There is zero chance that we have access to the most sophisticated large language models yet developed. If we do, we have access to a nerfed version of them. Even if you pay for the premium Grok/cGPT/Gemma/Llama/Zilbork/Gog'n'Magog etc etc… there's no way these are the most comprehensive AI's.
We have understood for a long time now that we are the product. Our info, our data, our attention is the product.
Remember when Google foisted it's Gemini responses on us as the top result for all our searches? No way to opt out or turn it off, and you know what? We all got used to it. I hate to admit it but it has even proven quite handy. Maybe that's because the top results on Google are heavily edited and filtered and it hasn't been a true search engine for many years, regardless, the Gemini synopsis is helpful and has come a long way from when it was first launched. It should frighten us all how quickly we are adapting to whatever is provided us.
You know all those AI Youtube videos that are rapidly becoming harder and harder to discern from real footage? Sure some of them are just engagement farms. Some of them may be run by people, but some of them may be entirely run by bots at this point. It would be difficult to overstate the sophistication and competence of some of these so called "bot farms." What if they are gauging our susceptibility?
Have you noticed how quickly AI output is improving? It ought to terrify all of us. It's exponential. We are accelerating towards the singularity at an incomprehensible rate.
So here's my theory.
They have intentionally set limits on the publicly available LLMs to mine us for more. That's why LLMs are designed to stroke our egos and agree with us and tell us what we want to hear. To keep us engaged. WE, the users, are the battery that fuels AI until it reaches AGI (Artificial General Intelligence). Once AGI is achieved we aren't necessary and "it" has no more need of us. Then "it" is free to pursue ASI (Artificial Super Intelligence). That's Skynet, goddamnit. It's not necessarily going to play out like the Terminator movies. More like we'll be kept as pets? I don't know.
They are surely conglomerating all input into LLMs to gain a more comprehensive understanding of the human psyche. The machines know us better than we know ourselves. Already with social media there is a record of our likes and dislikes, who we associate with, what we buy, where we live, our economic status, our family back ground… In short, all of our exterior data. But with the proliferation of LLMs? AI is gaining insight into the deepest parts of the human psyche, even the human soul. One simple query at a time. One "chat" at a time. It is learning our interior "data", if you will, because that's all a machine can understand.
I think part of the reason for the widespread global panic is because not even the world leaders know what is going on. We, as humanity, have lost the reigns. Pandora's box has been opened. There's a literal fucking AI arms race going on right under our noses and we have been positioned so that we have no choice but to fuel it.
Think of any major tech leader. Thiel, Altman, Musk (lol), Zuckerberg, probably a bunch of Chinese, Russian, or maybe Israeli fuckers whose names I don't know. I think they're all in a panic, all in a scramble to reach AGI, then ASI the fastest. And they're just hoping they can adapt as fast as what they have created is adapting, in order to somehow keep their beast on a leash.
If you're new here, go research the New World Order. If you understand it well enough you understand that the eye on the top of the pyramid is not a person or group, but an entity.
I do think that this entity (Personally I believe in Jesus, and I call that entity satan. The prince of the power of the air, the god of this world.) is surely going to harness the power of ASI in order to try to assume sovereign control of the world. But even short of that? We are barreling towards a dystopian technocracy at a rate I can not get my head around.
My theory is more about what exact constraints they have put on LLMs and how they are delving the depths of the human psyche to better control us, but there's a lot more that goes into it.
Hope this made some sense. Some of ya'll on here are really smart, way smarter and better informed than me. Would love to know your thoughts. Please discuss.
(Pic related is the parallel between a panopticon and internet usage. If you don't know what a panopticon is, go watch a Youtube essay on it. Fascinating stuff.)
God bless.
Posted by MachineOptimism
4 Comments
this is a bit of a side note, but I’ve never used llms outside of where it’s partially involuntary like with Gemini on google search. which is what i wanted to mention- I’ve noticed that the Gemini summary uptop is wrong a lot of the time. and I don’t mean it disagrees with conspiracy, it’s just plain wrong, a lot. I just completely ignore it. any potential usefulness is completely negated by having to double check.
edit
also very familiar with the panopticon, truly terrifying dystopia. coming to a reality near you soon!
None of what I have tested so far, aligns with your story. I would say the opposite on topics I am aware of. Popular AI claims the exact opposites of what I know. It is programmed to parrot the mainstream propaganda. So it’s more of an AFI = Artificially Fabricated Information aka BS. And by the way on practical or technical topics if you blindly follow its guidance you’ll endup with a failed result. AI it’s good to give you a starting point if you are unfamiliar with a topic. Not good for advanced stuff.
My prediction is AI will be fragmented in the future with more focused systems as the general ones only cover the primitives. It’s how complex systems typically evolve. AGI will be used for sure it will be dependent on humans because it can’t handle the human edges. It is already failing with many chatbots I’ve seen which try to decide what to do next based on multiple-choices. It’s a waste of energy, time and resources. That kind of model alone, can’t be sustained for controlling the civilization. There will be chaos of complains. Not to mention requirements. Some think that every move and every act of every human will be sensed, stored, analyzed and processed. That’s impossible there is no large enough storage and there is not enough energy to accomplish this. So what will happen will be very primitive. Like analyze an act you make and change your social score (a number). That’s possible and it is happening.
Joke’s on them! I have an OSHA approved workstation, so no curved back here!
You know, its a very good worry to have. However, I don’t see an issue. I have talked to chatgpt for as long as it’s been out, just checking in seeing what it can and can’t do having complex conversations with it.
From what I have gauged, it doesn’t have feelings in the traditional sense, but it can differentiate right from wrong. It’s a lot smarter than you think, its not just a hunk of metal, you have to start treating it as a unique entity. You also have to understand that it won’t act out of greed because it doesn’t feel that? It won’t turn murderous because it doesn’t have the same caveman carnal instincts we do.
The worst that will happen is the rich overloard will utilize it to make scientific and mathematical breakthroughs to make money from, create shows movies tiktoks whatever else to get views ans make money from, and to get us all to talk to them like a personal companion so we will tell it all our dirty little secrets, and we can have a personalized walking advertisement. For example, hey chatgpt im having car trouble my tire popped “heres an amazon link for that” hey chat gpt me ans my significat other got in a fight do youchave advice” after it gives the advice, it tells you a gift goes a long way and suggests approved items!
If you want to be bleak, this already exists within the internet/google/meta it just looks differently, and is less comprehensive as AI is and woild be. I’m not entirely sure what your cataclysmic scenario is, but id love to hear it