A Practical Exercise in Design Detection

Recently I described a possible alternative Turing test that looks for certain non-person like behavior. I’d like to try it out and see if it is robust and has any value.

Here is some data represented in an ordinary control chart.

 

It’s easy to see there is a recognizable pattern here. The line follows a downward trajectory till about observation 16 then it meanders around till about observation 31. At which point it begins an upward track that lasts till almost the end of the chart.

The data in question is real and public but obsolete. I won’t say what it’s source is right now to avoid any bias in your attempts see if we can infer design. I can provide the actual numbers if you like.

The question before the house is. Is the clear overall pattern we see here a record of intention or does it have a “non-mental” cause?

Using the criteria discussed in my “poker” thread I would suggest looking for the following behaviors and excluding a mental cause if they are present.

1) large random spikes in the data
2) sudden changes in the overall pattern of the data that appear to be random
3) long periods of monotony
4) unexplained disjunction in the pattern.

I have some other tests we can look at as well .

What do you say design or not? Are you willing to venture a conjecture?

 

peace

 

 

 

 

575 thoughts on “A Practical Exercise in Design Detection

  1. Erik: Your test is supposed to test that. Quote from an earlier FMM, “mental processes. You know the ones we are looking to identify in the test”

    Do you see the word infallibly there?

    I did not think so.
    Perhaps you also need to understand the difference between know and prove.

    For example I know that Eric is a person and not a bot but I can’t prove it

    peace

  2. fifthmonarchyman: Do you see the word infallibly there?

    I did not think so.
    Perhaps you also need to understand the difference between know and prove.

    For example I know that Eric is a person and not a bot but I can’t prove it

    peace

    What relation does this have to your OP? Is your OP about “I can’t prove anything about anything but I’m detecting design here in an obvious pattern”?

    Nevermind minds, that’s obviously over your head. Define “pattern”. This should be simple.

  3. Erik: Is your OP about “I can’t prove anything about anything but I’m detecting design here in an obvious pattern”?

    no the pattern is obvious but that is not what this is all about.

    It’s really simple

    I’m looking for non-mental behaviors in the pattern that might help with the design inference. This might be useful in all sorts of ways.

    Erik: Define “pattern”.

    pattern- A regular and intelligible form or sequence discernible in the way in which something happens or is done.

    from here

    https://en.oxforddictionaries.com/definition/pattern

    peace

  4. GlenDavidson: I’m not surprised that Dennett would write that, given that his “explanation” for consciousness is that it’s an illusion.

    Still, I think the more important fact is simply that one can’t look at behavior that is similar to ours and assume that subjective experience necessarily accompanies that behavior. That’s assuming the answer for what is actually in question.I do tire of hearing of “zombies,” to be sure, since it’s really about not assuming that one already knows the answer to the question, and the rest is pretty much useless blather.But it still does make the reasonable point that you don’t get to assume the conclusion you want (Dennett should learn).

    I hear what you say but I think Dennett’s

    If, *ex hypothesi*, zombies are behaviorally indistinguishable from us normal folk, then they are really behaviorally indistinguishable!

    is a killer punch. If the transporter actually spits out two Captain Kirks, and they both claim to be the real Captain Kirk (and they are identical to the last molecule) who’s to say they’re not both right?

  5. Alan Fox: . If the transporter actually spits out two Captain Kirks, and they both claim to be the real Captain Kirk (and they are identical to the last molecule) who’s to say they’re not both right?

    The point is that no one can “say” one way or the other.
    But the real Captain will certainly know who he is

    peace

  6. Richardthughes: Can you just paste the values as plain text, maybe pipe delimited?

    keith’s told me not to and I’m trying to be a courteous guest

    They do take up a lot of space and are not too pretty but I’m fine with it as long as the moderators are.

    peace

  7. fifthmonarchyman: keith’s told me not to and I’m trying to be courteous guest

    They do take up a lot of space and are not too pretty but I’m fine with it as long as the moderators are.

    You can create a file with the values, upload that, and link to it. Alternatively you can use PasteBin.

  8. fifthmonarchyman: no the pattern is obvious

    How? How did you arrive at that there is a pattern and that it’s obvious?

    It’s really your job to demonstrate the pattern. You have not even begun to do it.

    fifthmonarchyman: pattern- A regular and intelligible form or sequence discernible in the way in which something happens or is done.

    The graph in the OP is not “regular” or “intelligible” and not “something that happens or is done”. It’s a graph, a static representation of random-looking data.

    Please try something relevant to the OP. You are demonstrating a clear pattern here – consistent irrelevance and cluelessness about your own topic.

  9. Erik: How? How did you arrive at that there is a pattern and that it’s obvious?

    Do you know about my game? I could have run the graph through it if I was uncertain.

    However in this case the pattern jumps out at you when you look at the graph.

    I could if I needed to reproduce the pattern using different values and see if you would consider the data representation in the chart to be similar to what I see.

    we would need several controls to make it all work.

    I just don’t think it’s necessary in this case. But we could do it if you wanted to be difficult.

    Erik: The graph in the OP is not “regular” or “intelligible” and not “something that happens or is done”. It’s a graph, a static representation of random-looking data.

    I am certain that the majority of respondents would be able to reliably differentiate the chart I posted from randomized groupings of the same data.

    Do you really want to dispute this?

    peace

  10. Alan Fox: I hear what you say but I think Dennett’s

    is a killer punch. If the transporter actually spits out two Captain Kirks, and they both claim to be the real Captain Kirk (and they are identical to the last molecule) who’s to say they’re not both right?

    It depends on what you’re talking about. If it’s FMM’s “test,” and there’s nothing distinguishable between the human and the (advanced) computer, then there’s clearly no actual test for “mind” vs. machine.

    If it’s just what FMM was saying about zombies being “possible” (in the sense of being a logical possibility, I gather), then the problem is that zombies are conceivable and can’t be ruled out a priori. That they’re indistinguishable is the problem, not the solution, the latter of which Dennett oddly seems to be saying in that short piece.

    Glen Davidson

  11. Neil Rickert: How will the real Captain know?

    the same way you know that you are the real Neil Rickert

    Neil Rickert: Is there a real Captain?

    For the real Captain there is

    Neil Rickert: Maybe there are now two real Captains.

    Maybe for you but not for the Captain.

    We can test this by seeing if he would object to his doppelganger going on his date with a Martian hottie while he stayed home catching up on the ship’s log

    peace

  12. Alan Fox: If the transporter actually spits out two Captain Kirks, and they both claim to be the real Captain Kirk (and they are identical to the last molecule) who’s to say they’re not both right?

    The point of philosophical zombies is not to conclude which one is the right Kirk, but to point out the fact that behavior is distinct from cognition/experience/consciousness. If you think (as a materialist would) that consciousness or self-awareness is behavior, then certain uneasy implications follow for you: You cannot distinguish between death and a recoverable halt of physiological processes (because halt of physiological processes is death to you). You cannot distinguish between brain processes and thought processes (thoughts are about all sorts of things, whereas brain processes are nothing but brain neurons firing, regardless of conveying any thoughts or ideas or awareness). If you were consistent, you would not distinguish between knowing something and doing something. Etc.

    Given the two-Kirks-from-the-transporter scenario, the dualist would be in the tough position of positively asserting that one of the Kirks must be a zombie or a fake or a liar, but the materialist is in the tough position of asserting that the scenario yields the exact same person twice.

  13. fifthmonarchyman: I am certain that the majority of respondents would be able to reliably differentiate the chart I posted from randomized groupings of the same data.

    Do you really want to dispute this?

    Dispute what? That the same data represented in different ways looks different? And this is meant to prove a mental process in the process which is really data presented in a graph made to look designed but still looking random because it has zero context? Please try to make sense.

  14. GlenDavidson: If it’s FMM’s “test,” and there’s nothing distinguishable between the human and the (advanced) computer, then there’s clearly no actual test for “mind” vs. machine.

    maybe you also need to understand the difference between know and prove.

    Clearly the poker players identified non-personlike behavior in the bot. If a software patch was added to remove these behaviors then they would be able to identify other behaviors. If they can’t then the bot will have passed the test.

    That does not mean the bot is a person

    GlenDavidson: That they’re indistinguishable is the problem, not the solution

    exactly.

    If they can be distinguished then my test or something like it is in fact definitive and certain.

    If not then it’s useful but not infallible

    it’s a win win

    peace

  15. fifthmonarchyman: I am certain that the majority of respondents would be able to reliably differentiate the chart I posted from randomized groupings of the same data.

    Do you really want to dispute this?

    I said initially that it was non-random–which I think is ok for a first, if tentative, judgment (that is all that I meant the statement to be). It doesn’t appear to be random.

    That said, clearly it may very well be random, and anything that identifies the data as potentially representing non-random phenomena is going to have to check that tentative idea against context. It really wouldn’t be hard for a data-mining computer bot to find a very similar pattern in purely random data. That’s one of the issues in data-mining, of course “patterns” will appear in the data, but they may be just statistical flukes.

    The data are too few to determine definitively that it is non-random. That’s why context is crucial to determining whether or not it is truly non-random. The majority of respondents will indeed see it as non-random, the trouble is that it could very well be random in fact.

    Glen Davidson

  16. fifthmonarchyman: I have a ton of curiosity about minds and persons. I just don’t think it is possible or at all helpful to define them with mathematical precision.

    I asked, how does the five year excel at solving the problem of the unexpected? Your hypothesis is humans act in non algorithmic pattern sometimes . Describe generally how it works. Then if we contrast that with a machine we might be able to clarify what you are detected. Elimination of the similarities should make the differences more pronounced.

    It occurs to me while you might not be able the describe one person’s mental behavior mathematically, if it was shown you could for a group of people would that affect your hypothesis?

    I’m very curious about my wife but I don’t try to reduce and specify her essence in that way. To do so would be to exclude the very things that I find interesting about her in the first place

    Can you predict how she will behave under certain situations? If you are lucky you have developed an algorithm. When she asks why you did what you did can you describe your reasoning? Or do you say, mental processes?

  17. Erik: Dispute what? That the same data represented in different ways looks different?

    My game shows that you can’t distinguish one set of data from another if it is random unless you memorize each individual data point.

    Erik: And this is meant to prove a mental process in the process which is really data presented in a graph made to look designed but still looking random because it has zero context?

    1) Who said anything about prove?
    2) It does not look random there is a pattern as can be demonstrated.
    3) You might not have all the context you desire but you never do, if you did we would not need a test. You would already know whether there was a mind behind it

    peace

  18. GlenDavidson: The data are too few to determine definitively that it is non-random.

    That is fair observation.
    The more data we have the more confident we can be in our inference. The fact is we are often asked to make determinations with limited data.

    peace

  19. fifthmonarchyman: 1) Who said anything about prove?

    Right, sorry. The word is test. As it is, your test is not testing anything discernible, so it’s not really a test as far as I can tell.

    fifthmonarchyman: 2) It does not look random there is a pattern as can be demonstrated.

    Where did you demonstrate that? Did you demonstrate it when you said, “the pattern jumps out at you when you look at the graph.” That’s your demonstration?

    fifthmonarchyman: 3) You might not have all the context you desire but you never do, if you did we would not need a test. You would already know whether there was a mind behind it

    False. A test or exercise or experiment (which are all words you have used for whatever you are doing here) is a test or exercise or experiment worth the name if it is REPLICABLE. We do not have enough context to replicate this, so this is not a test. Not even enough context to know what you are going on about. For six pages now, it’s worthless nonsense.

  20. newton: Describe generally how it works.

    That is what I’m trying to do with my list of non-person like behaviors. “Generally” a person would behave in ways that are not reflected on that list

    newton: Elimination of the similarities should make the differences more pronounced.

    bingo

    newton: It occurs to me while you might not be able the describe one person’s mental behavior mathematically, if it was shown you could for a group of people would that affect your hypothesis?

    It is interesting and I have thought about that as well. It’s something I’d like to explore in the future. I think that at times a group of people can behave as a single mind but that is often not the case.

    What I’d like to look into is what makes a group behave like a single mind verses a process that can be statistically described.

    peace

  21. Erik: As it is, your test is not testing anything discernible, so it’s not really a test as far as I can tell.

    discernible is not the same thing as infallible.
    A cancer screening can not prove you have cancer it simply shows whether you have certain biological markers that are associated with cancer.

    My test is no different

    Erik: Did you demonstrate it when you said, “the pattern jumps out at you when you look at the graph.” That’s your demonstration?

    No I can demonstrate it with my game by showing that we can distinguish the chart I posted from a randomized grouping of same data. The only way you could do that if there were no pattern is by memorizing every data point

    Try and pay attention

    Erik: We do not have enough context to replicate this, so this is not a test.

    Are you kidding me.

    I have already performed the test and have my conclusion everyone of you that takes a crack at deciding if there is a mind behind it is doing a replication.

    why can’t you see that?

    peace

  22. fifthmonarchyman: discernible is not the same thing as infallible.
    A cancer screening can not prove you have cancer it simply shows whether you have certain biological markers that are associated with cancer.

    My test is no different

    There is a difference between a test and non-test. A test tests something defined or discernible. Yours doesn’t.

    fifthmonarchyman: No I can demonstrate it with my game by showing that we can distinguish the chart I posted from a randomized grouping of same data. The only way you could do that if there were no pattern is by memorizing every data point

    When will we get to see that demonstration? And, while at it, what does the demonstration demonstrate? A (non-)mental process? How? What is the relevance of the distinction of a chart of data from a different chart of the same data?

    fifthmonarchyman: I have already performed the test and have my conclusion everyone of you that takes a crack at deciding if there is a mind behind it is doing a replication.

    why can’t you see that?

    Can you try it again in English? Or perhaps better in your native language. It could help to clarify where you went wrong.

  23. fifthmonarchyman:
    newton: Elimination of the similarities should make the differences more pronounced.

    bingo

    To do that you ,like a Turing test, need both viewpoints solving the same problem or whatever your graph represents. Apples to apples.

  24. Neil Rickert: Yes, I’m sure that is correct — for both of the real Captains. Except that the two of them won’t agree with each other.

    and only one of them can be correct the other just thinks he Captain but is in fact the result of a transporter accident all his memories up to that point are false they did not happen to him.

    We call the correct one with the real memories the real Captain

    Peace

  25. I have a question for those who are demanding more context before they can make a determination

    A recently conducted standard Turing test consisted in a five minute “cold” conversation conducted with both a human and a machine after which the judge was expected to determine which is the person.

    Is that enough context for you? please explain your answer

    peace

  26. newton: To do that you ,like a Turing test, need both viewpoints solving the same problem or whatever your graph represents. Apples to apples.

    interesting,

    Do you think all persons will behave exactly the same way in solving a problem? What about all algorithms? How do you know you have the best representative of both?

    The poker players were able to identify the non-personlike behaviors in the bot by comparing it’s behavior to a mental generalization of all the games that they had played with humans up to that point.

    Why is this sort of thing not sufficient in your view?

    peace

  27. fifthmonarchyman:
    I have a question for those who are demanding more context before they can make a determination

    A recently conductedstandard Turing test consisted in a five minute “cold” conversation conducted with both a human and a machine after which the judge was expected to determine which is the person.

    Is that enough context for you? please explain your answer

    peace

    Source, please.

  28. Erik: And you don’t see how your test fails to be a test compared to that one?

    My test is not the same as that one. I never claimed it was.

    In fact the whole idea is to develop a test that is not constrained by verbal communication. If you have suggestions for improvement I’m all ears

    peace

  29. fifthmonarchyman: and only one of them can be correct the other just thinks he Captain but is in fact the result of a transporter accident all his memories up to that point are false they did not happen to him.

    You seem to be expecting the world to fit your ideas of logic.

    I have no such expectation.

  30. fifthmonarchyman: If you have suggestions for improvement I’m all ears

    No you aren’t, because you weren’t when I gave you suggestions. Namely: Context and definitions!

    In the test you linked to, it says, “‘Eugene’ simulates a 13 year old boy…” Now that’s something that really puts the subjects to the test. In comparison, what are you giving to us? A graph from nowhere claiming a “recognizable pattern” which is conspicuously not there. There’s no pattern in your graph. Certainly not recognizable or “easy to see”. Moreover, patterns are patterns of something, just like in the linked test the simulation was a simulation of someone, as it should be. Yours is a pattern of nothing to nothing for nothing.

    With this test of yours, you are testing only people’s patience here. And insulting intelligence at the same time. And it seems it’s by design.

  31. Erik: In the test you linked to, it says, “‘Eugene’ simulates a 13 year old boy…” Now that’s something that really puts the subjects to the test. In comparison, what are you giving to us?

    the judges did not know this till after they completed the test

    Erik: There’s no pattern in your graph. Certainly not recognizable or “easy to see”.

    Of course there is, You are the only one here who is not acknowledging that. If this was an actual test I would be happy to demonstrate the presence of a pattern by the method I detailed. It’s not that difficult to do so but it does require some effort.

    Erik: Yours is a pattern of nothing to nothing for nothing.

    When I reveal the source of the data in question the silliness of this comment will be obvious

    peace

  32. raw data

    1-0.035398,2-0.09434,3-0.018692,4-0.018519,5-0.035608,6-0.090361,7-0.053097,8-0.070588,9-0.016713,10-0.015873,11-0.030691,12-0.029703,13-0.014184,14-0.013575,15-0.026374
    16-0,17-0,18-0,19-0.04008,20-0.014354,21-0.006455,22-0.023881,23-0,24-0.020339,25-0.009467,26-0.008856,27-0.02106,28-0.040984,29-0.011502,30-0.007177,31-0.024014,32-0.006466
    33-0.012301,34-0.008772,35-0.034302,36-0.024954,37-0.026991,38-0.042743,39-0.034106,40-0.012626,41-0.016956,42-0.042994,43-0.035088,44-0.03914,45-0.100921,46-0.037736,47-0.092721
    48-0.061175,49-0.023669,50-0.061856,51-0.044688,52-0.018141,53-0.031384,54-0.049643,55-0.054763,56-0.048128,57-0.053189,58-0.05614,59-0.054299,60-0.077922,61-0.04881,62-0.021437
    63-0.016207,64-0.034404

  33. I will post the source of the data as soon as Richardthughes has a chance to chime in

    peace

  34. fifthmonarchyman: the judges did not know this till after they completed the test

    False. The people in the test knew that they were dialoguing either with a human or not. This is much more than we can say about your alleged pattern. What is it a pattern of? We are not told. What does it mean to say that there is a pattern? We are not told again. I can agree that it has a shape, but any shape can be called a pattern, so what makes your pattern special or relevant to anything? To solve this, you would need to say what it is a pattern of. And what relation does all of this have to detecting or inferring “design”? Is “design” the same thing as “pattern” or a different thing? We are not told again.

    The fact that you are not telling these absolutely necessary things demonstrates that you don’t know what a Turing test is. Or a test of anything, for that matter. You fail to provide the minimal necessary context and content to make this a test.

    fifthmonarchyman:
    If this was an actual test I would be happy to demonstrate the presence of a pattern by the method I detailed

    Correction: The method that you have not detailed despite of multiple requests over many pages now. But I get it: This is not an actual test. And you are not all ears to help you make it a test.

  35. fifthmonarchyman: interesting,

    Do you think all persons will behave exactly the same way in solving a problem? What about all algorithms? How do you know you have the best representative of both?

    There better be some common ” personness ” else you are measuring something else

    The poker players were able to identify the non-personlike behaviors in the bot by comparing it’s behavior to a mental generalization of all the games that they had played with humans up to that point.

    They also studied other bots, the schedule favored a player who did not require sleep and was able to focus consistently, and the programmers had a strategy to maximize their strengths . No known person is capable of that

    Why is this sort of thing not sufficient in your view?

    Because the goal was to win at poker not create a personlike player. If one of those players adopted the strategy and mastered the strategy, would the strategy become personlike?

    If you graph the strategy somehow, would his graph match the bot?

    peace

  36. Erik: The method that you have not detailed despite of multiple requests over many pages now.

    what are you talking about?

    I’ve repeatedly said we can simply see if we distinguish the data string in the chart from among a group of strings of the same data that has been randomized. If there is no pattern then the only way we will be able to identity the “chart” string is by memorizing every single piece of data.

    I think this is the forth time I’ve said that

    Erik: I can agree that it has a shape, but any shape can be called a pattern,

    So no you agree that the data can be described as a pattern

    Erik: what makes your pattern special or relevant to anything?

    It’s simply a pattern we are testing to see if we can decide whether is was produced by a “mindful” process.

    Erik: To solve this, you would need to say what it is a pattern of. And what relation does all of this have to detecting or inferring “design”?

    To solve what????

    Erik: Is “design” the same thing as “pattern” or a different thing? We are not told again.

    once again—– we are looking for behavior that is non-personlike in the pattern and if we find it we are saying that we can’t determine that the pattern’s source was a person. I explained this probably 5 times

    Erik: The fact that you are not telling these absolutely necessary things demonstrates that you don’t know what a Turing test is.

    since I have repeatedly “told” these things does that mean that you are not paying any attention?

    Erik: The method that you have not detailed despite of multiple requests over many pages now. But I get it: This is not an actual test. And you are not all ears to help you make it a test.

    do you need a nap?

    peace

  37. GlenDavidson: That they’re indistinguishable is the problem, not the solution…

    My understanding is that if you claim your zombie is identical to the last molecule with your real person, then there is no way to distinguish the zombie from the real person. As I am not a dualist, I have no problem in thus dismissing the thought experiment as useless. If someone else is convinced there is an immaterial aspect to consciousness, then it’s their problem.

  38. newton: There better be some common ” personness ” else you are measuring something else

    I think there are as witnessed by the fact the we are not rushing to marry our toaster ovens

    newton: The poker players were able to identify the non-personlike behaviors in the bot by comparing it’s behavior to a mental generalization of all the games that they had played with humans up to that point.

    exactly they had an idea of what common personness looks like
    we all do

    newton: If one of those players adopted the strategy and mastered the strategy, would the strategy become personlike?

    Did you watch the video?

    they say that they think that it’s impossible for persons to consistently behave in the manner that the computer behaved

    newton: If you graph the strategy somehow, would his graph match the bot?

    If I graphed the bots strategy it would match the bot.

    I think It would be pretty easy to do so by simply recording the bet’s relationship to the value of the cards dealt in each hand.

    peace

  39. Alan Fox: My understanding is that if you claim your zombie is identical to the last molecule with your real person, then there is no way to distinguish the zombie from the real person.

    Then you acknowledge the possibility of zombies. If zombies are possible then you can’t prove that a bot is conscious and not an unconscious “zombie”. This is not difficult

    Alan Fox: As I am not a dualist, I have no problem in thus dismissing the thought experiment as useless.

    I don’t think you understand,

    By saying that zombies are indistinguishable from conscious persons you agree exactly with the position I am taking.

    peace

  40. Erik: Given the two-Kirks-from-the-transporter scenario, the dualist would be in the tough position of positively asserting that one of the Kirks must be a zombie or a fake or a liar, but the materialist is in the tough position of asserting that the scenario yields the exact same person twice.

    That state of identity holds only at the instant of the second transporter download. From that instant on the two captains become more and more different. They’re in different positions even if they were chained together, they will perceive events from slightly different viewpoints, look in different directions and if separated by no more than a short distance will have unique experiences. Two different but real and complete Captain Kirks. If transporters were really real!

  41. fifthmonarchyman: By saying that zombies are indistinguishable from conscious persons you agree exactly with the position I am taking.

    My basic point is the zombie argument is nonsense. But you can’t have it both ways. Distinguishable or not distinguishable – but not both.

Leave a Reply