Hacker News Reader

Comment by mikemarsh | Hacker News Reader

The idea of replicating a consciousness/intelligence in a computer seems to fall apart even under materialist/atheist assumptions: what we experience as consciousness is a product of a vast number of biological systems, not just neurons firing or words spoken/thought. Even considering something as basic as how fundamental bodily movement is to mental development, or how hormones influence mood ultimately influencing thought, how could anyone ever hope to to replicate such things via software in a way that "clicks" to add up to consciousness?

View on HN

Conflating consciousness and intelligence is going to hopelessly confuse any attempt to understand if or when a machine might achieve either.

(I think there's no reasonable definition of intelligence under which LLMs don't possess some, setting aside arguments about quantity. Whether they have or in principle could have any form of consciousness is much more mysterious -- how would we tell?)

Defining machine consciousness is indeed mysterious, at the end of the day it ultimately depends on how much faith one puts in science fiction rather than an objective measure.

Seems like a philosophy question, with maybe some input from neuroscience and ML interpretability. I'm not sure what faith in science fiction has to do with it.

I don't see a strong argument here. Are you saying there is a level of complexity involved in biological systems that can not be simulated? And if so, who says sufficient approximations and abstractions aren't enough to simulate the emergent behavior of said systems?

We can simulate weather (poorly) without modeling every hydrogen atom interaction.

The argument is about causation or generation, not simulation. Of course we can simulate just about anything, I could write a program that just prints "Hello, I'm a conscious being!" instead of "Hello, World!".

The weather example is a good one: you can run a program that simulates the weather in the same way my program above (and LLMs in general) simulate consciousness, but no one would say the program is _causing_ weather in any sense.

Of course, it's entirely possible that more and more people will be convinced AI is generating consciousness, especially when tricks like voice or video chat with the models are employed, but that doesn't mean that the machine is actually conscious in the same way a human body empirically already is.

>but that doesn't mean that the machine is actually conscious in the same way a human body empirically

Does it matter? Is a dog/cow/bird/lizard conscious in the same way a human is? We're built from the same basic parts, and yet humans seem to have a higher state of consciousness than other animals around us.

For example the definition of the word conscious is

>aware of and responding to one's surroundings; awake.

I'll give that we likely mean this in a general sense, but I'd say we're pretty close to this with machines. They can observe the real world with sensors of different types, and then either directly compute, or use neural nets to make generalized decisions on what is occurring around them, then proceed to act on those observations.

If you simulate rainy weather, does anything get wet?

(Not my original quote, but can't remember right now where I read it.)

It's similar asking about whether silicon computers performing intelligent tasks is "conscious".

I guess it depends, can you tell the difference between a weather simulation and the actual world?

Can you?

You have weather readouts. One set is from a weather simulation - a simulated planet with simulated climate. Another is real recordings from the same place at the same planet, taken by real weather monitoring probes. They have the same starting point, but diverge over time.

Which one is real though? Would you be able tell?

They're not asking about telling the difference in collected data sets, data sets aren't weather.

The question is can you tell the difference between the rain you see outside your window, and some representation of a simulated environment where the computer says "It's raining here in this simulated environment". The implied answer is of course, one is water falling from the sky and one is a machine.

>The implied answer is of course, one is water falling from the sky and one is a machine.

Lets say you're in a room a good distance from the window. Suddenly you hear what sounds like thunder and rain falling. From a distance it appears that it's raining outside.

Is the rain real? Or is it simulated on a screen well enough you can't tell?

You have input output devices just like a computer. They don't see reality, they filter out huge amounts of data and your brain just interprets it. If our machines get good enough we may be able to blast signals directly to the brain that say it's raining and the brain wouldn't have any idea if it was simulated or not. Much in the same way it feels like we exist and not a 3d hologram an infinite distance away (or whatever other weirdness physics may or could do).

So how do we get from "machines may get really good at faking things to human perception" to "machines themselves can have human-like (or 'better') perception"?

The question is so-called consciousness arising out of machines, not machines deceiving human consciousness.

Even then, that deception still doesn't prove equivalency of simulated and real things, unless we're adopting an extreme and self-contradictory subjectivist epistemology.

People seem to way overcomplicate consciousness, especially in machines.

Where does a running video game exist? It's a simulation in the hardware. Where is the consciousness in a human brain, again it's electrical signals in our brain.

At the end of the day a system is what it does. Once it starts simulating the human mind in ways that appear human then we're at the point of saying a plane has to flap its wings or it's not flying.

You can't look at the "real weather" though. You can only look at the outputs. That's the constraint. Good luck and have fun.

A human brain is a big pile of jellied meat spread. An LLM is a big pile of weights strung together by matrix math. Neither looks "intelligent". Neither is interpretable. The most reliable way we have to compare the two is by comparing the outputs.

You can't drill a hole in one of those and see something that makes you go "oh, it's this one, this one is the Real Intelligence, the other is fake". No easy out for you. You'll have to do it the hard way.

Even granting all of your unfounded assertions; "the output" of one is the rain you see outside, "the output" of the other is a series of notches on a hard drive (or the SSD equivalent, or something in RAM, etc.) that's then represented by pixels on a screen.

The difference between those two things (water and a computer) is plain, unless we want to depart into the territory of questioning whether that perception is accurate (after all, what "output" led us to believe that "jellied meat spread" can really "perceive" anything?), but then "the output" ceases to be any kind of meaningful measure at all.

there is no "real weather". the rain is the weather. the map is not the territory. these are very simple concepts, idk why we need to reevaluate them because we all of a sudden got really good at text synthesis

Everyone's a practical empiricist until our cherished science fiction worldview is called into question, then all of a sudden it's radical skepticism and "How can anyone really know anything, man?"

You experience everything through digital signals. I dont see why those same signals cant be simulated. You are only experiencing the signal your skin sends to tell you there is rain, you dont actually need skin to experience that signal.

Bundling up consciousness with intelligence is a big assumption, as is the assumption that panpsychism is incorrect. You may be right on both counts, but you can't just make those two assumptions as a foregone conclusion.

AGI won't replicate our experience.

But it could be more powerful than us.

Honestly replicating our experience would be rather wasteful. Much like making planes the same way birds work to carry cargo.

Reply on news.ycombinator.com