that could play chess.
That’s always been the standard hasn’t it? To the Western Analytical Mind chess, a game of near infinite possibilities and complex calculations, has long been the gold-standard for proof of intelligence. Regardless of how someone performs in other areas of life, regardless of if they are slovenly or poor or don’t know how to tie their own shoes… if they can play chess, we know they’re “smart.”
This isn’t without reason obviously. You do have to have a very high IQ to be a leading chess player. Bobby Fischer for example is said to have scored a 195 on his I.Q. test, far above the scores of your average theoretical physicists, who usually sit squarely in the 150s. Whatever it is that we’re measuring with IQ, it correlates highly with one’s ability to play chess. Thus, when Wolfgang unveiled his chess playing robot to the world… the world took notice.
Here it was.
The first proof of concept.
The first advance towards the goal of the perfect slave.
“The Turk” or “The Mechanical Turk” was a complex device that may or may not have been on wheels. It consisted of a large box which housed a myriad of whirling gears and pistons and a moving mannequin that existed from the waist up and was designed to look like an Arab. Why “Middle Eastern” was chosen for the automaton’s ethnicity I can’t say, but I suspect it had something to do with the allure of exotic mystery that all things “from the East” had at the time. The Turk mannequin wore a great big turban on its head and had a mustache and each of the Turk’s arms could be moved about in space by rods which connected at the wrist to the inner mechanical workings of the box. Atop the box, just in front of where the Turk sat, was a chess board, and visitors could pay a little money and sit down in front of the Turk and play a game, just as they would with any human. The Turk would even move his own pieces! Yes! If all this were not a marvel in and of itself The Turk was also good. Like, really good. In its 84-year career touring the world, it beat loads of people… including the likes of Napoleon and Benjamin Franklin. Of course, it wasn’t unbeatable mind you. It wasn’t like Deep Blue built by IBM or anything, which could beat all possible human challengers…
Still though.
It was mighty impressive.
It was, also, a giant fraud.
You see, The Turk, while legitimately impressive as a work of engineering (having all that clockwork to move the mannequin is no small feat)… was also fake. It was a trick. A deception. A scam. It was a way for a couple of clever fellows to go on the road and make money. Because inside that box… the one with all the whirling rods and gears, was a little compartment. A small room where a little person (perhaps one so small that they would today be considered as having dwarfism) could cram themselves in and control the motions of the Turk’s playing hand via a magnet under the board. Using a complicated series of mirrors too, this tiny operator could also see the board from more or less the Turk’s vantage point, the porthole for his mirror system opening just beneath the Turk’s turban. And all the gears and rods and belts that were in constant motion while the Turk played its games?
They didn’t do anything.
They were just for show.
And the crazy thing is that this story is all 100% true.
I mean, you know, you’d think given almost a century of touring someone would’ve caught on to the act. They didn’t though. It wasn’t until the creators of the scam retired and the machine was sold off to others that people got a good look at its inner workings. The fact that Wolfgang could never explain to anybody how his machine was doing all these complex calculations never mattered. People wanted it to be true badly enough that they didn’t question it too hard.
They wanted to believe.
That’s all it ever takes to pull off a scam you know.
That’s all it’s taking right now.
The New Turks
For decades the public has been inundated with the word “algorithm.” The Algorithm has, in fact, become a kind of pseudo-religion, with everyone trying to make a buck online attempting to figure out exactly which incantations will please these disembodied gods. “If I make the length of my video longer than 10 minutes will YouTube’s algorithm promote me?” “I wonder if this post will get flagged by Twitter’s algorithm?” “What special words or tags can we put on our website to get ranked higher on Google’s algorithm?”
Honestly, it’s all a bit grotesque.
Grotesque and made worse by the fact that I’m fairly certain most people, despite living their lives in service to these things, couldn’t even really tell you what an algorithm is. In practice “algorithm” means about as much to a man living in the 2020s as the word “magic” meant to a man living in Tudor England. If that’s you, let it be so no longer and allow me to enlighten you as to what an algorithm actually is.
You ready?
Okay.
It’s a flow chart.
An algorithm is a flow chart.
At base algorithms are simply a giant series of If-Then statements designed to arrive at specific conclusions based on their starting inputs. Yes, you can make these into incredibly complicated If-Then statements, sure.
But they’re still just If-Then statements.
If, for example, a video is uploaded to YouTube with the Title “Why We Should Kill Black People”… Then… block it.
Simple.
If the video title contains a topic that is currently popular and likely to get a lot of views… Then… promote it heavily to other users.
Again. Simple.
Now of course the actual decision tree matrix of YouTube’s algorithm is probably a lot more complicated than that and takes in many different factors and pieces of information to try and drive eyeballs and revenue as high as possible, but, at base, that’s all it is. It’s just a series of If-Then statements that are automatically followed.
…Sort of.
Companies like Amazon, Google, and Facebook have gone to great links to convince you that their systems are more or less autonomous. Yes, sure, human beings have to work on the back end to keep the servers up and do maintenance and so on, but, they’d have you believe, “The Algorithms” are running mostly by themselves. And in a way this is true. What they don’t tell you though, is that in almost all cases “The Algorithm” is partly a bunch of poor people in India.
Yes.
Okay, so let’s say someone posts a dick-pic on Facebook.
That’s not ideal. That’s not the kind of environment Facebook is seeking to foster. They would like you to not post those sorts of things publicly because it might offend people and make them stop using the app as frequently which would drive down their opportunities to serve ads. Fortunately, “The Algorithm” can detect dick-pics, and filter them out, blocking them from being seen by the user base at large.
Unfortunately, “The Algorithm” in this case is not entirely automated.
No.
In fact it usually involves middle-aged housewives.
You see, “A.I.” is not actually intelligent enough to recognize a penis. That said, in general penises have a certain basic shape and you can run analysis on an image to try and detect that basic shape. It’s not a perfect process though and the computer often gets it wrong, missing real penis photos and accidently flagging someone’s thumb or banana or bald head as one instead. To compensate for this what the computer actually does is flag pictures that might be inappropriate as “questionable” and then throw them into a giant pool of images for real flesh and blood human beings to comb through. This, in modern parlance, is called “crowd sourcing”, and when you hear a company use that word, that’s what they’re talking about. The questionable images are put in a big pile for people to fight over.
Yes. Fight over.
You see, each questionable image that a person verifies or rejects they get paid for, and competition to get as many images as you can is fierce. Mind you, the pay isn’t great, in some cases only a penny or two per image verified, but, if you live in a third world country and are willing to spend twelve hours a day at it, you can sift through enough questionable content to eke out a living. The process goes something like this:
The images are placed in a large pool and everyone on the crowd source server can see them, often only as small thumbnails. Users then click on the images they’d like to try and verify. If they click before someone somewhere else does, they’ll get about 5 seconds to look at the picture. The thumbnail will expand, and a giant penis (or not) will fill their screen and they will have to click the green button or the red button before the 5 second timer runs out to get paid. Users are occasionally reviewed for their performance. If they are deemed to be performing poorly and rejecting images that ought not be rejected or accepting images that ought not be accepted, they will get downgraded and given less preference in the race to get images out of the pool. There is no appeals process for any of this and these people do not receive health care or any benefits and are more or less an exploited underclass of old women in Bangladesh who exist so Westerners can have moderated social media content.
That’s “The Algorithm”.
Almost every major tech company does some version of this.
How do you think Google’s search results, which everyone is trying to game, are never dominated by terrorist organizations or porn or hate speech? How do you think Amazon knows what you mean when you type in “Cast Aluminum End Table”? How do you imagine Reddit knows what to filter when you say you don’t want to see any posts which “aren’t safe for work”?
Did you think computers were that smart?
Did you think A.I. was that good?
If you did it’s understandable. Silicon Valley has a great vested interest in making you think that with their computers they’re more or less all-knowing. They actively try to make the crowd-sourced components of their algorithms as opaque as possible to the end user so that everything looks sleek and cool and futuristic. The reality is though that behind every algorithm is a small army of people working for slave wages trying to verify that your amazon search results are relevant.
Amazon even calls its crowd sourcing software MTurk.
Guess what that’s short for?
Now I hear the rebuttals.
“But Yoshi! Yoshi!” I hear you say. “Okay, maybe those old tech companies like Google and Amazon don’t have real A.I., but… but… Stable Diffusion! ChatGPT!”
Same stuff.
More nuanced and cleverly put together but… same stuff.
The New New Turks That Definitely Require Regulation
Here’s how ChatGPT probably actually works.
Just a guess mind you. Nobody really knows how it works because just like old Wolfgang and his Turk ChatGPT’s creators don’t really let anybody look under the hood of their device. In fact they like to claim it’s all so smart and complicated that even they, the makers, don’t know how it functions… Sure. [sarcasm] Okay. I definitely believe that.
That said, knowing what I know about software, here’s my best guess as to how this miracle device that some people are convinced is bordering on being conscious actually works:
For years, probably five, seven, maybe even ten, what would become ChatGPT was a crowdsourced project where they asked a bunch of third world people to write questions and writing prompts for other third world people. “Anything you want” the execs told them. “Really. Just go nuts. Talk about anything.” And they did. The people on the other end of the line, the ones receiving the questions and prompts, were then asked to write intelligent responses. This went on for several years, until several hundred thousand or perhaps several million such prompt / response conversations were had, and each and every one of them was recorded and tagged and cataloged. Then an algorithm was designed to more or less “replay” these conversations to new users. There are only so many topics to talk about in the world and most people are highly unoriginal so the odds that once they put it online people would be asking it about things it didn’t have a recording for would be pretty small. Combine those recorded conversations with transcripts from YouTube videos and Podcasts and so on, and you’ve got yourself a pretty darn good mimic of human conversation.
But that’s all it is.
A mimic.
There’s a little human inside the box just like there always is. He’s just had his responses all pre-recorded this time.
Now of course that’s a simplistic breakdown. As I said, these algorithms, these If-Then Flowcharts, can get very complicated and it’s probably manipulating these conversations via some pre-defined metrics to create the illusion of originality but what I described above is probably the base of what’s happening under the hood.
Because computers don’t think.
They can’t think.
They aren't alive.
Human beings can think however, and we can devise very very clever machines to mimic the choices that we would make. Such things don’t even have to be digital, on a computer. You can make analog computers, mechanical ones, and I don’t think anyone has ever been fooled into thinking that those are alive or conscious somehow. It’s the trick of the digital format, where all the lever switching and moving parts are hidden “on the cloud” that allows for the illusion of consciousness. If you had the same machine but built out of wires and springs instead of electric circuits… I doubt so many would be fooled.
“They” want you to be fooled though.
“They” being the tech giants, Elon Musk, and in general all the creeps trying to enslave the world for Power. They want you to believe that they’ve created an artificial lifeform, one that’s a hundred times smarter than you or very soon to be, and which will take over the world if left unchecked and fire the nukes at us or turn our coffee machines into blenders of death. And because of that you see, you need to allow them to have unprecedented levels of power. You need to Vote, to encourage Congress, to do whatever it takes to give people like them special oversight powers to decide what new tech can come on the market. For Safety you understand. FOR SAFETY. For safety they should get to decide if any new competitors are allowed to compete, and to get to spy on your social media so you don’t feed the A.I. wrong and make it a Nazi, and to have total control over the whole internet lest their Frankensteins run wild!
“For your safety please give us power to police all the imaginary monsters we unleashed on you.”
I know Elon knows better.
I know he’s not a dummy. He doesn’t really believe A.I. can think or pose a threat to humanity.
He just wants power because he’s a megalomaniac who wants to put a chip in your brain.
That’s all.
Don’t be fooled my friends. You have the Divine spark. Not the machines. You’re a living thing. They’re just wires. Fancier but no more alive than a clockwork duck that could swallow kernels of corn.
Amor Vincit Omnia.
I keep saying this at work! (I write for a living, so all my colleagues are extremely flustered by our impending defenestration by some bot called Claude.)
I recognise that I'm more Thomistically inclined than most, but you don't have to believe that the intellect is a faculty of the soul – or even to believe you have a soul – to be able to see that calculation is not intelligence. It is very rapid calculation dressed up as human words because that's how you make rapid calculation look really cool and get investors hot under the collar.
'Whoa, the Youtube algo knows me better than I know myself!!' is the same as 'Whoa, that lady in the tent on the beach JUST KNEW I've had my heart broken and would like to earn more money', except the lady on the beach is a lot more impressive because she's actually cold-reading you and doesn't have access to your search history.
Excellent reminder and so true. The disturbing thing is that we humans have been set up for this by modernity working hard to turn our brains off as much as possible. People’s jobs have been dumbed down and replaced by systems and protocols which are analogues of those flow charts. People are fed social media to become addicted to mindless escape.
Humanity has been marginalized and reduced while the machine and the algorithm have been elevated to minor gods. Ultimately the plan of the elites is for they and our tools to become our gods, or at least the puppet that appears to rule as the wizard used in Oz, to be worshiped and feared.
Modernity is seeking the annihilation of humanity as a goal and as an emergent consequence. It is virtually impossible to separate the two. But the essential nature of humanity is not so easily extinguished and it will not be defeated. Bravo for this reminder.