arthwollipot
Observer of Phenomena, Pronouns: he/him
Has anyone got a p-shotgun?
I need to headshot all the p-zombies in this thread.
I need to headshot all the p-zombies in this thread.
Artificial Intelligence thinks mushroom is a pretzel
- Thread starter William Parcher
- Start date
TragicMonkey
Poisoned Waffles
And yet it still wouldn't be the same as the original artwork. Illusion is not reality, even if it's so detailed you can't tell. Thinking that something's real will not make it so. Unless you are a magical wizard.
arthwollipot
Observer of Phenomena, Pronouns: he/him
It doesn't matter whether it's the original or not - if it's identical in every way, it is indistinguishable from the original, and there is no reason to treat it as though it is not.
theprestige
Penultimate Amazing
Does it have the same provenance as the original?
If it wasn't brought into existence by an intentional creative act of the artist, then it's not the same as the original, and there's no reason to treat it as such.
arthwollipot
Observer of Phenomena, Pronouns: he/him
Why would provenance have anything to do with it? We're talking about the object itself, and not anything that may or may not have happened before it came into existence.
theprestige
Penultimate Amazing
Part of the value of art comes from its provenance. A perfect counterfeit isn't the same as the original. It's not like someone manufacturing an AK-47 to the designer's specs, but without licensing the design.
Art is probably a terrible example of the idea we're trying to get at here. The whole point of art is that provenance matters, and a perfect replica is still just a replica.
P-zombies may be indistinguishable from real people, but p-art is always just a forgery of the real thing.
arthwollipot
Observer of Phenomena, Pronouns: he/him
I think you're probably right that we're using the wrong analogy here.
Dave Rogers
Bandaged ice that stampedes inexpensively through
OK then. Suppose, with sufficient computing power, we are able to simulate an entire human brain down to the atomic level, and simulate the operation of that brain in accordance with the complet set of laws of physics. We then construct the appropriate inputs to that brain, interpret the outputs, and teach it to translate between English and Chinese. Is the simulation thinking? If not, what is it doing? If it's doing something that is a perfectly accurate model of thinking, with the same inputs and outputs, I would argue that any attempt to describe what it is doing as something other than thinking is a distinction without a difference.
We can then simplify the model by aggregating groups of processes in accordance with emergent rules, to produce a higher level simulation. At some point this will become sufficiently high level to be expressed as a fairly simple rule set. The question is then, at what level of abstraction does thinking cease to take place? And that, I think, is the more interesting, and much harder to answer, question.
Dave
TragicMonkey
Poisoned Waffles
Would you then be okay with replacing your loved ones with exact android duplicates?
TragicMonkey
Poisoned Waffles
Is a simulation reality? No.
Is a fancy simulation reality? No.
Is a very fancy simulation reality? No.
Is a very very fancy simulation reality? No.
Is a very very very fancy simulation reality? No.
You can add an infinity of verys but it's not going to change the answer. Live in the holodeck or the Matrix, hallucinate or coma dream, illusion doesn't become reality because you think it's good enough.
Dave Rogers
Bandaged ice that stampedes inexpensively through
I'm not talking about whether the simulation is reality. I'm talking about whether the process going on in the simulation is the same process as is going on in reality. Suppose I simulate a computer down to the atomic level, then run a program on the simulation; is the program running?
Dave
Guybrush Threepwood
Trainee Pirate
Obviously. Everybody would be, since no-one could ever know.
TragicMonkey
Poisoned Waffles
Sure the program is running. But programming is not thought. Programming is a set of instructions. No matter how elaborate you make the instructions anything following them is not thinking. The mouse goes through the maze you built it. If you shape the maze so it spells out words would you say the mouse is writing to you?
jrhowell
Muse
Adding is also a process that people do and have been doing for thousands of years. If I pull out a calculator instead of doing the math in my head, will the calculator be doing real addition of only the illusion of addition? What if the calculator is a simulated calculator?
TragicMonkey
Poisoned Waffles
The calculator will be following the instructions given to it. If those instructions include the error that 2+2=5 the calculator will obediently return that result. It is not thinking. Addition is being done but not as a result of thought on the part of the device, it's a result of the efficacy of the instructions built into it by a human.
HansMustermann
Penultimate Amazing
- Joined
- Mar 2, 2009
- Messages
- 23,741
The point is that when talking processing data, adding a layer of emulation doesn't change anything.
The computer you read this on doesn't natively understand JavaScript. It's run in an interpreter, in a browser, which in turn may run in a C# sharp virtual machine, which in turn is run on the actual CPU. Hell, let's even add one more layer of emulation: the whole operating system may run in a virtual machine emulating a whole other CPU. But when the browser is open, then the computer can run JavaScript, period.
I mean, Jesus F Christ, if you asked anyone something like "can your Mac run JavaScript", they'd say some version of "yes" or "what's JavaScript?" It would have to be someone either trying to be funny by splitting hairs above and beyond the call of duty, or being genuinely dysfunctional to answer something like, "no. No computer can. It can only run a virtual machine which in turn runs the JavaScript."
Ditto for thinking, vs emulating a brain who thinks. WTH difference does it make if there's another layer of emulation or two?
And ditto for Chinese characters. Whether the computer natively understands Chinese (hint: none can) or it's emulating something that can, then that's that.
Frankly, the whole objection subthread strikes me as flippin' daft.
The computer you read this on doesn't natively understand JavaScript. It's run in an interpreter, in a browser, which in turn may run in a C# sharp virtual machine, which in turn is run on the actual CPU. Hell, let's even add one more layer of emulation: the whole operating system may run in a virtual machine emulating a whole other CPU. But when the browser is open, then the computer can run JavaScript, period.
I mean, Jesus F Christ, if you asked anyone something like "can your Mac run JavaScript", they'd say some version of "yes" or "what's JavaScript?" It would have to be someone either trying to be funny by splitting hairs above and beyond the call of duty, or being genuinely dysfunctional to answer something like, "no. No computer can. It can only run a virtual machine which in turn runs the JavaScript."
Ditto for thinking, vs emulating a brain who thinks. WTH difference does it make if there's another layer of emulation or two?
And ditto for Chinese characters. Whether the computer natively understands Chinese (hint: none can) or it's emulating something that can, then that's that.
Frankly, the whole objection subthread strikes me as flippin' daft.
Last edited:
Dave Rogers
Bandaged ice that stampedes inexpensively through
[citation needed]
I can sum up all I have to say quite simply. Either the process we call thought is a process regulated by the laws of physics, or it isn't. If it is, then it is in principle possible to construct a system capable of doing it. If it isn't, then the laws of physics are not universally applicable. We can't say for certain which of those is true, clearly; but I'm not prepared to give up on the universal applicability of the laws of physics just because we haven't yet constructed a machine capable of doing something that corresponds to a suitable definition of thinking.
Dave
TragicMonkey
Poisoned Waffles
In principle it should be possible, I agree. But I don't see it being done with current computer technology, or even by hypothetical vastly more sophisticated versions of current computer technology.
Dave Rogers
Bandaged ice that stampedes inexpensively through
I suspect you're right, but until we have a better definition of thinking it's difficult even to guess at what it might take.
Dave
theprestige
Penultimate Amazing
In humans, it seems to be electrochemical feedback loops, giving rise to sensations, stimulated by external phenomena mediated by sensory extensions of the electrochemical core's biological substrate.
So it'll probably take some sort of hardware emulation. A very fine bimetallic lattice in a chemical bath, perhaps. Something along those lines.
Wire it up with input and output feeds, and experiment with feeding in different stimuli until it starts feeding something coherent back out.
It works in humans, but humans have millions of of years of evolution setting up today's babies for success.
I wonder how long it would take to retrace that evolutionary journey in a lab. I wonder at what point the prototype would be sentient enough that continuing to experimemt with different kinds of stimulus would count as torture.