Several regulars have requested that I put together a short OP and I’ve agreed to do so out of deference to them. Let me be clear from the outset that this is not my preferred course of action. I would rather discuss in a more interactive way so that I can learn from criticism and modify my thoughts as I go along. OPs are a little too final for my tastes.
I want to emphasize that everything I say here is tentative and is subject to modification or withdraw as feedback is received,
It’s important to understand that I speak for no one but myself it is likely that my understanding of particular terms and concepts will differ from others with interest in ID. I also want to apologize for the general poor quality of this piece I am terrible at detail and I did not put the effort in I should have due mainly to laziness and lack of desire.
With that out of the way:
Background
For the purpose of this discussion I would like to expand upon the work of Phill Mcguire found here and stipulate that cognition can be seen as lossless data compression in which information is integrated in a non-algorithmic process. The output of this process is a unified coherent whole abstract concept that from here forward I will refer to as a specification/target. Mcguire’s work thus far deals with unified consciousness as a whole but I believe his incites are equally valid when dealing with integrated information as associated with individual concepts.
I am sure that there are those who will object to the understanding of cognition that I’m using for various reasons but in the interest of brevity I’m treating it an an axiomatic starting point here. If you are unwilling to accept this proviso for the sake of argument perhaps we can discuss it later in another place instead of bogging down this particular discussion.
From a practical perspective cognition works something like this: in my mind I losslessly integrate information that comprise the defining boundary attributes of a particular target; for instance,”house” has such information as “has four walls”, “waterproof roof”, “home for family”, “warm place to sleep”, as well as various other data integrated into the simple unified “target” of a house that exists in my mind. The process by which I do this can not be described algorithmically. from the outside it is a black box but it yields a specified target output: the concept of “house”.
Once I have internalize what a house is I can proceed to categorize objects I come across into two groups: those that are houses and those that are not. You might notice the similarity of this notion to the Platonic forms in that the target House is not a physical structure existing somewhere but an abstraction.
Argument
With that in mind, it seems reasonable to me to posit that the process of design would simply be the inverse of cognition.
When we design something we begin with a pre-existing specific target in mind and through various means we attempt to decompress it’s information into an approximation of that target. For instance I might start with the target of house and through various means proceed to approximate the specification I have in my mind into a physical object. I might hire a contractor nail and cut boards etc . The fruit of my labor is not a completed house until it matches the original target sufficiently to satisfy me. However, no matter how much effort I put into the approximation, it will never completely match the picture of an ideal house that I see in my mind. This is I believe because of the non-algorithmic nature of the process by which targets originate. Models can never match their specification exactly.
Another good example of the designing process would be the act of composing a message.
When I began to write this OP I had an idea of the target concept I wanted to share with the reader and I have proceeded to go about decompressing that information in a way that I hoped that could be understood. If I am successful after some contemplation a target will be present in your mind that is similar to the one that exists in mine. If the communication was perfect the two targets would be identical.
The bottom line is that each designed object is the result of a process that has at its heart an input that is the result of the non-algorithmic process of cognition (the target). The tee shirt equation would look like this
CSI=NCF
Complex Specified Information is the result of a noncomputable function. If the core of the design process (CSI) is non-computable then the process in its entirety can not be completely described algorithmically,
This insight immediately suggests a way to objectively determine if an object is the result of design. Simply put if an algorithmic process can fully explain an object then it is not designed. I think this is a very intuitive conclusion, I would argue that humans are hardwired to tentatively infer design for processes that we can’t fully explain in a step by step manner. The better we can explain an object algorithmically the weaker our design inference becomes. If we can completely explain it in this way then design is ruled out.
At some point I hope to describe some ways that we can be more objective in our determinations of whether an object/event can be fully explained algorithmically but as there is a lot of ground covered here so I will put it off for a bit. There are also several questions that will need to be addressed before this approach can be justifiably adopted generally such as how comprehensive an explanation must be to rule out design or conversely when we can be confident that no algorithmic explanation is forthcoming.
If possible I would like to explore these in the future perhaps in the comments section. It will depend on the tenor of feed back I receive.
peace
fifthmonarchyman,
Evidence, please.
I’ve unstickied this – please bookmark it if you want to continue the conversation!
I’ve been following the discussion. You have made no attempt to define “explain” in the context of explaining an object. Nor have you attempted to define “losslessly integrating”. If I’m wrong, demonstrate with a link or quote.
Obviously you can’t.
The meaning of “expand” is a little fuzzy there, but it’s a try. How do you detect a correct “expansion”? Since the algorithm to produce a given string is not unique, why can’t an observer make up his/her own algorithm?
fifthmonarchyman: Google is your friend
I see nothing there about tests to determine whether integrating is happening or not; it’s about consciousness. Does “integrating” mean “conscious” and “not integrating” mean “unconscious”?
So your proposed definition is “If a human does not recognize it as a pattern but a current neural network does, it’s not losslessly integrating; anytihng else is losslessly integrating”?
Seems to lack a certain je ne sais quoi,… maybe it’s the fact that it’s going to turn up a lot of false positives.
Again, what tests do wwe apply? An example is useless, series of tests is. You need ot look up “operational defnitiion”. I’ll help:
“An operational definition is a result of the process of operationalization and is used to define something (e.g. a variable, term, or object) in terms of a process (or set of validation tests) needed to determine its existence, duration, and quantity.[1][2] Since the degree of operationalization can vary itself, it can result in a more or less operational definition.[3] The procedures included in definitions should be repeatable by anyone or at least by peers.”
JonF says,
I say,
You honestly don’t have a clue what we are talking about do you?
Why don’t you go back and actually read the original paper I posted and while you are at it Google Integrated Information Theory.
After you have done that if you still have questions and can demonstrate that you are at least trying to understand by giving a summery of my method and why I think it is valuable it I will be happy to answer any questions you have.
It’s nothing personal and I don’t want to sound harsh but if you want to talk you need to put forth at least some effort.
peace
EL says,
I knew it was just a matter of time. I only wish we could have had more productive back and forth while we had the chance.
Thank you for the opportunity
peace
All,
Apparently Google drive does not allow the sharing of charts. The moving line chart is the heart of the entire game. So this will not work for us
below is the address if someone want’s to take a crack at it
https://docs.google.com/spreadsheets/d/1GhHTyOoACPHBv9XXjDbr-SZs_FnSixDd90Obr6DZxwQ/edit?usp=sharing
peace
Patrick says,
I say
Since you asked I would like longer strings. It will minimize the chances for cheating on the observers part or a inclusive result. I can run the strings you gave me if you want me to see what we get but longer is better.
Numeric character strings work better than binary. So far I’ve mostly been working with strings at least 500 characters long
I can use binary it just needs to be longer still to see the patterns.
you say,
I say,
give me a way to reach you and I will be happy to share my spreadsheet with you so you can see for your self. My method is pretty simple.
Maybe OMagain will help me to make it public.
peace
fifthmonarchyman,
Not quite. They make a much weaker conjecture in the conclusions:
Human accuracy averages around 73%, so people cannot “always tell the difference.”
I don’t see anything in the paper that suggests there were two different randomization processes. Can you point to the page where this is stated? They used eight different sets of price data, but it looks like the same randomization process was used for each.
This claim is not supported by the paper.
That was not done in the paper and it makes your claim untestable. Convenient if your goal is to not be proven wrong, but useless if you are genuinely interested in testing your ideas.
This leads into an earlier part of our discussion. You’ve emphasized this quote from the paper:
I proposed a five step procedure to test the hypothesis that software is unable to demonstrate the same capabilities as the humans in the paper. You’ve suggested adding additional steps, but have been unable to justify doing so on logical or mathematical grounds.
So please explain to me why following exactly the same procedure described in the paper, using a software agent instead of a human agent, and getting equal or better accuracy would not refute the claim you are making based on this paper.
As far as can tell you have two options. Either:
1) Your claim based on this paper is not an essential component of your argument, in which case we can drop any discussion of that claim and the paper
OR
2) Demonstrating that software can meet or exceed human performance on this problem refutes your claim and therefore your argument.
Which is it?
fifthmonarchyman,
Looks like one bucket to me.
fifthmonarchyman,
You can only expand Pi if you know an algorithm to do so.
An algorithm that produced the string “Pi, calculable by foo” where foo is your favorite algorithm for computing Pi would therefore meet your definition of “explain”.
What would it mean to “expand” a sonnet?
So an algorithm that produces the string 8675309 meets your definition of “explains”, since the extra information is in your head regardless of how the string got to you.
More generally, since any finite string can be generated algorithmically, any non-infinite string that meets your definition of “explain” can be so generated.
What are you really trying to say when you use the word “explain”? It seems like you are trying very hard to exclude anything other than human brains without explicitly putting that in your definition.
fifthmonarchyman,
Hey, the thread is still active. Just bookmark the title and try to keep up!
fifthmonarchyman,
That made my head hurt.
I’ll try to generate longer strings tomorrow. How many bits are optimal?
Patrick says,
I say,
You are right. I misremembered
What the paper did was vary the data sets and not the randomization process. When I do it I vary both so I confused my stuff with theirs. Sorry about that
So I take it back. All you will need to do is work with one randomization process and but vary the input data. It should be easier for you
you say,
How about this, The paper demonstrates how my method works.
If your software can meet or exceed human performance in this task it will in my opinion demonstrate that it is capable of Information integration as detailed in the first paper and I will need to either acknowledge that the software is conscious or change my understanding of what consciousness entails.
That would take a lot of the wind out of the sails of my argument
peace
Patrick says,
I say
There is no hard and fast rule
It seems like optimal would be at least 1500 numeric characters. whatever you can do to approximate that.
Thanks this should be fun
peace
You can post a description of what is required as a comment. However if the spreadsheet represents what would be required, that is obviously sufficient now it is shared.
Binary would be numeric. LOL.
Patrick says,
I say
Or if I’ve memorized the digits, or have them written down somewhere, or know some one who knows them or an algorithm to produce them
I say,
Yes it would,
Notice how the “target” Pi is in the algorithm itself. That is what I’m getting at. The only algorithms that are explanatory are those that contain the target and if you have the target you can use any algorithm you choose so the algorithm that actually produced the string is unimportant.
So it’s the target not the algroythym that does the explaining.
you say,
I say,
I could produce another one in the same style or I could find another one by the same author
you say,
I say,
no the algorithm does not explain the string the information in the target explains the string.
The extra information is not in necessarily in my head it might be in an old CD or on youtube. I would say it ultimately exists outside the cave but I understand you don’t want to go there.
You say,
I say,
No I’m trying hard to exclude the algorithmic. My approach is not anti-materialism or exclusively anthropomorphic. It’s anti algorithmic. My hope is to develop a way to detect design that is generalized beyond “made by a human”
peace
petrushka says,
I say,
10 thousand comedians out of work and there you go cracking wise.
lol
peace
OMagain says.
The problem is that Google does not share the macros or the chart itself.
All we have is some assorted data that I have been playing with and the non functioning interface It seems to be rather worthless.
What I need is a game like that in the financial paper I posted but that allows you to input any strings you choose it seems like it would be easy breezy but I’m finding it to be very tedious.
peace
Been reading intently, but still don’t get it. Can someone explain what fifthmonarchyman’s project is supposed to do?
I think it involves eyes and wool.
As a programmer I need a detailed specification 🙂 and I’d prefer one written in your own words as otherwise, what is it I’m implementing?
If you can write out a series of steps, leaving nothing out, as to exactly what the process is then I can implement that in a browser. Referencing a paper and saying “something like that” I feel would just be a waste of my time (as you are not a paying client! Paying clients can be a vague as they like) as it will never be what it should be done like that.
It’s your argument, and it’s clear that nobody here has a clue about what you mean by those terms. It’s your responsibility to be clear. You give examples of what you mean, but those are not operational definitions, and nobody has been able to figure out from your examples what such a definition might be.
Examples are not definitions.
Obviously you don’t have a clue what you mean. You’re just handwaving.
It’s impossible to memorize the digits of pi or write them down. (You can, of course, memorize or write down N digits, but that helps you not at all in expanding pi to N+1 digits). Knowing someone who has an algorithm is exactly equivalent to having an algorithm. Therefore the only way to expand pi to any arbitrary length is with an algorithm.
The Internet is choked with fre file sahring services such as https://mega.co.nz,
Nope, especially fifthmonarchyman.
OMagain says,
Do you want the step by step of the method? Ive posted that repeatedly here just scroll back a little bit and you will see it
Do you want a step by of the game itself. here goes
The Chart
1) I need a 2 parallel line charts like an ordinary process control charts but moving so that only a few points are visible at a time. Sort of like a medical EKG monitor
2) I need a start stop button to halt the advancement of the charts when the observer thinks he can distinguish the difference between the strings
3) I need the ability to load any numeric string into the charts that I choose
the control
1) I need a separate page to enter my guess as to which is the real string
2) The game should then tell me if I chose correctly and keep track of how many tries it takes to learn the difference between the strings
3) finally I need a reset button that randomizes the charts and the starting position. when I repeat the process
peace
PS I’m telling you this would all be so much simpler if I just sent you the spread sheet I’m using.
JonF says,
I say,
It’s not an argument and they are not my terms they are the terms that are used in the field of IIT . If you want to know what they mean you can easily look at any of the literature floating around about IIT.
After that if you think I’m using them in a way that is not standard just ask for clarification like Patrick did with the term explain.
The very fact that you call this an argument shows that you could care less about what I have to say anyway. I have repeatedly stated my purpose here. I do it every couple of posts
I am not trying to convince you or anyone else of anything. If you don’t feel that I’m making sense feel free to ignore me,
I honestly don’t give a rip. If however you feel like you would like to discuss get off your duff and look it up.
peace
fifthmonarchyman,
Just to be clear, we’re talking about the process I described earlier:
1. Create up to 1000 pairs of time-series data as described in the paper.
2. Create either a recurrent neural network that accepts the time series data directly or a support vector machine that takes in statistics directly generated from the data.
3. Train the network with 800 or so pairs of time-series data and validate it with the remainder.
4. Create a few hundred new pairs of time-series data for testing.
5. Demonstrate that performance on the test set meets or exceeds the human performance described in the paper.
Correct?
Thank you for making a testable claim. You’re the first ID proponent I’ve encountered who is willing to do so.
fifthmonarchyman,
But the fact remains that any finite string can be produced alorithmically. Since you agree that the finite string I described “explains” Pi, by your definition, you must recognize that an algorithm can “explain” Pi.
Any finite string can be produced algorithmically. The only way to exclude algorithms is to restrict “explain” to infinite strings.
Why are you trying to “exclude the algorithmic”? Why is it important to your argument?
Patrick says,
The only qualification I would make is I would want the strings to represent different data sets and not the same “market” at different times.
you say,
I say,
Because if we can say that cognition is non algorithmic and that design is the inverse of cognition then this gives us a concrete objective way to distinguish things that are designed from things that are not.
peace
fifthmonarchyman,
Hmm. We need to have a talk about the definition of “inverse” soon. In the meantime, I’ll generate some longer strings and start hacking some neural nets.
Patrick says,
I say,
There you go again with “by your definition”.
My definition is tentative don’t hold me to it yet please.
I understand what you are saying but I’m not just sure I agree. Pi is not like an ordinary finite constituent part of an algroythym but instead is an transcendental irrational constant. It’s addition to the algorithm changes the “finiteness” of it somehow
I think you are on to something with your mention of infinite strings.
That is really the Idea I’m trying to get across here, Even though the string 3.14159 is finite it represents the infinite string of Pi to the observer. That is what I was getting at earlier when I clumsily said that the Kolmogorov complexity of Pi was infinite.
I’m not sure how to get this idea across but I know there is something to it.
I’ll buy that the definition needs more work but the concept is sound imho.
You say,
I assure you this is my ultimate goal here. I like Philosophy as much as the next guy but I’m interested in science with this endeavor.
peace
fifthmonarchyman: Even though the string 3.14159 is finite it represents the infinite string of Pi to the observer.
It would be clearer if you added ellipses, 3.14159…
fifthmonarchyman: That is what I was getting at earlier when I clumsily said that the Kolmogorov complexity of Pi was infinite.
Pi’s Kolmogorov complexity is finite, however, it is transcendental, meaning it’s not the root of any integer polynomial. However, there are a number of infinite expansions that equal Pi.
http://mathworld.wolfram.com/images/equations/PiFormulas/NumberedEquation3.gif
Zach, 5th seems to have trouble understanding that the only way to have any actual digits of pi is to apply an algorithm, and that regardless of how many digits you have to compare, the algorithm can always supply one more. One can easily make the procedure halt by placing a limit on the number of digits or iterations.
Guys perhaps a more concrete example might help you you to understand what I’m getting at.
Suppose using a scanner and 3D printer we built an algorithm to reproduce the Antikythera mechanism. Would the algorithm explain the Antikythera mechanism? Of course not
now look at my tentative definition
Explain- To describe the patterns in the string in such a way as enable an observer to reproduce and expand them with out reference to the actual steps of the algorithm.
Would the algorithm help you to build a device that was similar but not identical to the original. I don’t think so.
On the other hand if you understood what the Antikythera mechanism was supposed to do you could easily build something like it using any process you choose to.
that is what I mean by explain
peace
Would listing the incremental steps leading to the string, along with the physical causes of each modification qualify as an explanation?
petrushka asks,
No,
That is just another way of saying algorithm.
Would listing the incremental steps leading to a reproduction of the Antikythera mechanism , along with the physical causes of each modification qualify as an explanation of the Antikythera mechanism ?
peace
This makes no sense.
You talk about strings, then you talk about artifacts. Those are very different things. “Similar” does not apply to strings. There is a “compare” instruction in your computer, but there is no “similar” instruction. Similarity is a matter of human perceptual judgment, and is sensitive to context and to cultural traditions.
Based on your examples, your poorly defined “explain” seems to depend on cultural tradition. So talk of an algorithm explaining is likely nonsensical.
Need I point out that it is typical for humans to have 130 copy errors in their individual genome strings. What explains this?
Neil Rickert,
Soundex is close.
fifthmonarchyman,
The problem you are facing is that any finite string that you agree “explains” the Antikythera mechanism can be generated algorithmically. So either you’re saying that it can’t be explained by a human or it can be explained by an algorithm. Which is it?
Neil Rickert says,
Any phyiscal object can be represented by numeric string. simply measure the perimeter of the object at discrete distances and record the measurements and you have a string.
peace
Patrick says,
I’m not sure I agree with that. I will agree that an algorithm can produce any finite string but produce does not equal explain.
The only algorithm that can explain a designed object is one that contains the target itself. And if you have the target the algorithm is unnecessary and superfluous
peace
fifthmonarchyman,
If the string produced “explains” the artifact according to your definition of “explains”, on what basis do you claim that the algorithm does not explain the artifact?
You need to demonstrate this, not just assert it.
Representing depends on human conventions.
Do you expect the algorithm to “explain” the string as an entity in its own right? Or do you want it to “explain” the original thing or the representation as formed from the original thing.
You cannot have the target for pi.
Pi is the target,
the ratio of a circle’s circumference to its diameter
peace
ps I hate block quotes