The architectute of LLM neurons is an incredibly simplified and bayesean in nature. It can interact with other neurons and maintain an activation weight and some other parameters, but it's not a physical object.
Biological neurons are independent organisms capable of self organization, migration, communication, and interaction either directly with the world or abstractly through nerve senses.
The general concept of LLM architecture (something that has been around for decades now, I think all the way back to the 50s) is a reduced and simplified facsimile of that biological function.
I think because we interact with LLMs through an interface that's been basically exclusively limited to other human interactions forever, it can be easy to forget that they aren't the system they're emulating. They're no more a sentient machine than a dialysis machine is a kidney.
The very first chatbots had a similar effect on users, even though those were more expert machines and didn't use large natural language training sets. And in the end I believe that replicating the biological function of kidneys and livers and lungs is a much more important step in human history than replicating the function of the mind. Especially because any simulation of the mind trained on a natural language dataset is not something that can ever help us.
It will at best begin to placate us, we will have a mirror held up to ourselves because the training of the model isn't done for the sake of creating intelligence, but for making something that resembles intelligence enough to make us happy. The training is done entirely on our terms.
And again, LLMs and more broadly statistical models do have tons of uses, and using them to discover hidden patterns in data that would take forever for a human to find by hand. They can also be used in planning to simplify economic forecasting and detect possible shortages and future labor allocation needs (this was done by hand for GOSPLAN, and TANS proposes using these models for cybernetic planning systems).
But it's still just a machine, it's still just a programming language. A language where the syntax is a giant matrix of floating point numbers and relationship rules, but a programming language nonetheless.
Idk if we can ever see eye to eye here.. if we were to somehow make major advances in scanning and computer hardware to the point where we could simulate everything that biologists currently consider relevant to neuron behavior and we used that to simulate a real person's entire brain and body would you say that A) it wouldn't work at all, the simulation would fail to capture anything about human behavior, B) it would partly work, the brain would do some brain like stuff but would fail to capture our full intelligence, C) it would capture human behaviors we can measure such as the ability to converse but it wouldn't be conscious, or D) something else?
Personally I'm a hard core materialist and also believe the weak version of the church turing thesis, I'm quite strongly wedded to this opinion, so the idea that being made of one thing vs another or being informational vs material says anything about the nature of a mind is quite foreign. I'm aware that this isn't shared by everyone but I do believe it's the most common perspective inside the hard sciences, though not universal, Roger Penrose is a brilliant physicist who doesn't see this way.
I understand your perspective, and I don't necessarily disagree or think that there's anything innately spiritual or unique about biological intelligence. I do also agree that you could hypothetically scan every aspect of a brain or build a system that exactly mimics the behavior of neurons and probably pretty accurately recreate human intelligence.
I really think our only disconnect is that I don't think the current LLM model is anything close to complex or developed enough to be considered that.
That's a perfectly reasonable position, the question of how complex a human brain is compared with the largest NNs is hard to answer but I think we can agree it's a big gap. I happen to think we'll get to AGI before we get to human brain complexity, parameter wise, but we'll probably also need at least a couple architectural paradigms on top of transformers to compose one. Regardless, we don't need to achieve AGI or even approach it for these things to become a lot more dangerous, and we have seen nothing but accelerating capability gains for more than a decade. I'm very strongly of the opinion that this trend will continue for at least another decade, there's are just so many promising but unexplored avenues for progress. The lowest of the low hanging fruit has been, while lacking in nutrients, so delicious that we haven't bothered to do much climbing.
Would love to see more development in this field, but it's clear that you don't need to have complex or biologically accurate systems to manipulate other humans. This fact alone means that machine learning models will never be advanced beyond that basic goal under capitalism.
They've been used for economic modeling and stock forecasting for decades now, since the 80s, and the modern implementations of these systems is nothing more than the application of those failed financial modeling systems on human social interactions. Something that wasn't possible before because until the widespread adoption of the internet, there just wasn't enough digital communication data to feed into them.
Since these systems are not capable of self development that isn't a negative feedback loop, they literally can't improve without more and more data from different human activity being fed to them.
That alone shows that they aren't a new form of intelligence, but instead a titular interface for the same type of information you used to be able to get with "dumb" indexing engines.
There's a reason that search engine companies are the primary adopters of this technology, and it's because they already have been using it for 20+ years in some form, and they have access finally to enough indexed information to make them appear intelligent.
The architectute of LLM neurons is an incredibly simplified and bayesean in nature. It can interact with other neurons and maintain an activation weight and some other parameters, but it's not a physical object.
Biological neurons are independent organisms capable of self organization, migration, communication, and interaction either directly with the world or abstractly through nerve senses.
The general concept of LLM architecture (something that has been around for decades now, I think all the way back to the 50s) is a reduced and simplified facsimile of that biological function.
I think because we interact with LLMs through an interface that's been basically exclusively limited to other human interactions forever, it can be easy to forget that they aren't the system they're emulating. They're no more a sentient machine than a dialysis machine is a kidney.
The very first chatbots had a similar effect on users, even though those were more expert machines and didn't use large natural language training sets. And in the end I believe that replicating the biological function of kidneys and livers and lungs is a much more important step in human history than replicating the function of the mind. Especially because any simulation of the mind trained on a natural language dataset is not something that can ever help us.
It will at best begin to placate us, we will have a mirror held up to ourselves because the training of the model isn't done for the sake of creating intelligence, but for making something that resembles intelligence enough to make us happy. The training is done entirely on our terms.
And again, LLMs and more broadly statistical models do have tons of uses, and using them to discover hidden patterns in data that would take forever for a human to find by hand. They can also be used in planning to simplify economic forecasting and detect possible shortages and future labor allocation needs (this was done by hand for GOSPLAN, and TANS proposes using these models for cybernetic planning systems).
But it's still just a machine, it's still just a programming language. A language where the syntax is a giant matrix of floating point numbers and relationship rules, but a programming language nonetheless.
Idk if we can ever see eye to eye here.. if we were to somehow make major advances in scanning and computer hardware to the point where we could simulate everything that biologists currently consider relevant to neuron behavior and we used that to simulate a real person's entire brain and body would you say that A) it wouldn't work at all, the simulation would fail to capture anything about human behavior, B) it would partly work, the brain would do some brain like stuff but would fail to capture our full intelligence, C) it would capture human behaviors we can measure such as the ability to converse but it wouldn't be conscious, or D) something else?
Personally I'm a hard core materialist and also believe the weak version of the church turing thesis, I'm quite strongly wedded to this opinion, so the idea that being made of one thing vs another or being informational vs material says anything about the nature of a mind is quite foreign. I'm aware that this isn't shared by everyone but I do believe it's the most common perspective inside the hard sciences, though not universal, Roger Penrose is a brilliant physicist who doesn't see this way.
I understand your perspective, and I don't necessarily disagree or think that there's anything innately spiritual or unique about biological intelligence. I do also agree that you could hypothetically scan every aspect of a brain or build a system that exactly mimics the behavior of neurons and probably pretty accurately recreate human intelligence.
I really think our only disconnect is that I don't think the current LLM model is anything close to complex or developed enough to be considered that.
That's a perfectly reasonable position, the question of how complex a human brain is compared with the largest NNs is hard to answer but I think we can agree it's a big gap. I happen to think we'll get to AGI before we get to human brain complexity, parameter wise, but we'll probably also need at least a couple architectural paradigms on top of transformers to compose one. Regardless, we don't need to achieve AGI or even approach it for these things to become a lot more dangerous, and we have seen nothing but accelerating capability gains for more than a decade. I'm very strongly of the opinion that this trend will continue for at least another decade, there's are just so many promising but unexplored avenues for progress. The lowest of the low hanging fruit has been, while lacking in nutrients, so delicious that we haven't bothered to do much climbing.
Would love to see more development in this field, but it's clear that you don't need to have complex or biologically accurate systems to manipulate other humans. This fact alone means that machine learning models will never be advanced beyond that basic goal under capitalism.
They've been used for economic modeling and stock forecasting for decades now, since the 80s, and the modern implementations of these systems is nothing more than the application of those failed financial modeling systems on human social interactions. Something that wasn't possible before because until the widespread adoption of the internet, there just wasn't enough digital communication data to feed into them.
Since these systems are not capable of self development that isn't a negative feedback loop, they literally can't improve without more and more data from different human activity being fed to them.
That alone shows that they aren't a new form of intelligence, but instead a titular interface for the same type of information you used to be able to get with "dumb" indexing engines.
There's a reason that search engine companies are the primary adopters of this technology, and it's because they already have been using it for 20+ years in some form, and they have access finally to enough indexed information to make them appear intelligent.