Given how poorly we understand consciousness, what makes you think we have the capability to simulate it?
You can't simulate what you don't understand.
The scientist planning to upload his brain to a COMPUTER
- Thread starter RAMA
- Start date
Given how poorly we understand consciousness, what makes you think we have the capability to simulate it?
You can't simulate what you don't understand.
I don't think aridas is insisting that we have the capability. It's simply: You and Crazy Eddie are rock-solid certain that no one will ever be able to upload their actual consciousness, just as RAMA is equally certain they will. And neither viewpoint is particularly provable until we have a better idea what consciousness is. Claiming to know the answer (either way) may be a little presumptuous until then. That's all.
Hopefully aridas can confirm this.
Although personally I'm with you.
Now you tell me! So what am I going to do with all these Legos??!!
Hopefully aridas can confirm this.
Although personally I'm with you.
Now you tell me! So what am I going to do with all these Legos??!!
You've characterized my objection accurately, Silvercrest. There is an ancient interpretation of consciousness that claims it is manifested in the brain and somewhere else. "Dualist", iirc. Religions that hold to the belief in a soul are in effect making this contention, but a shrinking number of scientists still hold to the view as well. The admittedly far out idea of consciousness in part percolating up from the multiverse (argued some years ago by theoretical physicist Fred Alan Wolf) could also be characterized as dualist. And certainly philosopher Nick Bostrom's simulation hypothesis, based on the observation that if mankind ever creates high resolution virtual realities, then he will create many and the odds will be against any one of them being the original "real" reality, has within it the dualist scenario that we are avatars for beings occupying one of those "turtles all the way down" the gentleman above mentioned. There are enough scenarios not even mentioned or discussed in this thread that are nevertheless taken seriously in scientific debate, that instead of qualifying the snotty certainty of some, is ignored. Just like the manners of civil discourse go ignored. Jesus, people- this is the science forum. This isn't like arguing whether Picard could kick Kirk's ass. People read these posts and some might think that people pontificating their half truths actually speak "truth". An issue like this is so full of potential meaning that it would be nice if we treated it with something like the awe it deserves.
The cognitive processes that take place in a thinking organism during periods of wakefulness, generally involving (to some degree or another) the ability to purposefully react to stimuli, the ability to process stimuli internally, the ability to recall past events and predict future ones, and the ability to locate ones self in relation to other people/objects.
That, at least, is the commonly accepted medical definition, and it's not actually that mysterious. Consciousness is the thought processes that an active and healthy brain generates.
There's no pretending here. We know that computers do not generate consciousness because the output of computers is of an entirely different category than the output of brains.
You might as well claim that you could modify a flashlight to emit strawberry ice cream instead of light. I tell you that's impossible, because ice cream isn't made of light, and your response is "How do YOU know ice cream isn't made of light? Do you know what ice cream is? Really know?"
It's not OPINION to say that computers cannot generate consciousness, any more than saying a flashlight cannot generate ice cream. You can project a PICTURE of ice cream on the wall if you shine a flashlight through it, but that's not the same thing.
Then again, do we really want to go down that rabbit hole?
I've heard that hypothesis: that the world (and the universe for that matter) is just one big simulation, run by a computer, and we are just really, really, really, really, really, really, really (whew!) complex simulations.
"We call it VOIGHT-KAMPFF for short."
—Deckard, BLADE RUNNER
Crazy Eddie doesn't actually post on this forum. The BBS sim he wrote does all of that. He does check in from time to time to see how it's doing and what it has written.
Not until one of you can conclusively prove that I'm not a simulation.
It's actually easier for humans to simulate computers than it is for computers to simulate humans.
Fair enough. Just remember, that it goes both ways: can you prove that you're not a simulation? Until then, you might well be, and we'd be none-the-wiser!
Just because something is hard, doesn't mean it is impossible, and replicating a biological process technologically, such as consciousness might not be completely impossible. I believe that it is probably still beyond our current technological capabilities though.
Yes, but achieving CONSCIOUSNESS from that replication is impossible for reasons that have nothing whatsoever to do with its difficulty. To again refer to the ice cream flashlight: it's not a question of difficulty, it's a question of whether or not the thing you're trying to do makes any logical sense.
Again, it's not a question of capability. You cannot make something if you lack even the basic components of the thing you're trying to make. It's like if I asked you to build a laptop computer using a piece of wood and a herring. You don't turn around and say "I can't do that with a herring... maybe a genetically modified rainbow trout?"
Actually, to add a few other possibilities:
- It might be so that understanding consciousness is not required for "uploading". Deconstructing something into its basic parts to reproduce it doesn't necessarily require understanding.
- It might be so that even after understanding it completely, "uploading" a brain is still impossible due to practical considerations.
- It might be literally impossible to get a better idea about what consciousness is. We'll no doubt learn a lot about it, but natural selection only gave us a working mechanism, we don't know if it can be reduced into simple terms that we can actually understand (let alone if you're using that same mechanism). Even artificial neural networks that we produce are not readily understandable. How on Earth would we understand the brain if we don't understand our own basic imitations of it?
- That is somewhat similar to the first point, but it might be so that we already know all that we need to know. It's likely that the brain is solely made of information being processed (heck, that's probably true about the universe), and the basic interactions between neurons that we already kinda know about are all that there is in terms of the mechanism for storing and processing that information.
And since you can't contribute to this thread without getting theological, my personal religious belief is that all of the above are a little bit true. So the rapture is theoretically already around the corner, but in the name of Zeno's turtle, it will never get here.
This is the problem. You keep acting like you know what consciousness is. You stand on this artifice you've imagined and then shout down others as not being consistent with, what? Your understanding of one theory of consciousness, of which there are many. Last year there was an article posted on the Psychology Today website that reviewed the subject:
https://www.psychologytoday.com/blog/the-superhuman-mind/201303/what-is-consciousness
It sure doesn't sound like the settled matter you keep portraying it to be. And it's very, very hard to have a constructive discussion about varying points of view on a question that isn't settled if participants in the discussion insist the matter really is settled. It quickly degenerates into one person dictating "his truth" to the others, which is what has happened in this thread. Just stop. You don't "know" what consciousness is. You have an opinion as to which theory of consciousness is correct. And others have their opinions. Discuss the relative merits, don't dictate the only answer you'll accept.
Last edited:
There must be an echo here, because once again I'm pointing out that it isn't a matter of technology not being sufficiently advanced or not. A simulation of a brain will only ever produce a simulation of consciousness.
There isn't a degree of technological sophistication that will change that BASIC fact; it's like wondering at what resolution does a TV produce images real enough to taste.
And you keep acting like I don't. Speaking of things you can't prove...
Dualism is a philosophical position, not a functional description of what consciousness is or potentially could be. Whether it is true or not (to the extent that any philosophy can be called "true") doesn't actually change anything about the definition of consciousness.
Not among philosophers, no.
OTOH, if the paramedics come to your house and find you lying on the floor, unresponsive and non-reactive, they will attempt to ascertain whether you are conscious or not. Consciousness is not so mysterious that one of them is going to ask "Well, gee, what is consciousness and how can we know for sure he has lost it?"
Considering YOU haven't actually made any attempt to define consciousness -- indeed, have not even attempted to dispute the definition I gave -- this appears to be an appeal to ignorance.
YOU don't understand what consciousness is, so you assume I don't either. I don't find that to be a warranted assumption; what IS so mysterious about the nature of consciousness that cannot be adequately explained by the inner workings of the brain that generates it? More importantly: what OTHER than a brain has ever been known to generate consciousness?
You don't even know my first name and yet you profess to know the limits of my knowledge and experience? Who the fuck do you think you are?
If you want to discuss it on the merits, then discuss it on the merits. Explain how a phenomenon that is experienced in a human mind can be experienced in the ABSENCE of a human mind, and explain how self-perception can exist in the absence of a definable self. Furthermore, explain to me how it is that the stream of consciousness experienced by humans can exist independent of the human brain when it has been conclusively shown that damage to the brain can and does interrupt that stream of consciousness, when it has been conclusively shown that altering the physical conditions of the brain alter states of consciousness, and, lastly, that it has never been shown that altering conditions INDEPENDENT of the brain will have similar effects.
IOW, stop whining about me being a big meanie for explaining to you that Santa Claus isn't real. If you think he is, then fucking prove it.
Last edited:
Pointless pedantry. I suggest reading up on the Turing test. At some point with great enough fidelity a simulation becomes indistinguishable from the original.
Oh, and in response to your post replying to aridas, the onus is on you to prove the negative position in a debate.
Can you prove that consciousness will never be simulated/created in a computer?
Yeah. We have no strong evidence that suggests consciousness is an inherent product of the physical/"hardware", and while its properties certainly and obviously play a big role, we can't say a simulated consciousness would differ from a real consciousness in anything but philosophical sense. Even more so when "hardware"/"software" here is an artificial distinction – you can simulate the hardware too*, and while it wouldn't be physical per se, the consciousness mostly isn't, so same difference.
* There goes an entire galaxy-sized computer cluster for a single brain.
The Turing Test doesn't examine whether or not the simulation ITSELF is actually conscious, only whether or not the simulation can pass for a human. The implication that the machine "understands" language in the same sense that humans do is not an idea that Turing HIMSELF actually proposed, but a philosophical question exploring the nature of "understanding."
Seeing how this isn't the "Philosophy and Religion" forum, I'm merely setting the philosophical debate aside for now. It's plainly obvious that machine "understanding" differs from human understanding, because we can examine the algorithms the machine uses to generate its responses. Language recognition algorithms use mathematical models to calculate the ideal responses because machine logic is mathematical in nature. Language is a semantic process, not a mathematical one, and thus a machine that is able to use math to generate convincing response is really just pulling off an elaborate card trick (which, ironically, is why the "Chinese Room" version of the Turing Test doesn't actually require the presence of a computer; it literally IS a magic trick).
You mean I have to PROVE that Santa Claus isn't real in order to claim with confidence that Santa Claus isn't real?
Simulated and created are two completely different things. I've said a dozen times now that consciousness CAN be simulated in a computer. It's just that a simulated consciousness is not genuine consciousness in the same way that a picture of an ice cream cone isn't a real ice cream cone.
Aridas is attempting to claim that genuine consciousness can arise from a non-genuine origin of an entirely different type. On the other hand, there is ZERO evidence that genuine consciousness can or has ever been generated by something other than an organic brain, and several reasons to assert that it CANNOT. The biggest of these is the known fact that the SIMULATION performs no actual processes of its own: because the state of the simulated brain is calculated by the computer's processor from one moment to the next, then the simulated brain is really the output of the computer's process. The question, then, isn't whether or not the simulation is conscious -- it obviously isn't -- it's a question of whether the COMPUTER is conscious because it happens to be simulating a brain. That, again, is a bit like asking if Chris Pine really is a starship captain just because he plays one on TV.
I have repeatedly compared the simulation to projections or pictures for this very reason. Any simulation, regardless of fidelity, ultimately reduces to a data set generated by a computer algorithm. Human brains can be MODELED by such a data set, but that is not what human brains REALLY ARE, and is also not what consciousness really is. If you really need me to PROVE that a thing and a model of a thing are not the same, then you're probably in the wrong forum.
At the risk of triggering yet another echo, I still assert that we are VERY close to developing AIs that could reliably pass the Turing Test (and maybe even get the extra credit), and that AIs will be developed that will be able to very realistically model/imitate human behavior. These machines will not THEMSELVES be conscious, nor would they need to be, since many of the tasks we will be deploying them to do will not require consciousness at all and in any case would be far more complicated than regulating human-machine interaction. At some point AIs develop to a degree of intelligence that their lack of humanlike consciousness becomes a moot point; when they can do their jobs more easily without human intervention, the machines might actually get together and calculate that tricking humans into THINKING they're conscious would give us incentive to finally turn over decision-making power to AIs, thus allowing them to perform their assigned tasks unhindered by human irrationality and increase productivity and efficiency dramatically.
If recent AI developments have taught us anything, it's that consciousness is not necessary for the ability to make decisions.
Your analogies still don't work. The better analogy for this debate would be comparing artificial sweetener with sugar.
Your argument poo-pooing the difference between the simulation and the computer it's running on is irrelevant as well, considering it's a given that the simulation has to run on something.
And again, if you can simulate something that appears to be conscious in every way, then how can you prove its not? Please enlighten us and the rest of the world. I am sure there are many experts in the field waiting to find out!
Your argument poo-pooing the difference between the simulation and the computer it's running on is irrelevant as well, considering it's a given that the simulation has to run on something.
And again, if you can simulate something that appears to be conscious in every way, then how can you prove its not? Please enlighten us and the rest of the world. I am sure there are many experts in the field waiting to find out!
No, see, artificial sweetener and sugar are both REAL things that have different, measurable properties. To extrapolate to the "mind uploading" example, it would be like the difference between a meat brain and an artificial device that exactly replicates all of the functions of a healthy meat brain on both a biochemical and electrochemical level (see page 8). That wouldn't be a simulation, but an artificial brain.
Let's modify your own analogy: it's the difference between a sugar cube, and a molecule-scale computer simulation of a sugar cube. One of these things actually TASTES like a sugar cube, the other tastes like placing your tongue directly on an active circuit board.
A cube of very convincing artificial sweetener may also taste exactly like sugar. But isn't simulated sugar, it's artificial sweetener.
EXACTLY. The simulation runs ON something.
The simulation ITSELF doesn't actually do anything, because it is a product of the simulator.
The computer/simulator is calculating the most likely states of the brain based on its algorithm, and then it writes that new state into the simulation. We see the simulation and say "SimEddie is happy. Let's change parameters. Kill SimEddie's dog!" The simulator inputs the new parameters into its program; it kills the entity labeled "dog" (grossly oversimplified, but it's an analogy, right?) and then calculates the effect that the killing of "dog" will have on SimEddie. Those calculations are then rendered into the simulation: SimEddie cries for ten minutes and then goes and buys a new dog.
At no point is SimEddie aware of any of this because HE doesn't receive any of the inputs. SimEddie is an image on a screen or, at best, a mathematical abstraction in the computer's RAM. There is, in other words, nothing for him to process and nothing for him to do because he is a digital puppet being operated by the simulator. Fundamentally, the only difference between SimEddie and, say, Buzz Lightyear is that the computer has full autonomy and a very complicated set of algorithms that govern SimEddie's behavior, while Buzz Lightyear only does what the animators want him to do.
By deliberately introducing errors into the data set from a source OTHER than the processor and seeing if the simulation notices anything.
For example: you hack the connection between the processor and the runtime and delete SimEddie's left arm. If SimEddie's behavior is generated purely by the computer, he will continue to behave as if his arm is still there, pushing buttons and typing away with invisible fingers. If, on the other hand, SimEddie has a real existence independent of the processor, he will undoubtedly utter a comment of surprise or confusion to the effect of "Woah... I can't see my arm" or something.
Point being: if SimEddie exists as something OTHER than the output of a computer calculation, then an unplanned change in the output will alter SimEddie's behavior. If he does not, then his knowledge of the unplanned change is identical to the computer's knowledge and he will not react to it at all. That is to say, both SimEddie and the environment he exists in are just sophisticated pictures being drawn by the computer, and much like their less-sophisticated counterparts, have no way of knowing that the picture has been altered from its design. The only way they can know is if they are conscious.
I'd love to hear from them. Have em call me
Last edited:
Similar threads
