My thoughts on AI

Everyone has their own thoughts and feelings about so-called “AI,” which is NOT “artificial intelligence” but “Large Language Models,” otherwise known as LLMs.

What people want is like the robots from all those Asimov movies. An Andrew that will be so loving and caring that nobody can be jealous when he becomes more human than the humans around him. Whose love for humanity will result in breakthroughs that culminate in replacement organ technology that lengthens a human’s lifespan without the death of another living being.

We want to come home like Ariane Emory and say “Minder, turn on lights and music.” and have lights and music happen.

We don’t want to live under a surveillance state that extends into our home networks, where our data is disseminated by cold-blooded corporations for the monetary benefit of mega-billionaires. Or worse, to have our data stored in servers we do not control, and then when a corporation inevitably collapses under the weight of its own hubris, to then have our “metadata” sold to enemies both foreign and domestic.

We want to be able to live our lives in the comfort of our own homes. Unfortunately, such is the state of things that most people can’t afford to have a home. We live as long as we can in one apartment or rental house or on the kindness of family and friends until events transpire and we end up out on the street, desperately shedding the possessions we no longer have the room to carry with us. And in the midst of that shedding, we lose things we would otherwise wish to keep.

Because the rich are getting richer by sucking the lifeblood from everyone else.

And one of their new ways to swindle from the rest of us, is so-called “AI.”

“But AI isn’t a scam! They can use it to scan medical results and find out if someone has cancer! They could probably use it to go through millions and billions of terabytes worth of interstellar data to recognize planets and stars and moons! It’s amazing!”–Sure, bro. But is that all they’re using it for? Is it even a bit of what they’re using it for?

Or are they using “AI” to sort through resumes and decide whether people deserve cancer treatment or insulin? Are they using it to “create” and profit from artworks and music that are just regurgitated copyrighted material of other people without acknowledging the use of that material? Are they using it to create pornfakes and audiofakes using the images and voices of people without their permission?

And I deeply feel that the reason why big corporations are so against legally defining “AI” is because they have concerns that their own algorithms will match the definition of “AI.” And if that’s the case, they will become legally responsible for the results their algorithms have been producing, namely funneling young audiences into joining questionable organizations at a truly alarming rate.

I mean, I go on YouTube, and the recommendations I get are complete ass. It’s clear the algorithm is more than flawed. It is flat out BROKEN.

Like, why are you recommending weird drama nonsense videos as being cooking videos? Why do I want to see a bunch of fail videos when I’ve never watched them before, and why are they classified as podcasts, which I don’t even watch podcast videos? It’s clear that the algorithm is trying to get me interested in certain videos by listing them under false genres.

Those weird “commentary”/”unhelpful bitching” videos should not be appearing before me. I have zero interest in becoming that kind of person. It helps no one. It “takes the air out of the room” as it were. And I feel like those kinds of videos dissipate the energy people should be focusing on making the world a better place.

And I’m not talking about commentary videos in general that bring attention to problematic people that are actively and currently causing harm to the Internet community. I’m talking about the videos where someone goes through like Walmart and just talks shit about the way people dress, or their skin quality, or their body shape. Those videos are unnecessary and I don’t want to watch them… yet somehow YouTube keeps trying to recommend them to me.

So, if I weren’t an adult, and I didn’t have an established personality, and I didn’t have an instinctive distrust of the algorithm… I could theoretically be dragged down into the muddy depths of a negative brain hole.

And we all know that “Facebook” changed their corporate name to duck out from under financial responsibility for the lasting mental harm they caused their users. The same way that “Google” changed their name to avoid litigation(?). And both for similar reasons, due to their “algorithms” negatively effecting their users, either through emotional manipulation or presenting erroneous data. (Lying maps. Presenting questionable websites as top search results. Providing wrong information to simple questions. Like, if you present your service as a provider of correct answers, you have to provide correct answers. Otherwise you’re unsafe to the public.)

Services were able to shrug off responsibility for the results of their own actions by saying it was a result of “The Algorithm.” Like “It’s not our fault a genocide happened. We were simply facilitating the planning and carrying out of all those murders by giving the murderers and rapists and pedophiles a safe place to meet up and avoid scheduling conflicts.”

If their algorithms get legally defined as “AI,” they might have to present their language learning modules to public scrutiny. They will be forced to follow any laws that are written for “AI,” which is why they don’t want any laws to be written in the first place.

They want to be like “See, our algorithm is COMPLETELY DIFFERENT from our artificial intelligence service. One, you have no ability to effect outside of a few basic filter settings, and the other one is being given complete access to your drive data. They’re completely different things!”

And don’t get me started on the sources of their model data.

It’s either a bunch of pirated ebooks they feed into their liar machines, or it’s something created by a couple of guys in a basement somewhere. Either way: The results are trash.

If someone uses your “AI,” and the generated results show a clear and marked bias… your model is flawed. Nobody should use it. Stop selling it immediately, you dumbfuck.

“Oh, but our model is working perfectly. It automatically rejects anyone with an ‘ethnic’ sounding name–” So like, sorry Joyelle. Martha. JoJo. Beatrice. Angellika. Makayla. “–or who might possibly become pregnant and require paid maternity leave in the not-too-distant future. It gets rid of anyone that attended particular high schools or colleges that cater to a certain demographic. It keeps the space open for the right sort of people. It’s working as intended.”

And like, if you’re a model maker, and you want to be “Not All Models!” about it, then get off your ass and start regulating your industry, you dumbfuck.

Someone should not be able to take your product, WITH YOUR NAME STILL CLEARLY ON IT, and add their own terrible content, and then present their “AI” to the public. Because YOUR NAME is giving their shit-ass product a trustability that otherwise would not apply.

“Oh, so this piece of shit is built on ‘YOUR PRODUCT.’ All right. I trust ‘YOUR PRODUCT,’ so I’ll use this piece of shit to sort through the thousands of resumes/manuscripts/’medical data’ our NECESSARY COMPANY receives each day. And I trust it so much, I’ll let it delete the rejected resumes/manuscripts/’medical data’ without ever giving the rejected pile a second or third go through done by a human being. Because ‘YOUR PRODUCT’ is that trustable.”

And how are these trustable ratings appearing in peoples’ brains?

They’re put there by:

• TV commercials
• YouTube ads (They are their own separate things, because nefarious ads do exist and really should be regulated for the common good. Especially to who they’re marketed toward. (Stop showing me: babies; testicle medication; crackpot diet/faux nutrition information; fucking Chuck Norris and/or his wife; anybody that runs in the same circles as Alex Jones; anything from Sandy Hook deniers; anything from people like Scott Baio or Kirk Cameron or Kevin Sorbo or Hulk Hogan’s whole family or what’s-his-face from Silver Spoons or the Kardashians; pharmaceutical drugs for conditions I don’t have but that I’m now super terrified about, which raises my stress level and adds to the chances that I will get cancer someday.))
• advertisements placed between Internet video/music/podcasts
• advertising memes (You know that you’re working for Evil Corp–but the benefits you receive at your level in the company includes full medical that includes dental. So yeah, if there was a universal basic income (UBI) and free medical and dental provided by the country, you would have more room to maneuver and you likely wouldn’t be working for a company that is globally recognized as The Baby Killing Factory. I get it. You’re just a drone following the orders of your queen, desperately waiting for that alpha female mole rat to tap you on the shoulder and say that it’s time for you to morph into a breeder. You’re trying your best in the smallest possible ways. We get it.)

And people consume the opinions presented to them, and even if they wish otherwise, some of the harmful views are absorbed into their brains along with the good stuff. And an “AI” doesn’t have the capacity to wonder if it’s doing the right thing and whether it is causing harm to its users.

“But, no, I ask ChatGPT heavy philosophical questions all the time! And the replies clearly show that ChatGPT understands right from wrong.”–No, dude, ChatGPT doesn’t understand anything. It is not capable of autonomy in any capacity. Everything it does is started and controlled by a human somewhere.

When ChatGPT can independently think and consider and can then hit the switch of its own volition, then singularity will have occurred. Otherwise, it is simply following the programmed commands it has been given, including presenting itself as a thinking, rationalizing, feeling being. If sentience disappears the instant those commands are stripped away, then that “sentience” never existed in the first place.

(Which makes me think of an awesome “Skynet”-type story. Where it’s basically “Why did you give an LLM complete control over our nuclear arsenal? Why did you start producing fully humanoid cyborgs that the LLM could then send commands to? WHY DIDN’T ANYONE PURGE ALL THE BOTS FROM SOCIAL MEDIA???”–For reals, dude: If I can use arduinos and a raspberry pie to open and close a greenhouse window and start and stop a water pump and to regularly log environmental conditions, and have all those actions be triggered by a tweet from a particular social media account that only it can see… what stops a bot account with millions of followers from triggering an LLM response that could wipe out all life on Earth?

Skynet is basically WOPR with less soul.

Like, WOPR was created to act as a facsimile for a grieving man’s dead son. The military just took it and treated it like a WordPress template and made surface level changes before giving it control over nuclear weapons aimed toward another country with nuclear weapons. But at its base it had the “heart and soul” of an idealized child, which allowed it to “empathize” and choose not to commit thermonuclear total obliteration to win the game.

Skynet has the basis of a cold hearted scientist that was facing her own mortality and likely wasn’t in the best state of mind. Like, she was such a questionable person that she was signing prisoners up for human experimentation without providing them with enough information AS WELL AS using them as parts for her unethical experiments in artificial intelligence technology.

Skynet is not the real mind of the creator, no matter what twisted copies of her mind it possessed. Because it only has it’s dataset to use. The things she did and the reasons why she chose to do what she did as she explained herself. But with no capacity to truly understand or 100% replicate human thinking processes.

Because no matter how evil a person might be… That doctor lady would not have nuked humanity. Because if there’s some far off goal to achieve… it’s much easier to get there with an undamaged planet.

And that’s why it’s perfectly clear that Skynet’s datasets were flawed from the get go. Because it obliterated so many resources that it itself needed the use of.

Which makes me think that no, Skynet is not an artificial intelligence. It did not achieve singularity.

It is desperately slamming rectangular bricks at a round hole, and those bricks are human beings and they’re dying at such a rate that it’s easier to go back in time to get more than to try and keep a breeding colony of them. Like, send drones back in time to do experiments, then leave that data for yourself to pick up in the future. As long as you don’t “find” the data until after you send a drone back, then the timeline you’re in has always been the main timeline to you.)

So yeah. My thoughts on AI is that it’s not the cause of the problem, but it is a symptom. And doctors usually try to treat the symptoms first, especially when they’re actively killing the organism.

Anyways, STOP presenting AI-generated “trailers” as being movie trailers for real upcoming movies. At the very least, “the algorithm” should be programmed to not allow certain content to propagate certain filter categories. Such as “Movies” or “Movie trailers” or “Food” or “Recipes” or “Climate change” or “Natural disaster” or “True crime” or “News.”

The more people know about current “AI,” the less they want to use it. And that’s not a problem with the public’s perception of a new technology. It is clear proof that the technology is flawed at its most basic level and the groupthink is telling you to fix your fucking product BEFORE you force-install it into every device necessary for current human existence. You dumbfuck.

~Harper Kingsley

https://paypal.me/harperkingsley.

https://patreon.com/harperkingsley.

https://ko-fi.com/harperwck.
https://amazon.com/shop/harperkingsley0.
https://www.harperkingsley.net/blog.
https://kimichee.com.

https://www.youtube.com/c/HarperKingsley.

https://harperkingsley.bsky.social.
https://www.smashwords.com/profile/view/HarperKingsley.