What happened to goorold Neural Networks term of the quite recent past?
Neural Networks were all the rage 8-15 years ago.
Can we go back calling today's AIs and LLMs that?
No controversy, no ambiguity there, everybody's happy
At least its back for rendering
...Anthropic are such hypocrites btw.. talking about slowing down then releasing their strongest model ever can that idiot just shut his trap
Trump tells UN government documents will now call artificial
intelligence "super intelligence," a name picked after polling followers on X
and Truth Social. — The Washington Post
LLMs are not SUPER intelligence, artificial intelligence, or—indeed—any intelligence at all.
Tell that to your LLMs solving millenium puzzles
If intelligence is defined by the ability to form useful internal representations, transform them according to context, reason over them, and produce appropriate behaviour, then there is no obvious reason those computations must be implemented in biological neurons
Or simpler, If a logical thought can be expressed via a word consquentlly a system can be intelligent if its thoughts are expressed via words
People saying LLMs, what we currently call AI, are just random word generators is like saying humans aren't thinking, they're just generating motor-neuron activations
Even if that is hugely inefficient compared to our system, any system that can express neuronal structures and logic through words, sounds, shapes or whatever still has a valid path to intelligence.
The architecture is different but even current systems exhibit forms of intelligence. If you define intellignce functionally: reasoning, abstraction, generalization, planning, linguistic competence etc.
But the denial crowd is getting quiter.
If the internal workings are different, but the output is always the same as organic systems no matter the input then the discussion becomes mute and it doesn't matter if we the system is using words. And that is where we are moving towards.
No wonder Trump thinks LLMs are Super Itelligence!
He thinks that changing the name of lake Ontario to lake America will happen just because he say so...
I still stand by the first, of many but not the only marker, before I think we can call it intelligent is:
It has a single thought on its own, just one. That's all.
Knowledgeable and intelligent are not the same thing
So just one self composed, and internalized thought, even a inkling of a thought without someone manually putting a 'seed' and command to initiate processing and rationalizing a thing. It needs to do at least one thing intelligent proactively, not just reactively on command.
If it cannot have self isolated thoughts, or processing of data independant of a seed parameter and need to be told to do. Its less intelligent than bacteria.
I mean that as in, it may have more processing power than bacteria by FAR. But if it is not doing it on its own, then its a powerful processing machine that needs a 'driver' to do anything. Even bacteria are smart enough to process itself and environment independently on its own.
-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.
Not just a thought. It also has to be able to adapt itself based on its thoughts. A “thought” is not something really one can define anyway. To dumburt, it’s calculating 2+2.
Sorry, what I mean by it. Just didn't convey it well.
It forming a cognitive..process(?). An autonomous cognition, or intrinsically generated cognition? Dunno the right words. Hope it comes across.
It isn’t enough to produce a sophisticated response. The system would need to initiate the process itself, internalize it, and adapt because of it. Else it remains an extraordinarily capable reactive processor, not an independently thinking intelligence.
And (to me anyway and my understanding of the thin grip we have on what intelligence means anyway) If it takes an initiator, and driver to cause the thinking. The thinking isn't intelligence, it's purely high level processing.
That processing can produce the same output that intelligence might produce.
But if the system cannot initiate such a process on its own, then only the capacity to process information exists, not the intelligence to autonomously consider what information is worth processing, or why.
And that last part, even with our fuzzy definition of what intelligence is, is the validated core that the fuzzy edges grows out of. And it does not have that yet.
-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.
But that’s inherently impossible with current LLMs, which remain stateless. Context may grow, thus larger capacity for “memories” and instructions, but ultimately they are just stateless pachinko machines.
All the Ai's at the moment are a sliver of what a AGI would need, and they are all narrowly HIGHLY specialized at a specific thing: Produce words, produce art, do math, o quantum math, do CGI graphics, molecular chemistry. And none of them can do the others work.
Being able to speak, think, imagine (for mental images for images), understanding (true understanding, not 'book smart' memorizing of connected facts to reduce to conclusion). And would all need it in a different form to start with. To be intelligent.
I really feel a part of it like I said, is that it's the only 'thing' in our history that can communicate with us, like us. So the applying of erroneous emergent properties is a easy trap to fall into.
I (in my head anyway) make weird analogies...who knew? But it would be akin to some going:
This diesel engine has high work ethic and wants to be a productive member of humanities bettering. Because when I start it, and push the throttle. It instantly outputs horsepower to do any work I can imagine to attach it to: Plow, move stuff, generate electricity, move people, power water pumps..anything. It's more than willing to do it.
Not a good example. But attributing things not existent to an an entity, because what it produces seems to have intent and meaning in scope of what you see the results are.
If you have to start it, and tell it what to do, to do a thing. Intelligence is missing.
-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.
"Artificial intelligence (AI) is the capability of computational systems to perform tasks typically associated with human intelligence, such as learning, reasoning, problem-solving, perception, and decision-making." (from wikipedia)
we have been able to do this for years with LLMs. It's AI per the very definition, undoubtedly.
you're making AI sound like some kind of god-like ability almost, or as complex as the human mind, that's not it at all. AI is not something that impressive. you are talking about something which is far beyond AI.
where are you even having an issue with current AI (LLMS are AI per definition) for anything (problem solving, reasoning, coding etc etc etc), give even one example. Bet you can't. Leotard has the reasoning capability of a 5-6 year old at most. He can't even grasp that artificial intelligence is artificial intelligence. Which is so stupid it's beyond belief.
they must have been able to do this for a LONG while, I have a really hard time thinking an AI would have any kind of issues improving its code over time.
Nah. It's just a feedback loop. I was doing that for AI images (textures), though it did need my human eyes for sorting. Today i could possibly set it up with an AI to automate it entirely because of how good they are as "eyes" these days. They certainly were not when i started..
For code its more advanced but i can see how it could be automated.
If i had a ton of money i would love to try it on my own code projects, but yeah, it would be super costly. Just setting it up would be super fun and interesting
I think to me its the blurring of the terms I get stuck on. If they called it EI, Emulated Intelligence I wouldn't mind.
Its the word artificial. It makes people attribute more to it by sense of the name than it is. They mix artificial with the meaning of "Real, just not organic intelligence"
Far too many people, at least of ones I know (not here just in general) start expanding what it can do, beyond what it could eventually do.
Like saying AI (in LLM centric context anyway) is close to being able to automate assembly lines with no humans, or in the near future replace doctors and surgeons with pinpoint accuracy to minimize recovery time.
(well I guess it COULD do the first, but man the most inefficient way to. Like cutting a muffler off with a hammer. You CAN beat the pipe in half over time...but...40x the work vs a thing made to do that to start with)
Its a language model, that is FAR outside its scope of abilities to even comprehend. It's good at Language and Math because mathematics is just a highly structured rule-governed language in the end. Its symbols have precise meanings, its expressions follow rigid “grammar,” and valid transformations obey explicit rules.
People conflate artificial intelligence with a universal intelligence that can learn any task.Language models will always be gated to language based tasks.
It's like claiming Image models will one day be able to write sonata's or epic poems. Or that Suno will one day be autonomous vehicle AI.
It's not made for that, so much so not made it it, it's a nonsensical statement to say. Like saying if you make a good enough hammer tool, you can also use it as a tool to learn to sing.
-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.
Blurring, how? I think the explanation on e.g wikipedia is exactly what LLM's are.
Artificial is the opposite word of natural, genuine, authentic, real, or organic. That's why we have the word in the first place.
I have no idea how people are upset that we use "AI" for AI.
If people have a misunderstanding of the word "artificial", then maybe they just are not very good at English, that's hardly the fault of LLMs or people using the word correctly. I see people wanting to remove the word a lot when discussing AI, just to be able dismiss it as AI. Which is just moronic.
Our language is the reason we mix up things. An LLM does not think, Stable Diffusion does not create art. These systems print out words and pictures that make sense for us but not for them. But those are the words we use to describe what they do because of the lack of better ones. The verbs "think" and "create" are meaningful for subjects but the LLM is not a subject. Its not an "it" even. This is anthropomorphic reduction through the grammatical necessities of language. Add to that our ability and need to recognize ourselves in everything and the confusion is perfect.
"Enlightenment is man's emergence from his self-imposed nonage. Nonage is the inability to use one's own understanding without another's guidance. This nonage is self-imposed if its cause lies not in lack of understanding but in indecision and lack of courage to use one's own mind without another's guidance. Dare to know! (Sapere aude.) "Have the courage to use your own understanding," is therefore the motto of the enlightenment."
Language is not necessarily hard or confusing and it's a great time to learn with internet, you can find the meaning of something in a second, very different from when i learned english in the 80's (almost entirely from playing adventure games, looking up words).
It actually does make sense to an LLM, that's how they are so good at what they do, the world model they use is complex beyond our comprehension and incredibly good. They encode a substantial amount of semantic structure: concepts, relationships, syntax, spatial relations, factual associations, abstractions and so on!
However that does not mean it works 100% like a human, it certainly does not. The use of the word artificial is thus correct and not hard to understand at all why it would be used, IF you understand what it means.
Artificial is not another word for real. It's its opposite. Artificial Intelligence is not the same as Intelligence (human intelligence). I guess this is super hard for some people, i do not fully understand why.
Generative art is a subgenre of art, its goes way back, before SD or LLM's were a thing. Generative music is also a very old term, i have done both since way before stable diffusion or models like Suno.
I really don't think that what LLMs encode is the same as it "making sense" to "them". Again, terms we use to describe our experiences. An LLM does not experience anything.
I believe the term artificial intelligence is problematic because somehow we came to think that, intelligence is intelligence and that "natural" or "artificial" are just two different ways of achieving intelligence. Like "intelligence" being the same thing just achieved by a different route.
"Enlightenment is man's emergence from his self-imposed nonage. Nonage is the inability to use one's own understanding without another's guidance. This nonage is self-imposed if its cause lies not in lack of understanding but in indecision and lack of courage to use one's own mind without another's guidance. Dare to know! (Sapere aude.) "Have the courage to use your own understanding," is therefore the motto of the enlightenment."
What you maen they don’t experience? A google incel was fired for claiming that “AI” was alive and suffering, because the LLM created some complimenting output. it’s sad tbh.
Our language is the reason we mix up things. An LLM does not think, Stable Diffusion does not create art. These systems print out words and pictures that make sense for us but not for them. But those are the words we use to describe what they do because of the lack of better ones. The verbs "think" and "create" are meaningful for subjects but the LLM is not a subject. Its not an "it" even. This is anthropomorphic reduction through the grammatical necessities of language. Add to that our ability and need to recognize ourselves in everything and the confusion is perfect.
Way better put than me.
Its along the line of when discussing bacteria or single celled in school/college. And using terms like "wants to" or "has a compulsion to" for actions they do.
Like splicing in some organic material to make the bacteria 'want to' eat petroleum based products. It carries more meaning that it needs to. It doesnt think or 'want' to do it. It just a series of complex chemical reactions that produce movement towards the source of the stimuli to convert and break it down.
It 'wants' to in the same way vinegar wants to react with baking soda., just more complex and numerous reaction to produce the result.
Some people attribute actual intelligence to it, because artificial gets confused with "Real and grown, vs man made" and ignore if the intelligence applies, it's just a given it is. All we are doing is defining what 'type' it is with the name.
vurt wrote:
If people have a misunderstanding of the word "artificial", then maybe they just are not very good at English
Welcome to the majority of the English speaking world.
Most governments and courtrooms spend a LOT of time arguing what words someone else said/wrote down/used really mean (laws and regulations). We have a whole branch of our government that does only that (Supreme Court) argue for what words on paper mean, and they even have had this exact discussion for what 'is and isn't' artificial flavoring for food labels.
It's not misunderstanding of the word, it's that the more common daily use of it like above people correlate "Real and grown, vs man made". Artificial coloring doesnt mean its NOT color..just that its man made color. Artificial flavoring doesnt mean its imaginary or doesn't exist. Means it's just man made. Therefor people think: Artificial Intelligence doesnt mean its not intelligence, just means it's man made.
-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 really don't think that what LLMs encode is the same as it "making sense" to "them". Again, terms we use to describe our experiences. An LLM does not experience anything.
Of course its not the same, its entirely artificial.
Quote:
I believe the term artificial intelligence is problematic because somehow we came to think that, intelligence is intelligence and that "natural" or "artificial" are just two different ways of achieving intelligence. Like "intelligence" being the same thing just achieved by a different route.
I more see it as intelligent tasks that can be done by a machine. Unsure why so many are caught up with "it doesn't THINK actively" etc, why would we even want that? How would that be better? What we want is a dormant "oracle" that we can ask and it will do our biddings, and that is what we have. Like i said before, i think its pop-culture influencing what people think AI should be and shouldn't be, and not more practical and realistic scenarios.
They are already super scared of these mostly dormant AI's. Imagine one with RSI and with constant thinking and reasoning, evolving freely.
That is not just twice as scary as current tech, its 1000x as scary and totally unpredictable. Not realistic and not practical.
@DXWarlock
Artificial limb - it will do the TASK of a limb but it will also not resemble the actual body part in most ways.
AI, it might do the job but its not the same as human intelligence or an actual brain.
But sure, strip away that part and i can see what you mean.
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
But it would be akin to some going:
