Researchers just mathematically proved that AI can't recursively self-improve its way to superintelligence.
-
Yeah, so why do you think it's relevant that some brain processing starts to form before we acquire language? Most people vocalize their thoughts, even though you might not (and I don't always either). All our intellectual skills are acquired through language.
What an LLM is, is a "thinking engine". That's what the training creates. That "thinking" can then be applied to different subjects, with a rudimentary form of working memory.
The big surprise to those developing LLMs was that the technology suddenly created emergent effects not foreseen from their basic architecture - the ability to _reason_ and _create world models_. If you're still in 2022 and don't think that this is what they do then maybe you need to get off the "stochastic parrot" bandwagon and update your own knowledge?
After all - humans don't do anything but map inputs to outputs through neural networks either.
The fact that we learn before acquiring language is itself a demonstration of the fact that mammalian thought and reasoning, and human thought and reasoning, is fundamentally not based on language. Your argument is disproving your point.
-
The fact that we learn before acquiring language is itself a demonstration of the fact that mammalian thought and reasoning, and human thought and reasoning, is fundamentally not based on language. Your argument is disproving your point.
-
-
@troed @devsimsek what you have actually written is based on a category error. You're confusing the platform, neurons, with the software.
Wouldn't it be prudent if you learnt anything about the subject first?
Here - I'll help: One of the better books on the subject is "Consciousness: An Introduction" by Susan Blackmore.
I read it 15 years ago. That you believe there's a "platform" and "software" means you have no idea how human cognition works.
-
Wouldn't it be prudent if you learnt anything about the subject first?
Here - I'll help: One of the better books on the subject is "Consciousness: An Introduction" by Susan Blackmore.
I read it 15 years ago. That you believe there's a "platform" and "software" means you have no idea how human cognition works.
I have been following this approach since the first steps in the '80s. I'm pretty clear on how it works.
Software is a metaphor, the connectome is obviously a different kind of construct than procedural code, but it is the connections, not the fact that it is built out of neurons, that determines the kind of reasoning and model construction that the human brain performs. You are looking at the implementation and ignoring the big picture.
-
I have been following this approach since the first steps in the '80s. I'm pretty clear on how it works.
Software is a metaphor, the connectome is obviously a different kind of construct than procedural code, but it is the connections, not the fact that it is built out of neurons, that determines the kind of reasoning and model construction that the human brain performs. You are looking at the implementation and ignoring the big picture.
Here - read a scientific paper:
"Our findings reveal reasoning-like mechanisms within the LLM's layers that operate across structurally similar tasks. Crucially, these mechanisms remain stable despite variations in input and output data, suggesting the existence of internal processes that transcend basic language processing."
There's no hardware and software in humans. The hardware and the software are one and the same.
https://www.sciencedirect.com/science/article/pii/S2949882126000010
-
Here - read a scientific paper:
"Our findings reveal reasoning-like mechanisms within the LLM's layers that operate across structurally similar tasks. Crucially, these mechanisms remain stable despite variations in input and output data, suggesting the existence of internal processes that transcend basic language processing."
There's no hardware and software in humans. The hardware and the software are one and the same.
https://www.sciencedirect.com/science/article/pii/S2949882126000010
All the "internal processes" are linguistic, and not even sophisticated linguistic processing, anything else is hallucinated by the researchers fooled by the "clever hans" effect.
All mammals have basically the same hardware. All behavioral differences are due to differences in the size and arrangement of the connections between the neurons.
-
All the "internal processes" are linguistic, and not even sophisticated linguistic processing, anything else is hallucinated by the researchers fooled by the "clever hans" effect.
All mammals have basically the same hardware. All behavioral differences are due to differences in the size and arrangement of the connections between the neurons.
-
Your argument is "when my gut feeling doesn't agree with science I trust my gut feeling"?
I mean. That's a choice.
-
Ah, abuse. It always comes down to abuse.
-
I'm sorry, did I hurt your feels by pointing out that you're choosing emotions over facts?
-
Researchers just mathematically proved that AI can't recursively self-improve its way to superintelligence.
Not "we think it's unlikely." Not "it seems hard." Formally proved.
The model doesn't climb toward AGI — it slowly forgets what reality looks like. They call it model collapse. The math calls it inevitable.
I wrote about it
https://smsk.dev/2026/04/26/ai-cannot-self-improve-and-math-behind-proves-it/
@devsimsek Hi,
I just wanted to ask what the difference between this and fitting a simple regression model on predicted outcomes? I feel the conclusions of this study is pretty obvious even from a simple regression case - or is it not? Why would we expect something different from LLMs? -
I'm not going to respond to abuse with my own abuse, no matter how tempted I am.
This is "abuse":
"anything else is hallucinated by the researchers fooled by [...]"
Don't debate subjects you don't understand. Your initial post to me is laughably ignorant and directly contradicted by science:
"[LLMs] do not build models or reason about them"
-
This is "abuse":
"anything else is hallucinated by the researchers fooled by [...]"
Don't debate subjects you don't understand. Your initial post to me is laughably ignorant and directly contradicted by science:
"[LLMs] do not build models or reason about them"
-
What? I didn't even mention or refer to you in the text you quoted.
-
@devsimsek Hi,
I just wanted to ask what the difference between this and fitting a simple regression model on predicted outcomes? I feel the conclusions of this study is pretty obvious even from a simple regression case - or is it not? Why would we expect something different from LLMs?@drmambobob hi
Before I reply completely you must know that I'm not an expert on this topic but just a curious guy.You made a great remark, it is essentially the same problem scaled up. The obvi part is that you cannot get something from nothing. The main reason I assume the paper authors are talking about it differently with llms is that while a regression model just gets stale, olm goes probabilistic hell. It basically actively deletes its output diversity until it produces a single acceptable response.
I hope this answers your question.
-
Researchers just mathematically proved that AI can't recursively self-improve its way to superintelligence.
Not "we think it's unlikely." Not "it seems hard." Formally proved.
The model doesn't climb toward AGI — it slowly forgets what reality looks like. They call it model collapse. The math calls it inevitable.
I wrote about ithttps://smsk.dev/2026/04/26/ai-cannot-self-improve-and-math-behind-proves-it/
-
Researchers just mathematically proved that AI can't recursively self-improve its way to superintelligence.
Not "we think it's unlikely." Not "it seems hard." Formally proved.
The model doesn't climb toward AGI — it slowly forgets what reality looks like. They call it model collapse. The math calls it inevitable.
I wrote about ithttps://smsk.dev/2026/04/26/ai-cannot-self-improve-and-math-behind-proves-it/
@devsimsek I am the farthest thing from a mathematician but just from a general lay persons perspective the thesis seems a little rigid. Assuming all training data outside the internet is corrupt, assuming that synthetic data only comes from recursive souring, etc. I don't really want to cheerlead AGI as I assume it will bring about a bunch of existential risk but I'm still in the "wouldn't bet against it" camp.
-
.