I want to thank OMagain in advance for doing the heavy lifting required to make my little tool/game sharable. His efforts will not only speed the process up immeasurably they will lend some much needed bipartisanship to this endeavor as we move forward. When he is done I believe we can begin to attempt to use the game/tool to do some real testable science in the area of ID . I’m sure all will agree this will be quite an accomplishment.
Moving forward I would ask that in these discussions we take things slowly doing our best to leave out the usual culture warfare template and try to focus on what is actually being said rather than the motives and implications we think we see behind the words.
I believe now would be a good time for us to do some preliminary definitional housework. That way when OMagain finishes his work on the gizmo I can lay out some proposed Hypotheses and the real fun can hopefully start immediately.
It is always desirable to begin with good operational definitions that are agreeable to everyone and as precise as possible. With that in mind I would like to suggest the following short operational definitions for some terms that will invariably come up in the discussions that follow.
1. Random– exhibiting no discernible pattern , alternatively a numeric string corresponding to the decimal expansion of an irrational number that is unknown to the observer who is evaluating it
2. Computable function– a function with a finite procedure (an algorithm) telling how to compute the function.
3. Artifact– a nonrandom object that is described by a representative string that can’t be explained by a computable function that does not reference the representative string
4. Explanation –a model produced by a alternative method that an observer can’t distinguish from the string being evaluated
5. Designer– a being capable of producing artifacts
6. Observer– a being that with feedback can generally and reliably distinguish between artifacts and models that approximate them
Please take some time to review and let me know if these working definitions are acceptable and clear enough for you all. These are works in progress and I fully expect them to change as you give feedback.
Any suggestions for improvement will be welcomed and as always please forgive the spelling and grammar mistakes.
peace
so it is interesting but not especially relevant to the game
peace
fifthmonarchyman,
Support your claim or retract it. It’s not up to other people to make your case for you.
It’s a good thing I preemptively explained that I am terrible at spelling and grammar.
Show me a person preoccupied with spelling and grammar and I’ll show you some one who often misses the big picture 😉
peace
It’s not that I need help to support my case. I want help to make the tool shareable.
Once that happens I expect the case will be obvious.
peace
fifth,
Right. Your memory for spelling and grammar is extremely lossy.
Yet your case rests on the assumption that memory is “vastly non-lossy”. You are a walking counterexample.
fifthmonarchyman,
You have yet to explain the connection between your string game and your claim that snowflakes are designed. Please support that claim or retract it.
Maybe not to the game, but very relevant to design detection on the basis of distinguishing non-computable functions. Software produces (‘computes’) undesigned music that causes a false positive design detection, so your basic assumption is failing already.
And before you go there, the fact that the software is designed doesn’t mean that its output is therefore also designed. If you maintain it is, you should be consistent and accept evolution as a designing entity too.
fG
You missed when science supports hypothesis with evidence.
Your premises are your hypothesis, you can’t just claim your game counts as evidence because your game relies on your premises/hypothesis so it can be evidence for itself
I don’t intend to go to the tired old culture war script. I find it boring and unfruitful.
I would hope we could stick to science as much as possible.
I suppose If you wanted to test music styles against software models in my game you would need to represent a musical style as a numeric string.
I think that is doable. and it would be an interesting exercise
peace
Every result in the game is evidence,
Every time an observer distinguishes between an artifact and a fake it supports the hypothesis that observers can always distinguish between artifacts and fakes
Every time a model fails to fool an observer it supports the hypothesis that artifacts are noncomputable
I love science
peace
Use mp3 files. Of course your software won’t be anywhere as sophisticated as Iamus or Melomics109 so it would be a poor test of the hypothesis that design is non-computable. But why bother? The fact that knowledgeable people can’t tell the difference anymore between music composed by humans or by software has put that theory to bed anyway.
fG
FG
I’ve discussed modifying the game to look at 4 strings instead of 2.
I suppose we could load 3 real pieces of music of a particular style and one fake designed to mimic them.
I would guess that with feedback an observer could pick out the fake.
Would you be willing to bet that an observer could not distinguish the fake from the real ones?
again I think it would be an interesting exercise.
peace
As I said, your game is a very crude tool to synthesise compositions so I am sure people could tell the difference.
However, as per my earlier reference people fail to distinguish the output from the latest high-end composition software from actual human compositions. What you propose has been done and your hypothesis has failed the test. What is the point of repeating it with an inferior tool?
fG
Are you defining “poor” as any test that software can not pass?
It seems to me that if I can distinguish the real from the fake using my method and you can not then my method is superior to yours
peace
The point is, your fakes are not the best that computer generated designs can offer nowadays.
If you want to test the hypothesis that fakes (i.e. computable functions) can be distinguished reliably from originals (i.e. human designs), you should use the most sophisticated fakes possible . Not crude ones that anybody can instantly spot as artifacts. All that demonstrates is that you are a poor designer of design software.
fG
Yet more reason to take this thread as pure comedy.
I think he has automated “moving the goal posts.”
No, that would only show that dictionaries are lossy. The original is what is in FFM’s memory (or so he tells us).
LOL.
OK figure out how to represent the original and your sophisticated fake as a numeric string and we can test it.
Any phyisical object or phenomena can be represented that way you just need figure out how to do it.
If you are claiming that your “sophisticated fake” can’t be represented as a numeric string then we have a different sort of dilemma on our hands,
Is that your claim?
peace
I think perhaps instead of it being a case of me moving the goal posts it’s a case of you imaging goalposts that are nothing of the sort.
Sort of like this
http://www.nfl.com/videos/buffalo-bills/0ap2000000148850/27-Leon-Lett-fumble-in-Super-Bowl-XXVII
peace
You can download mp3 files of the computer composed music, and of similar music in the same style composed by humans. Then run them through your software and see what happens.
Of course I already reported that audiences couldn’t tell them apart, so what is it again that you want to demonstrate?
No, that is not my claim. Computer composed music is available as mp3 files. Look on Youtube under Iamus. The piece ‘Hello World’ is a good place to start.
Actually, what would be interesting is if you could pretend that Hello World is a human design, run you game with fakes of it, and see if it would classify them correctly viz a viz the original.
Design comes in shades.
fG
I’m still basically drinking till next week, so you all carry on having a wonderful time and I’ll finish up my part (the magic APP) sometime next week.
you are the man baby
peace
Of course it often takes some effort and feedback to tell the real thing from a clever fake. If it did not counterfeiting would be impossible. This is not news
What (among other things) I want to demonstrate is that algorithmic mimicry only gets you so far and we can sleep easy knowing that Skynet and the Singularity are science fiction.
Whether my silly little game actually does that is an open question.
we shall see
peace
By the way, fmm, I want you to know that I am enjoying this conversation. Not because I think you are onto something, because frankly I don’t, but because it makes me look up stuff and discover things I didn’t know yet. The fact that you are arguing in good faith and without snark also helps a lot.
Let me be honest – until this afternoon I had never heard of Iamus. I am now downloading a complete album of its compositions and enjoying it. It is amazing how computer-composed music has come on in the last 20 years!
The way this is going it won’t be long before we will have software to create music like that of J.S. Bach. I won’t be the one writing it, far from it, but I will now stake my claim on what the software will be called:
C.P.U. Bach
fG
My game does not look for “human design” it looks for artifacts that are nonrandom and noncomputable.
If we can distinguish between “hello world” and nonrandom models that are close it only means that it is an artifact.
Something I’m sure the software developers would heartily agree to
peace
ditto back at you.
peace
By the way, Iamus’s Adsum is amazing.
fG
ETA: my link doesn’t show up? Anyway, search it on Youtube.
Umm… ‘Hello World’ is computable. It was written by a computer!
fG
you are confusing algorithmic with computable. Any finite string can be produced by an algorithm you just need to know what algorithm to use.
For example the string 8675309 (Jenny’s number) can be easily produced by taking 8675308 and adding 1.
The hard part is deciding to use that particular algorithm out of the infinite possible algorithms available.
When we say a particular string is noncomputable we mean that there is no finite recipe (algorithm) telling us which algorithm to use to produce it.
understand?
peace
I am slowly working my way through the 900+ comments on the thread “design as the Inverse of Cognition”, and I read this comment from fmm:
the Iamus’ composition ‘Hello World’ (or ‘Adsum’, which I prefer) is written by a computer and can therefore be fully explained by an algorithmic process. So, fmm would say that it is not designed.
Yet on this thread, fmm says:
So what gives? Is ‘Hello World’ designed, or not?
fG
No.
And see my comment above where you say that when an algorithmic process can fully explain an object, it is not designed. Iow, Iamus’s music is not designed. Yet, you call it an artifact.
fG
Designed
You are missing the the definition of explanation given at the top of this thread,
A big part of the confusion on that last thread was in not clarifying our definitions
that is why I wanted to start this one out right
peace
This might help
https://en.wikipedia.org/wiki/Computable_function
The function in this case is what we do to get to something that mimics a particular music style.
The software engineers spent lots of mental effort coming up with that particular algorithm.
That is the hallmark of design.
understand now?
peace
You are making no sense.
I love you man but you are going to have to elaborate a little if you want an answer. What part do you find nonsensical?
hint If you say “everything” you haven’t clarified anything
peace
Since when is the hallmark of design ‘the amount of mental effort spent in coming up with something’? That must rank as one of the most useless design inference criteria ever proposed!
But anyway, to take this forward, you now consider Iamus’s compositions designed – because a programmer put some effort in writing the program. Have you considered that by that criterion your ‘fakes’ are also designed? So your game distinguishes designed things from…other designed things. Very useful.
fG
By this definition, anything we produce will be designed.
oh the irony
Yes. You are more confused than I had realized.
Of course fakes are designed that is why counterfeiting is a crime and not an accident
peace
No, observers use my game to distinguish real strings from random strings and nonrandom models that are close.
If an object can’t be distinguished from a model we can’t infer design. There are lots of things that fail to pass the test.
peace
how so?
peace
I am getting tired of this. What now is a ‘real string’?
Look, the only reason I am posting in this thread is your claim that you have a design detection tool. I Now you say that lots of things could fail the test, even if they are designed like Iamus’s compositions, which are designed (according to you – not according to me, btw).
So give me some examples of things that pass the test, and preferably not things that we may reasonably expect to be computable in the not too distant future. Keep in mind that nobody 30 years ago would have foreseen that computers could write music that people can’t tell apart from music composed by humans.
I expect you will now fall back on ‘meaning’, and I will point out again that patterns and meaning are two very different things, and it does no good at all to confuse the two.
fG
Examples or STFU.
Finally. So if there’s always an algorithm that will produce any given sequence, what’s the point of all this nonsense?
My take is that it is all about detecting meaning and pretending that this can be done by analysing patterns stand-alone by themselves. This totally ignores the reality that meaning in strings only exists because of pre-existing conventions between sender and receiver, and this is the only reason why observers can tell ‘designed’ strings from ‘non-designed’ ones.
In cases where such conventions are unknown, or do not exist at all, it is impossible to reliably conclude ‘design’ from analysing mere patterns. This means that there is no such thing as a universal design detection tool, no matter how much ID enthusiasts wish it to be otherwise.
All the rest appears to be smoke and mirrors.
fG
you don’t quite understand
If what I do when I learn the pattern of a string is lossless information integration it’s my position that something that passes the test today will never ever be computable.
That is why mine is a testable hypothesis
If a particular style of music passes the test today computers will never be able to create a fake that is indistinguishable from the real thing. Provided the modeling algroythym does not target the actual “real” string itself.
you can pick any object you choose run it through the game and if it passes today you can consider it a prediction that it will pass forever.
how is that for sticking my neck out there?
peace
more later ….family fun and all that
Because there is no finite procedure for finding the magic algroythym.
peace
By using “non-computable” in a way that is different from what is normally meant by that term, yet expecting this to communicate your ideas.