Knox Hill - The Internet (Initial Thoughts)

 I loved it.

Great lyrics (for what I can hear from them, since he's still not posting them with the song), great rapping, great video.

He has three outfits in the video:

• a joker? at least that's what the red coat and the diamond shaped makeup on his cheeks reminded me of
• Knox in his everyday clothes, in a tattoo studio
• a half Knox, half machine having a dialogue

He takes a photo with a man who's badly injured and unconscious and the woman who's giving the man CPR smiles for the camera as well. These situations do happen IRL. If there's a crime or an accident happening on the street, people don't call for help (911 or 112 or whatever the number is in your country) or try and help themselves - they take out their phones and start to film or livestream it. People also fake injuries for social media, or actually hurt animals to "rescue" them on camera... The world is crazy. If the internet didn't exist, I don't think this would happen, at least to this extent.

There's people arguing with each other, while Knox raps about "left and right" - the political sides and how the internet brings out the extremes. Most people have moderate leftist or right-wing views, and those people can usually have a somewhat reasonable and calm conversation with each other without arguing. But then there's the extremes, and those are what get other people talking. Extremists find each other, for one thing, and when they find validation for their extremist views, they feed off each other and their views grow more and more extreme and elaborate. At some point other people also notice the extremists and start talking to them (opposing them, trying to reason with them) and about them, with other people. "Look at this crazy person!" They create more buzz. And the internet algorithms LOVE buzz. They are designed to keep you in there as long as possible and by showing you content that people have opinions about, that you have an opinion about, they achieve that. So, they push that content to you. And that's why you only keep seeing the extreme views and not the moderate ones. The people on the moderate left see the extreme right and people on the moderate right see the extreme left. Or at least that's what it seems to me. I am a moderate leftist (I'd say about halfway between centrist and extreme left) and all I see are other moderate leftists, because I engage with their content and the internet algorithms like that, and those moderate leftists reacting to extreme right-wingers, who are sometimes reacting to the extreme leftists (so I see a reaction to a reaction).

In example the Oli London case. I only heard about them because extreme right wing anti-trans people used them as an argument against trans people, and then the leftists I follow reacted to those right wing people. In case you don't know, Oli London was a "trans" person, who claimed to be "trans-racial", being born Caucasian but identifying as Asian, because they were obsessed with a certain kpop star. They had plastic surgeries and changed their name to appear more Asian. Later they backtracked and identify as British again. There was also an aspect of them thinking they're transgender and then not being transgender anyway (backtracking on that as well). I can't remember which gender they identify as now (it was either a man or a woman) so I'm using they/them pronouns. At first the right wing used Oli as an example of "this is the future liberals want! if we accept transgender people, soon we'll have to accept trans-racial people and trans-species as well!" and when Oli backtracked, they said "see, trans is NOT a real thing!" and pushing anti-trans kids stuff (wrongly claiming that liberals want kids to eat hormones and have surgeries, etc). Sure, there is a teeny tiny percentage of people who think they're trans and then realize they're not, but it's a very small portion of people. And Oli London happens to be one of those people. And Oli has ever since been very anti-trans themselves, agreeing with the right wing and blaming the internet and the leftists for what they themselves said online and did to their own body. It was a whole big ass mess, I hope I explained it well enough.

Then the joker character is sitting on a sofa and is offered red and blue pills, reference to the Matrix. I think he took the red pills, although it's left a bit open as to which ones he chose. In the movie, the entire world we see around us is a computer simulation and we don't know we're living in it. The main character Neo learns the truth (and it's not pretty) and is offered a red and a blue pill - blue pill will allow him to forget what he's learned and return to the simulation, the red pill will allow him to see the world as it really is. He takes the red pill.

In the end we see the dialogue, but the camera is spinning around Knox and because of the lighting, sometimes it's hard to tell when we're looking at the machine and when we're looking at Knox. I mean, one side of the face is painted, so the painted side is the machine, right, but sometimes the camera happens to be directly in front of or directly behind Knox, so we don't know which side is talking at that moment. If they wanted to keep it clear, they're just film him directly from the front and he'd turn his head, or they'd film from both sides separately, altering them according to who's talking atm.

I'd like to see the lyrics, I can't catch most of the bars. From what I do catch, they say about the same things I was thinking while I was watching the video.

In the dialogue, Knox asked the machine something about how the AI (the machine) are created to resemble humans, and how does that work exactly, since humans are a) not always good b) flawed.

Does that mean machines are going to turn against people as well, because people have already turned against people multiple times in history, and if the AI is just copying us? Doesn't sound promising. Sounds scary.

How can a machine be flawed like a human? A machine is supposedly perfect, it doesn't make mistakes. Will AI eventually learn to mimic humane flaws as well? If we program them to copy our behavior and allow them to develop on their own, I think they might. They'll calculate the frequency an average human makes mistakes where it makes them seem endearing and relatable to other people, but that doesn't make them annoyed with the person, and then they start making minor mistakes and handling them as humans would like other people handling their problems - bringing them up on their own, apologizing and then fixing them. From there, I think the problem would be that the apology wouldn't seem genuine (because it's not, AI doesn't have feelings) and this would make the human not trust the AI.

AI are basically psychopaths. They think the same as people, but they don't have emotions. They can mimic emotions to seem more human(e), but they don't really feel anything. And that's why they're dangerous.

I also liked the bar "I am not morals but I am not wrong".

AI don't have feelings or morals. It can copy and adopt yours if either you or it wants to, but that's it. You can ask AI anything and it will not judge you, but it will learn from you - what YOUR morals are, what you consider good or bad, what you feel about certain things, what are your triggers... AI and algorithms have freaked me out a few times by how well they know me. Sometimes it feels like they can read my mind...

AI doesn't have a concept on what's right or wrong in the sense of morality. But it can easily give you correct answers to concrete questions. And even less concrete ones, by copying what people have said about the matter before. For example, if you ask AI the moral question "is theft wrong", it cannot say "yes, because..." or "no, because..." but it can pull up questionnaires conducted on people and what those results were, what academic papers written on this moral dilemma have said about it and what are some famous quotes on the matter, for example. Human generated information that it's analyzing mathematically (how many people said yes, how many said no, what the experts said on their papers, which quotes have most likes, etc) and then delivering you the results. And therefore, it is not wrong. It's repeating what most people consider the truth to be, whether we're talking about science or moral dilemmas. 

This also means, though, that the internet doesn't judge you. Certain words will trigger certain things, because humans programmed those things to happen (you google a store name --> you soon get ads from them, you google certain buzzwords --> the police takes notice of your actions, you loan certain books from the library --> the government will start tracking you), but the internet itself doesn't judge you. You can be 100% yourself with it. As long as you keep your browser history a secret from all humans and don't google those certain buzzwords, you're good. No one is judging you. And that's both a good and a bad thing. On one hand, it's liberating and inspirational to not be judged. But on the other hand, no is questioning your questionable morals if you don't talk to another human about them, and this might lead someone to think their abnormal and problematic thoughts are in fact normal and ok. But then again, the internet consists only of things other people have put there, or programmed to appear there. I think. I might be wrong though. If a human designs a game designing AI for example, are those games considered to be something the human made or that the AI made? Because the AI needs to have some data to be able to make the games, and the data comes from the human. But the exact combination the AI decides to use on that certain game kind of is something the AI made. Just like AI generated music. There's only a certain number of instruments and sounds those instruments can make, and if the AI is given all those sounds and some principles on what kind of sounds humans like hearing together, it can make music. Music that hasn't existed before. Even better than a human can, because if you also input the data of all the songs ever made, it can intentionally avoid using any parts of those songs in it's own songs. So I guess it is technically possible for a human to be interacting with only AI and not data made by other humans, especially if they themselves limit their own interaction to just AI. And this can escalate, as I theorized above.