Mechanism for free will I don't see anyone has proposed

[Sculptor]

It does halt. It seems to me that you have never written a program. Have you?

Yes I have written a program. Several in fact.
I also have a solution to Buridan's Ass.


And ... BTW

https://www.nature.com/news/2008/080411 ... 8.751.html

If you have written a program then you know that the program halts if there are two outputs that both have equal weight. In simple word, there is no algorithm which can resolve such a situation when outputs have equal weight. You can try it.

Not true. It depends on the time and the conditions.
Buridan's Ass problem is only difficult because it is conceived synchronically. Reality is diachronic. The Ass does not starve, but makes his decision upon which ever bail of hay he regards WHEN the hunger value exceeds the indecision variable.
I think your problem is that you are thinking in integer terms. In reality NO values can ever be exactly the same, in terms of both time and space. No decision is ever equal.
But even if they were, this would be the same problem for so-called "free" will as for determinism.

Did you take a look at the article I posted?

[bahman]

Yes I have written a program. Several in fact.
I also have a solution to Buridan's Ass.


And ... BTW

https://www.nature.com/news/2008/080411 ... 8.751.html

If you have written a program then you know that the program halts if there are two outputs that both have equal weight. In simple word, there is no algorithm which can resolve such a situation when outputs have equal weight. You can try it.

Not true. It depends on the time and the conditions.
Buridan's Ass problem is only difficult because it is conceived synchronically. Reality is diachronic. The Ass does not starve, but makes his decision upon which ever bail of hay he regards WHEN the hunger value exceeds the indecision variable.
I think your problem is that you are thinking in integer terms. In reality NO values can ever be exactly the same, in terms of both time and space. No decision is ever equal.
But even if they were, this would be the same problem for so-called "free" will as for determinism.

Did you take a look at the article I posted?

Yes, I am aware of Libet's experiment and experiments afterward. So you have never been in a situation when you like two options equally. How about the situation when you are uncertain about the outcome of your decision in the future? You want to buy something, invest in the market, but you are uncertain whether the price goes up or down.

[Sculptor]

If you have written a program then you know that the program halts if there are two outputs that both have equal weight. In simple word, there is no algorithm which can resolve such a situation when outputs have equal weight. You can try it.

Not true. It depends on the time and the conditions.
Buridan's Ass problem is only difficult because it is conceived synchronically. Reality is diachronic. The Ass does not starve, but makes his decision upon which ever bail of hay he regards WHEN the hunger value exceeds the indecision variable.
I think your problem is that you are thinking in integer terms. In reality NO values can ever be exactly the same, in terms of both time and space. No decision is ever equal.
But even if they were, this would be the same problem for so-called "free" will as for determinism.

Did you take a look at the article I posted?

Yes, I am aware of Libet's experiment and experiments afterward. So you have never been in a situation when you like two options equally. How about the situation when you are uncertain about the outcome of your decision in the future? You want to buy something, invest in the market, but you are uncertain whether the price goes up or down.

Indecision is a red herring here. Did you not read what I said?
People can be indecisive, since different values are given for different considerations. There are always pros and cons. But time changes them and decisions get made. This is cause and effect.
Equal values are a human construct and do not occur in nature, since they are only representative of things more complex that occupy of necessity different places in space/time.
Look at any two things ; oranges, apples. They are different. Even if they have the same number of molecules (unlikley - since even that number is changing), they still have differences in place. Time presses on and nothing stays the same. This is cause and effect.

[bahman]

Not true. It depends on the time and the conditions.
Buridan's Ass problem is only difficult because it is conceived synchronically. Reality is diachronic. The Ass does not starve, but makes his decision upon which ever bail of hay he regards WHEN the hunger value exceeds the indecision variable.
I think your problem is that you are thinking in integer terms. In reality NO values can ever be exactly the same, in terms of both time and space. No decision is ever equal.
But even if they were, this would be the same problem for so-called "free" will as for determinism.

Did you take a look at the article I posted?

Yes, I am aware of Libet's experiment and experiments afterward. So you have never been in a situation when you like two options equally. How about the situation when you are uncertain about the outcome of your decision in the future? You want to buy something, invest in the market, but you are uncertain whether the price goes up or down.

Indecision is a red herring here. Did you not read what I said?
People can be indecisive, since different values are given for different considerations. There are always pros and cons. But time changes them and decisions get made. This is cause and effect.
Equal values are a human construct and do not occur in nature, since they are only representative of things more complex that occupy of necessity different places in space/time.
Look at any two things ; oranges, apples. They are different. Even if they have the same number of molecules (unlikley - since even that number is changing), they still have differences in place. Time presses on and nothing stays the same. This is cause and effect.

Ok, so to your conscious decision is an illusion. If it is so, I mean, if your decision is made before you are aware of it then it must happen in your unconscious mind. Here there are a few questions: Why does the unconscious mind create the feeling of conscious decision if the decision is made unconsciously and is useless? How does the unconscious mind know what it has to produce? How possibly the unconscious mind can know the difference between apple and orange?

[Skepdick]

If you have written a program then you know that the program halts if there are two outputs that both have equal weight. In simple word, there is no algorithm which can resolve such a situation when outputs have equal weight. You can try it.

This is not true. Randomness/entropy is considered a computational resource, having a source of entropy available allows you to implement a random choice-function.

Here is one: https://repl.it/repls/SoggyCalmSorting

Code: Select all

from numpy.random import choice

options = ['hay', 'water']

for _ in range(20):
  print(choice(options))

Run it and see that it chooses randomly between 'hay' and 'water'. The magic lies in this thing called a Pseudo-random number generator

The above example is an unbiased (e.g unweighted) choice function. It will fairly choose between the two options.
Or you could bias the function like this: https://repl.it/repls/AliveBonyAutocad

Code: Select all

from numpy.random import choice

options = ['hay', 'water']
weights= [0.95, 0.05]

for _ in range(20):
  print(choice(options, p = weights))
  

[bahman]

If you have written a program then you know that the program halts if there are two outputs that both have equal weight. In simple word, there is no algorithm which can resolve such a situation when outputs have equal weight. You can try it.

This is not true. Randomness/entropy is considered a computational resource, having a source of entropy available allows you to implement a random choice-function.

Here is one: https://repl.it/repls/SoggyCalmSorting

Code: Select all

from numpy.random import choice

options = ['hay', 'water']

for _ in range(20):
  print(choice(options))

Run it and see that it chooses randomly between 'hay' and 'water'. The magic lies in this thing called a Pseudo-random number generator

The above example is an unbiased (e.g unweighted) choice function. It will fairly choose between the two options.
Or you could bias the function like this: https://repl.it/repls/AliveBonyAutocad

Code: Select all

from numpy.random import choice

options = ['hay', 'water']
weights= [0.95, 0.05]

for _ in range(20):
  print(choice(options, p = weights))
  

The result of your program depends on when I run the program.

[Skepdick]

The result of your program depends on when I run the program.

And why is that a problem? That's precisely how non-determinism works.

You don't know whether Buridan's ass will choose hay or water but you know it will choose SOMETHING.
And if you made the donkey choose often enough it will choose water 50% of the time and hay 50% of the time.

That's exactly what the code does. It chooses between hay and water 20 times.

if you only want it to make one choice do this: https://repl.it/repls/ScaredLooseBookmarks

Code: Select all

from numpy.random import choice

options = ['hay', 'water']
print('I choose {}.'.format(choice(options)))

[bahman]

The result of your program depends on when I run the program.

And why is that a problem? That's precisely how non-determinism works.

You don't know whether Buridan's ass will choose hay or water but you know it will choose SOMETHING.
And if you made the donkey choose often enough it will choose water 50% of the time and hay 50% of the time.

That's exactly what the code does. It chooses between hay and water 20 times.

if you only want it to make one choice do this: https://repl.it/repls/ScaredLooseBookmarks

Code: Select all

from numpy.random import choice

options = ['hay', 'water']
print('I choose {}.'.format(choice(options)))

First, the pseudo-random generator is a deterministic function. The only source of randomness could be from the agent who runs the program. There is no randomness if you accept that the agent who runs the program is a deterministic being too. So the whole, the agent plus the program are not in a marginal situation. By marginal I mean that the situation contains two options that are equally liked. Why the program doesn't halt? Because the program is not in a marginal situation anymore.

[Skepdick]

First, the pseudo-random generator is a deterministic function.

In your previous comment you literally admitted that the program is NOT deterministic. You said:

The result of your program depends on when I run the program.

So is it deterministic or isn't it deterministic? If it's deterministic - then go ahead and determine the next output for me.

Deterministic IN PRINCIPLE doesn't mean deterministic TO YOU. The function is only deterministic if you have complete knowledge e.g the random seed used to initialise the PRNG. You don't have that knowledge - so to you it's practically random.

The only source of randomness could be from the agent who runs the program.

Obviously! The donkey doesn't die so it clearly has a mechanism in place for breaking deadlocks.

Q.E.D

There is no randomness if you accept that the agent who runs the program is a deterministic being too.

Deterministic from whose perspective?

So the whole, the agent plus the program are not in a marginal situation. By marginal I mean that the situation contains two options that are equally liked. Why the program doesn't halt? Because the program is not in a marginal situation anymore.

Exactly! The program has chosen from two EQUAL options. Which you claim is "impossible" to have happened. And yet I demonstrate it to you.

Do you not believe your eyes?

[bahman]

First, the pseudo-random generator is a deterministic function.

In your previous comment you literally admitted that the program is NOT deterministic. You said:

The result of your program depends on when I run the program.

So is it deterministic or isn't it deterministic? If it's deterministic - then go ahead and determine the next output for me.

Deterministic IN PRINCIPLE doesn't mean deterministic TO YOU. The function is only deterministic if you have complete knowledge e.g the random seed used to initialise the PRNG. You don't have that knowledge - so to you it's practically random.

The only source of randomness could be from the agent who runs the program.

Obviously! The donkey doesn't die so it clearly has a mechanism in place for breaking deadlocks.

Q.E.D

There is no randomness if you accept that the agent who runs the program is a deterministic being too.

Deterministic from whose perspective?

So the whole, the agent plus the program are not in a marginal situation. By marginal I mean that the situation contains two options that are equally liked. Why the program doesn't halt? Because the program is not in a marginal situation anymore.

Exactly! The program has chosen from two EQUAL options. Which you claim is "impossible" to have happened. And yet I demonstrate it to you.

Do you not believe your eyes?

Think of it this way. The random generator is based on an algorithm so the outcome is determined if you give him the same side. Now think of the situation that a device has to decide between two options which are equally liked. How the first choice is done? By an agent (this depends on the time at which the agent runs the program). Is the first choice is random? That depends. There are four cases: 1) The agent is free but he is determined (no decision is required since he, for example, likes to run the code at specific time), 2) The agent is free and not determined (he has to make a decision between two options which are equally liked, options being to run or not run the code), 3) the agent is not free and he is determined and 4) the agent is not free and he is not determined. The first case is doable and the first choice of the program intrinsically is made by the agent so the first choice is not random (consequently, the rest of the choices made by the program are not random too since they are algorithmically related to the first choice). The second is doable and the result is random. The third case is doable and the result is not random. The fourth case is not doable so we need another agent.

[Skepdick]

Think of it this way. The random generator is based on an algorithm so the outcome is determined if you give him the same side. Now think of the situation that a device has to decide between two options which are equally liked. How the first choice is done? By an agent (this depends on the time at which the agent runs the program). Is the first choice is random? That depends. There are four cases: 1) The agent is free but he is determined (no decision is required since he, for example, likes to run the code at specific time), 2) The agent is free and not determined (he has to make a decision between two options which are equally liked, options being to run or not run the code), 3) the agent is not free and he is determined and 4) the agent is not free and he is not determined. The first case is doable and the first choice of the program intrinsically is made by the agent so the first choice is not random (consequently, the rest of the choices made by the program are not random too since they are algorithmically related to the first choice). The second is doable and the result is random. The third case is doable and the result is not random. The fourth case is not doable so we need another agent.

You are over-complicating everything. Lets just stick with the scientific method.

If the function is deterministic then you should be able to determine (predict!) what the computer is going to guess next. So lets play a guessing game.

https://repl.it/repls/LightpinkTechnoUsername

Yes, in the strictest sense a PRNG is deterministic, but that is only true IF you have the initial seed and IF you have a clock perfectly synchronised with that of the PRNG. You don't have either of those things.

If this is your strongest objection though, we can always replace the PRNG with a TRNG. It would costs us just $50 to buy one: http://ubld.it/products/truerng-hardwar ... generator/.

After we install it - not a whole lot is going to change from your point-of-view.

The way you play the game is not going to change.
The fact that the computer is choosing from two equally-weighted options is not going to change.
The rate at which you predict the computer's guesses correctly is not going to change (± 50%)

In fact, you can't even tell whether the computer you are playing against right now has a PRNG or a TRNG inside.

[bahman]

Think of it this way. The random generator is based on an algorithm so the outcome is determined if you give him the same side. Now think of the situation that a device has to decide between two options which are equally liked. How the first choice is done? By an agent (this depends on the time at which the agent runs the program). Is the first choice is random? That depends. There are four cases: 1) The agent is free but he is determined (no decision is required since he, for example, likes to run the code at specific time), 2) The agent is free and not determined (he has to make a decision between two options which are equally liked, options being to run or not run the code), 3) the agent is not free and he is determined and 4) the agent is not free and he is not determined. The first case is doable and the first choice of the program intrinsically is made by the agent so the first choice is not random (consequently, the rest of the choices made by the program are not random too since they are algorithmically related to the first choice). The second is doable and the result is random. The third case is doable and the result is not random. The fourth case is not doable so we need another agent.

You are over-complicating everything. Lets just stick with the scientific method.

If the function is deterministic then you should be able to determine (predict!) what the computer is going to guess next. So lets play a guessing game.

https://repl.it/repls/LightpinkTechnoUsername

Yes, in the strictest sense a PRNG is deterministic, but that is only true IF you have the initial seed and IF you have a clock perfectly synchronised with that of the PRNG. You don't have either of those things.

The fact that I don't know what would be the outcome of PRNG doesn't mean that it is random.

If this is your strongest objection though, we can always replace the PRNG with a TRNG. It would costs us just $50 to buy one: http://ubld.it/products/truerng-hardwar ... generator/.

After we install it - not a whole lot is going to change from your point-of-view.

The way you play the game is not going to change.
The fact that the computer is choosing from two equally-weighted options is not going to change.
The rate at which you predict the computer's guesses correctly is not going to change (± 50%)

In fact, you can't even tell whether the computer you are playing against right now has a PRNG or a TRNG inside.

PRNG gives numbers that look like random. A programmer can know the seed and have what PRNG produces. Therefore can guess what game reaction would be.

[Skepdick]

The fact that I don't know what would be the outcome of PRNG doesn't mean that it is random.

It doesn't mean that it isn't either. Welcome to epistemology.

PRNG gives numbers that look like random. A programmer can know the seed and have what PRNG produces. Therefore can guess what game reaction would be.

But you aren't the programmer... I am. I know whether it's a PRNG or a TRNG producing those numbers.

You don't.