A working version of FMM’s design detection game is available.
Download and install the applicable version of “Processing”.
https://processing.org/download/?processing
Get the fifthmonarchyman progam code from here, and paste it into the Processing script area.
Sample data here
http://pastebin.com/raw/MjV8RmvW
You need two files in the same folder as the Processing executable.
real.txt and fake.txt
The testing and such starts here
test strings
I see no reason to conclude that conscious contents require lossless memory,
On the other hand I would say that we humans integrate information losslessly and this is something computers don’t do.
Yes if the mathematics are disproved they will be disproved. Your point is?
The paper explores this possibility and concludes that it can happen because it would result in a contradiction, Do you disagree?
No a function is notcomputable if it is noncomputable.
Integrating functions are mathematically shown to be noncomputable
Recall the definition of an integrating function.
the knowledge of m(z)does not help to describe m(z′),when z and z′are close.
in this context
The knowledge of the real newton does not help to describe African nun newton.
They are not remotely the same person. They are different We would never confuse the two.
The person we call newton is a unified consciousness and as such he is noncomputable.
Peace
A pattern is “right” if it allows you to distinguish between the two strings
peace
The W is characteristic of what?
Then you probably have a broken game. To be sure click the space bar a few times to see if the position of the strings change.
They should randomly flip every time the game is reset
I’ve provided instructions.
You can try them and ask for clarification as needed or you can wait till the shareable game is available or you can forget it all and go back to the same old culture war discussions.
It’s really up to you
peace
In this particular iteration of the game the W tells you that you are looking at the “real” string.
If you choose the string with the W pattern you will be right 100 percent of the time
peace
I see the W in only one of my data sets. I provided a second submission with for sets. One real and three with the same data scrambled in various ways. I don’t see that full swing W in any of them.
Ok,
As I said the W pattern is valid in only this iteration of the game.
Every single game will be different. If the pattern held for all iterations then a computer could generally and reliably distinguish between the strings. It’s the unpredictability and uniqueness of the pattern that makes this enterprise an alternative Turing Test.
The “right” pattern will probably be different with each of your other data sets.
To find it simply make a guess and see if you are correct.
It usually takes a few wrong guesses before you “learn” the pattern but once you get it you get it
peace
I did some quick looking at your second set of data found here
If I had to guess I would say that the third string is the “real” one. The pattern I saw was extended plateaus at the lowest level of 10 that were not present in the other strings. What did you see?
I could not easily distinguish the other strings so I am guessing that they are random. Once I found one with a clear pattern I quit
Keep in mind this is not what the game is designed to do.
I’m not used to looking at 4 strings at once and I did not spend near enough time to come to a firm conclusion. Most importantly I am not used to looking at 3 random and one real It would not surprise me if I’m all wet with my inference
but the third string seemed different than the others.
I’m really interested in what you see when you compare these strings that is why I want a shareable version.
peace
fmm
That doesn’t sound like a very science way of analyzing/collecting data. Are you keeping track of all of your misses and factoring that into some form of accuracy/precision QA/QC for your ‘game’?
fmm
I imagine so that after making a number of guesses you pick some pattern that you missed all those times you guessed wrong and declare “Eureka! there is the pattern”. Your method does not seem to have any objective component to it atl and strikes me as the antithesis of any form of scienctific methodology/analysis.
I’m doing some more playing and I have even less confidence, I’m not sure right now, give me some more time.
peace
Yes but it’s not nearly as formal as It should be.
That is why we need to make the game shareable so that the results can be public and the test double blind
That is close to how it works. There is no generalized algorithmic process for learning the pattern of a string.
That is the whole point.
If there was an algorithmic process to accomplish the task of learning a pattern a computer could do it.
The objective component is that multiple independent observers can distinguish between the “real” strings and randomized copies or models that are close
peace
fmm
Why not? Seems to me that would be a critical component of any methodology. Why keep guessing? One not one shot at each data set and see how accurate/precise you are in your determinations? Making multiple guesses is nonsense. Of course you are going to find the ‘pattern” at some point in time. Especially if you are only comparing two strings at a time.
fmm
That is sad way of doing any data analysis.
fmm
Seems to me that there is a algorithmic process for learning the pattern of a string and it is: Keep guessing until the game tells you that you are correct. IMO you need to reassess your methodology and come up with something more robust. The ‘keep guessing’ methodology is nonsense.
fmm
Hardly. The only component to the game is that multiple independent observers keep guessing until they can distinguish between the real and the others. Also this is only a hypothesis of yours since there is no supporting data for the assertion.
If you limit each observer to a single assessment and track those results I imagine a very different picture would immediately spring to the forefront.
It is.
And it will be before we can come to any conclusions. That is why it’s important to have a shareable version of the game
Because that is how we learn
This is about learning (information integration) not psychic readings
1) I don’t think a computer will be able to do this.
2) often there is no pattern to be found.
Nope that is not what is going on at all. you should be correct about 50% by chance.
You are not simply guessing you are hypothesizing. Looking for patterns the game simply gives you feedback.
Have you played the game?
No, multiple observers look for patterns in the data until they find them. Just like we learn stuff in everyday life.
Correct I have only my intuition and a couple of papers that provided the inspiration. The data will come or not. That is how science works
Yes in that case you would be measuring psychic ability and it would be completely irrelevant to ID or AI
peace
Well, I’m not impressed by the formulation of the hypothesis, but i will concede at this point that I will continue playing.
There is an unnecessary pattern in my data that apparently gets broken by scrambling. I haven’t double checked, but if my memory is correct, fmm has called the first two rounds.
Could be fun.
I hope it turns out that I got it but looking at four strings has proven to be very difficult. So I won’t be be surprised if I missed the second one. I would need to play a lot more to be sure.
I want to emphasize yet again that this is not about some special ability that I have.
It is about a general ability that humans have to losslessly integrate the information in an object.
That is why I want everyone to play the game. It’s cool there is a sense of accomplishment when you learn the pattern.
It’s almost as if you are noticing something that you have in common with the process that produced the string.
peace
I would say it’s about pattern recognition. There are quite a few ways of displaying data that reveal patterns.
There are regularities in my data that i think can be seen without graphing.
But I am intrigued.
fmm
Actually, no. You could have (should have) ground truth all this before presenting it for scrutiny.
fmm
Sorry I don’t buy into that at all. The only reason you keep guessing is because you are unable to determine any pattern to the data and it is only through iterative guessing and the game telling you if you are correct or not. If I am learning a mathematical concept I don’t keep guessing until I am right. I learn the methodology and apply that correctly to the problem. Guessing is a poor way of learning, IMO. Understanding basic concepts and applying them appropriately is a much better method.
fmm
What a poor attempt at dodging the problem. If you have the ability to assess the strings and compare them getting one shot at determining which one is real or not has nothing to do with psychic ability but rather reflects on the robustness of the methodology.
fmm>blockquote>Nope that is not what is going on at all. you should be correct about 50% by chance.
That is what I would expect from your game given a no multiple guessing scenario.
fmm
You are simply guessing each and every time you play the game. the only reason you keep looking for a pattern is because the game tells you you are out in left field with your determinations. Of course you are going to get the correct answer eventually given enough guesses as long as there is something to tell you that you are correct.
fmm
No need to play a rigged game where you continually guess until you stumble onto the correct answer. Hardly worth the effort, IMO.
fmm
I might believe that if the multiple observers weren’t making continuous guesses until they found the solution..It is all post hoc determinations.
fmm
Not really. Repeated guesses tells me you haven’t learned anything and are just making queries to the oracle until it says :Bingo” you finally got it right.
fmm
Not in the least. You seem to be just dancing around the issue with glib comments like the above.
If you have the opportunity to visualize the data, as you state, then the determination is not based on psychic ability but the ability of the observer to determine patterns in the data. This is after all what you wish to test (or so it seems) and if the observers need multiple guesses and guidance from an oracle then that suggests that humans aren’t that good at this type of pattern recognition. The number of guesses tells us how bad humans are at this task and it appears you are not keeping track of all the misses and focusing only on the successes. That rigs the game from the outset.
I have gone back to my original data download files, and the third set in my second submission is character scrambled, not the original.
I did three kinds of scrambling, using online text scramblers. One scrambler works character by character and another scrambles word order. My original data had six character groups separated by spaces. (The spaces are arbitrary and have no meaning in the original data. They were just inserted to make reading by humans easier.)
I also submitted a fourth file that was scrambled both ways. The original data is a genome. It has frequent character repeats, sometimes as many as five repeating characters. It does not look to the eye to be random, but it passes an online test for randomness.
I’m not presenting it for scrutiny. I’m presenting it so that both sides can have a hand in developing the idea. I do this to try and minimize the tired old culture war tripe and and move the discussion forward a bit.
How do you learn the methodology? How do you know if you understand it?
Nuff said. I guess that puts you in the company of Patrick et al.
If you don’t have basic curiosity I really can’t help you
Actually we are testing for lots of things
1) Are there patterns in the data?
2) When we learn patterns are we losslessly integrating the information in the string?
3) Is there something special about objects that are nonrandom and noncomputable?
4) can computers generally and reliably do what humans do when we learn the patterns in the data?
etc etc etc
In the game observers are not “determining the patterns” we are learning the patterns with feedback
peace
lets do it like this.
Start with an original and then randomize it then compare the two (not four) strings. See if you can distinguish between the two.
That is how the game is supposed to work.
When I looked at your 4 strings I copied and pasted your data directly into the game. I think that the formatting was making a glitch.
When I went back to double check I did not see the same pattern as strong
In the mean time don’t tell me which one is the original one.
I’m going to start over and I will look at each string individually and then randomize it to see if I can distinguish the strings.
That was my intention all along but I quit when I thought I saw a pattern.
peace
PS thanks for the interaction. Remember there is no reason to be adversarial,
Just think of this as exploring pattern recognition. and having a little fun.
I’m still confused. Your program only allows comparison of pairs. But you strongly imply that you can spot patterns in “real” data, as opposed to randomized data.
I don’t see the point in comparing.
fmm
You first presented this as a design detection tool that you had great success in using. It comes across to me that you have presented it for scrutiny and it seems that I am not the only who has come to this conclusion based on your prior posts.
From your posts it seems that you have already developed the idea as far as it is and I am only posting what I view as severe shortcomings of your design detector.
fmm
The same way most folks do by standing on the shoulders of those who have come before us. I know I understand it when I can successfully solve problems with the methodology.
fmm
When you come up with something that has been ground-truthed and doesn’t depend on repeated guesses as the means to find the solution I’ll express a bit more interest in a positive light. Right now I am expressing a great deal of interest in the game. Enough so that I am taking the time to respond to you and give you my opinion on some aspects of the game methodology that I see as potentially fatal flaws.
That you don’t have enough curiosity to keep track of the basics statistics of game trials you’ve run, which you say is a critical issue, is sad and puzzling to me in some ways and revealing in others.
fmm
Every time you query the oracle you have made a determination of something you thought was a pattern. That is, after all, why you chose that string. You are either correct in your determination or you are not. If you must use repeated guessing that should ‘learn’ you that you aren’t very good at spotting patterns after all. Which is of course why the success/failure stats are important enough that you have no curiosity to find what that distribution looks like.
This is why I wish you could play the original game.
randomized data is indistinguishable that is part of what random means
If you have two randomized strings you should not be able to tell them apart.
If you have a string that is not random it should have patterns in it that are not present in a randomized copy.
Peace
PS
Did you try hitting the space bar in my game to see if the strings shuffle?
not sure how you got this idea. I use the game to evaluate the strength of models that is all.
I have not used it to detect design. Why would I? I believe that everything is designed and true randomness does not exist.
In a best case scenario the game will be able to show that a string is nonrandom and noncomputable, That knowledge will not necessarily require you to infer design.
I have repeatedly said that in the end this will come down to the problem of other minds and design deniers will always have an escape hatch. You can relax
peace
This is exactly what happens in the game
What would you think of a person who wanted to point out what they thought were flaws in the game of tic tac toe but who could not be bothered to actually play it?
I do keep track of the statistics of the game trials. It’s one of the ways that I use to determine if there is a pattern or not.
however I don;t do this as formally as I will need to when we begin testing hypothesis.
What needs to happen is that the game should keep track of this automatically to ease reporting and minimize cheating
peace
fmm
It doesn’t appear that is the case at all.
fmm
Why would not playing the game make any difference at all to the veracity of the alleged flaws that were being pointed out. The points are either valid or not and independent of any game playing by the individual.
fmm
What would be your estimated mean number of guesses required to successfully find a solution in the game?
fmm
Then you should get to work on that aspect of the game.
That must have been part of the instructions I asked you to publish. But as far as I know, you never provided any instructions.
Since I have asked at least half a dozen times, I’d expect at least a link, if the have appeared on this site.
What??? You now say that your game is not a design detection tool???
Have you forgotten that you posted an OP, on December 27th, titled “Working Definitions for the Design Detection Game/Tool”?
Now I am getting suspicious that there is something fishy going on with this entire discussion. Are you taking us for a ride?
I think I am done wasting my time here.
fG
BK:
fifth:
Bullshit. You wrote:
You made your bed. Now lie in it.
fG 29,
I’m not sure how someone who posts an OP can completely forget what they wrote in a few short days…..especially after all this lossless memory stuff they make claims about. I certainly agree with you that something doesn’t smell quite right about the entire enterprise.
keiths @30,
Your observation of ‘Bullshit’ is spot on and has, apparently, filled
fmm’s bed to overflowing capacity!
curious to see what comes next…….if anything at all.
Thump, thump.
Is this microphone still on?
barely,
Ive been very busy this week and Ive also come to realize that you all are not particularly interested in helping to develop the game.
That means that the heavy lifting will fall to me. That is the shame but not surprising
It means that I need to devote some time and effort to develop a shareable version on a website.
It might take a while as I’m pretty clueless when it comes to coding.
Anyway thanks for the interaction.
I intend to ask again for some help when it comes time for hypothesis testing.
peace
Keith’s I am nothing short of flabbergasted that after all this time that you don’t understand the difference between inferring design and detecting design.
peace
fifth,
fifth,
I am nothing short of flabbergasted at your ability to trip over your shoelaces again and again.
Your own words, from the Design as the Inverse of Cognition thread (emphasis added):
fifth:
fifth:
OMagain:
fifth:
fifth:
fifth:
How many bullet holes can your poor foot take?
Really keiths.?
This is not difficult if you think about it for a second
I don’t use the game to detect design simply because I believe everything is designed. I infer design in some cases and assume it in others.
On the other hand you might try and use it to detect design since you believe that things can be divided into two piles those that are designed and those that are not.
I can only laugh at your inability to understand that there are different perspectives and different ways at looking at the world.
And that you interpret your own conceptual shortcomings as errors on the part of those your are communicating with.
Peace
I have no idea what the game is trying to demonstrate. Yo certainly can’t use it toseparate real data from scrambled data.
That sounds like two piles.
Two piles yes.
But the names of the piles are not “designed” and “not designed” IMO.
They are “nonrandom noncomputable” and “undetermined”.
Peace
I have come to realize that the only way to show you what the game is demonstrating is to actually demonstrate it.
I had hoped that you would spend some time evaluating strings for yourself on your own and that we could explore them together and that in doing so we could work together to develop the platform.
I hoped for the same from OMagain but alas it was not to be.
No worries I will work on the shareable game myself and let you know when it is completed.
peace
fifthmonarchyman,
Excellent idea! I wish multiple people had repeatedly suggested it sooner.
I have spent quite a few hours looking at strings in your program. I see no consistent differences that correlate to the string origin. Based on your feedback from my samples, neither do you.
I had hoped you would demonstrate your ability.
Non-computable is mathematically nonsensical. You need an independent definition.
Works for me, I never thought the purpose of the game was design detection ID style. By the way, you still haven’t said what conclusions one might draw from having a pile of non computable,non random strings is.
Based on the rather limited demonstration, fifth can’t do even that.
fifth:
Patrick:
🙂
newton,
Fifth did:
And:
fifth, then:
fifth, now:
BK:
fifth:
Fifth is vying with WJM for the title of Backpedaling Champion of TSZ.