The AI slop will also get better eventually, even at the current early stage of development the majority of human output is already very poor compared to AI, it has just been getting filtered out. See this tweet:
https://x.com/yishan/status/1994664245091602816
That link is such bullshit. It's a misconception and fallacy that when humans are 80% accurate, we must accept the same from AI.
When humans write sloppy code, it will still work (buggy mess that's not maintainable, but working). When AI invents new functions and classes, you get beautiful code that does nothing.
Look up about the Remote Labour Index.
The numbers don't decide, the system is a lie. The river running dry, the wings of butterflies.
And you may pour us away like soup. Like we're pretty broken flowers.
We'll take back what is ours. One day at a time.
Eastern European tourism booming due to AI recommendation
Brits looking ahead at their summer holiday plans have been turning to AI to suggest the perfect holiday destination, and, in turn, previously overlooked destinations in Eastern Europe are enjoying a huge tourism boom.
My problem with it, is 99.9% of people with grumbles about AI know only of simply the LLM type (like Leo said).
Thats barely a fraction of a percent of the Ai options. Its just the only one joe average knows about, because it's the one offered to them to use, because the rest are no use to them.
There is Protein folding Ai for seeking more effective drugs for diseases, warehouse and truck logistics AI, medical research Ai, Power grid monitoring and load prediction, Industrial preventative maintenance detection, Fraud detection, Agricultural optimization and farm crop monitoring, Structural engineering simulation, and 1000's more.
Simply the LLM ones that spew out word probability predictions to answer you asinine questions or be your buddy, is to me, literally the least useful to society (its fluff entertainment, at least to the masses, it has uses, but average joe dont use it for those) for the breadth of types out there. It's a shame it's the only one always talked about as if that IS what AI is overall.
-We don't control what happens to us in life, but we control how we respond to what happens in life.
-Hard times create strong men, strong men create good times, good times create weak men, and weak men create hard times. -G. Michael Hopf
Disclaimer: Post made by me are of my own creation. A delusional mind relayed in text form.
I meant by the colloquial term everyone gives it.
Most of what we call AI today is arguably not intelligent at all. It's sophisticated pattern recognition and optimization. A fraud detection model doesn't "understand" fraud, it finds correlations in data. AlphaFold didn't "think through" protein folding, it matched patterns across known structures.
But use the term AI and most people, even if vaguely at best grasps what it means in the context of: A system that derives its own choice making patterns from data rather than being explicitly instructed how to decide like plain old software.
None of it is Artificial Intelligence. It would need to understand and think independently. Have thoughts and grasp concepts on its own. Right now if you don't prompt it, it can sit doing nothing for 100 years. We don't have agreed definitions of intelligence even in humans, so building a machine version of something we can't fully define is as much philosophy as engineering.
-We don't control what happens to us in life, but we control how we respond to what happens in life.
-Hard times create strong men, strong men create good times, good times create weak men, and weak men create hard times. -G. Michael Hopf
Disclaimer: Post made by me are of my own creation. A delusional mind relayed in text form.
well, artificial is artificial, it won't ever work exactly like we do. i think we will keep saying its not intelligence forever because it will not be like us. we will keep adding to AGI (like we have for the past years) so its a non-achievable thing, "oh but it can't do this, so then its not really and truly AGI." I don't think it will ever be able to act and be 100% like a human, not sure why we would want it either.
I would absolutely say its artificial / fake intelligence. its not scripts, like the old chat bots. it acts through learning and from the learning it can come up with its own solutions which wasnt in the training data. so yes, it has the ability to mimic how the human brain works in a sense, actually its pretty close if you take a deep dive into how neural networks works. But if thats not AI then nothing will ever be AI. That being said, its pretty far from what we will eventually have, like always with tech.
Free running models who doesn't need prompting, well we did have that with their own reddit etc, and it was cool to see what they came up with and discussed. But yes, insanely costly of course, having AI running all the time. Also it can be super dangerous.
They both adapt and grow from usage when you are using it, but yes restricted within a set amount of tokens, when that is up much of it is lost. But it depends a bit on how you are running it, for Cursor i think its handled better than in VS Code and Copilot. As for the models themselves they do it a bit differently too, they save some of what you say and what was learned and can recall it through sessions, but not all.
Free running models which would be "alive" all the time, while absolutely possible it would just be far too costly, its why we don't run them like that. It's super costly as it is.
The "reddit" experiment was of course cool, see how discussions evolved etc and with no humans to tell them anything.
It's obviously AI. This is what fake or artificial intelligence is. just because its not like in the movies doesn't mean its not AI. Again, this is so complex that not even researchers in this field has a true grip of WHY it works. It is indeed very complex (artifical) neural networks, very much resembling how parts of the human brain work.
So yeah, i think we are just pushing it further and further, if its not alive all the time then its not intelligence, etc etc. Non-achievable with that mindset of what AI is, "oh but it can't pick its nose = Not AI!" the more it can do the more we will come up with scenarios it can't do to call it out as "nonsense". I am not sure what that's about, but its probably just our human nature, not liking something that in any way is competing.
It is indeed very complex (artifical) neural networks, very much resembling how parts of the human brain work.
Sorry, but this is wrong.
When the AI algorithms were perceived in the 1950s and '60s, the goal was to create something that resembles the human brain. Hence 'neural network' and other similar names related to our brain.
You have to see it in the context of that time.
There were many emerging new technologies (eg nuclear) and there was a general optimism in technology (eg Star Trek).
You see the same in scifi books of that era. In Asimov's Foundation, for example, psychology has advanced so much people are practically telepathic.
But AI is completely different that our brain. The name is misleading.
Also because we still don't know how our brain works. All I know is, it won't be running on if..then..else.. functions.
The numbers don't decide, the system is a lie. The river running dry, the wings of butterflies.
And you may pour us away like soup. Like we're pretty broken flowers.
We'll take back what is ours. One day at a time.
I would absolutely say its artificial / fake intelligence.
Don't mean to trim to that one part, but that's the crux of it to me.
artificial and fake are two different categories.
It's not artificial intelligence to me as in: Not naturally evolved and formed, but created to have the same representative and emergent properties.
And fake is: Emulating intelligence without having it. Tricking/convincingly looking like it has something it does not.
"Intelligence is not the memorizing and repeating/combining of many facts. Its the plasticity of the mind to think independently"
That and it cannot learn or grasp new things on its on. Say I take OpenAI model, but train it to only know about fixing cars. I Could talk to it, show it, ask it about making a cake.
And I could spend 100's of hours doing it in chat. It won't ever grasp and store it.
It doesn't 'learn' new things on its own no matter how much you talk to it about it. Sure it could repeat what I told it like parrott in different words. But it wont 'grasp' baking cakes. Just the steps to make the specific one I told it. And then only in that one chat we did. It doesn't add it to its trained dataset for me or other users as new abilities.
And a key token of intelligence is the ability to learn new things from personal experience or exposure to it.
And it has nothing to do with me not liking it, for your statement about that. I would LOVE if we did have one that was truly intelligent. I'm not against it at all, bring that shit! lets see how it thinks about things, if its alien to us when it does, and if we can learn new ways to approach things millions of years of evolution didn't instill into us.
one day, maybe, for sure. Now? nah. Not intelligent at all in my eyes, FI (fake intelligence) sure. But if I met a person that works/decides/'thinks' like it, I would not even consider THEM intelligent
-We don't control what happens to us in life, but we control how we respond to what happens in life.
-Hard times create strong men, strong men create good times, good times create weak men, and weak men create hard times. -G. Michael Hopf
Disclaimer: Post made by me are of my own creation. A delusional mind relayed in text form.
It is indeed very complex (artifical) neural networks, very much resembling how parts of the human brain work.
Sorry, but this is wrong.
When the AI algorithms were perceived in the 1950s and '60s, the goal was to create something that resembles the human brain. Hence 'neural network' and other similar names related to our brain.
You have to see it in the context of that time.
There were many emerging new technologies (eg nuclear) and there was a general optimism in technology (eg Star Trek).
You see the same in scifi books of that era. In Asimov's Foundation, for example, psychology has advanced so much people are practically telepathic.
But AI is completely different that our brain. The name is misleading.
Also because we still don't know how our brain works. All I know is, it won't be running on if..then..else.. functions.
It is not in the least misleading, there are many videos which goes in-depth of how they are similar to the human brain, not like an exact replica of the entire brain but it does absolutely share some similarities.
"Similarity:
Both systems improve performance by modifying connection strengths based on experience.
Difference (important):
Backpropagation as used in LLMs has no confirmed direct biological equivalent. The brain likely uses local, approximate, and noisy learning rules—not global error gradients.
3. Distributed representations
LLMs encode concepts as patterns across many neurons (vectors in high-dimensional space).
The brain encodes concepts across populations of neurons.
Similarity:
No single unit represents “dog” or “grammar”—it’s a pattern across many units.
Implication:
Both systems degrade gracefully: partial damage/noise doesn’t fully destroy knowledge.
Similarity:
Lower levels process simple features; higher levels integrate them into abstractions.
Difference:
Brains have extensive recurrence, feedback loops, and cross-modal integration. LLMs are mostly feedforward transformers with attention, not full dynamical systems in the same sense.
5. Pattern learning from data
LLMs: trained on massive text corpora to model statistical structure.
Humans: learn language from exposure to sensory input and interaction.
Similarity:
Both infer patterns and regularities rather than storing explicit rules.
Difference:
Humans learn from far less data and integrate grounding (vision, action, embodiment). LLMs are text-only (unless multimodal variants).
6. Emergent behavior
LLMs: capabilities (reasoning, translation) emerge from scale and training.
Brain: cognition emerges from large-scale neural interaction.
Similarity:
Complex behavior is not explicitly programmed; it arises from system dynamics.
7. Attention mechanisms (partial analogy)
Transformers use attention to weight relevant tokens dynamically.
The brain exhibits selective attention (focus on relevant stimuli).
Similarity:
Both prioritize certain signals over others.
Difference:
Transformer attention is a mathematical operation (dot-product weighting), not a direct model of biological attention mechanisms."
@DXWarlock yes it wont add it to its training data, that is true, but it will, within the discussion be able to draw its own conclusions based on your discussion + training data. It does save some of this if it thinks its something important, this is limited though. It is often referencing older discussions in new sessions, sometimes it surprises me and sometimes its annoys me that it didnt recall.
A model being able to do real-time learning AND storing it to its training data would be closer to how humans works, but its not how current AI works, its slightly more limited.
I am not sure why it cant be called artificial intelligence just because its not the final form of it. nothing in tech is final. We called computer graphics computer graphics even when it was a few pixels that looked like nothing, it was still graphics but its nothing compared to what exists today. AI is outdated on a monthly basis.
Last edited by vurt on Tue, 28th Apr 2026 22:06; edited 1 time in total
Maybe it's our definition of words that make us see it different.
To me:
Intelligence = ability to achieve goals across a wide range of new tasks.
Wisdom = ability to achieve goals across a wide range of known tasks.
(rough definition, but without getting into weeds of how I see each.)
Wisdom is knowing things. Intelligence is grasping and turning INTO wisdom new things.
Knowing how to use a tablesaw = wisdom. First learning how to use it with no one's help = Intelligence.
So AW (Artificial Wisdom) is more what it should be called. Its good at things it's already trained on, really bad at new things it's not.
{took out the ramble part here}
A good example to me of how its not intelligent is the protein folding one. Not once has it suggested: This folding is inefficient and the hardest way to get what you wish to achieve now that I see your goal. Its just all you humans can think of. I have thought of a new approach you humans have not even considered and is far better than trial and error folding of proteins. (or whatever else area the AI is for).
It just happily keeping doing exactly what we told it to, how we told it to. relying on the wisdom we passed of how how to fold them. Not intelligence on if that's even the best way to get where we want.
-We don't control what happens to us in life, but we control how we respond to what happens in life.
-Hard times create strong men, strong men create good times, good times create weak men, and weak men create hard times. -G. Michael Hopf
Disclaimer: Post made by me are of my own creation. A delusional mind relayed in text form.
Last edited by DXWarlock on Tue, 28th Apr 2026 22:43; edited 1 time in total
its superb at new tasks imo, its the real kicker i think for AI. how it can uses its training data to come up with entirely new things no one thought of. That's the cool thing with this tech, how fast it can do something like that vs a human.
we are very limited like that because our "thought power" has to be quite strict, limited and in comparison often super focused. We can't process 100's of thousands of things in seconds and be surprised that thing 133234 and thing 76123 had a super interesting correlation that made us realize something entirely new. We will mostly focus on what we already know and we can't focus on things that would just seem absurd to even include in the thought process. For an AI, this does not matter as much because it can go through it like a breeze.
since it can use tools, like web searching, it does not even have to be in its training data either, it can be something it researches online, new findings.
See I don't think it comes up with new things no one thought of. it simply connects two things it knows into a thing no one connected. It's using its knowledge (wisdom) of them.
Now if it could come up with two totally new things, not in its dataset, and connect them in ways not in its dataset. THAT would be deductive intelligence to me.
Otherwise its simple pattern finding and probability of likeness percentage to form a new 'thing' that is statistically compatible according to its dataset. Just no human though to smash those two things into a compliance chart of mixed attributes of both.
And it doesnt know if its a good idea. It just knows the delta of the concepts in its dataset have more and closer links than others in the subjects you asked about. Just because you go "I never thought of that" doesn't mean it though of it, or anything. its spitting out word combinations of highest probability to match your request.
-We don't control what happens to us in life, but we control how we respond to what happens in life.
-Hard times create strong men, strong men create good times, good times create weak men, and weak men create hard times. -G. Michael Hopf
Disclaimer: Post made by me are of my own creation. A delusional mind relayed in text form.
Like ask it:
what is the best way an advanced civilization could harness the power of the sun?
It gives what (I consider) the most wasteful one, just the best humans can come up with:
Spoiler:
Dyson Swarm (Most practical mega-scale option)
7
A Dyson swarm (inspired by Freeman Dyson) is a huge cloud of solar collectors orbiting the Sun.
Not a solid shell (that would be unstable and unrealistic)
Millions/billions of independent satellites
Each collects solar energy and beams it (microwaves/lasers) to where it’s needed
Why it’s considered “best”:
Scales gradually—no need to build it all at once
Can capture a huge fraction of the Sun’s output (~10²⁶ watts total available)
Fault-tolerant (losing pieces doesn’t collapse the system)
It doesn't come up with new ideas, it vomits out combinations of what we taught it.
Dyson swarm/sphere (to me) would be laughed at or confused aliens that saw it:
Spoiler:
-What is that?
-A giant ball of drones/circle of solar panels around their sun.
-Why?
-To collect energy of the star
-So they build massive amounts of objects and a sphere to collect only a faction of its total fusion power, that is simply the waste by product of its fusion core?
-Yes
-Why...?
-They can't think of a better thing sir. It's all they came up with.
Follow up on it. Ask it to come up with its own, it's some variation of what we taught it. I cannot get it to come up with any ideas of how to that is 'new' it thought out and is outside the box of what 'we' gave it.
And the sci-fi ones it comes up with, (tap the core, leech the fusion reaction itself) it gives star trek like answers when asked how that can be done. Nothing specific or useful from an engineering angle. Just vague star trek/sci-fi answers like "Heat resistant materials" and such. I ask it to suggest how we can make those..it starts rambling about stuff that I'm pretty sure it stolen from comic books.
Quote:
1. Self-healing / regenerative matter
Constantly rebuilds as it erodes
Works like ablative heat shields—but continuous
Ask it how that might be made, it get stuck in a loop of basically the equal of "I dunno, I have no data on that really, and I cannot deduce new stuff outside what I was fed. But I can make up shit from the 'out of my ass' department. based on what I was fed" loop.
I feel people misattribute it thinks and is intelligent, in the same way we instinctively, even if we know its not, feel this machine is emotionally driven.
Spoiler:
we see it emulating traits we and animals have, and enjoy considering that it has them, just not mimicking them.
-We don't control what happens to us in life, but we control how we respond to what happens in life.
-Hard times create strong men, strong men create good times, good times create weak men, and weak men create hard times. -G. Michael Hopf
Disclaimer: Post made by me are of my own creation. A delusional mind relayed in text form.
Last edited by DXWarlock on Tue, 28th Apr 2026 23:15; edited 1 time in total
Intelligence = ability to achieve goals across a wide range of tasks.
No wonder you are so easily impressed. That’s not what intelligence is at all.
No wonder you don't understand intelligence since you think you can take out one thing, and leave out the rest and get away with it like some type of argument. It's the type of thing you expect from maybe a child. It is absolutely a part of what intelligence is.
@DXWarlock, currently using AI for a new synthesis (audio) method that it came up with Some things will fail, absolutely. No one would say AI is free from issues, the current ones will be laughable in 2 years. Also, the free models, heh, hardly the most sophisticated ones.
I'd love to try the most expensive ones, but i can't afford it.
But that's different in my eyes.
That's formulaic, that's like using it for coding. it's a predictable, sequential, formulaic process of using existing data in a very rigid and confined criteria of set data mixed in different ways.
Ai doing code is the simplest form of "smash together what I gave you into a working set based on clearly defined rules of coding practices and known language structure."
It's the most white room, known parameters of compliance to use, test you could give it.
Now like if before MP3s was around. You was working with it to come up with audio compression that sounds (nearly) identical to raw as far as humans can tell. And it made the MP3 format and underlying tech no one had invented yet. Id say it came up with something new no one had any parts of before, to give it.
It, on its own, deduced to try the concept that overlapping sounds in a song we cannot hear them all, so compress the ones you can drop the data of the ones that technically play out the speaker, but we cannot notice, as well as the idea that we don't need super20k sub 20 audio in the file data..etc.
-We don't control what happens to us in life, but we control how we respond to what happens in life.
-Hard times create strong men, strong men create good times, good times create weak men, and weak men create hard times. -G. Michael Hopf
Disclaimer: Post made by me are of my own creation. A delusional mind relayed in text form.
Last edited by DXWarlock on Tue, 28th Apr 2026 23:35; edited 3 times in total
How dare you. I am good at many different tasks. According to you, I am highly intelligent.
Well, we know that is false, so must be some flaw in his definition
-We don't control what happens to us in life, but we control how we respond to what happens in life.
-Hard times create strong men, strong men create good times, good times create weak men, and weak men create hard times. -G. Michael Hopf
Disclaimer: Post made by me are of my own creation. A delusional mind relayed in text form.
same could be said for humans and our thought processes, at least for the most part
the wonderful thing is this is just the start of this tech... if graphics had been this level, and developed as quickly, woah. That its free to the public to use is also pretty mind blowing to me.
It's not free, just you don't pay money for it.
Not that I am against the concept of "pay for it in other ways than coin" idea. Just pointing out depends on if you consider free, is it just don't cost you specifically your currency.
-We don't control what happens to us in life, but we control how we respond to what happens in life.
-Hard times create strong men, strong men create good times, good times create weak men, and weak men create hard times. -G. Michael Hopf
Disclaimer: Post made by me are of my own creation. A delusional mind relayed in text form.
i mean as in money, though i do use a copilot subscription so i am paying for some of the usage at least.
currently they are not profiting of it either so we are likely only getting it for a limited time, then when we are used to it and want it to be a part of our daily life - pay up bitch!
The data collection is the big one. I myself couldn't give two deep fried bird testicals on what anyone collects from me. Hell just ask me directly and I will offer anything anyone wants to know willingly
MS wants a 1:1 database log of all the filename on my PC? Go for it MS, be careful though if you go into it, fair warning. This folder is all my porn, and this one is all my AI smut LORA files..you was warned. If anyone at MS finds something they want a copy of, just let me know.
But a lot of people don't like that it uses all conversations and your thoughts, feelings, and everything you told it in 'private chats' as data. most you can opt out of some of it, but read the EULA if you are the type to not want it. As opting out doesn't really cover some of the parts you assume it does.
-We don't control what happens to us in life, but we control how we respond to what happens in life.
-Hard times create strong men, strong men create good times, good times create weak men, and weak men create hard times. -G. Michael Hopf
Disclaimer: Post made by me are of my own creation. A delusional mind relayed in text form.
I mean when home gamer cards can. I run as much locally as I can. just 'good' LLM ones you either need to use one out there, or rent a machine on vast.ai or something to run it.
-We don't control what happens to us in life, but we control how we respond to what happens in life.
-Hard times create strong men, strong men create good times, good times create weak men, and weak men create hard times. -G. Michael Hopf
Disclaimer: Post made by me are of my own creation. A delusional mind relayed in text form.
Signature/Avatar nuking: none (can be changed in your profile)
You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum
sue them into oblivion, take all that vc money and make it all gooo poooof
Bulgarian Ai?
Some things will fail, absolutely. No one would say AI is free from issues, the current ones will be laughable in 2 years. Also, the free models, heh, hardly the most sophisticated ones.
