I see a huge amount of confusion around terminology in discussions about Artificial Intelligence, so here’s my quick attempt to clear some of it up.
Artificial Intelligence is the broadest possible category. It includes everything from the chess opponent on the Atari to hypothetical superintelligent systems piloting spaceships in sci-fi. Both are forms of artificial intelligence - but drastically different.
That chess engine is an example of narrow AI: it may even be superhuman at chess, but it can’t do anything else. In contrast, the sci-fi systems like HAL 9000, JARVIS, Ava, Mother, Samantha, Skynet, or GERTY are imagined as generally intelligent - that is, capable of performing a wide range of cognitive tasks across domains. This is called Artificial General Intelligence (AGI).
One common misconception I keep running into is the claim that Large Language Models (LLMs) like ChatGPT are “not AI” or “not intelligent.” That’s simply false. The issue here is mostly about mismatched expectations. LLMs are not generally intelligent - but they are a form of narrow AI. They’re trained to do one thing very well: generate natural-sounding text based on patterns in language. And they do that with remarkable fluency.
What they’re not designed to do is give factual answers. That it often seems like they do is a side effect - a reflection of how much factual information was present in their training data. But fundamentally, they’re not knowledge databases - they’re statistical pattern machines trained to continue a given prompt with plausible text.
What they’re not designed to do is give factual answers
or mental health therapy
To add to the confusion, you also have people out there thinking it’s “Al” or “A1”. It’s a real mess.
And “intelligence” itself isn’t very well defined either. So the only word that remains is “artificial”, and we can agree on that.
I usually try to avoid the word “AI”. I’ll say “LLM” if I talk about chatbots, ChatGPT etc. Or I use the term “machine learning” when broadly speaking about the concept of computers learning and doing such things. It’s not exactly the same thing, though. But when reading other people’s texts I always think of LLMs when they say AI, because that’s currently what they mean almost every time. And AGI is more sci-fi as of now, so it needs some disclaimers and context anyway.
In computer science, the term AI at its simplest just refers to a system capable of performing any cognitive task typically done by humans.
That said, you’re right in the sense that when people say “AI” these days, they almost always mean generative AI - not AI in the broader sense.
Yeah, generative AI is a good point.
I’m not sure with the computer scientists, though. It’s certainly not any task, that’d be AGI. And it’s not necessarily connected to humans either. Sure they’re the prime example of intelligence (whatever it is). But I think a search engine is AI as well, depending how it’s laid out. And text to speech, old-school expert systems. A thermostat that controls your heating with a machine learning model might count as well, I’m not sure about that. And that’s not really like human cognitive tasks. Closer to curve fitting, than anything else. The thermostat includes problem-solving, learning, perception, knowledge, and planning and decision making. But on the human intelligence score it wouldn’t even be a thing that compares.
Usually the reason we want people to stop calling LLMs AI is because there has been a giant marketing machine constructed designed to (and successfully) tricking laymen into believing that LLMs are adjacent to and one tiny breakthrough away from becoming AGI.
From another angle, your statement that AI is not a specific term is correct. Why, then, should we keep using it in common parlance when it just serves to confuse laymen? Let’s just use the more specific terms.
In defense of people who say LLMs are not intelligent: they probably mean to say they are not sapient, and I think they’re loosely correct if you consider the literal word “intelligent” to have a different meaning from the denotative “Intelligence” in the context of Artificial Intelligence.
I remember when “heuristics” were all the rage. Frankly that’s what LLMs are, advanced heuristics. “Intelligence” is nothing more than marketing bingo.
‘Intelligence’ requires understanding, understanding requires awareness. This is not seen in anything called “AI”, not today at least, but maybe not ever. Again, why not use a different word, one that actually applies to these advanced calculators? Expecting the best out of humanity, it may be because of the appeal of the added pizzazz and the excitement that comes with it or simple semantic confusion… but seeing the people behind it all, it probably is so the dummies get overly excited and buy stuff/make these bots integral parts of their lives. 🤷
The term “Artificial Intelligence” has been around for a long time, 25 years ago AI was an acceptable name for NPC logic in videogames. Arguably that’s still the case, and personally I vastly prefer “Artificial Intelligence” to “Broad Simulation Of Common Sense Powered By Von Neumann Machines”.
“Understanding requires awareness” isn’t some settled fact - it’s just something you’ve asserted. There’s plenty of debate around what understanding even is, especially in AI, and awareness or consciousness is not a prerequisite in most definitions. Systems can model, translate, infer, and apply concepts without being “aware” of anything - just like humans often do things without conscious thought.
You don’t need to be self-aware to understand that a sentence is grammatically incorrect or that one molecule binds better than another. It’s fine to critique the hype around AI - a lot of it is overblown - but slipping in homemade definitions like that just muddies the waters.
Do you think “AI” KNOWS/UNDERSTANDS what a grammatically incorrect sentence is or what molecules even are? How?
You’re moving the goalposts. First you claimed understanding requires awareness, now you’re asking whether an AI knows what a molecule is - as if that’s even the standard for functional intelligence.
No, AI doesn’t “know” things the way a human does. But it can still reliably identify ungrammatical sentences or predict molecular interactions based on training data. If your definition of “understanding” requires some kind of inner experience or conscious grasp of meaning, then fine. But that’s a philosophical stance, not a technical one.
The point is: you don’t need subjective awareness to model relationships in data and produce useful results. That’s what modern AI does, and that’s enough to call it intelligent in the functional sense - whether or not it “knows” anything in the way you’d like it to.
Intelligence, as the word has always been used, requires awareness and understanding, not just spitting out data after input, as dynamic and complex the process might be, through a set of rules. AI, as you just described it, does nothing necessarily different from other computational tools: they speed up processes that can be calculated/algorithmically structured. I don’t see how that particularly makes “AI” deserving of the adjective ‘intelligent’, it seems more of a marketing term the same way ‘smartphones’ were. The disagreement we’re having here is semantic…
Do most humans understand what molecules are? How?
Everything I know about molecules I got from textbooks. Am I just regurgitating my “training data” without understanding? How does one really understand molecules?
None of this is AI if it doesn’t have the ability to become self-aware.
Consciousness - or “self-awareness” - has never been a requirement for something to qualify as artificial intelligence. It’s an important topic about AI, sure, but it’s a separate discussion entirely. You don’t need self-awareness to solve problems, learn patterns, or outperform humans at specific tasks - and that’s what intelligence, in this context, actually means.
So… not intelligent. In the sense that when someone without enough knowledge of computers and/or LLMs hears “LLM is intelligent” and sees “an LLM tells me X”, they will be likely to believe that X is true, and not without a reason. Exactly this is my main reason against all the use of intelligence-related terms. When spoken by knowledgeable people who do know the difference - yeah, I am all for that. But first we need to cut the crap of advertisement and hype
“Intelligent” is itself a highly unspecific term which covers quite a lot of different things.
What you’re think is “reasoning” or “rationalizing”, and LLMs can’t do that at all.
However what LLMs (and most Machine Learning implementations) can do is “pattern matching” which is also an element of intelligence: it’s what gives us and most animals the ability to recognize things such as food or predators without actually thinking about it (you just see, say, a cat, and you know without thinking that it’s a cat even though cats don’t all look the same), plus in humans it’s also what’s behind intuition.
PS: Way back since when they were invented over 3 decades ago, Neural Networks and other Machine Learning technologies were already very good at finding patterns in their training data - often better than humans.
The evolution of the technology has added to it the capability of creating content which follows those patterns, giving us things like LLMs or image generation.
However what has been made clear by LLMs is that using patterns alone (plus a little randomness to vary the results) in generating textual content is not enough to create useful content beyond entertainment, and that’s exactly because LLMs can’t rationalize. However, the original pattern matching stuff without the content generation is still widely used and very successfully so, in things from OCR to image recognition.
So… not intelligent.
But they are intelligent - just not in the way people tend to think.
There’s nothing inherently wrong with avoiding certain terminology, but I’d caution against deliberately using incorrect terms, because that only opens the door to more confusion. It might help when explaining something one-on-one in private, but in an online discussion with a broad audience, you should be precise with your choice of words. Otherwise, you end up with what looks like disagreement, when in reality it’s just people talking past each other - using the same terms but with completely different interpretations.
They ain’t intelligent
AGI itself has been made up as a marketing term by LLM companies.
Let’s not forget that the official definition of AGI is that it can make 200 billion dollars.
The term AGI was first used in 1997 by Mark Avrum Gubrud in an article named ‘Nanotechnology and international security’
By advanced artificial general intelligence, I mean AI systems that rival or surpass the human brain in complexity and speed, that can acquire, manipulate and reason with general knowledge, and that are usable in essentially any phase of industrial or military operations where a human intelligence would otherwise be needed. Such systems may be modeled on the human brain, but they do not necessarily have to be, and they do not have to be “conscious” or possess any other competence that is not strictly relevant to their application. What matters is that such systems can be used to replace human brains in tasks ranging from organizing and running a mine or a factory to piloting an airplane, analyzing intelligence data or planning a battle.
What would you call systems that are used for discovery of new drugs or treatments? For example, companies using “AI” for Parkinson’s research.
Both that and LLMs fall under the umbrella of machine learning, but they branch in different directions. LLMs are optimized for generating language, while the systems used in drug discovery focus on pattern recognition, prediction, and simulations. Same foundation - different tools for different jobs.
Is there a specific name? Or just “non-LLM ML systems”?
They’re generally just referred to as “deep learning” or “machine learning”. The models themselves usually have names of their own, such as AlphaFold, PathAI and Enlitic.
I still think intelligence is a marketing term or simply a misnomer. It’s basically an advanced calculator. Intelligence questions, creates rules from nothing, transforms raw data from reality into ideas, has its own volition… And the same goes for a chess engine, of course, it’s just more visible because it’s not spitting out text but chess moves. Intelligence and consciousness don’t seem to be computational processes.
You’re describing intelligence more like a soul than a system - something that must question, create, and will things into existence. But that’s a human ideal, not a scientific definition. In practice, intelligence is the ability to solve problems, generalize across contexts, and adapt to novel inputs. LLMs and chess engines both do that - they just do it without a sense of self.
A calculator doesn’t qualify because it runs “fixed code” with no learning or generalization. There’s no flexibility to it. It can’t adapt.
I could follow everything you said up until the conclusion. If consciousness is not computational, then what is going on in our brains instead? I know of course that even neuroscientists don’t know exactly, but just in broad principle. I always thought our brains are still doing computation, just with a different method to computers. I don’t mean to be contrarian, I’m just genuenly curious what other kind of process could support consciousness?
I’m not gonna claim to ‘know’ things here, and I’m too groggy to even attempt to give you a satisfying answer but: applied formal logic as seen in any machine based on logic gates is just an expression/replication of simplified thought and not a copy of our base mental processes. The mind understands truths that cannot even be formalized or derived, such as axiomatic truths. Even if something can be understood and predicted, it doesn’t mean the process could be written down in code. It certainly isn’t today…
My understanding of the topic is closer to Roger Penrose’s postulates so please check this wiki page and maybe watch a couple of vids on the topic, I’m just a peasant with a hunch when it comes to “AI”. 🤷
