Earlier quoted context omitted.
we didn't really previously NEED to, but we still tried bc it was an interesting question. today the NEED seems more urgent
Can you spell out more why we need it now?
To settle endless discussions like these! ;-)
The Case That A.I. Is Thinking
931–940 of 1001 posts
Re: The Case That A.I. Is Thinking
#932Earlier quoted context omitted.
Well, "the same consciousness" it's not, as for example it occupies a different position in spacetime. It's an identical copy for a split second, and then they start diverging. Nothing so deep about any of this. When I copy a file from one disk to another, it's not the same file, they're identical copies for some time (usually, assuming no defects in the copying process), and will likely start diverging afterwards.
It might be deeper than you think. Qualia exist "outside" spacetime, e.g. redness doesn't have a position in spacetime. If consciousness is purely physical, then how can two identical systems (identical brains with identical sensory input) giving rise by definition to the same qualia not literally be the same consciousness?
> Qualia exist "outside" spacetime, e.g. redness doesn't have a position in spacetime.
I'm sensing redness here and now, so the sensation of redness exists very clearly tied to a particular point in spacetime. In what sense is the qualia of redness not firmly anchored in spacetime? Of course, you could talk about the concept redness, like the concept Pi, but even then, these concepts exist in the mind of a human thinking about them, still tied to a particular location in spacetime.
> If consciousness is purely physical, then how can two identical systems (identical brains with identical sensory input) giving rise by definition to the same qualia not literally be the same consciousness?
The two brains don't receive the same sensory inputs, nothing in the experiment says they do. From the second right after the duplicate is created, their sensory inputs diverge, and so they become separate consciousnesses with the same history. They are interchangeable initially, if you gave the same sensory inputs to either of them, they would have the same output (even internally). But, they are not identical: giving some sensory input to one of them will not create any effect directly in the other one.
Re: The Case That A.I. Is Thinking
#933Earlier quoted context omitted.
Consistency is one aspect, but it is not enough. I believe (and this is somewhat based in other arguments from neuroscience and discussions with alignment researchers) that two more are necessary: compression, which demonstrates algorithmic development; and linear representation capacity, as this is the only way that we really interpret the world, and therefore will only define another as intelligent if it can distil…
I think compression is probably a natural consequence of coherent self models? Isn't requiring other minds to package their intelligence in human interpretable linear narratives is like requiring dolphins to demonstrate intelligence through written language?
> compression is probably a natural consequence of coherent self models
I don't think so. The case in the Chinese Room argument is coherent by construction but is not a compression.
> dolphins to demonstrate intelligence through written language
Because LLMs are artificial and not a product of physical evolution, our standards for proving intelligence are much, much higher.
Re: The Case That A.I. Is Thinking
#934Earlier quoted context omitted.
I think it has a practical, easy definition. Can you drop an AI into a terminal, give it the same resources as a human, and reliably get independent work product greater than that human would produce across a wide domain? If so, it's an AGI.
Doesn't sound like AGI without physical capabilities. It's not general if it's bound to digital work.
Any AGI capable of this wouldn't have much trouble with physical operation of equipment, of all things.
Re: The Case That A.I. Is Thinking
#935Earlier quoted context omitted.
The vast majority of human “thinking” is autocompletion. Any thinking that happens with words is fundamentally no different to what LLMs do, and everything you say applies to human lexical reasoning. One plus one equals two. Do you have a concept of one-ness, or two-ness, beyond symbolic assignment? Does a cashier possess number theory? Or are these just syntactical stochastic rules? I think the problem here is the d…
> Any thinking that happens with words is fundamentally no different from what LLMs do. This is such a wildly simplified and naive claim. "Thinking with words" happens inside a brain, not inside a silicon circuit with artificial neurons bolted in place. The brain is plastic, it is never the same from one moment to the next. It does not require structured input, labeled data, or predefined objectives in order to learn…
The instantiation of models in humans is not unsupervised, and language, for instance, absolutely requires labelled data and structured input. The predefined objective is “expand”.
See also: feral children.
Re: The Case That A.I. Is Thinking
#936Earlier quoted context omitted.
LLMs are increasingly trained on images for multi-modal learning, so they too would have seen one object, then two.
They never saw any kind of object, they only saw labeled groups of pixels – basic units of a digital image, representing a single point of color on a screen or in a digital file. Object is a material thing that can be seen and touched. Pixels are not objects.
My friend, you are blundering into metaphysics here - ceci n’est pas une pipe, the map is the territory, and all that.
We are no more in touch with physical reality than an LLM, unless you are in the habit of pressing your brain against things. Everything is interpreted through a symbolic map.
Re: The Case That A.I. Is Thinking
#937Earlier quoted context omitted.
If you go around believing all things no one has disproved yet, you will have a very busy belief system.
By now, if you claim "LLMs aren't thinking", the burden of proof is on you. They obviously and openly do. The pile of capabilities that were previously exclusive to human intelligence is ever growing - and the amount of claims of "LLMs can't do X" that are followed up by examples of LLMs doing X shortly after is staggering.
I respect your beliefs of course but I am not religious myself
Re: The Case That A.I. Is Thinking
#938Having seen LLMs so many times produce coherent, sensible and valid chains of reasoning to diagnose issues and bugs in software I work on, I am at this point in absolutely no doubt that they are thinking. Consciousness or self awareness is of course a different question, and ones whose answer seems less clear right now. Knee jerk dismissing the evidence in front of your eyes because you find it unbelievable that we c…
>Knee jerk dismissing the evidence in front of your eyes because you find it unbelievable that we can achieve true reasoning via scaled matrix multiplication is understandable, but also betrays a lack of imagination and flexibility of thought.
You go ahead with your imagination. To us unimaginative folks, it betrays a lack of understanding of how LLMs actually work and shows that a lot of people still cannot grasp that it’s actually an extremely elaborate illusion of thinking.
Re: The Case That A.I. Is Thinking
#939Earlier quoted context omitted.
You might have lost me but what you're describing doesn't sound like an LLM. E.g: > each new token selection aims to bring predicted sensory input (the next word) into alignment with the model’s expectations. what does that mean? An llm generates the next word based on what best matches its training, with some level of randomisation. Then it does it all again. It's not a percepual process trying to infer a reality fr…
> An llm generates the next word based on what best matches its training, with some level of randomisation. This is sort of accurate, but not precise. An LLM generates the next token by sampling from a probability distribution over possible tokens, where those probabilities are computed from patterns learned during training on large text datasets. The difference in our explanations is that you are biasing towards LLM…
I'm not sure why you'd describe "sampling from a probability distribution over possible tokens" as "minimize surprise in predictions" other than to make it sound similar to the free energy thing.
The free energy thing as I understand it has internal state, makes predictions, evaluates against new input and adjusts it internal state to continuously learn to predict new input better. This might if you squint look similar to training a neural network, although the mechanisms are different, but it's very distinct from the inference step
Re: The Case That A.I. Is Thinking
#940Earlier quoted context omitted.
> Neither can humans. We also just brute force "autocompletion" I have to disagree here. When you are tasked with dividing 2 big numbers you most certainly don't "autocomplete" (with the sense of finding the most probable next tokens, which is what an LLM does), rather you go through set of steps you have learned. Same as with the strawberry example, you're not throwing guesses until something statistically likely to…
> I have to disagree here. When you are tasked with dividing 2 big numbers you most certainly don't "autocomplete" (with the sense of finding the most probable next tokens, which is what an LLM does), rather you go through set of steps you have learned. Why do you think that this is the part that requires intelligence, rather than a more intuitive process? Because they have had machines that can do this mechanically…
Good points - mechanically just following algorithms isn't thinking, and neither is "predicting the next tokens".
But would a combination of the 2 then be close to what we define as thinking though?