Artificial Intelligence, Computer Learning and Adaptive Programming, NOT Synonyms!
The term Artificial Intelligence has become a sort of talismanic chant to ward off the evil spirits of technophobia. Soon, or even now, we can all stop worrying about the next world war, because the robots are taking over and mankind is doomed. Artificial Intelligence®™© (always with the capitals, amen) will do everything for us, apparently, all you need is the right App. This nonsense has progressed to the point where people try to solve every “challenge” with an App, and then they get despondent when they cannot find an App to get a stain out.
Technology is now so wonderful, we don’t need jobs anymore, or salaries, or bonds, no, we can all live in momma’s basement, from where we will develop all these Apps that we will sell each other and we all become Tech Millionaires. The minds that came up with that theory, actually got paid for it. This proves that, even though we as yet don’t have artificial intelligence, we still have fake intellectualism. The problem is, these fake intellectuals are the ones tasked with building that artificial intelligence. The kind of intellect that spells things with capitals. To make sure we know how important their work is, I guess.
Now, here’s the thing about AI: It does not exist. Internet platforms are vast collections of computers and file servers, thousands and thousands of them per square whatsisnames, with more computing power than most governments. You can ask Google anything, and get an answer, isn’t that clever? No it is not. The Yanks like to tell the story of the idiot misfit that saves the day because he remembers the score for every baseball league game ever played. The movie usually ends with a little moral on how overrated an education is, as long as you have a good heart, and love sports. Intelligence is for those clever buggers in white coats, the ones making the science that is warming our planet. We should stop them. Teach them some ball sense…
Remembering numbers from a list is just remembering, exactly what Google does; it remembers where it saw the words you are typing in last, and poops out a web address, a couple of million different ones, usually. There is no intelligence, only programs containing instructions telling computers which data to assemble and present. It will answer you, all right, by dredging up the answers provided by other people, with no care for the correctness thereof, and no way to check. It is just not clever enough to make impartial decisions on truth, it can only present all the truths it has been supplied with, conflicting ‘facts’ and all.
Facebook, Twitter, Yahoo and every other social platform is presented as an artificial, sorry, Artificial Intelligence, and we are told of the filtering, and the scoring, and the automatic banning, all overseen by their AI. In reality, their AI is an office building full of otherwise unemployable meatbags trying to look at everything and deciding what is acceptable or not. We are often told how stressful this job is, what with company guidelines being non-existent or vague, and oh my, all that radicalising content out there, block, block, shadow-ban, suspend, lock account and why don’t you smear your excrement all over your workstation before you commit suicide in the parking lot? No, really, that is apparently life at the office for these “stressed-out moderators”.
If you want to see the intelligence behind those anonymous arbiters of your moral standards, go find a video of Facebook staff when they heard Hillary Clinton lost to Trump. Disregarding your own political views, just look at their reactions, their emotional outbursts, the virtual writs-slashing, and tell me that thing has the right to suppress your research, because it “might influence the susceptible”. Videos of little girls getting gang-raped is okay, a dude praising Trump is fascist. This they call Artificial Intelligence.
Come to think of it, the capitals in A I might also be simple Orwellian doublespeak. In the world of warfare and genocide, the term ‘Intelligence’ is often spelled with a capital, and it is used as a synonym for ‘information’. Spies gather Intelligence, radios broadcast Intelligence to other units in code, the Intelligence community gather Intelligence for use by those with intelligence. Now add the word ‘artificial’ to ‘intelligence’, spell it with Capitals, and what you get? That’s right, ‘Artificial’ Intelligence, Fake Information. Please note, and never forget, Google, Facebook, Twitter and every other little atom of the Internet, actually belongs to the American Army. They invented it, they built it, they grew it, and then they gave us a bit of it to play around on, to spew disrespect and blasphemy against those who are taking down names for kicking ass later. There is no privacy on the Internet, and very little truth, just a lot of Artificial Intelligence.
Making fun of words still does not address the actual issue of A I. What is A I supposed to be? Those with distorted views on humanity are on their collective knees, praying for the day computers outpace human thought. There are those who say it already has, computers can think so much faster than humans already. No, they do not, they merely look up the data you request, provided that data has been provided in the first place. The very first calculator can add and multiply a lot faster than any human, yet no calculator can be accused of being intelligent, it just looks up pre-programmed answers, the average calculator does not actually compute much, they work off huge lists of answers hard-wired in, called look-up tables. Like that Ohio kid with learning disabilities but access to reams of sports statistics. Whereas the kid might have to rack his memory to find that odd score, computers have their data all neatly arranged and categorised. Computers do not have to remember, only dish up others’ memories.
To decide whether A I is real, one first has to define ‘intelligence’. The usual requirement of true intelligence is not the repeating of information, or even the processing thereof, but the creation of new information. People can be programmed, just like computers (refer again the kiddies weeping for Hillary) but here is the difference: an intelligent being can exceed the limitations of its programming. An intelligence can think of new things, without referencing the old, and still retain the old faithfully. When a computer program starts acting outside the design limits, it is usually because of some error, causing the original programmed behaviour to be compromised in some way, abandoned even. Broken. A computer can draw very pretty pictures, but a computer cannot create art.
To set a million monkeys upon a million typewriters has always been an joking example of futility in uninspired art, yet the computer crowd does exactly that: they have set millions of processors to task drawing a billion ‘pictures’, then they show us the three best-looking examples of their attempt, and scream “Look, a computer made art”. Picasso out if focus through a wet window, boo-hah. I never liked Picasso, but he did feel called upon to quantify art, so I guess he wins that one. It still is not art, though… at least Picasso could focus his eyes. Mostly, computer art is merely pseudo-random arrangements of tones, colours or shapes, chosen from pre-programmed examples. Or you can tell your computer to search for all other data sets to see what is most popular, skim the most recurring themes, stitch those together, say your computer did it all by itself, and call this artificial intelligence, or Machine Learning.
Machine learning is how our tin brains are supposed to find enough facts from which they can synthesize a useful core of data, and apply that to their task. There are many ways to do this, the most popular form you know, is probably the Google ranking thing, where the computer knows which is the most likely answer you expect, because that was the one the majority looked at. For Hillary, they actually told their A I to hide bad things, repeat all the good things, and make sure she’s at the top of every page. They also do this for shampoo companies, dog food and medical advice about poisonous vaccines, which suddenly don’t exist no more, at least not before page 97 or so.
The machine can only learn what you tell it to learn. It is a useful and powerful tool, learning from old data, but popularity is not truth, especially on subjects most people tend to misunderstand. Machine learning is actually just machines gathering data they were told they like, possibly by other data they collected. In the end, the action that machine takes upon that data, was programmed in at the beginning, possibly with a wide variety of choices, but always limited by the intellect that programmed it at first.
Another form of machine learning is Heuristics. That’s the one where you remember what happened yesterday, see what happens today, and tomorrow you know exactly where and when to pull the trigger. In humans, this is seen as a serious defect, because this sort of thinking leads to Hubris, the bit where you pull the trigger at the exact moment required, even if there is a kid walking past the target.
Another trap with heuristics, is the forming of bias, where you give certain things more importance than others, purely by own experience. Every time you see something that conforms to your beliefs, it confirms your bias, your preferred explanation, Every time you see a Myscopcher hitting a kid, it confirms your opinion of Myscopchers being child molesters. Racism is one form of heuristic learning gone wrong for humans.
Or how about the A I autopilot that kept crashing the plane simulator straight into the ground, over and over, because when you overload all the stress sensors at once, the computer does not register your mistake, and the machine was programmed to try all the ways possible until it finds a way that records no mistakes. The program was biased towards the wrong goal, and every crash confirmed to the machine its own success at finding a solution to the stated problem.
There is also something called Adaptive Programming. The idea is that the computer will find new data, then change its own programming to process this new data. Once again, that program will essentially come from the original programmer/s, with all their biases, mistakes and hang-ups included. The Facebook computer hates political radicalism, and it was programmed to suppress all bad political commentary. “Bad” in this case, was defined by a youngster still so pumped up with Liberalist vigour, it actually hates Conservatives. People have been scrubbed out for saying something patriotic, because Liberals feel patriotism is divisive. Adaptive programming, therefor, is not intelligence, it is merely the collection of new subroutines as part of the data it needs to process. Those new subroutines suffer all the shortcomings and benefits of the programmer’s intelligence, no more, no less.
So why the harping on Artificial Intelligence? If it does not exist, why is everyone talking about it? Firstly, it is a great marketing tag. Just like you can make them eat worm poop by calling it Superfood, or sell them the latest environmental disaster by telling them your plan is Green, now you can sell any technology by pretending it is intelligent. Most people have no idea, and even less interest in the science or even language of things like robotics, so whatever the news says, is what we know. The question should therefor be ; why is the news telling us that A I is News? What are they trying to accustom us to? Why are they lying so hard about the state of technology, hiding so many things, while exhibiting things not possible yet? Why must we believe in A I?
Many tasks are being done by machines, allowing humans the free time to apply for unemployment benefits and visiting charity food collectives. In the olden days, men and children were sent out to work while the womenfolk puttered around the house trying to get everyone fed end cleaned up. As machinery improved, it became necessary to reprogram the people for living with their new neighbours. Men fed and cleaned the machines, while the kids went to school, to learn the things needed to properly care for the machines. Once the newly-educated youth started building more and better machines, the men were sent off to eradicate each other in war, the women were set to tending the machines, and the children were sent off to school to keep them occupied with the learning of anything except what is needed to remain masters of their environment. Education turned into indoctrination, kiddies had to learn where they fit into the social structure, to “prepare them for modern life” where nobody is supposed to do anything outside the Program.
So that is where we are now: The womenfolk are working, the men have a choice between crime and financial scamming, and the children are attending classes in accepting this setup as normal. In the meantime, our exalted leadership discuss the legalities around robotics and Artificial Intelligence. If the men were earning a living decent enough to look after a family, and the mothers had the time to see what their children are being put up to, and the children were not busy using their expensive techno-toys to replace human contact, we might have wondered what that was about. But we did not, even this writer can find nary a trace of those high-level meetings and what was decided. Half of Europe’s bestest and brighterest came together, discussed the legal standing of artificial intelligence, and walked away with a decision they are not sharing with us.
To see what the legal status of a robot is, we can look at corporate law: corporations have been given legal personae, you can sue and get sued by a corporation for as little as a slight insult. As we all know, the legal weight of your case before the court is determined by the social weight of you BAR representative, and the funds you make available to said lawyer type. The individual’s chances of winning a court case against a mega corporation is almost nil, and the chances of a corporation actually making good on the verdict, less than zero. The best one can hope for, is that the corporation will expel one of its junior administrators as a show of legal compliance, but the corporation itself, its shareholders and subsidiaries shall carry on as per usual. The corporation itself, as a supplier of “much-needed employment” and producer of goods with “strategic national importance” is above you and your petty squabbles.
Now transfer this mentality to the robots: The machine was built with Artificial Intelligence, therefor the decisions it takes are informed by current situational wareness as gathered by the machine’s sensors. Any harm that comes to you or yours, is thus the fault of the robot, not the owner, not the manufacturer or programmer, it was the robot that gone kill your child, let’s wipe its programming, that will learn it! Unless, of course, we can prove say that you provoked the robot, possibly by acting unpredictably, doing something near that robot, causing a failure of execution, you were the one acting outside the confines of the machine’s programming! Are you insured against overstepping the bounds of a computer's programming?
The most obvious danger in all this fake cleverness nonsense is, of course, the excuses it allows the executives; If the machine acted upon its own intelligence, then any mistakes by the machine, is not the fault of the owner. When a robot beheads your child, your case will be with the robot, and good luck suing a robot! Even if you win, what revenge will you visit upon a machine? You certainly have no claim on the corporation that built, programmed, owne, hired, rented, deployed, tasked, maintained or otherwise caused the robot to be where it was when it was cutting the head off a child, drowning fish in oil, burning butterflies with radiation...