Two big ways in which human intelligence is different from LLM intelligence are:
1) human intelligence makes no sharp distinction between training and generation. Every time you ask a human a question it modifies its neural structure a little.
2) continuous operation: human intelligence deals with a continuous stream of multimedia data for sixteen hours a day and starts hallucinating when deprived of it.
There's also the fact that you can't branch or roll back human intelligence, but this is something most sci-fi novels tackle when discussing mind uploading first.
Are these two differences critical aspects of human intelligence or unfortunate limitations of its biological hardware? I do not know. If we somehow manage to simulate a human brain on silicon, we will get "computer" intelligence that learns like a human, but will we have to simulate the whole virtual world for it 16/7 and let it sleep for eight hours each day just to stop it from going mad?
Or will it be cheaper to fork and kill an uploaded math genius a billion times, pumping the same recycled sensory data into his or her mind, slipping a question into the auditory data, getting the answer and then switching the simulation off and trashing the copy? Will we consider this a bigger atrocity than doing the same to an LLM right now in 2026?
Structurally a transformer model is so unrelated to the shape of the brain there's no reason to think they'd have many similarities. It's also pretty well established that the brain doesn't do anything resembling wholesale SGD (which to spell it is evidence that it doesn't learn in the same way).
>Structurally a transformer model is so unrelated to the shape of the brain there's no reason to think they'd have many similarities.
Substrate dissimilarities will mask computational similarities. Attention surfaces affinities between nearby tokens; dendrites strengthen and weaken connections to surrounding neurons according to correlations in firing rates. Not all that dissimilar.
Given the large visible differences in behavior and construction, akin to the difference between a horse and a pickup truck, I would ask the reverse question: In what ways do LLMs meet the definition of having consciousness and agency?
Veering into the realm of conjecture and opinion, I tend to think a 1:1 computer simulation of human cognition is possible, and transformers being computationally universal are thus theoretically capable of running that workload. That being said, that's a bit like looking at a bird in flight and imagining going to the moon: only tangentially related to engineering reality.
> In what ways do LLMs meet the definition of having consciousness and agency?
Agency: an ability to make decisions and act independently. Agentic pipelines are doing this.
Consciousness: something something feedback[1] (or a non-transferable feeling of being conscious, but that is useless for the discussion). Recurrent Processing Theory: A computation is conscious if it involves high-level processed representations being fed back into the low-level processors that generate it.
Tokens are being fed back into the transformer.
> that's a bit like looking at a bird in flight and imagining going to the moon: only tangentially related to engineering reality.
Is it? Vacuum of space is a tangible problem for aerodynamics-based propulsion. Which analogous thing do we have with ML? The scaled-up monkey brain[2] might not qualify as the moon.
These questions really vex me. The appearance of intelligence is almost orthogonal to "consciousness and agency." If a human has a stroke and forgets how to speak, or never learns, or has some severe form of learning disorder, they still have exactly the same rich inner life full of subjective qualititative experience known only to them as the rest of us. Similar to an array of GPUs. If you remove the text encodings from the rest of the computing system it is a part of, outputs will appear as gibberish to you and it will no longer appear to be intelligent at all, but whatever is happening at the level of electrons meeting silicon would still be exactly the same. If it's having conscious experience at all, it should be having it regardless of whether the outputs it computes are interpreted as text or as textures on a game background.
I just don't see why "I can talk to it now" changes anything. We don't give humans less moral consideration when they're dreaming, hallucinating, tripping on LSD. The brain is just as conscious when it's having nothing but completely abstract nonsense thoughts as when it's writing The Republic.
I understand why it feels different to people. Shit, this thing can talk to me; maybe it's alive and I should treat it like such. But that's a conservative reaction to a black box known only by its behavior. The problem is these things are not actually black boxes. We don't understand the functions being computed or we'd just hard-code them and not need statistical learning techniques, but we do understand how computers work. We know process state is saved off and restored billions of times per second because of context switching. We know that state is simply a stored byte sequence that can be copied, backed up, restored endlessly. Servers and computing hardware can be destroyed but software cannot and LLMs are software. It's not at all like a brain. There are animals that go into various levels of reduced or suspended function that appear like dormancy, but there is no stream of personal subjective experience that can survive the complete destruction of its own physical body. The fact that it pays off evolutionarily to tacitly encode that reality into our instincts at an extremely deep, core level is why we have fear and pain in the first place, to nudge us toward predictive modeling of the world that keeps us alive, able to find food, and able to reproduce. Software needs none of that. There is no reason whatsoeve that, assuming a processor has subjective experience, that the subjective experience of having some gates fire versus others gets interpreted by humans programmers as "loss" and "training" and some is numerically approximating a PDE solution. Why should those feel different to the machine when the firing patterns are exactly the same and only the human interpretation of the output is different?
It just feels like a vast, vast category error for people to be speculating about machine consciousness and moralizing about how we "treat" software systems.
If platonic representation hypothesis holds across substrates, then it might matter very little, in the end. It holds across architectures in ML, empirically.
The crowd of "backpropagation and Hebbian learning + predictive coding are two facets of the very same gradient descent" also has a surprisingly good track record so far.
I don't know which direction you're going with this, but predictive coding has a pretty obvious advantage when it comes to continuous learning. Since predictive coding primarily encodes errors, it can distinguish between known and novel data and therefore reduce the damaging effects of catastrophic forgetting by having a very obvious regularisation scheme for avoiding forgetting.
It is hypothesized that the human brain uses predictive coding for obvious biological reasons such as energy efficiency (spiked error coding means only differences need to be transmitted) and biological plausibility (only local communication is permitted, meanwhile backpropagation is a global algorithm).
Transformers have a thing called a context window which doesn't really have a biological equivalent, since the brain has a fixed size and doesn't grow or shrink in response to the amount of information being processed.
LLMs consist of several layers that communicate at fixed points between the layers, whereas neurons can form feedback loops and communicate with any neighbour in any direction.
Humans do not consume or produce tokenized information. The brain controls the human body which is a biomechanical system. Spoken or written language is the result of controlling muscles via an internal model of the biomechanical system, not something that was designed via a software tokenizer that compresses character sequences.
The equivocation just doesn't seem appropriate. Try again in 2050.
For starters, natural brains have the innate ability to differentiate between things that it knows and things that it have no possibility of knowing...
Does it matter? Evolution is the brain's very own "pre-training". Hundreds of millions of years of priors hardwired.
We can do that for AIs too - pre-train on pure low Kolmogorov complexity synthetics. The AI then "knows things" before it sees any real data. Advantageous sometimes. Hard to pick compute efficient synthetics though.
Because today it will be used as a first responder.
Tomorrow a police officer will suggest that these drones (that we are already using successfully) could be very useful for checking up on that "dangerous" neighborhood.
In the city I live in, there is a chronic shortage of police officers and a lot of dangerous neighborhoods. If a drone could be used to do the same or a substantially-similar policing job in those neighborhoods that a human cop would, without having to pay for a human cop (not just their salary in and of itself, but also in terms of making the police department a place people are willing to work for at that salary), this would be an improvement to public safety and quality of life.
Also remote-operated drones don't need to fear that they will get suddenly shot or stabbed to death by a criminal suspect whose potential crimes they are investigating, like a human cop does; and this would itself have some beneficial effects on policing.
The "same or substantially-similar policing job" is the key to this argument. Which it can't. A drone can't de-escalate a tense conflict between neighbors, it can't provide traffic redirection after an accident, or even rescue a kitten from a tree.
It can't be a calm, reassuring presence, offer a kind smile, or give directions. It only disconnects the police force from the policed community. Its presence will only raise tensions and paranoia. And that's with unarmed drones!
Hi, I'm in Denver. They're already doing this over on Colfax. It's a significant change vs the existing halo cameras, because they use the drones to follow people.
I'm not really worried about the police. There's mountains of well reinforced legal precedent restricting them. Sure they have violence, but they kind of need to show up to do that. All the other stuff they do runs up against your rights which are really well established. Even the "civil" traffic stuff is pretty hard fought, comparatively.
Every other civil enforcer can basically fine you on a whim and then your appeal goes into a system that makes jim crow look impartial. So yeah, I'm not worried about the cops. I'm worried about the zoning office "fixing" a budget shortfall by fining people for unpermitted kiddie pools or whatever and in the 10yr it takes to get smacked down in court they'll have stolen the property of a ton of people. I'm dead serious. However bad you think it could be reality is worse. These non-LEO departments make the most sloppy podunk sheriff's office look like the FBI.
The actual problem is that there's a law requiring a permit for kiddie pools with a fine attached, that was mostly unenforced previously, so a lot of people got into the habit of breaking it.
I disagree. I like to read articles and explore Show HN posts, but in the past 6 months I’ve wasted a lot of time following HN links that looked interesting but turned out to be AI slop. Several Show HN posts lately have taken me to repos that were AI generated plagiarisms of other projects, presented on HN as their own original ideas.
Seeing comments warning about the AI content of a link is helpful to let others know what they’re getting into when they click the link.
For this article the accusations are not about slop (which will waste your time) but about tell-tell signs of AI tone. The content is interesting but you know someone has been doing heavy AI polishing, which gives articles a laborious tone and has a tendency to produce a lot of words around a smaller amount of content (in other words, you’re reading an AI expansion of someone’s smaller prompt, which contained the original info you’re interested in)
Being able to share this information is important when discussing links. I find it much more helpful than the comments that appear criticizing color schemes, font choices, or that the page doesn’t work with JavaScript disabled.
> you’re reading an AI expansion of someone’s smaller prompt, which contained the original info you’re interested in
This got me thinking: what if LLMs are used to do the opposite? To condense a long prompt into a short article? That takes more work but might make the outcome more enjoyable as it contains more information.
> This got me thinking: what if LLMs are used to do the opposite? To condense a long prompt into a short article? That takes more work but might make the outcome more enjoyable as it contains more information.
You're fighting an uphill battle against the inherent tendency to produce more and longer text. There's also the regression to the mean problem, so you get less information (and more generic) even though the text is shorter.
Yes. These HN guidlines already basically cover it:
> Please don't post shallow dismissals, especially of other people's work. A good critical comment teaches us something.
> Please don't complain about tangential annoyances—e.g. article or website formats, name collisions, or back-button breakage. They're too common to be interesting.
Its not a person's work. It reads like an LLM's work. If you can't be bothered to write an article yourself, its incredibly arrogant to ask me to read it.
Speaking of the HN guidelines, they also say this:
> Don't post generated comments or AI-edited comments. HN is for conversation between humans.
> Yes. These HN guidlines already basically cover it:
>> Please don't post shallow dismissals, especially of other people's work. A good critical comment teaches us something.
>> Please don't complain about tangential annoyances—e.g. article or website formats, name collisions, or back-button breakage. They're too common to be interesting.
HN has gotten to the point where it’s not even worth clicking the link because of course it’s ai slop.
There is some real content in the haystack, but we almost need some kind of curator to find and display it rather than a vote system where most people vote on the title alone.
If you’re looking for a place that surfaces only human-written content regardless of whether it’s interesting, rather than interesting content regardless of how it was written, HN is not the place.
There might be a market for your alternative though. Should be easy enough to build with Claude Code.
If the content was interesting, the author would've written about it himself.
By asking AI to write the article for you, you're asserting that the subject matter is not interesting enough to be worth your time to write, so why would it be worth my time to read?
I know the author personally. He's hardly the type of person to publish AI slop. Read his other articles and watch his talks, this is very much Henry's literary style.
I remember trying out some voice-to-text around 2002 that I believe was included with Windows XP.. or maybe Office?
You had to go through some training exercises to tune it to your voice, but then it worked fairly well for transcription or even interacting with applications.
Ah yes, getting access to your own data would be a massive problem, can you imagine such a world?! /s
Such data should be put in (or encrypted by) the hardware-backed keystore. You get to have full access to what the OS does, including seeing what data gets passed into this secure element for encryption or signing (you retain visibility and control), and yet secrets can't be leaked to you or an attacker who tries to extract those secrets
See e.g. your bank card: it's yours, you can choose where to stick it and what transactions it authorizes, but you can't get at the token that serves as proof of possession nor reset the PIN attempts counter. Your phone('s banking app) could work in the same way and has the hardware on board that makes this possible. So you see, it's a choice that you don't get to see what apps are doing and people are scared into believing that access to their own phone is bad. It's a matter of conflicting incentives on the vendor side, not technical risk
There is an API for backing up all app data that requires authorization. This is different from giving the user root, so any malicious can back up all app data at any time.
Oh, that useless thing. I was very confused about something which can "app data that requires authorization" (thought maybe it's some Google service that extracts your secrets for device migrations) but you just mean the old adb backup that the security industry (that I'm part of, and fighting from within :p) destroyed in the name of people's own good
Like, yes this exists, but it doesn't back up half the things you need :(
Anecdotal, but I've spoken with many taxi and ride-share drivers, and my impression is that their decision to seek out and continue that line of work is almost always driven by outside economic considerations. I've never heard someone base their decision on their ability to perform the job.
Exactly - I'm thinking the bad spatial navigators have a higher probability of washing out of driving and pursue some other career. They may not say "I'm bad at figuring out where I am", but the economics of the job are just a little bit worse for these people.
Why regulate? Look at the failure that is the "war on drugs".
The solution is education. The government should be educating society and especially parents on how to protect their children.
Education worked to cut cigarette use, and is starting to lower alcohol consumption as well. It can work for social media without all the negative impacts on civil liberties that come with regulations.
Ah, yes, it was education, and certainly not strong regulation of how cigarettes are produced, marketed, and sold. Educating the youth is why vaping and smoking is not cool anymore.
The government can do a whole lot more than embarrass CEOs and powerful people they don't like. Look at how China controls its tech CEOs by making them disappear until their views align.
2. You're on a site with a bunch of programmers who regularly use weird words for stuff that already has a name. Reading through HN is wading through a swamp of made up names and tech neologisms, you're just used to it already. I once told a software guy that our team's SWEs had migrated away from React and Node to Stork.JS and Blackadder. He nodded like that meant anything.
devs have really got to start using NSA style naming conventions where they use the Joycean compound with random stuff that sounds cool e.g. BANNANADAIQUIRI or FOXACID.
Have you ever actually heard it used in conversation or writing where the speaker's intention was a meaning that included people outside of the USA? I haven't.
What makes you think natural brains are doing something so different from LLMs?