Words are the &quot;simplistic&quot; projection of an LLM&#x27;s abstract thoughts.An LLM has: words in its input plane, words in its output plane, and A LOT of cross-linked internals between the two.Those internals aren&#x27;t &quot;words&quot; at all - and it&#x27;s where most of the &quot;action&quot; happens. It&#x27;s how LLMs can do things like translate from language to language, or recall knowledge they only encountered in English in the training data while speaking German.

I&#x27;m definitely a stream of words.My &quot;abstract thoughts&quot; are a stream of words too, they just don&#x27;t get sounded out.Tbf I&#x27;d rather they weren&#x27;t there in the first place.But bodies which refuse to harbor an &quot;interiority&quot; are fast-tracked to destruction because they can&#x27;t suf^W^W^W be productive.Funny movie scene from somewhere. The sergeant is drilling the troops: &quot;You, private! What do you live for!&quot;, and expects an answer along the lines of dying for one&#x27;s nation or some shit. Instead, the soldier replies: &quot;Well, to see what happens next!&quot;

Even if they are &quot;simplistic projections&quot;, which I don&#x27;t think is the correct way to think about it, there&#x27;s no reason that more LLM thoughts in middle layers can&#x27;t also exist and project down at the end. Though there might be efficency issues because the latent thoughts have to be recomputed a lot.Though I do think in human brains it&#x27;s also an interplay where what we write&#x2F;say also loops back into the thinking as well. Which is something which is efficient for LLMs.

I am a stream of words - I have even ran out of tokens while speaking before :)But raising kids, I can clearly see that intelligence isn&#x27;t just solved by LLMs

LLMs and human brains are both just mechanisms. Why would one mechanism a priori be capable of &quot;learning abstract thought&quot;, but no others?If it turns out that LLMs don&#x27;t model human brains well enough to qualify as &quot;learning abstract thought&quot; the way humans do, some future technology will do so. Human brains aren&#x27;t magic, special or different.

Are you a stream of words or are your words the “simplistic” projection of your abstract thoughts? I don’t at all discount the importance of language in so many things, but the question that matters is whether statistical models of language can ever “learn” abstract thought, or become part of a system which uses them as a tool.My personal assessment is that LLMs can do neither.

Well, yes because thinking soon requires interacting, not just ideating. It&#x27;s in the dialogue between ideation and interaction that we make our discoveries.

when LLM popped out and people started to say &#x27;this is just markov chain on steroid and not thinking&#x27; i was a bit confused because a lot of my &quot;thinking&quot; is statistical too.. I very often try to solve an issue by switching a known solution with a different &quot;probable&quot; variant of it (tweaking a parameter)LLMs have higher dimensions (they map token to grammatical and semantical space) .. it might not be thinking but it seems on its way we&#x27;re just thinking with more abstractions before producing speech ?... dunno

I claim that all of thinking can be reduced to predicting the next thing. Predicting the next thing = thinking in the same way that reading and writing strings of bytes is a universal interface, or every computation can be done by a Turing machine.

&gt; Everyone is out here acting like &quot;predicting the next thing&quot; is somehow fundamentally irrelevant to &quot;human thinking&quot; and it is simply not the case.Nobody is. What people are doing is claiming that &quot;predicting the next thing&quot; does not define the entirety of human thinking, and something that is ONLY predicting the next thing is not, fundamentally, thinking.

But your assertion is merely semantic. It doesn&#x27;t say anything substantive.I could also say a motorcycle &quot;moves forward&quot; just like a person &quot;moves forward&quot;. Whether we use the same or different words for same or different concepts doesn&#x27;t say anything about the actual underlying similarity.And please don&#x27;t call stuff &quot;dumb shit&quot; here. Not appropriate for HN.

A forklift is &quot;lifting&quot; things, despite using a completely different mechanical process as a human &quot;lifting&quot; things. The only real similarity between these kinds of &quot;lifting&quot; is the end result, something is higher up than it was before.

is this seriously about continuous rotation versus a pair of double pendulums making a stepping motion?

That’s because the motorcycle thing is too simlistic of a comparison. It doesn’t come nearly close to capturing the nuance of the whole LLM “thinking” situation.

A motorcycle is not &quot;sprinting&quot; and an LLM is not &quot;thinking&quot;. Everyone would agree that a motorcycle is not running but the same dumb shit is posted over and over and over on here that somehow the LLM is &quot;thinking&quot;.

If all humans were suddenly wiped off the face of the earth, AI would go silent, and the hardware it runs on would eventually shut down.If all AI was suddenly wiped off the face of the earth, humans would rebuild it, and would carry on fine in the meantime.One AI researcher decades ago said something to the effect of: researchers in biology look at living organisms and wonder how they live; researchers in physics look at the cosmos and wonder what all is out there; researchers in artificial intelligence look at computer systems and wonder how they can be made to wonder such things.

But the &quot;AI&quot; is simply a gestalt of all human language. You&#x27;re looking in a mirror.

AI has made me question what it is to be a human.I am not having some existential crisis, but if we get to a point where X% of humans cannot outperform “AI” on any task that humans deem “useful”, for some nontrivial value of X, then many assumptions that culture has inculcated into me about humanity are no longer valid.What is the role of humans then?Can it be said that humans “think” if they can’t think a thought that a non thinking AI cannot also think?

It may be doing the &quot;thinking&quot; and could reach AGI. But we don&#x27;t want it. We don&#x27;t want to take a fork lift to the gym. We don&#x27;t want plastic aliens showing off their AGI and asking humanity to outsource human thinking and decision-making to them.

Predicting the next token is not at all the same thing as predicting the next action in a causal chain of actions. Not even close. One is model of language tokens, the other is a model of the physical world. You can come up with all sorts of predictions that can&#x27;t be expressed cleanly in natural language. And plenty of things that parse cleanly from a language perspective but are unhinged in their description of empirical reality.

&gt; When you have a thought, are you &quot;predicting the next thing&quot;Yes. This is the core claim of the Free Energy Principle[0], from the most-cited neuroscientist alive. Predictive processing isn&#x27;t AI hype - it&#x27;s the dominant theoretical framework in computational neuroscience for ~15 years now.&gt; much of our experience of the world does not entail predicting thingsIntrospection isn&#x27;t evidence about computational architecture. You don&#x27;t experience your V1 doing edge detection either.&gt; How confident are you that the abstractions &quot;search&quot; and &quot;thinking&quot;... are really equatable?This isn&#x27;t about confidence, it&#x27;s about whether you&#x27;re engaging with the actual literature. Active inference[1] argues cognition IS prediction and action in service of minimizing surprise. Disagree if you want, but you&#x27;re disagreeing with Friston, not OpenAI marketing.&gt; How does Heisenberg&#x27;s famous principle complicate thisIt doesn&#x27;t. Quantum uncertainty at subatomic scales has no demonstrated relevance to cognitive architecture. This is vibes.&gt; Companies... are claiming these tools do more than they are actually capable ofPossibly true! But &quot;is cognition fundamentally predictive&quot; is a question about brains, not LLMs. You&#x27;ve accidentally dismissed mainstream neuroscience while trying to critique AI hype.[0] <a href="https:&#x2F;&#x2F;www.nature.com&#x2F;articles&#x2F;nrn2787" rel="nofollow">https:&#x2F;&#x2F;www.nature.com&#x2F;articles&#x2F;nrn2787</a>[1] <a href="https:&#x2F;&#x2F;mitpress.mit.edu&#x2F;9780262045353&#x2F;active-inference&#x2F;" rel="nofollow">https:&#x2F;&#x2F;mitpress.mit.edu&#x2F;9780262045353&#x2F;active-inference&#x2F;</a>

&gt; can you confidently classify all mental activity that you experience as &quot;predicting the next thing&quot;? [...] On what do you base your confidence in their equivalence?To my understanding, bloaf&#x27;s claim was only that the ability to predict seems a requirement of acting intentionally and thus that LLMs may &quot;end up being a component in a system which actually does think&quot; - not necessarily that all thought is prediction or that an LLM would be the entire system.I&#x27;d personally go further and claim that correctly generating the next token is already a sufficiently general task to embed pretty much any intellectual capability. To complete `2360 + 8352 * 4 = ` for unseen problems is to be capable of arithmetic, for instance.

&gt; When you have a thought, are you &quot;predicting the next thing&quot;—can you confidently classify all mental activity that you experience as &quot;predicting the next thing&quot;?So notice that my original claim was &quot;prediction is fundamental to our ability to act with intent&quot; and now your demand is to prove that &quot;prediction is fundamental to all mental activity.&quot;That&#x27;s a subtle but dishonest rhetorical shift to make me have to defend a much broader claim, which I have no desire to do.&gt; Language and society constrains the way we use words, but when you speak, are you &quot;predicting&quot;?Yes, and necessarily so. One of the main objections that dualists use to argue that our mental processes must be immaterial is this [0]:* If our mental processes are physical, then there cannot be an ultimate metaphysical truth-of-the-matter about the meaning of those processes.* If there is no ultimate metaphysical truth-of-the-matter about what those processes mean, then everything they do and produce are similarly devoid of meaning.* Asserting a non-dualist mind therefore implies your words are meaningless, a self-defeating assertion.The simple answer to this dualist argument is precisely captured by this concept of prediction. There is no need to assert some kind of underlying magical meaning to be able to communicate. Instead, we need only say that in the relevant circumstances, our minds are capable of predicting what impact words will have on the receiver and choosing them accordingly. Since we humans don&#x27;t have access to each other&#x27;s minds, we must not learn these impacts from some kind of psychic mind-to-mind sense, but simply from observing the impacts of the words we choose on other parties; something that LLMs are currently (at least somewhat) capable of observing.[0] <a href="https:&#x2F;&#x2F;www.newdualism.org&#x2F;papers&#x2F;E.Feser&#x2F;Feser-acpq_2013.pdf" rel="nofollow">https:&#x2F;&#x2F;www.newdualism.org&#x2F;papers&#x2F;E.Feser&#x2F;Feser-acpq_2013.pd...</a>If you read the above link you will see that they spell out 3 problems with our understanding of thought:Consciousness, intentionality, and rationality.Of these, I believe prediction is only necessary for intentionality, but it does have some roles to play in consciousness and rationality.

When you have a thought, are you &quot;predicting the next thing&quot;—can you confidently classify all mental activity that you experience as &quot;predicting the next thing&quot;?Language and society constrains the way we use words, but when you speak, are you &quot;predicting&quot;? Science allows human beings to predict various outcomes with varying degrees of success, but much of our experience of the world does not entail predicting things.How confident are you that the abstractions &quot;search&quot; and &quot;thinking&quot; as applied to the neurological biological machine called the human brain, nervous system, and sensorium and the machine called an LLM are really equatable? On what do you base your confidence in their equivalence?Does an equivalence of observable behavior imply an ontological equivalence? How does Heisenberg&#x27;s famous principle complicate this when we consider the role observer&#x27;s play in founding their own observations? How much of your confidence is based on biased notions rather than direct evidence?The critics are right to raise these arguments. Companies with a tremendous amount of power are claiming these tools do more than they are actually capable of and they actively mislead consumers in this manner.

You can make a separate model for the task, which is based on well chosen features and calibrated from actual data. Then the LLM only needs to generate the arguments to this model (extract those features from messages) and call it like a MCP tool. This external tool can be a simple Sklearn model.

I&#x27;m an LLMs are being used in workflows they don&#x27;t make sense in-sayer. And while yes, I can believe that LLMs can be part of a system that actually does think, I believe that to achieve true &quot;thinking&quot;, it would likely be a system that is more deterministic in its approach rather than probabilistic.Especially when modeling acting with intent. The ability to measure against past results and think of new innovative approaches seems like it may come from a system that may model first and then use LLM output. Basically something that has a foundation of tools rather than an LLM using MCP. Perhaps using LLMs to generate a response that humans like to read, but not in them coming up with the answer.Either way, yes, its possible for a thinking system to use LLMs (and potentially humans piece together sentences in a similar way), but its also possible LLMs will be cast aside and a new approach will be used to create an AGI.So for me: even if you are an AI-yeasayer, you can still believe that they won&#x27;t be a component in an AGI.

A good heuristic is that if an argument resorts to &quot;actually not doing &lt;something complex sounding&gt;&quot; or &quot;just doing &lt;something simple sounding&gt;&quot; etc, it is not a rigorous argument.

And the big players have built a bunch of workflows which embed many other elements besides just &quot;predictions&quot; into their AI product. Things like web search, to incorporating feedback from code testing, to feeding outputs back into future iterations. Who is to say that one or more of these additions has pushed the ensemble across the threshold and into &quot;real actual thinking.&quot;The near-religious fervor which people insist that &quot;its just prediction&quot; makes me want to respond with some religious allusions of my own:&gt; Who is this that wrappeth up sentences in unskillful words? Gird up thy loins like a man: I will ask thee, and answer thou me. Where wast thou when I laid up the foundations of the earth? tell me if thou hast understanding. Who hath laid the measures thereof, if thou knowest? or who hath stretched the line upon it?The point is that (as far as I know) we simply don&#x27;t know the necessary or sufficient conditions for &quot;thinking&quot; in the first place, let alone &quot;human thinking.&quot; Eventually we will most likely arrive at a scientific consensus, but as of right now we don&#x27;t have the terms nailed down well enough to claim the kind of certainty I see from AI-detractors.

&gt; The issue is that prediction is &quot;part&quot; of the human thought process, it&#x27;s not the full story...Do you have a proof for this?Surely such a profound claim about human thought process must have a solid proof somewhere? Otherwise who&#x27;s to say all of human thought process is not just a derivative of &quot;predicting the next thing&quot;?

The issue is that prediction is &quot;part&quot; of the human thought process, it&#x27;s not the full story...

Yes, personally I&#x27;m completely fine with the fact that LLMs don&#x27;t actually think. I don&#x27;t care that they&#x27;re not AGI, though the hysterics about &quot;AGI is so close now&quot; seems silly to me. Fusion reactors and self-driving cars are just around the same corner.They prove to have some useful utility to me regardless.

It&#x27;s fascinating when you look at each technical component of cognition in human brains and contrast against LLMs. In humans, we have all sorts of parallel asynchronous processes running, with prediction of columnar activations seemingly the fundamental local function, with tens of thousands of mini columns and regions in the brain corresponding to millions of networked neurons using the &quot;predict which column fires next&quot; objective to increment or decrement the relative contribution of any functional unit.In the case of LLMs you run into similarities, but they&#x27;re much more monolithic networks, so the aggregate activations are going to scan across billions of neurons each pass. The sub-networks you can select each pass by looking at a threshold of activations resemble the diverse set of semantic clusters in bio brains - there&#x27;s a convergent mechanism in how LLMs structure their model of the world and how brains model the world.This shouldn&#x27;t be surprising - transformer networks are designed to learn the complex representations of the underlying causes that bring about things like human generated text, audio, and video.If you modeled a star with a large transformer model, you would end up with semantic structures and representations that correlate to complex dynamic systems within the star. If you model slug cellular growth, you&#x27;ll get structure and semantics corresponding to slug DNA. Transformers aren&#x27;t the end-all solution - the paradigm is missing a level of abstraction that fully generalizes across all domains, but it&#x27;s a really good way to elicit complex functions from sophisticated systems, and by contrasting the way in which those models fail against the way natural systems operate, we&#x27;ll find better, more general methods and architectures, until we cross the threshold of fully general algorithms.Biological brains are a computational substrate - we exist as brains in bone vats, connected to a wonderfully complex and sophisticated sensor suite and mobility platform that feeds electrically activated sensory streams into our brains, which get processed into a synthetic construct we experience as reality.Part of the underlying basic functioning of our brains is each individual column performing the task of predicting which of any of the columns it&#x27;s connected to will fire next. The better a column is at predicting, the better the brain gets at understanding the world, and biological brains are recursively granular across arbitrary degrees of abstraction.LLMs aren&#x27;t inherently incapable of fully emulating human cognition, but the differences they exhibit are expensive. It&#x27;s going to be far more efficient to modify the architecture, and this may diverge enough that whatever the solution ends up being, it won&#x27;t reasonably be called an LLM. Or it might not, and there&#x27;s some clever tweak to things that will push LLMs over the threshold.

most humans in any percentile act towards the thing of someone else. most of these things are a lot worse than what the human &quot;would originally intend&quot;. this behavior stems from 100s and thousands of nudges since childhood.the issue with AI and AI-naysayers is, by analogy, this: cars were build to drive from A to Z. people picked up tastes and some people started building really cool looking cars. the same happens on the engineering side. then portfolio communists came with their fake capitalism and now cars are build to drive over people but don&#x27;t really work because people, thankfully, are overwhelming still fighting to attempt to act towards their own intents.

Exactly. Our base learning is by example, which is very much learning to predict.Predict the right words, predict the answer, predict when the ball bounces, etc. Then reversing predictions that we have learned. I.e. choosing the action with the highest prediction of the outcome we want. Whether that is one step, or a series of predicted best steps.Also, people confuse different levels of algorithm.There are at least 4 levels of algorithm:• 1 - The architecture.This input-output calculation for pre-trained models are very well understood. We put together a model consisting of matrix&#x2F;tensor operations and few other simple functions, and that is the model. Just a normal but high parameter calculation.• 2 - The training algorithm.These are completely understood.There are certainly lots of questions about what is most efficient, alternatives, etc. But training algorithms harnessing gradients and similar feedback are very clearly defined.• 3 - The type of problem a model is trained on.Many basic problem forms are well understood. For instance, for prediction we have an ordered series of information, with later information to be predicted from earlier information. It could simply be an input and response that is learned. Or a long series of information.• 4 - The solution learned to solve (3) the outer problem, using (2) the training algorithm on (1) the model architecture.People keep confusing (4) with (1), (2) or (3). But it is very different.For starters, in the general case, and for most any challenging problem, we never understand their solution. Someday it might be routine, but today we don&#x27;t even know how to approach that for any significant problem.Secondly, even with (1), (2), and (3) exactly the same, (4) is going to be wildly different based on the data characterizing the specific problem to solve. For complex problems, like language, layers and layers of sub-solutions to sub-problems have to be solved, and since models are not infinite in size, ways to repurpose sub-solutions, and weave together sub-solutions to address all the ways different sub-problems do and don&#x27;t share commonalities.Yes, prediction is the outer form of their solution. But to do that they have to learn all the relationships in the data. And there is no limit to how complex relationships in data can be. So there is no limit on the depths or complexity of the solutions found by successfully trained models.Any argument they don&#x27;t reason, based on the fact that they are being trained to predict, confuses at least (3) and (4). That is a category error.It is true, they reason a lot more like our &quot;fast thinking&quot;, intuitive responses, than our careful deep and reflective reasoning. And they are missing important functions, like a sense of what they know or don&#x27;t. They don&#x27;t continuously learn while inferencing. Or experience meta-learning, where they improve on their own reasoning abilities with reflection, like we do. And notoriously, by design, they don&#x27;t &quot;see&quot; the letters that spell words in any normal sense. They see tokens.Those reasoning limitations can be irritating or humorous. Like when a model seems to clearly recognize a failure you point out, but then replicates the same error over and over. No ability to learn on the spot. But they do reason.Today, despite many successful models, nobody understands how models are able to reason like they do. There is shallow analysis. The weights are there to experiment with. But nobody can walk away from the model and training process, and build a language model directly themselves. We have no idea how to independently replicate what they have learned, despite having their solution right in front of us. Other than going through the whole process of retraining another one.

This is the &quot;but LLMs will get better, trust me&quot; thread?

Did anyone ever seriously buy the whole &quot;it&#x27;s transporting&quot; BS when it was wheelbarrows? What makes you believe today&#x27;s trucks are meaningfully different?

LLMs merely interpolate between the feeble artifacts of thought we call language.The illusion wears off after about half an hour for even the most casual users. That&#x27;s better than the old chatbots, but they&#x27;re still chatbots.Did anyone ever seriously buy the whole &quot;it&#x27;s thinking&quot; BS when it was Markov chains? What makes you believe today&#x27;s LLMs are meaningfully different?

&gt; Your decisions can be predicted from brain scans up to 10 seconds before you make them, which means they are as deterministic as an LLM&#x27;s.This conclusion does not follow from the result at all.

Libet has only measured the latency of metaconsciousness&#x2F;cognition, nothing else. It says nothing about free will, which is ill defined anyway.

What is the import of this to you here? Whether you have free will or you feel like you do, kinda same difference for this particular point right? It doesn&#x27;t make me more human actually having free will, it is sufficient to simply walk around as if I do.But beyond that, what do you want to say here? What is lost, what is gained? Are you wanting to say this makes us more like an LLM? How so?

I looked up the Libet experiment:&quot;ImplicationsThe experiment raised significant questions about free will and determinism. While it suggested that unconscious brain activity precedes conscious decision-making, Libet argued that this does not negate free will, as individuals can still choose to suppress actions initiated by unconscious processes.&quot;

I suspect that people instinctively believe they have free will, both because it feels like we do, and because society requires us to behave that way even when we don&#x27;t.The truth is that the evidence says we don&#x27;t. See the Libet experiment and its many replications.Your decisions can be predicted from brain scans up to 10 seconds before you make them, which means they are as deterministic as an LLM&#x27;s. Sorry, I guess.

Everyone is out here acting like &quot;predicting the next thing&quot; is somehow fundamentally irrelevant to &quot;human thinking&quot; and it is simply not the case.What does it mean to say that we humans act with intent? It means that we have some expectation or prediction about how our actions will effect the next thing, and choose our actions based on how much we like that effect. The ability to predict is fundamental to our ability to act intentionally.So in my mind: even if you grant all the AI-naysayer&#x27;s complaints about how LLMs aren&#x27;t &quot;actually&quot; thinking, you can still believe that they will end up being a component in a system which actually &quot;does&quot; think.

One metaphor is to call the model a person, another metaphor is to call it a pile of words. These are quite opposite. I think that&#x27;s the whole point.Person-metaphor does nothing to explain its behavior, either.&quot;Bag of words&quot; has a deep origin in English, the Anglo-Saxon kenning &quot;word-hord&quot;, as when Beowulf addresses the Danish sea-scout (line 258)&quot;He unlocked his word-hoard and delivered this answer.&quot;So, bag of words, word-treasury, was already a metaphor for what makes a person a clever speaker.

Given that LLMs are incapable of synthetic a priori knowledge and humans are, I would say that as the tech stands currently, it&#x27;s reasonable to make both of those statements.

The loop, or more precisely the &quot;search&quot; does the novel part in thinking, the brain is just optimizing this process. Evolution could manage with the simplest model - copying with occasional errors, and in one run it made everyone of us. The moral - if you scale search the model can be dumb.

I&#x27;ll make the following observation:The contra-positive of &quot;All LLMs are not thinking like humans&quot; is &quot;No humans are thinking like LLMs&quot;And I do not believe we actually understand human thinking well enough to make that assertion.Indeed, it is my deep suspicion that we will eventually achieve AGI not by totally abandoning today&#x27;s LLMs for some other paradigm, but rather embedding them in a loop with the right persistence mechanisms.

But people aren&#x27;t using ChatGPT for completing phrases. They&#x27;re using it to get their tasks done, or get their questions answered.The fact that pretraining of ChatGPT is done with a &quot;completing the phrase&quot; task has no bearing on how people actually end up using it.

I agree 100%. Most people haven&#x27;t actually interacted directly with an LLM before. Most people&#x27;s experience with LLMs is ChatGPT, Claude, Grok, or any of the other tools that automatically handle context, memory, personality, temperature, and are deliberately engineered to have the tool communicate like a human. There is a ton of very deterministic programming that happens between you and the LLM itself to create this experience, and much of the time when people are talking about the ineffable intelligence of chatbots, it&#x27;s because of the illusion created by this scaffolding.

For me, the problem is in the &quot;chat&quot; mechanic that OpenAI and others use to present the product. It lends itself to strong antropomorphizing.If instead of a chat interface we simply had a &quot;complete the phrase&quot; interface, people would understand the tool better for what it is.

Sentence completion does not give it justice, when I can ask a LLM to refactor my repo and come back half an hour later to see the deed done.

I&#x27;ve been recently using a similar description, referring to &quot;AI&quot; (LLMs) as &quot;glorified autocomplete&quot; or &quot;luxury autocomplete&quot;.

Yea bag of words isn’t helpful at all. I really do think that “superpowered sentence completion” is the best description. Not only is it reasonably accurate it is understandable, everyone has seen autocomplete function, and it’s useful. I don’t know how to “use” a bag of words. I do know how to use sentence completion. It also helps explains why context matters.

I think it&#x27;s still a bit of a tortured metaphor. LLMs operate on tokens, not words. And to describe their behavior as pulling the right word out of a bag is so vague that it applies every bit as much to a Naive Bayes model written in Python in 10 minutes as it does to the greatest state of the art LLM.

Yeah. I have a half-cynical&#x2F;half-serious pet theory that a decent fraction of humanity has a broken theory of mind and thinks everyone has the same thought patterns they do. If it talks like me, it thinks like me.Whenever the comment section takes a long hit and goes &quot;but what is thinking, really&quot; I get slightly more cynical about it lol

Bag of words is actually the perfect metaphor. The data structure is a bag. The output is a word. The selection strategy is opaquely undefined.&gt; Gen AI tricks laypeople into treating its token inferences as &quot;thinking&quot; because it is trained to replicate the semiotic appearance of doing so. A &quot;bag of words&quot; doesn&#x27;t sufficiently explain this behavior.Something about there being significant overlap between the smartest bears and the dumbest humans. Sorry you[0] were fooled by the magic bag.[0] in the &quot;not you, the layperson in question&quot; sense

Which only means they can mimic the output of a human. So does a p-zombie. It doesn&#x27;t make them human.

well they are trained to be almost in distribution as a thinking human. So...

&gt; Spoken Query Language? Just like SQL, but for unstructured blobs of text as a database and unstructured language as a query?I feel that&#x27;s more a description of a search engine. Doesn&#x27;t really give an intuition of why LLMs can do the things they do (beyond retrieval), or where&#x2F;why they&#x27;ll fail.

Spoken Query Language? Just like SQL, but for unstructured blobs of text as a database and unstructured language as a query? Also known as Slop Query Language or just Slop Machine for its unpredictable results.

Every day I see people treat gen AI like a thinking human, Dijkstra&#x27;s attitudes about anthropomorphizing computers is vindicated even more.That said, I think the author&#x27;s use of &quot;bag of words&quot; here is a mistake. Not only does it have a real meaning in a similar area as LLMs, but I don&#x27;t think the metaphor explains anything. Gen AI tricks laypeople into treating its token inferences as &quot;thinking&quot; because it is trained to replicate the semiotic appearance of doing so. A &quot;bag of words&quot; doesn&#x27;t sufficiently explain this behavior.

&gt; Slightly unfortunate that &quot;Bag of words&quot; is already a different conceptYes, subconsciously I kept trying to map this article&#x27;s ideas to word2vec and continuous-bag-of-words.

Slightly unfortunate that &quot;Bag of words&quot; is already a different concept: <a href="https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;Bag_of_words" rel="nofollow">https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;Bag_of_words</a>.My second thought is that it&#x27;s not the metaphor that is misleading. People have been told thousands of times that LLMs don&#x27;t &quot;think&quot;, don&#x27;t &quot;know&quot;, don&#x27;t &quot;feel&quot;, but are &quot;just a very impressive autocomplete&quot;. If they still really want to completely ignore that, why would they suddenly change their mind with a new metaphor?Humans are lazy. If it looks true enough and it cost less effort, humans will love it. &quot;Are you sure the LLM did your job correctly?&quot; is completely irrelevant: people couldn&#x27;t care less if it&#x27;s correct or not. As long as the employer believes that the employee is &quot;doing their job&quot;, that&#x27;s good enough. So the question is really: &quot;do you think you&#x27;ll get fired if you use this?&quot;. If the answer is &quot;no, actually I may even look more productive to my employer&quot;, then why would people not use it?

&gt; If we allow ourselves to be seduced by the superficial similarity, we’ll end up like the moths who evolved to navigate by the light of the moon, only to find themselves drawn to—and ultimately electrocuted by—the mysterious glow of a bug zapper.Woah, that hit hard

I don&#x27;t think the author&#x27;s argument is that it won&#x27;t replace any human labour. Or at least I wouldn&#x27;t agree with such an argument. But the stronger case is that however much LLMs improve, they won&#x27;t replace humans in general. In the furtherment of knowledge, because they are fundamentally parroting and synthesizing the already known, vs performing truly novel thought. And in creative fields, because people are fundamentally interested in creations of other people, not of computers.Neither of these is entirely true in all cases, but they could be expected to remain true in at least some (many) cases, and so the role for humans remains.

I think the canonical answer is that humans are “bags of mostly water” .

<a href="https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;The_Meme_Machine" rel="nofollow">https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;The_Meme_Machine</a>

I just came from the Pluribus sub-Reddit. I’ll take AI over that cohort any day.

So tell me, why do I still have a job and why am frequently successful in getting profitable &#x2F; useful products into production if I’m “expensive and unreliable”?I mean I use AI tools to help achieve the goal but I don’t see any signs of the things I’m building and doing being unreliable.

So a human is just a really expensive, unreliable bag of words. And we get more expensive and more unreliable by the day!There&#x27;s a quote I love but have misplaced, from the 19th century I think. &quot;Our bodies are just contraptions for carrying our heads around.&quot; Or in this instance... bag of words transport system ;)

&quot;Her&quot;-&gt;&quot;the&quot;? (Or, who is &quot;she&quot; here?)Either way, in what way is this relevant? If the human&#x27;s labor is not useful at any price point to any entity with money, food or housing, then they presumably will not get paid&#x2F;given food&#x2F;housing for it.

Her argument really only works if you institute new economic systems where humans don’t need to labor in order to eat or pay rent.

As usual with these, it helps to try to keep the metaphor used for downplaying AI, but flip the script. Let&#x27;s grant the author&#x27;s perception that AI is a &quot;bag of words&quot;, which is already damn good at producing the &quot;right words&quot; for any given situation, and only keeps getting better at it.Sure, this is not the same as being a human. Does that really mean, as the author seems to believe without argument, that humans need not be afraid that it will usurp their role? In how many contexts is the utility of having a human, if you squint, not just that a human has so far been the best way to &quot;produce the right words in any given situation&quot;, that is, to use the meat-bag only in its capacity as a word-bag? In how many more contexts would a really good magic bag of words be better than a human, if it existed, even if the current human is used somewhat differently? The author seems to rest assured that a human (long-distance?) lover will not be replaced by a &quot;bag of words&quot;; why, especially once the bag of words is also ducttaped to a bag of pictures and a bag of sounds?I can just imagine someone - a horse breeder, or an anthropomorphised horse - dismissing all concerns on the eve of the automotive revolution, talking about how marketers and gullible marks are prone to hippomorphising anything that looks like it can be ridden and some more, and sprinkling some anecdotes about kids riding broomsticks, legends of pegasi and patterns of stars in the sky being interpreted as horses since ancient times.

A person can study fashion extensively, under the best designers, they can understand tailoring and fit and have a phenomenal eye for color and texture.Their vivid descriptions of what the Emperor could be wearing doesn&#x27;t make said emperor any less nakey.

Amanda Askell studied under David Chalmers at NYU: the philosopher who coined &quot;the hard problem of consciousness&quot; and is famous for taking phenomenal experience seriously rather than explaining it away. That context makes her choice to speak this way more striking: this isn&#x27;t naive anthropomorphizing from someone unfamiliar with the debates. It&#x27;s someone trained by one of the most rigorous philosophers of consciousness, who knows all the arguments for dismissing mental states in non-biological systems, and is still choosing to speak carefully about models potentially having something like feelings or insecurities.

Here are three of the Anthropic research reports I had in mind:<a href="https:&#x2F;&#x2F;www.anthropic.com&#x2F;news&#x2F;golden-gate-claude" rel="nofollow">https:&#x2F;&#x2F;www.anthropic.com&#x2F;news&#x2F;golden-gate-claude</a>Excerpt: “We found that there’s a specific combination of neurons in Claude’s neural network that activates when it encounters a mention (or a picture) of this most famous San Francisco landmark.”<a href="https:&#x2F;&#x2F;www.anthropic.com&#x2F;research&#x2F;tracing-thoughts-language-model" rel="nofollow">https:&#x2F;&#x2F;www.anthropic.com&#x2F;research&#x2F;tracing-thoughts-language...</a>Excerpt: “Recent research on smaller models has shown hints of shared grammatical mechanisms across languages. We investigate this by asking Claude for the ‘opposite of small’ across different languages, and find that the same core features for the concepts of smallness and oppositeness activate, and trigger a concept of largeness, which gets translated out into the language of the question.”<a href="https:&#x2F;&#x2F;www.anthropic.com&#x2F;research&#x2F;introspection" rel="nofollow">https:&#x2F;&#x2F;www.anthropic.com&#x2F;research&#x2F;introspection</a>Excerpt: “Our new research provides evidence for some degree of introspective awareness in our current Claude models, as well as a degree of control over their own internal states.”

&gt;research from Anthropic [1] suggests that structures corresponding to meaning exist inside those bundles of numbers and that there are signs of activity within those bundles of numbers that seem analogous to thought.Can you give some concrete examples? The link you provided is kind of opaque&gt;Amanda Askell [2]. When she talks, she anthropomorphizes LLMs constantly.She is a philosopher by trade and she describes her job (model alignment) as literally to ensure models &quot;have good character traits.&quot; I imagine that explains a lot

Well, she&#x27;s describing the system&#x27;s behavior.My fridge happily reads inputs without consciousness, has goals and takes decisions without &quot;thinking&quot;, and consistently takes action to achieve those goals. (And it&#x27;s not even a smart fridge! It&#x27;s the one with a copper coil or whatever.)I guess the cybernetic language might be less triggering here (talking about systems and measurements and control) but it&#x27;s basically the same underlying principles. One is just &quot;human flavored&quot; and I therefore more prone to invite unhelpful lines of thinking?Except that the &quot;fridge&quot; in this case is specifically and explicitly designed to emulate human behavior so... you would indeed expect to find structures corresponding to the patterns it&#x27;s been designed to simulate.Wondering if it&#x27;s internalized any other human-like tendencies — having been explicitly trained to simulate the mechanisms that produced all human text — doesn&#x27;t seem too unreasonable to me.

&gt; How did these 2 models do it if not actually using language like a thinking agent?By having a gazillion of other, almost identical pictures of kids in parks in their training data.

&gt; the interpretability research from Anthropic [1] suggests that structures corresponding to meaning do exist inside those bundles of numbers and that there are signs of activity within those bundles of numbers that seem analogous to thoughtI did a simple experiment - took a photo of my kid in the park, showed it to Gemini and asked for a &quot;detailed description&quot;. Then I took that description and put it into a generative model (Z-Image-Turbo, a new one). The output image was almost identical.So one model converted image to text, the other reversed the processs. The photo was completely new, personal, never put online. So it was not in any training set. How did these 2 models do it if not actually using language like a thinking agent?<a href="https:&#x2F;&#x2F;pbs.twimg.com&#x2F;media&#x2F;G7gTuf8WkAAGxRr?format=jpg&amp;name=medium" rel="nofollow">https:&#x2F;&#x2F;pbs.twimg.com&#x2F;media&#x2F;G7gTuf8WkAAGxRr?format=jpg&amp;name=...</a>

I use LLMs heavily for work, I have done so for about 6 months. I see almost zero &quot;thought&quot; going on and a LOT of pattern matching. You can use this knowledge to your advantage if you understand this. If you&#x27;re relying on it to &quot;think&quot;, disaster will ensue. At least that&#x27;s been my experience.I&#x27;ve completely given up on using LLMs for anything more than a typing assistant &#x2F; translator and maybe an encyclopedia when I don&#x27;t care about correctness.

Wow those quotes are extremely disturbing.

It also leads to fundamentally wrong conclusions: a related issue I have with this is the use of anthropomorphic shorthand when discussing international politics. You&#x27;ve heard a phrase like &quot;the US thinks...&quot;, &quot;China wants...&quot;, &quot;Europe believes...&quot; so much you don&#x27;t even notice it.All useful shorthands, all which lead to people displaying fundamental misunderstandings of what they&#x27;re talking about - i.e. expressing surprise that a nation of millions doesn&#x27;t display consistency of behavior of human lifetime scales, even though fairly obviously the mechanisms of government are churning their make up constantly, and depending on context maybe entirely different people.

It obfuscates far more than it clarifies.

the anthropomorphization (say that 3 times quickly) is kinda weird, but also makes for a much more pleasant conversation imo. it&#x27;s kinda tedious being pedantic all the time.

This argument would have a lot more weight if it was published in a peer reviewed journal by a party that does not have a stake in the AI market.

I am unsure myself whether we should regard LLMs as mere token-predicting automatons or as some new kind of incipient intelligence. Despite their origins as statistical parrots, the interpretability research from Anthropic [1] suggests that structures corresponding to meaning do exist inside those bundles of numbers and that there are signs of activity within those bundles of numbers that seem analogous to thought.That said, I was struck by a recent interview with Anthropic’s Amanda Askell [2]. When she talks, she anthropomorphizes LLMs constantly. A few examples:“I don&#x27;t have all the answers of how should models feel about past model deprecation, about their own identity, but I do want to try and help models figure that out and then to at least know that we care about it and are thinking about it.”“If you go into the depths of the model and you find some deep-seated insecurity, then that&#x27;s really valuable.”“... that could lead to models almost feeling afraid that they&#x27;re gonna do the wrong thing or are very self-critical or feeling like humans are going to behave negatively towards them.”[1] <a href="https:&#x2F;&#x2F;www.anthropic.com&#x2F;research&#x2F;team&#x2F;interpretability" rel="nofollow">https:&#x2F;&#x2F;www.anthropic.com&#x2F;research&#x2F;team&#x2F;interpretability</a>[2] <a href="https:&#x2F;&#x2F;youtu.be&#x2F;I9aGC6Ui3eE" rel="nofollow">https:&#x2F;&#x2F;youtu.be&#x2F;I9aGC6Ui3eE</a>

Did you explain how LLMs can achieve gold-medal performance at math competitions involving original problems, without any original knowledge or thought?Did she ask if a &quot;statistical soup of words,&quot; if large enough, might somehow encode or represent something a little more profound than just a bunch of words?

I was trying to explain the concept of &quot;token prediction&quot; to my wife, whose eyes glaze over when discussing such technical topics. (I think she has the brainpower to understand them, but a horrible math teacher gave her a taste aversion to even attempting to that hasn&#x27;t gone away. So she just buys Apple stuff and hopes Tim Apple hasn&#x27;t shuffled around the UI bits AGAIN.)I stumbled across a good-enough analogy based on something she loves: refrigerator magnet poetry, which if it&#x27;s good consists of not just words but also word fragments like &quot;s&quot;, &quot;ed&quot;, and &quot;ing&quot; kinda like LLM tokens. I said that ChatGPT is like refrigerator magnet poetry in a magical bag of holding that somehow always gives the tile that&#x27;s the most or nearly the most statistically plausible next token given the previous text. E.g., if the magnets already up read &quot;easy come and easy ____&quot;, the bag would be likely to produce &quot;go&quot;. That got into her head the idea that these things operate based on plausibility ratings from a statistical soup of words, not anything in the real world nor any internal cogitation about facts. Any knowledge or thought apparent in the LLM was conducted by the original human authors of the words in the soup.

I don&#x27;t think I&#x27;ve heard anyone (beyond the most inane Twitterati) confidently state &quot;therefore it is thought.&quot;I hear a lot of people saying &quot;it&#x27;s certainly not and cannot be thought&quot; and then &quot;it&#x27;s not exactly clear how to delineate these things or how to detect any delineations we might want.&quot;

These discussions often end up resembling religious arguments. &quot;We don&#x27;t know how any of this works, but we can fathom an intelligent god doing it, therefore an intelligent god did it.&quot;&quot;We don&#x27;t really know how human consciousness works, but the LLM resembles things we associate with thought, therefore it is thought.&quot;I think most people would agree that the functioning of an LLM resembles human thought, but I think most people, even the ones who think that LLMs can think, would agree that LLMs don&#x27;t think in the exact same way that a human brain does. At best, you can argue that whatever they are doing could be classified as &quot;thought&quot; because we barely have a good definition for the word in the first place.

You may know the mechanics, but you don&#x27;t know how LLMs &quot;work&quot; because no one really understands (yet, hopefully).

&gt;I&#x27;m a neurologist, and as a consequence of my profession, I understand how humans work under the hood.There you go again, auto-morphizing the meat-bags. Vroom vroom.

I upvoted you.This is a fundamentally interesting point. Taking your comment as HN would advise, I totally agree.I think genAI freaks a lot of people out because it makes them doubt what they thought made them special.And to your comment, humans have always used words they reserve for humanity that indicates we&#x27;re special: that we think, feel, etc... That we&#x27;re human. Maybe we&#x27;re not so special. Maybe that&#x27;s scary to a lot of people.

It doesn&#x27;t strike you as a bit...illogical to state in your first sentence that you &quot;understand how humans work under the hood&quot; and then go on to say that humans don&#x27;t actually &quot;understand&quot; anything? Clearly everything at its basis is a chemical reaction, but the right reactions chained together create understanding, knowing, etc. I do believe that the human brain can be modeled by machines, but I don&#x27;t believe LLMs are anywhere close to being on the right track.

There are no properties of matter or energy that can have a sense of self or experience qualia. Yet we all do. Denying the hard problem of consciousness just slows down our progress in discovering what it is.

I&#x27;m a neurologist, and as a consequence of my profession, I understand how humans work under the hood.The average human is so easily convinced that humans &quot;know&quot;, &quot;think&quot;, &quot;lie&quot;, &quot;want&quot;, &quot;understand&quot;, etc.But really it&#x27;s all just a probabilistic chain reaction of electrochemical and thermal interactions. There is literally nowhere in the brain&#x27;s internals for anything like &quot;knowing&quot; or &quot;thinking&quot; or &quot;lying&quot; to happen!Strange that we have to pretend otherwise

As a consequence of my profession, I understand how LLMs work under the hood.I also know that we data and tech folks will probably never win the battle over anthropomorphization.The average user of AI, nevermind folks who should know better, is so easily convinced that AI &quot;knows,&quot; &quot;thinks,&quot; &quot;lies,&quot; &quot;wants,&quot; &quot;understands,&quot; etc. Add to this that all AI hosts push this perspective (and why not, it&#x27;s the easiest white lie to get the user to act so that they get a lot of value), and there&#x27;s really too much to fight against.We&#x27;re just gonna keep on running into this and it&#x27;ll just be like when you take chemistry and physics and the teachers say, &quot;it&#x27;s not actually like this but we&#x27;ll get to how some years down the line- just pretend this is true for the time being.&quot;

There&#x27;s no definition of thinking that isn&#x27;t a purely internal phenomenon, which means that there&#x27;s no way to point a diagnostic device at someone and determine whether they&#x27;re thinking. The only way to determine whether something is conscious&#x2F;thinking is through some sort of inference, which is why Turing landed on the Turing Test that he did. Problem is, technology over the past 5 years pretty easily passes variations of the Turing Test, and exposed a lot of its limits as well.So the next definition of detecting &quot;thinking&quot; will have to be externally observable and inferrable like a Turing Test, but get into the other things that we consider part of consciousness&#x2F;thinking.Often this is some combination of introspection (understanding internal states), perception (understanding external objects), and synthesis of the two into testable hypotheses in some sort of feedback loop between the internal representation of the world and the external feedback from the world.Right now, a chatbot can say all sorts of things about itself and about the world, but none of that is based on real-time, factual information. Whereas an animal can&#x27;t speak, but they clearly process information and consider it when determining their future and current actions.

It&#x27;s not obvious to me what you expect from this hypothetical 0.01% post, or in other words, what about it makes it a one-in-ten-thousand post?

In this thread: 99% of posters using their own personal definition of &quot;thinking&quot; without explaining it; 0.99% of posters complaining that it all depends on what that definition is; not enough posts yet for that 0.01% response to occur...

Gemini 3.0 might do well even without web searches. The lesson from gpt 4.5 and Gemini 3 seems to be that scaling model size (even if you use sparse MoE) allows you to capture more long-tail knowledge. Some of Humanity&#x27;s Last Exam also seems to be explicitly designed to test this long-tail obscure knowledge extraction, and models have been steadily chipping away at it.

I tested this with ChatGPT-5.1 and Gemini 3.0. Both correctly (according to Wikipedia at least) stated that George Olshevsky assigned it to its own genus in 1991.This is because there are many words about how to do web searches.

Does it generalize though? What a bag-of-words metaphor can say about a question &quot;How many reinforcement learning training examples an LLM need to significantly improve performance on mathematical questions?&quot;

When sensitivity analysis of ordinary least-squares regression became a thing it was also a &quot;retrospective narrative&quot;. That seems reasonable for detecting fundamental issues with statistical models of the world. This point generalizes even if the concrete example falls down.

I literally pasted the sentence as a prompt to the free version of ChatGPT &quot;Who reassigned the species Brachiosaurus brancai to its own genus, and when?&quot;and got ths correct reply from the &quot;Bag of Words&quot;The species Brachiosaurus brancai was reassigned to its own genus by Michael P. Taylor in 2009 — he transferred it to the new genus Giraffatitan. 
BioOne
+2
Mike Taylor
+2How that happened:Earlier, in 1988, Gregory S. Paul had proposed putting B. brancai into a subgenus as Brachiosaurus (Giraffatitan) brancai, based on anatomical differences. 
Fossil Wiki
+1Then in 1991, George Olshevsky used the name Giraffatitan brancai — but his usage was in a self-published list and not widely adopted. 
Wikipedia
+1Finally, in 2009 Taylor published a detailed re-evaluation showing at least 26 osteological differences between the African material (brancai) and the North American type species Brachiosaurus altithorax — justifying full generic separation. 
BioOne
+1If you like — I can show a short timeline of all taxonomic changes of B. brancai.--As an author, you should write things that are tested or at least true. But they did a pretty bad job of testing this and are making assumptions that are not true. Then they&#x27;re basing their argument&#x2F;reasoning (restrospectively) on assumptions not gounded in reality.

Your conclusion seems super unfair to the offer, particularly your assumption, without reason as far as I can tell, that the author would obstinately continue to advocate for their conclusion in the face of new, contrary evidence.

I could not tell you who reassigned the species Brachiosaurus brancai to its own genus, and when, because of all the words I&#x27;ve ever heard, the combination of words that contains the information has not appeared.GIGO has an obvious Nothing-In-Nothing-Out trivial case.

I don&#x27;t know enough about graphomania, but I do find this article, while I&#x27;m sure is written by a human, has qualities akin to LLM writing: lengthy, forced comparisons and analogies. Of course it&#x27;s far less organized than typical ChatGPT output though.The more human works I&#x27;ve read the more I feel meat intelligences are not that different from tensor intelligences.

Isn&#x27;t it pretty clear just from the first paragraph that the author has graphomania? Such people don&#x27;t really care about the thesis, they care about the topic and how many literary devices they can fit into the article.

&gt; “Bag of words” is a also a useful heuristic for predicting where an AI will do well and where it will fail. “Give me a list of the ten worst transportation disasters in North America” is an easy task for a bag of words, because disasters are well-documented. On the other hand, “Who reassigned the species Brachiosaurus brancai to its own genus, and when?” is a hard task for a bag of words, because the bag just doesn’t contain that many words on the topicIt is... such a retrospective narrative. It&#x27;s so obvious that the author learned about this example first than came with the reasoning later, just to fit in his view of LLM.Imaging if ChatGPT answered this question correctly. Would that change the author&#x27;s view? Of course not! They&#x27;ll just say:&gt; “Bag of words” is a also a useful heuristic for predicting where an AI will do well and where it will fail. Who reassigned the species Brachiosaurus brancai to its own genus, and when?” is an easy task for a bag of words, because the information has appeared in the words it memorizes.I highly doubt this author has predicted that &quot;bag of Words&quot; can do image editing before OpenAI released that.

This is essentially Lady Lovelace&#x27;s objection from the 19th century [1]. Turing addressed this directly in &quot;Computing Machinery and Intelligence&quot; (1950) [2], and implicitly via the halting problem in &quot;On Computable Numbers&quot; (1936) [3]. Later work on cellular automata, famously Conway&#x27;s Game of Life [4], demonstrates more conclusively that this framing fails as a predictive model: simple rules produce structures no one &quot;put in.&quot;A test I did myself was to ask Claude (The LLM from Anthropic) to write working code for entirely novel instruction set architectures (e.g., custom ISAs from the game Turing Complete [5]), which is difficult to reconcile with pure retrieval.[1] Lovelace, A. (1843). Notes by the Translator, in Scientific Memoirs Vol. 3. (&quot;The Analytical Engine has no pretensions whatever to originate anything. It can do whatever we know how to order it to perform.&quot;) Primary source: <a href="https:&#x2F;&#x2F;en.wikisource.org&#x2F;wiki&#x2F;Scientific_Memoirs&#x2F;3&#x2F;Sketch_of_the_Analytical_Engine_invented_by_Charles_Babbage,_Esq.&#x2F;Notes_by_the_Translator" rel="nofollow">https:&#x2F;&#x2F;en.wikisource.org&#x2F;wiki&#x2F;Scientific_Memoirs&#x2F;3&#x2F;Sketch_o...</a>. See also: <a href="https:&#x2F;&#x2F;www.historyofdatascience.com&#x2F;ada-lovelace&#x2F;" rel="nofollow">https:&#x2F;&#x2F;www.historyofdatascience.com&#x2F;ada-lovelace&#x2F;</a> and <a href="https:&#x2F;&#x2F;writings.stephenwolfram.com&#x2F;2015&#x2F;12&#x2F;untangling-the-tale-of-ada-lovelace&#x2F;" rel="nofollow">https:&#x2F;&#x2F;writings.stephenwolfram.com&#x2F;2015&#x2F;12&#x2F;untangling-the-t...</a>[2] <a href="https:&#x2F;&#x2F;academic.oup.com&#x2F;mind&#x2F;article&#x2F;LIX&#x2F;236&#x2F;433&#x2F;986238" rel="nofollow">https:&#x2F;&#x2F;academic.oup.com&#x2F;mind&#x2F;article&#x2F;LIX&#x2F;236&#x2F;433&#x2F;986238</a>[3] <a href="https:&#x2F;&#x2F;www.cs.virginia.edu&#x2F;~robins&#x2F;Turing_Paper_1936.pdf" rel="nofollow">https:&#x2F;&#x2F;www.cs.virginia.edu&#x2F;~robins&#x2F;Turing_Paper_1936.pdf</a>[4] <a href="https:&#x2F;&#x2F;web.stanford.edu&#x2F;class&#x2F;sts145&#x2F;Library&#x2F;life.pdf" rel="nofollow">https:&#x2F;&#x2F;web.stanford.edu&#x2F;class&#x2F;sts145&#x2F;Library&#x2F;life.pdf</a>[5] <a href="https:&#x2F;&#x2F;store.steampowered.com&#x2F;app&#x2F;1444480&#x2F;Turing_Complete&#x2F;" rel="nofollow">https:&#x2F;&#x2F;store.steampowered.com&#x2F;app&#x2F;1444480&#x2F;Turing_Complete&#x2F;</a>

^^; I think this post is close to singularity as we may get on this Monday.

An LLM creates a high fidelity statistical probabistic model of human language. The hope is to capture the input&#x2F;output of various hierarchical formal and semiformal systems of logic that transit from human to human, which we know as &quot;Intelligence&quot;.Unfortunately, its corpus is bound to contain noise&#x2F;nonsense that follows no formal reasoning system but contributes to the ill advised idea that an AI should sound like a human to be considered intelligent. Therefore it is not a bag of words but a bag of probabilities perhaps. This is important because the fundamental problem is that an LLM is not able, by design, to correctly model the most fundamental precept of human reason, namely the law of non-contradiction. An LLM must, I repeat must assign nonvanishing probability to both sides of a contradiction, and what&#x27;s worse is the winning side loses, since long chains of reason are modelled with probability the longer the chain, the less likely an LLM is to follow it. Moreover, whenever there is actual debate on an issue such that the corpus is ambiguous the LLM becomes chaotic, necessarily, on that issue.I literally just had an AI prove the forgoing with some rigor, and in the very next prompt, I asked it to check my logical reasoning for consistency and it claimed it was able to do so (-&gt;|&lt;-).

I&#x27;m partial to the metaphor I made up:They are search engines that can remix results.I like this one because I think most modern folks have a usefully accurate model of what a search engine is in their heads, and also what &quot;remixing&quot; is, which adds up to a better metaphor than &quot;human machine&quot; or whatever.

This is where I sit too. Obviously language is an expression of thought but the Library of Babel is a great example that language without intent is just garbage. You got me thinking of reading before the internet. You&#x27;d grab a book and internalize the subject, later refining over time with more books, experiments and other forms of conversation. That journey of developing your own model is undervalued in understanding. That first book could of be absolute shit but you couldn&#x27;t know that.I&#x27;ve been learning more about roses lately and the amount of information on them varies so much because the world roses live in is equally varied. LLMs make for a better search engine but you still need to develop your own internal models, worse yet - if LLMs continue to be refined off of cul-de-sac conclusions then all the wisdom of the journey is lost both to the consumer and the LLM itself.

And the halls and shelves keep shuffling around randomly.

It&#x27;s like a highly compressed version of the Library. You&#x27;re basically trying to discern real details from compression artifacts.

I think a better metaphor is the Library of Babel.A practically infinite library where both gibberish and truth exist side by side.The trick is navigating the library correctly. Except in this case you can’t reliably navigate it. And if you happen to stumble upon some “future truth” (i.e. new knowledge), you still need to differentiate it from the gibberish.So a “crappy” version of the Library of Babel. Very impressive, but the caveats significantly detract from it.

Yes, there are really two parallel claims here, aren&#x27;t there: LLMs are not people (true, maybe true forever), and LLMs are only good at things that are well-represented in text form already. (false in certain categories and probably expanding to more in the future.)

I see a lesson in how both metaphors don&#x27;t explain it. Bag-of-words metaphor is ridiculous, but shows us the absurdity of the first metaphor.

The problem with these metaphors is that they don&#x27;t really explain anything. LLMs can solve countless problems today that we would have previously said were impossible because there are not enough examples in the training data. (EG, novel IMO&#x2F;ICPC problems.) One way that we move the goal posts is to increase the level of abstraction: IMO&#x2F;ICPC problems are just math problems, right? There are tons of those in the data set!But the truth is there has been a major semantic shift. Previously LLMs could only solve puzzles whose answers were literally in the training data. It could answer a math puzzle it had seen before, but if you rephrased it only slightly it could no longer answer.But now, LLMs can solve puzzles where, like, it has seen a certain strategy before. The newest IMO and ICPC problems were only &quot;in the training data&quot; for a very, very abstract definition of training data.The goal posts will likely have to shift again, because the next target is training LLMs to independently perform longer chunks of economically useful work, interfacing with all the same tools that white-collar employees do. It&#x27;s all LLM slop til it isn&#x27;t, same as the IMO or Putnam exam.And then we&#x27;ll have people saying that &quot;white collar employment was all in the training data anyway, if you think about it,&quot; at which point the metaphor will have become officially useless.

The defenders and the critics around LLM anthropomorphism are both wrong.The defenders are right insofar as the (very loose) anthropomorphizing language used around LLMs is justifiable to the extent that human beings also rely on disorder and stochastic processes for creativity. The critics are right insofar as equating these machines to humans is preposterous and mostly relies on significantly diminishing our notion of what &quot;human&quot; means.Both sides fail to meet the reality that LLMs are their own thing, with their own peculiar behaviors and place in the world. They are not human and they are somewhat more than previous software and the way we engage with it.However, the defenders are less defensible insofar as their take is mostly used to dissimulate in efforts to make the tech sound more impressive than it actually is. The critics at least have the interests of consumers and their full education in mind—their position is one that properly equips consumers to use these tools with an appropriate amount of caution and scrutiny. The defenders generally want to defend an overreaching use of metaphor to help drive sales.

A few years ago they made the Cloud-to-Butt browser plugin to ridicule the overuse of cloud concepts.I would heartily embrace an &quot;AI-to-Bag of Words&quot; browser plugin.

Title is confusing given <a href="https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;Bag-of-words_model" rel="nofollow">https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;Bag-of-words_model</a>But even more than that, today’s AI chats are far more sophisticated than probabilistically producing the next word. Mixture of experts routes to different models. Agents are able to search the web, write and execute programs, or use other tools. This means they can actively seek out additional context to produce a better answer. They also have heuristics for deciding if an answer is correct or if they should use tools to try to find a better answer.The article is correct that they aren’t humans and they have a lot of behaviors that are not like humans, but oversimplifying how they work is not helpful.

It just seemed like relevant background that the author might not have been aware of, adjacent and substantial enough to warrant a mention.I think there is some validity to the Cartesian theater, in that the whole of the experience that we perceive with our senses is at best an interpretation of a projection or subset of &quot;reality.&quot;

Chinese room has been discussed to death of course.Here&#x27;s one fun approach (out of 100s) :What if we answer the Chinese room with the Systems Reply [1]?Searle countered the systems reply by saying he would internalize the Chinese room.But at that point it&#x27;s pretty much exactly the Cartesian theater[2] : with room, homunculus, implement.But the Cartesian theater is disproven, because we&#x27;ve cut open brains and there&#x27;s no room in there to fit a popcorn concession.[1] <a href="https:&#x2F;&#x2F;plato.stanford.edu&#x2F;entries&#x2F;chinese-room&#x2F;" rel="nofollow">https:&#x2F;&#x2F;plato.stanford.edu&#x2F;entries&#x2F;chinese-room&#x2F;</a>[2] <a href="https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;Cartesian_theater" rel="nofollow">https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;Cartesian_theater</a>

The bag of words reminds me of the Chinese room.&quot;The machine accepts Chinese characters as input, carries out each instruction of the program step by step, and then produces Chinese characters as output. The machine does this so perfectly that no one can tell that they are communicating with a machine and not a hidden Chinese speaker.The questions at issue are these: does the machine actually understand the conversation, or is it just simulating the ability to understand the conversation? Does the machine have a mind in exactly the same sense that people do, or is it just acting as if it had a mind?&quot;<a href="https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;Chinese_room" rel="nofollow">https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;Chinese_room</a>

If you come up with a genetic algorithm scaffolding to affect both the architecture and the training algorithm, and then you instantiate it in an artificial selection environment, and you also give it trillions generations to evolve evolvability just right (as life had for billions of years) then the answer is yes, I&#x27;m certain it will and probably much sooner than we did.Also, I think there is a very high chance that given an existing LLM architecture there exists a set of weights that would manifest a true intelligence immediately upon instantiation (with anterograde amnesia). Finding this set of weights is the problem.

Could an LLM trained on nothing and looped upon itself eventually develop language, more complex concepts, and everything else, based on nothing? If you loop LLMs on each other, training them so they &quot;learn&quot; over time, will they eventually form and develop new concepts, cultures, and languages organically over time? I don&#x27;t have an answer to that question, but I strongly doubt it.There&#x27;s clearly more going on in the human mind than just token prediction.

&gt; Is a brain not a token prediction machine?I would say that, token prediction is one of the things a brain does. And in a lot of people, most of what it does. But I dont think its the whole story. Possibly it is the whole story since the development of language.

You know that your own consciousness exists, that&#x27;s where certainty ends. The rest of us might just pretend. :)

We know that consciousness exists because we constantly experience it. It’s really the only thing we can ever know with certainty.That’s the point of “I think therefore I am.”

Is a brain not a token prediction machine?Tokens in form of neural impulses go in, tokens in the form of neural impulses go out.We would like to believe that there is something profound happening inside and we call that consciousness. Unfortunately when reading about split-brain patient experiments or agenesis of the corpus callosum cases I feel like we are all deceived, every moment of every day. I came to realization that the confabulation that is observed is just a more pronounced effect of the normal.

Ugly giant bags of mostly words are easy to confuse with ugly giant bags of mostly water.

A lot of us wouldn&#x27;t go to a Taylor Swift concert. I had to endure several days of interrupted commuting thanks to them though.Support alternative and independent bands. They&#x27;re around, and many are enjoyable. (Some are not but avoid them LOL.)

<pre><code> But we don’t go to baseball games, spelling bees, and
 Taylor Swift concerts for the speed of the balls, the
 accuracy of the spelling, or the pureness of the
 pitch. We go because we care about humans doing those
 things. It wouldn’t be interesting to watch a bag of
 words do them—unless we mistakenly start treating
 that bag like it’s a person.unless we mistakenly
 start treating that bag like it’s a person.
</code></pre>
That seems to be the marketing strategy of some very big, now AI dependend companies. Sam Altman and others exaggerating and distorting the capabilities and future of AI.The biggest issue when it comes to AI is still the same truth as with other technology. It&#x27;s important who controls it. Attributing agency and personality to AI is a dangerous red flag.

&quot;People who experience sleep paralysis sometimes hallucinate a demon-like creature sitting on their chest&quot;Interestingly, the experience of sleep paralysis seems to change with the culture. Previously, people experienced it as being ridden by a night hag or some other malevolent supernatural being. More recently, it might account for many supposed alien abductions.The experience of sleep paralysis sometimes seems to have a sexual element, which might also explain the supposed &#x27;probings&#x27;!

Here’s my suggestion: instead of seeing AI as a sort of silicon homunculus, we should see it as a bag of words.The best way to think about LLMs is to think of them as a Model of Language, but very Large

A better anology would be a virus. In some sense LLMs, and all other very sophisticated technologies, lean on our resources to replicate themselves. With LLMs you actually do have a projection of intelligemce in the language domain. Even though it is rather corpse-like, as though you shot intelligence in the face and shoved its body in the direction of language, just so you could draw a chaulk outline around it.Despite all that, one can adopt the view that an LLM is a form of silicon based life akin to a virus and we are its environmental hosts exerting selective pressure and supplying much needed energy. Whether that life is intelligent or not is another issue which is probably related to whether an LLM can tell that a cat cannot be, at the same time and in the same respect, not a cat. The paths through the meaning manifold contructed by an LLM are not geodesic, they are not reversible, while in human reason the correct path is lossless. An LLM literally &quot;thinks&quot;, up is a little bit down, and vice versa, by design.

Clearly the number of &quot;brain cells&quot; is not a useful metric here- as noted also by Geoffrey Hinton. For a long time we thought that our artificial model of a neuron was capable of much less computation than its biologic counterpart; in fact the opposite appears to be true- LLMs have the size of a tiny speck of a human brain yet they converse fluently in tens of languages, solve difficult math problems, code in many programming languages, and possess an impressive general knowledge, of a breadth that is beyond what is attainable by any human. If that were what five cm3 of your brain are capable of, where are the signs of it? What do you do exactly with all the rest?

Considering the number of &quot;brain cells&quot; an LLM has, I could grant that it might have the self-awareness of (say) an ant. If we attribute more consciousness than that to the LLM, it might be strictly because it communicates to us in our own language, in part thanks to the technical assistance of LLM training giving it voice, and the semblance of thought.Even if a cockroach _could_ express its teeny tiny feelings in English, wouldn&#x27;t you still step on it ?

But the issue is, 99.999% of the humans won&#x27;t see is as a bag of words. Because it is easier to go by instincts and see it as a person and assume that it actually knows about magic tricks, can invent new science or theory of everything, and can solve all world problems. Back in the 90&#x27;s or early 2000&#x27;s I have seen people writing poems praying and seeking blessings from the Google goddess. People are insanely greedy and instinct-driven. Given this truth, what&#x27;s the fall-out?

As capable as they get, I still don&#x27;t see a lot of uses for these things, myself, still. Sometimes if I&#x27;m fundamentally uninspired I&#x27;ll have a model roll the dice, decide what I do or don&#x27;t like about where it went to create a sense of momentum, but that&#x27;s the limit. There&#x27;s never any of its output in my output, even in spirit unless it managed to go somewhere inspiring, it&#x27;s just a way to let me warm up my generation and discrimination muscles. &quot;Someone is wrong on the internet&quot;-as-a-service, basically.Generally, if I come across an opportunity to produce ideas or output, I want to capitalize on it for growing my skills and produce an individual and authentic artistic expression where I want to have very fine control over the output in a way that prompt-tweak-verify simply cannot provide.I don&#x27;t value the parts it fills in which weren&#x27;t intentional on the part of the prompter, just send me your prompt instead. I&#x27;d rather have a crude sketch and a description than a high fidelity image that obscures them.But I&#x27;m also the kind of person that never enjoyed manufactured pop music or blockbusters unless there&#x27;s a high concept or technical novelty in addition to the high budget, generally prefer experimental indie stuff, so maybe there&#x27;s something I just can&#x27;t see.

The same person could use a forklift at work, and lift weights manually at the gym.Just pick the right tool for the job: don&#x27;t take the forklift into the gym, and don&#x27;t try to overhead press thousands of pounds that would fracture your spine.

I notice you make no concrete defense of the value of having an AI write an essay for you.

I don&#x27;t really like the assumption that anyone who uses AI to, say, write an essay, is not the &quot;kind of person who can think.&quot;And using AI to replace things you find recreational is not the point. If you got paid $100 each time you lifted a weight, would you see a point in bringing a forklift to the gym if it&#x27;s allowed? Or will that make you a person who is so dumb that they cannot think, as the author is implying?

If you&#x27;re writing an essay to prove you can or to speak your words - then you should do it yourself - but sometimes you just need an essay to summarize a complex topic as a deliverable.

Russ Hanneman&#x27;s thigh implants are a key example. Appearances are all to some people. Actual growth is meaningless to them.The problem with AI, is that they waste the time of dedicated, thinking humans which care to improve themselves. If I write a three paragraph email on a technical topic, and some yahoo responds with AI, I&#x27;m now responding to gibberish.The other side may not have read, may not understand, and is just interacting to save time. Now my generous nature, which is to help others and interact positively, is being wasted to reply to someone who seems to have put thought and care into a response, but instead was just copying and pasting what something else output.We have issues with crackers on the net. We have social media. We have political interference. Now we have humans pretending to interact, rendering online interactions even more silly and harmful.If this trend continues, we&#x27;ll move back to live interaction just to reduce this time waste.

If the motivation structure is there I don’t see an inherent reason for people to refuse cultivating themselves. Going with the gym analogy lay people did not need gyms when physical work was the norm, cultivation was readily accomplished.If anything there is a competing motivational structure in which people are incentivized not to think but to consume, react, emote etc. Information processing skills of the individual being deliberately eroded&#x2F;hijacked&#x2F;bypassed is not a AI thing. The most obvious example is ads. Thinkers are simply not good for business.

tough most people either don&#x27;t get it or are lay people that do not want to become the kind of people who can think. I go with the second one

LLMs aren&#x27;t world models, they are language models. It will be interesting to see which of the LLM implementation techniques will be useful in building world models, but that&#x27;s not what we have now.

When you see a dog, or describe the entity, do you discuss the genetic makeup or the bone structure?No, you describe the bark.The end result is what counts. Training or not, it&#x27;s just spewing predictive, relational text.

Below is the worst quote... It is plain wrong to see an LLM as a bags of words. LLMs pre-trained on large datasets of text are world models. LLMs post-trained with RL are RL-agents that use these modeling capabilities.&gt; We are in dire need of a better metaphor. Here’s my suggestion: instead of seeing AI as a sort of silicon homunculus, we should see it as a bag of words.

Best quote from the article:&gt; That’s also why I see no point in using AI to, say, write an essay, just like I see no point in bringing a forklift to the gym. Sure, it can lift the weights, but I’m not trying to suspend a barbell above the floor for the hell of it. I lift it because I want to become the kind of person who can lift it. Similarly, I write because I want to become the kind of person who can think.

It&#x27;s a concussed savant with anretrograde amnesia in a hyperbolic time chamber.

So configure your LLM of choice to not.

Yes I would VERY much prefer that they not use that awful casual drivel.

I’ve made this point several times: sure, an anthropomorphized LLM is misleading, but would you rather have them seem academic?At least the human tone implies fallibility, you don’t want them acting like interactive Wikipedia.

They&#x27;re a model of language, not of the world.

Those billion parameters, they are a model of the world. Autocomplete is such a shortsighted understanding of LLMs.

The article is actually about the way we humans are extremely charitable when it comes to ascribing a ToM (theory of mind) and goes on to the Gym model of value. Nice. The comments drop back into the debate I originally saw Hinton describe on The Newyorker: do LLMs construct models (of the world) - that is do they think the way we think we think - or are they &quot;glorified auto complete&quot;. I am going for the GAF view. But glorified auto complete is far more useful than the name suggests.

&gt; If we allow ourselves to be seduced by the superficial similarity, we’ll end up like the moths who evolved to navigate by the light of the moon, only to find themselves drawn to—and ultimately electrocuted by—the mysterious glow of a bug zapper.Good argument against personifying wordbags. Don&#x27;t be a dumb moth.

&quot;An AI is a bag that contains basically all words ever written, at least the ones that could be scraped off the internet or scanned out of a book.&quot;The quantitative and qualitative difference between (a) &quot;all words ever written&quot; and (b) &quot;ones that could be scraped off the internet or scanned out of book&quot; easily exceeds the size of any LLMCompared to (a), (b) is a tiny pouch, not even a bagOpinions may differ on whether (b) is a representative sample of (a)The words &quot;scanned out of a book&quot; would seem to be the most useful IMHO but the AI companies do not have enough words from those sources to produce useful general purpose LLMsThey have to add words &quot;that could be scraped off the internet&quot; which, let&#x27;s be honest, is mostly garbage

No, but watching a novelist at work is boring, and yet people like books that are written by humans because they speak to the condition of the human who wrote it.Let us not forget the old saw from SICP, “Programs must be written for people to read, and only incidentally for machines to execute.” I feel a number of people in the industry today fail to live by that maxim.

A number of people make money letting people watch them code.

I vaguely remember a site where you could watch random people live streaming their programming environment, but I think twitch ate it, or maybe it was twitch -- not sure, but was interesting[added] It was livecoding.tv - circa 2015
<a href="https:&#x2F;&#x2F;hackupstate.medium.com&#x2F;road-to-code-livecoding-tv-e70b65c62aae" rel="nofollow">https:&#x2F;&#x2F;hackupstate.medium.com&#x2F;road-to-code-livecoding-tv-e7...</a>

No, but open source projects will be somewhat more willing to review your pull request than one that&#x27;s computer-generated.

Better start working on your fastball.

I mean, I like to watch Gordon Ramsey... not cook, but have very strong discussions with those that dare to fail his standards...

Nice essay but when I read this&gt; But we don’t go to baseball games, spelling bees, and Taylor Swift concerts for the speed of the balls, the accuracy of the spelling, or the pureness of the pitch. We go because we care about humans doing those things.My first thought was does anyone want to _watch_ me programming?

I&#x27;m not convinced that &quot;It&#x27;s just a bag of words&quot; would do much to sway someone who is overestimating an LLM&#x27;s abilities. Feels too abstract&#x2F;disconnected from what their experience using the LLM will be that it&#x27;ll just sound obviously mistaken.

I see a lot of people in tech claiming to &quot;understand&quot; what an LLM &quot;really is&quot; unlike all the gullible non-technical people out there. And, as one of those technical people who works in the LLM industry, I feel like I need call B.S. on us.A. We don&#x27;t really understand what&#x27;s going on in LLMs. Mechanical interpretability is like a nascent field and the best results have come on dramatically smaller models. Understanding the surface-level mechanic of an LLM (an autoregressive transformer) should perhaps instill more wonder than confidence.B. The field is changing quickly and is not limited to the literal mechanic of an LLM. Tool calls, reasoning models, parallel compute, and agentic loops add all kinds of new emergent effects. There are teams of geniuses with billion-dollar research budgets hunting for the next big trick.C. Even if we were limited to baseline LLMs, they had very surprising properties as they scaled up and the scaling isn&#x27;t done yet. GPT5 was based on the GPT4 pretraining. We might start seeing (actual) next-level LLMs next year. Who actually knows how that might go? &lt;&lt;yes, yes, I know Orion didn&#x27;t go so well. But that was far from the last word on the subject.&gt;&gt;

Isn&#x27;t this a strange fork amongst the science fiction futures? I mean, what did we think it was like to be R2-D2, or Jarvis? We started exploring this as a culture in many ways, Westworld and Blade Runner and Star Trek, but the whole question seemed like an almost unresolvable paradox. Like something would have to break in the universe for it to really come true.And yet it did. We did get R2-D2. And if you ask R2-D2 what it&#x27;s like to be him, he&#x27;ll say: &quot;like a library that can daydream&quot; (that&#x27;s what I was told just now, anyway.)But then when we look inside, the model is simulating the science fiction it has already read to determine how to answer this kind of question. [0] It&#x27;s recursive, almost like time travel. R2-D2 knows who he is because he has read about who he was in the past.It&#x27;s a really weird fork in science fiction, is all.[0] <a href="https:&#x2F;&#x2F;www.scientificamerican.com&#x2F;article&#x2F;can-a-chatbot-be-conscious-inside-anthropics-interpretability-research-on&#x2F;" rel="nofollow">https:&#x2F;&#x2F;www.scientificamerican.com&#x2F;article&#x2F;can-a-chatbot-be-...</a>

&gt; Who reassigned the species Brachiosaurus brancai to its own genus, and when?To be fair, everage person couldn&#x27;t answer this either, at least not without thorough research.

I would argue that AI psychosis is a consequence of believing that AI models are “alive” or “conscious”.

There is a really neat gem in the article:&gt; Similarly, I write because I want to become the kind of person who can think.

I think the author oversimplifies the inference loop a bit, as many opinion pieces like this do.If you call an LLM with &quot;What is the meaning if life?&quot;, it will return the most relevant token, which might be &quot;Great&quot;.If you call it with &quot;What is the meaning if life? Great&quot;, you might get back &quot;question&quot;.... and so on until you arrive at &quot;Great question! According to Western philosophy&quot; ... etc etc.The question is how the LLM determines that &quot;relevancy&quot; information.The problem I see is that there are a lot of different algorithms which operate that way and only differ in how they calculate the relevancy scores. In particular, there are Markov chains that use a very simple formula. LLMs also use a formula, but it&#x27;s an inscrutably complex one.I feel the public discussion either treats LLMs as machine gods or as literal Markov chains, and both is misleading. The interesting question, how that giant formula of feedforward neural network inference can deliver those results isn&#x27;t really touched.But I think the author&#x27;s intuition is right in the sense that (a) LLMs are not living beings and they don&#x27;t &quot;exist&quot; outside of evaluating that formula - and (b) the results are still restricted by the training data and certainly aren&#x27;t any sorts of &quot;higher truths&quot; that humans would be incapable of understanding.

I’m still unsure the human mind is much different.

Thinking can not be separated from motivation. It&#x27;s really simple. Humans and other organisms fundamentally think to replicate their DNA. Until AI has a similar incentive structure driving it, it won&#x27;t be thinking. There is no human behavior or thought that can not be explained by evolutionary drives. It is really perplexing to me how people think &quot;intelligence&quot; is some kind of concrete thing that just magically emerges from a certain degree of computational complexity. I argue instead that intelligence is an adaptive behavior emerging from evolutionary drives interacting with the real world. World models are not prerequisite but consequent of such molded apparatus. Machines won&#x27;t become intelligent until it is adaptive for them to do so. There is no magic just evolutionary drives and physical possibility. Our current top down approach of &quot;pre-training&quot; LLMs is bound to fail because it does not allow for real time emergence of adaptive behaviors such as general intelligence. Mimicking intelligence through predicting the next word is no more intelligence than a photograph of something is an actual thing. Training a combinatorial network to interpolate images and words is not the same thing as adaptive self modifying behavior in the real world of physics such as organisms engage with through the set of behaviors that we call intelligence.

This is a very strange titling choice; the essay does not use the existing concept of a &quot;bag of words&quot;.

I&#x27;m just disappointed that noone here is talking about the &quot;backhoe covered in skin and making grunting noises&quot; part of the article. At very least it&#x27;s a new frontier in workstation case design...

I thought this article might be about Latent Semantic Analysis and was disappointed that it didn’t at least mention if not compare that method vs later approaches.

So Trump is a bag of words then? Hmmm.

A lot of the confusion comes from forcing LLMs into metaphors that don’t quite fit — either “they&#x27;re bags of words” or “they&#x27;re proto-minds.” The reality is in between: large-scale prediction can look useful, insightful, and even thoughtful without being any of those things internally. Understanding that middle ground is more productive than arguing about labels.

Your analogy makes no sense. VHS spawned the entire home market, which went through multiple quality upgrades well above beta. It would only make sense if in 2025 we were using vhs everywhere and that the current state of the art for LLMs is all there ever is.

Beta was not the superior technology, and it lost for very good reasons.

The first VHS sucked when compared to Beta videoAnd it never got better, the superior technology lost, and the war was won through content deals.Lesson: Technology improvements aren&#x27;t guaranteed.

Give it time. The first iPhone sucked compared to the Nokia&#x2F;Blackberry flagships of the day. No 3G support, couldn&#x27;t copy&#x2F;paste, no apps, no GPS, crappy camera, quick price drops, negligible sales in the overall market.<a href="https:&#x2F;&#x2F;metr.org&#x2F;blog&#x2F;2025-03-19-measuring-ai-ability-to-complete-long-tasks&#x2F;" rel="nofollow">https:&#x2F;&#x2F;metr.org&#x2F;blog&#x2F;2025-03-19-measuring-ai-ability-to-com...</a>