As I said, the Computer has to have an algorithm. It cannot do anything without an algorithm and it cannot do something that algorithm doesn't say to do. It's Limited by it's algorithm. — ssu

First of all, most computer programs are algorithms that process data, so it is not just an algorithm that you put in - it is algorithm plus data, and data can bring in potentially unlimited information. Deep learning programs are already pretty impressive, to the point that they can fool some of the people some of the time. Second, what is to stop a computer from creating new algorithms, or indeed from evolving its own algorithms in response to inputs? That sounds suspiciously like what the brain is doing, and indeed that is the direction that some of the more advanced machine learning is taking.

Your argument is: computers just follow predefined rules. But if you are a physicalist, i.e. you believe that the world we live in is regular through and through, with no place for magic and the supernatural, then everything in this world - including you - just follow predefined rules (whether or not those rules were predefined by some sentient being is irrelevant to this discussion, as far as I can see).

Now, whether everything in the world can be computed is still a hotly disputed thesis, but this conundrum cannot be resolved by pointing out that computers just follow rules - the question is much more complex than that.

I know that Pruss is pretty clever, but that argument was singularly bad. He should have just left it where Leibniz did.

In which case, as you can see, given infinite time we'll progress towards a limit -- wherever that happens to be -- but that limit will not be infinite. — Moliere

Although log(x) grows sublinearly, it doesn't have an upper limit ;) But I take your point.

Heh. It's been a few years. :D

EDIT: Just cuz it was bothering me.

y = -e^(-x) + a

That was the function I was thinking of. Superficially looks like a log curve, but has a number it approaches when you take its limit.

What does it mean for an algorithm to tell a computer how to react? — Michael

Again:

Definitions of an algorithm:

"A process or set of rules to be followed in calculations or other problem-solving operations, especially by a computer."

"An algorithm is a step by step method of solving a problem. It is commonly used for data processing, calculation and other related computer and mathematical operations."

"An algorithm is a set of instructions designed to perform a specific task."

Definitions of a computer:

"A computer is a machine or device that performs processes, calculations and operations based on instructions provided by a software or hardware Program"

So: computer follows instructions (algorithms). It doesn't do things that the instructions don't tell it to do.

You haven’t addressed the substance of my post.

First of all, most computer programs are algorithms that process data, so it is not just an algorithm that you put in - it is algorithm plus data, and data can bring in potentially unlimited information. — SophistiCat

Yet how the handle the data has to be in the algorithm. There surely can be feedback loops even in very simple computer programs, that in the old days were called cybernetic systems and there is a myriad of other ways how computers "learn" from the given data. Yet for that learning there has to be a specific algorithm.

Or let's put this another way. Give me an example of a computer that doesn't follow an algorithm, instructions provided by a software or hardware program as said above.

You haven’t addressed the substance of my post. — Michael

Neither have you my post.

You see the definition of a computer, a Turing Machine, matters. It's not a synonym for agent. Computation has it's limits. You simply cannot argue that because there is cause and effect, because there is this "Black box" in between input and output the two everything is computable. This actually isn't a thing about consciousness or free will, but the limitations of computation.

Or let's put this another way. Give me an example of a computer that doesn't follow an algorithm, instructions provided by a software or hardware program as said above. — ssu

Why? What would that prove?

Why? What would that prove? — SophistiCat

I would say then that computers really can think, but I assume that I would be just confusing you.

Ok. Assume a computer that you give a program to run. The computer follows first the program, yet later you find it running a totally different program, which wasn't at all described in the first program to be done.

You simply cannot argue that because there is cause and effect, because there is this "Black box" in between input and output the two everything is computable. — ssu

I’m not saying that. I’m saying that human brains are not in principle impossible to manufacture and that unless there really is some magic involved then if we reproduce the material and the behaviour then consciousness will result. We can then manipulate this artificial brain’s experiences by stimulating the relevant neurons, just as we can to a limited extent in real people already.

Whether or not you want to call this artificial brain a biological computer or its experiences a simulation is an irrelevant semantic matter and of no concern to the scenarios described by Bostrom’s trilemma.

I’m saying that human brains are not in principle impossible to manufacture and that unless there really is some magic involved then if we reproduce the material and the behaviour then consciousness will result. We can then manipulate this artificial brain’s experiences by stimulating the relevant neurons, just as we can to a limited extent in real people already. — Michael

Fair enough. I'm not implying that there is any magic either, only that our current Turing Machines called computers have severe limitations in being accurate models on how we function. Of course in many ways they can model us, that's for sure.

Whether or not you want to call this artificial brain a biological computer or its experiences a simulation is an irrelevant semantic matter — Michael

I wouldn't call it an irrelevant semantic matter as a computer does have a specific definition. Now, if you use the term AI, you aren't implying something specific on how the AI operates, but calling it a computer you do that, because (as I've said now many times) a computer has a definition. Just like in earlier historical times people just assumed humans to be just advanced mechanical devices.

I'm not sure where the neuroscientists fall on this on average, but I would guess they're a bit more reserved about making such assumptions. — Marchesk

Well the binding problem is still unresolved. (Well some like Dennett say it's not real.) So it's not like there is specific criteria for it.

Ok. Assume a computer that you give a program to run. The computer follows first the program, yet later you find it running a totally different program, which wasn't at all described in the first program to be done. — ssu

Yes, that's what evolutionary algorithms do: they modify part of their own code (the other part you may think of as the environment, which is subject to unchanging rules).

Yes, that's what evolutionary algorithms do: they modify part of their own code — SophistiCat

With the way the algorithm instructs them to do.

Notice the part "which wasn't at all described in the first program to be done". That part you see means that it's not following the instructions, it's not modifying it's code how it was instructed to do.

What's the difference between consciousness and simulated consciousness anyway?
Simulated suggests crafted intentionally by someone else.
If that's the only difference, then "simulated" has little bearing on consciousness itself, just the circumstances.

What's the difference between consciousness and simulated consciousness anyway? — jorndoe

Hi jorndoe. :)

None, as I see it.

With the way the algorithm instructs them to do.

Notice the part "which wasn't at all described in the first program to be done". That part you see means that it's not following the instructions, it's not modifying it's code how it was instructed to do. — ssu

I am still trying to understand where (if anywhere) you are leading with these requirements for programs that spring into existence fully formed out of the blue. Are you trying to say that consciousness is a miracle? Many do think so, but why beat around the bush? Just come out and say it and we will be done.

I am still trying to understand where (if anywhere) you are leading with these requirements for programs that spring into existence fully formed out of the blue. — SophistiCat

No. I'm just explaining the limitations of computation and using algorithms.

Are you trying to say that consciousness is a miracle? — SophistiCat

Again no. Look, if I were to say that not everything is purely mechanical and can be modelled to work as clock-work, would that mean that I'm implying that there are miracles?

I've always wondered, what reasons are there to suppose that life could be simulations or illusions? The arguments in favour of such are often well presented but none expressly identifies a reason for venturing into such propositions. Is it just a sort of way to balance the speculation about reality? So, instead of "what is reality?", we have "what is not reality?".

Look, if I were to say that not everything is purely mechanical and can be modelled to work as clock-work, would that mean that I'm implying that there are miracles? — ssu

Depends on how one defines miracles. If we assume the popular Humean view of miracles as violations of the laws of nature - which already implies that nature mostly behaves in law-like ("mechanical") fashion - then yes, that is what you are implying.

Only an idealised computer will follow its instructions and only those instructions.

A real computer will respond to input data according to its physical characteristics which may not follow the specified instructions (algorithms) perfectly. There can be problems with the hardware, problems with the software , intrinsic logical problems such as stack overflow all of which conspire to produce output that is not intended by the programmer.

The simulation theory doesn't hold up.

It attributes magical powers are the transistor-switchings in some computer.

That computer can duplicate and display, for its audience, a hypothetical world-story, but it doesn't make there be that world-story. It's already "there", as a system of inter-referring abstract implications.

Why is it that people are so prone to believe the simulation-theory, but rebel at the suggestion our world, and the objective physical facts in our experience, are a hypothetical story consisting of a complex system of inter-referring abstract implications?

Evidently people firmly believe that there must be, even in this logically-inter-dependent realm, some absolute objective existence at the basis of it all. ...that either this physical world has objective absolute existence, or that, at least, there is, at the bottom of all the hierarch of simulations, some absolutely objectively existent physical world.

...unless some people believe an infinite regress of simulations, with no definite objectively absolutely existent world at the bottom of it. But, if you believe in that, and if you agree that all of the apparent physical worlds don't have an objective absolute physical basis, then why do you need all those (nonexistent) computers to make it be?

I suggest that this physical world, as the setting for your life-experience-story, is a figment of logic. And it doesn't need a computer for its (non) existence.

Michael Ossipoff

See http://TheInterMind.com for more on the terminology of the following:

Even if Reality is a Simulation we obviously still have Conscious Experiences of that Reality. So there is probably still a Conscious Mind doing the Experiencing in Conscious Space. There is probably still an Inter Mind but it would now connect the Conscious Mind to the Simulation instead of to a Physical Mind. There are two basic types of Simulations that we can talk about. One type is a Simulation that just runs with us being helpless observers having no ability to affect things that are happening in the Simulation. This means that all our desires, strivings, and actions are just something we experience, but we really can't do anything about anything. The Simulation makes us think we have desires and strivings and that we can do things. In this type of Simulation the Conscious Mind would have no Volitional connections back to the Simulation and would only have connections from the Simulation to the Inter Mind and then to the Conscious Mind. In the other type the Conscious Mind can, through Volitional connections through the Inter Mind and to the Simulation, affect things in the Simulation similar to how the Conscious Mind can, through the Inter Mind, affect things in Physical Space. The Simulation will make us believe we are actually in Physical Space, but there would be no difference for us if we were in an Actual Physical Universe or a Simulated Physical Universe. The take away from this is that it doesn't matter if the Inter Mind is connected to a Brain or to a Simulation.

Depends on how one defines miracles. If we assume the popular Humean view of miracles as violations of the laws of nature - which already implies that nature mostly behaves in law-like ("mechanical") fashion - then yes, that is what you are implying. — SophistiCat

Yet we know that the reality cannot be at all times accurately modelled with the idea of a clock-work mechanical universe. Quantum Physics and relativity do have their merits in making better models of reality.

Only an idealised computer will follow its instructions and only those instructions.

A real computer will respond to input data according to its physical characteristics which may not follow the specified instructions (algorithms) perfectly. There can be problems with the hardware, problems with the software , intrinsic logical problems such as stack overflow all of which conspire to produce output that is not intended by the programmer. — A Seagull

Sure, we get that "syntax error" from time to time. But it's not intentional (or who knows, perhaps it's a clever marketing scheme that computers stop working after enough time).

Yet if the idealised Computer is basically a Turing Machine, then these problems exist. That's my basic point.

Yet we know that the reality cannot be at all times accurately modelled with the idea of a clock-work mechanical universe. — ssu

Do we? How?

Anyway, I am not going to argue for or against the laws of nature. If you believe that conscious beings are outside any general order of things, then obviously you will reject the simulation conjecture for that reason alone. So there is nothing to talk about.

Give a computer a Hex code of 000000, have it add FF, and the result is 0000FF. This is the hex code for blue, and it tells the computer to turn on the blue lamps that each make up part of a pixel.

Only in our scenario that biological computer isn't told to turn on a blue light but to activate the parts of its "brain" that are responsible for bringing about a blue colour experience. — Michael

Love it. A computer can be programmed to operate a light switch. Therefore a conscious computer is possible. [Hands wave furiously.]

So how is it that neural firing would "look blue"? How is this little trick achieved? What is it that we know "in principle" here that would warrant your extrapolation.

So how is it that neural firing would "look blue"? How is this little trick achieved? — apokrisis

We don't know how but we know it happens in us. What makes our brains so special that the same effect can't be achieved artificially? You're making the brain out to be a miracle as SophistiCat mentioned earlier.

Love it. A computer can be programmed to operate a light switch. Therefore a conscious computer is possible. [Hands wave furiously.] — apokrisis

It was an analogy, not a syllogism.

Getting back to the OP, the interesting thing is this idea of a simulation that would somehow be all our consciousnesses, plus the world we think we share. Is anyone stopping to think what this would entail?

What even is the hypothesis?

Is there one fake world and then somehow a whole lot of fake minds having private thoughts, feelings and understandings of it?

Or is there only one fake mind and that mind is the entire world as such, any others appearing in this world being merely fake furnishing?

Heh, that's a good point. I suppose that if you were only simulating one mind, you could make your simulation domain smaller than if you were, say, simulating the entire population of the earth.

We don't know how, but we know it happens in us. — Michael

If you can't say anything to bridge this explanatory gap then you can't claim anything "in principle" here. That's pretty straightforward.

I'm not denying that we can't take a biologically inspired approach to "computation". Neural network approaches already do.

But you can't offer a Turing Machine example - hex code to operate a switch - and freely extrapolate from that. You have to show that biology is in principle doing that kind of computation.

And as I said - and as you have ignored - we know enough about biology to see that it relies on material instability, while TMs, and machines in general, rely on material stability.

So biology is essentially relational. It is about informational constraints on material dissipation. The overall organisation is emergent.

While computation is essentially dualistic. The software is informationally isolated from the material hardware needed to implement it. Where biology is about an intimate sensitivity to the material conditions of its being, computing is the precise opposite - the ability to completely disregard those material conditions.

If you are wanting to make "in principle" claims, then that basic difference is where you have to start.

Computation is nothing more than rule-based pattern making. Relays of switches clicking off and on. And the switches themselves don't care whether they are turned on or off. The physics is all the same. As long as no one trips over the power cord, the machine will blindly make its patterns. What the software is programmed to do with the inputs it gets fed will - by design - have no impact on the life the hardware lives.

Now from there, you can start to build biologically-inspired machines - like neural networks - that have some active relation with the world. There can be consequences and so the machine is starting to be like an organism.

But the point is, the relationship is superficial, not fundamental. At a basic level, this artificial "organism" is still - in principle - founded on material stability and not material instability. You can't just wave your hands, extrapolate, and say the difference doesn't count.

I've been talking about using biological material rather than inorganic matter so the above is irrelevant.