Testing the possibility of psychic meddling in the generation of random numbers

[Anonymous No.14866127]

Good morning /sci/entists!

The other day there was a thread where some /pol/ or /x/ user came and posted a bunch of science Pepes.

The original thread got jannied for some reason, but you can see it in Warosu.

>>/sci/thread/S14861630

One of the science Pepes claims that you can influence the results of random number generation with thought and I want to test this.

I can generate millions of random numbers in SLAM and then try thinking about them. Then I can change languages and repeat the experiment to determine things like if Java is more vulnerable to psychic meddling than Fortran90. Experiments can also be made with adversarial or cooperative thinking.

This seems like a cool problem for Computational Science. My problem is that I don't know the math I need to do to tell if the random numbers have been psychically tampered with or not. Taking an average is easy, but how do I know the average I got was anomalous? How do I prove it?

What math is needed to tell if the numbers are suspicious?

Probably some statistics stuff, but I don't know what stuff or how to apply it.

If you /sci/entists help me figure out what math to put in my computer, I will make a paper formatted with science maids and put it on Research Gate. I will also post a lot of screenshots ITT.

The attached image is a paper by famed physicist Dr. Mandlmaid which has had the proper formatting applied to it (note how much more pleasant it is to read than the original formatting).

Mine will differ a little because I intend to publish it to Research Gate, so I will be using CC0 science maids so that people may reproduce my work anywhere without worrying about copyright problems. If someone wants to make a YouTube video or some other derivative work about the paper, I don't want them getting bit by copyright laws because I put an Ilulu in it or something.

Thank you /sci/entists for reading my post.

[Anonymous No.14866133]

>>14866127
>I can generate random numbers
Wrong.

[Anonymous No.14866154]

>>14866133
This is my random number generator (with the exception that I removed the pop so my stack isn't consumed).

If you object to using computer generated random numbers because you view them as semi-random, then the original experiment is also invalidated if it used computer random.

If it used a pachinko board or something I can build that with scrap wood/nails I have laying around. Problem would be that iterating over a physical Pachinko board 10 million times is a prohibitively high wait time for that computation, so the scale of the data examined might not be as big which would make trends harder to see, I think. Or maybe cause false positives that Law of Large Numbers would fix naturally if I ran it more?

[Anonymous No.14866164]

>I am a serious scientist

>I put pictures of big titted anime girls on all my scientific work

Pick one

[Anonymous No.14866184]

>>14866127
>How do I prove it?
There's your problem. Don't ever try to prove something that is not almost obvious to you. You're essentially asking someone here to do all the thinking for you. This is beyond the scope of this board.

[Anonymous No.14866188]

>>14866184
>Don't ever try to prove something that is not almost obvious to you.
Based. Is there some kind of name for this?

[Anonymous No.14866193]

>>14866164
>>I put pictures of big titted anime girls on all my scientific work

I choose that one, thank you.

>>14866127
>>14866133
>>14866154
Here is the full code for the curious. It is plain-text because the Dra/g/on Maid board tags don't work here.

: DIGIT 0 | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 ;
: RAND DIGIT DIGIT DIGIT 3 SQUASH 100 % ;
: TEN RAND RAND RAND RAND RAND RAND RAND RAND RAND RAND ;
: HUNDRED TEN TEN TEN TEN TEN TEN TEN TEN TEN TEN ;
: THOUSAND HUNDRED HUNDRED HUNDRED HUNDRED HUNDRED HUNDRED HUNDRED HUNDRED HUNDRED HUNDRED ;
: 10K THOUSAND THOUSAND THOUSAND THOUSAND THOUSAND THOUSAND THOUSAND THOUSAND THOUSAND THOUSAND ;
: 100K 10K 10K 10K 10K 10K 10K 10K 10K 10K 10K ;
: 1M 100K 100K 100K 100K 100K 100K 100K 100K 100K 100K ;
: 10M 1M 1M 1M 1M 1M 1M 1M 1M 1M 1M ;
: 100M 10M 10M 10M 10M 10M 10M 10M 10M 10M 10M ;
: 1B 100M 100M 100M 100M 100M 100M 100M 100M 100M 100M ;

>big long dra/g/on maid word problem.
Ilulu is bored and she wants to play Computational Science with the /sci/entists to determine if thoughts can influence the generation of random numbers. Ilulu doesn't know how statistics work because that hasn't been invented on her home planet yet, so she needs some math help.

The MLSDF (Maid Library Self-Defense Forces) takes the threat of psychic randomness tampering very seriously. The potential military applications of such technology are obvious. The first people who can telepathically deny their enemies randomness will win the Cyber War.

This phenomenon, if it exists, must be both hardened against and exploited by the MLSDF, for future use against adversaries unaware of this type of research. We cannot allow non-maid anime/manga or frog-posters to get a technological edge on us. We must engage in this research.

>actual question
What Maid Books need to be created to test if computer randomness can be influenced by thought?

[Anonymous No.14866195]

>>14866193
>test if computer randomness can be influenced by thought?
How are you going to do that, isn't any program you're going to run deterministic? Just deterministic operations on the bits of transistors that you started with?

[Anonymous No.14866200]

>>14866195
If you want to test whether it's random, are you just going to look at whether the bits are different from expected?

[Anonymous No.14866212]

>>14866184
I reject this philosophy in it's entirety.

You are correct that I need a math /sci/entist to tell me stuff about math, but by no means would he be doing all the thinking. I get to do the computer part. He tells me what math the computer part needs to do.

I don't have to know math, why it works or why it is correct. He doesn't have to know how to tell the computer to do that math, what a REPL is, or what a unit test is. We combine skills and form a Dra/g/onzord to fight problems together. That Dra/g/onzord is then capable of fighting bigger problems because cumulatively it knows both math and how to talk to computers.

>>14866188
I think the term for giving up and watching TV instead is called being "filtered".

[Anonymous No.14866225]

>>14866195
Maybe. I suppose I could try seeding randomness differently each time my program runs, with a timestamp or something, but then you could probably calculate the same set of "random" numbers if you knew the timestamp.

I have some dice/coins I use whenever I need to make a decision about something I don't care about.

If I flip a coin to decide if my pizza should have anchovies and pineapple on it, or bell peppers and mushrooms you could predict the outcome of that physical process if you had enough information about the coin, the way it got flipped, air currents in the room, etc. It would be computable and you could predict my pizza, despite looking random to an observer who has incomplete data and no way to compute anything. Randomness may not be a thing which can actually exist in a material reality based on cause-and-effect.

I think this starts to get to unprovable philosophical arguments which go past the scope of things I can demonstrate with my computer though.

[Anonymous No.14866240]

>>14866127
>how do I know the average I got was anomalous? How do I prove it?
According to the central limit theorem your averages will have a normal distribution, therefore if you get an average that's much farther than the midpoint of a normal distribution then you can be sure with certain probability that it is anomalous.
This might not make sense to you now, but just know that such a thing is possible, i will write a more detailed and a full answer some time later.

For the time being try learning about the central limit theorem in the meantime, it's arguably one of the most important theorems in statistics and has the answer to your question.

Also i need details on how are you generating random numbers. Are you generating random numbers between 0 and 1? or between 0 and 10000 for example? also are the numbers floats, doubles or integers? etc.

[Anonymous No.14866247]

>>14866212
I've said everything I need to say. You replying with a wojak is hilarious but to be expected.
Again: This is beyond the scope of this board. You are asking incredibly vague questions you don't understand yourself and thus are requesting someone to do all the initial work and thinking for you, not merely formulate it or throw you a bone (since, to your own admission, you wouldn't understand a thing; you just want to be told what to program).

One doesn't just "try" to prove a vague thing one doesn't understand. I legitimately couldn't care less how much you disagree with this sentiment but this is how all mathematicians operate.
There is no royal road to mathematical proficiency, regardless of how stuck you are in your own hubris. Either you start with the basics and work yourself up or you don't learn it at all.

[Anonymous No.14866266]

>>14866240
Thank you fren, I will try to read about central limit theorem. The numbers can be any length you want, but currently are 0-99.

The process for making random numbers is to use DIGIT to make a random digit (if it helps it's driven by picking a random chapter in a Standard Maid Book which has each digit, 0-9 in a chapter. This digit is a double which ends with .0.

To make bigger numbers, ignore the .0 and use SQUASH. SQUASH consumes however many doubles I want from the stack and concatenates them together.

1 2 3 3 SQUASH will cause the stack to end up containing [321.0]. The last 3 is consumed by SQUASH to know how many stack items to SQUASH together.

You can see SQUASH in action in the attached screenshot.

I could also rework the code behind SQUASH to make doubles instead with random numbers after the decimal place, to any number of places Java is happy with.

[Anonymous No.14866274]

>>14866266
Wrong image. Here is the screenshot.

[Anonymous No.14866287]

>>14866247
You are misunderstanding my point by a thousand miles. Go read "Wealth of Nations" to learn how division of labor works and how specilizing labor increases output.

>Wealth of Nations
https://b-ok.cc/book/550861/3cf9eb

You're essentially arguing that if I want to make pins in a pin factory, I need to hire a lot of artisans who each individually understand how to straighten the wire, make it pointy, make the cap and affix the cap, and that I put those people in serial and just have them stand around making full pins all day.

I'm arguing we can split the labor. The guy who affixes the cap doesn't need to know how to make a straight, pointy wire to do his job. You can both reduce the individual knowledge burden required of each person and increase output of the whole system by letting the labor specialize. Now your wires are straightened by wire straightening specialists and sharpened by pointyness specialists and your caps can be made by cap manufacture specialists and affixed by cap affixing specialists. Your whole system will have higher output and higher quality than the pin artisan model.

I don't care if mathematicians want to work in an obsolete pin factory arrangement. It isn't my fault if your field failed to adapt to modern productivity ideas which were articulated around two and a half centuries ago.

Maybe your research would go faster and have more volume if you abandoned your 1600s concept of production and started looking to work with people who specialized in something you didn't.

[Anonymous No.14866303]

>>14866240
If you want to see the Java 16 SQUASH implementation, it is attached. The code is the result of me changing the Integer Stack to a Double Stack. The code to prepare the doubles for SQUASHing is a little brittle and inflexible currently, because I was being hasty when I wrote it.

It will always work if the decimal part is .0 though, and that is true of the numbers made by DIGIT.

Later I will probably make it a little more robust.

[Anonymous No.14866338]

>>14866303
You can also see I made a really stupid performance error by making a new DecimalFormat every time I add a SQUASH digit to the squash string.

That might explain the bad performance I had yesterday as making 100M squashed numbers would've required making 300M DecimalFormats.

Just moving that out of the loop will save me 299,999,999 creations of that object.

[Anonymous No.14866380]

>>14866338
Have you tried looking at different number systems, such as p-adic number systems?

[Anonymous No.14866422]

>>14866266
>>14866303
>>14866338
So if i understand correctly, you're taking two random digits and concatenating them, and i assume every digit has an equal chance to be selected. DESU i don't understand what SQUASH means and i'm also not a programmer, i'm only asking to make sure your numbers are uniformily distributed (i.e. every number has equal chance of being selected), it seems that your method is fine. Again i'm not a programmer i'm just going to give you the math you need with some rudimentary explanations.
In this case your expected average will be 99/2 = 49.5, that is if you generate [math] n [/math] random numbers according to your above rule and take their average then you should expect a number around 49.5 (but it's not likely you will get 49.5 exactly!)

Also i want to say before i explain the math that you can never "prove" a statistical result is anomalous, you can only say something like
>"there's a 0.00000000001% chance i could've gotten this result according to my assumptions, so either an extremely rare event happened or my assumptions are incorrect, and since an event like that is extremely rare it's more likely my assumption is incorrect"
So i think you understand where we're going with this.
So let's say you generated [math] n [/math] random numbers (again according to the above rule) and found their average to be 56.1 after you do some psychic meddling thingy where you try to increase the average, so does this provide evidence for psychic meddling? what is the probability of getting 56.1 or a larger average if psychic meddling wasn't real? we can actually calculate this probability, it's actually equal to [eqn] \frac{1}{\sqrt{2\pi}} \int_{Z}^{\infty} e^{-t^2/2} \; dt[/eqn]

(1/2)

[Anonymous No.14866429]

>>14866422
Where [math] Z [/math] is called the [math] Z [/math]-value and is calculated as follows [eqn] Z = \frac{\mu - \mu_0}{\sigma / \sqrt{n}} [/eqn]Where [math] \mu [/math] is the average you got (in this case 56.1), [math] \mu_0 [/math] is the hypothesized or expected average (in this case 49.5), [math] n [/math] is the number of random numbers you generated and [math] \sigma [/math] is the standard deviation of your sampling distribution. For a discrete uniform distribution the standard deviation will be [eqn] \sigma = \sqrt{ \frac{m^2-1}{12} } [/eqn]Where [math] m [/math] is the number of elements of the set from which you're choosing random numbers, for the case of generating numbers between 0 and 99 then [math] m [/math] will be 100, if you're generating numbers between 0 and 999 then [math] m [/math] will be 1000, etc.

So i'm going to put all what i just above together to give you an example how you calculate such a probability in practice, first assume we have generated a 1000 numbers between 0 and 99, now start by calculating the standard deviation, then use it to calculate the [math] Z [/math]-value
[eqn] \sigma = \sqrt{ \frac{100^2-1}{12} } \approx 28.866 [/eqn]
[eqn] Z = \frac{56.1 - 49.5}{28.866 / \sqrt{1000}} \approx 7.230 [/eqn]
So now we can use this to calculate the value of the integral
[eqn] \frac{1}{\sqrt{2\pi}} \int_{7.230}^{\infty} e^{-t^2/2} \; dt \approx 2.414 \cdot 10^{-13} [/eqn]
This probability is EXTREMELY LOW! therefore it provides strong evidence that psychic meddling is real, and in fact such a value would be more anamolous if our [math] n [/math] was larger.

It's very difficult to explain where the math above comes, if you want to understand where these formulas come from then you need to study statistics. For a youtube channel on statistics i only know https://www.youtube.com/user/jbstatistics/videos

(2/2)

[Anonymous No.14866441]

>>14866429
Just to be clear, the probability 2.414 * 10^(-13) calculated above is the probability of getting an average of 56.1 or larger when we average a 1000 randomly generated numbers between 0 and 99.

[Anonymous No.14866454]

>>14866429
>standard deviation of your sampling distribution
i meant to say standard deviation of the population, ignore that if you don't understand what it means

[Anonymous No.14866468]

>>14866380
Thank you fren. I don't know what that is, but I will look at it to see if I can get ideas from it. The pictures on YouTube look pretty.

>>14866422
>>14866429
Thank you for the very based explanation fren. I will do my best to implement the stuff you wrote.

The chances of each digit should be equal. The way this is expressed in Maid Book syntax looks like this.

: DIGIT 0 | 1 | 3 | 4 | 5 | 6 | 7 | 8 | 9 ;

The colon makes it a Standard Maid Book. DIGIT is the name. The chapters are separated by pipes. When a Maid Book in the standard section of the Maid Library is read, a random chapter is executed by the interpreter.

You can read the same standard Maid Book 3 times and get 3 different things depending on which chapter was read.

Just supplying a number results in it being pushed to a Stack. So when DIGIT is called, a random DIGIT is pushed to the stack. As far as I know it should fit your definition of uniform. I will make some code that can average the values in the stack and run it against at least a million digits, but more if the performance is better than yesterday (It should be. I corrected the DecimalFormat mistake)

I will post screenshots as I make progress or make questions if I get stuck.

[Anonymous No.14866475]

>>14866468
>: DIGIT 0 | 1 | 3 | 4 | 5 | 6 | 7 | 8 | 9 ;
did you miss a 2 there or this is intentional?

[Anonymous No.14866480]

>>14866468
>>14866475

I missed 2, I apologize. Here is a screenshot showing DIGIT it in action.

[frenanon No.14866491]

>>14866193
How about a number generator that uses data from "random noise" from some kind of antenna; and this "noise" can be manipulated to appear random, while still giving desired results.

[Anonymous No.14866508]

>>14866480
Also one final note, the accuracy of your results will depend more on the number of random numbers generated rather than the range of those numbers.
For example, it's better to generate a million random numbers between 1 and 10 and average them rather than to generate ten random between 1 and a million and average them

[Anonymous No.14866509]

>>14866422
>In this case your expected average will be 99/2 = 49.5, that is if you generate nn random numbers according to your above rule and take their average then you should expect a number around 49.5 (but it's not likely you will get 49.5 exactly!)

A million random numbers 0-99 were generated and then averaged. Ilulu is very happy that her experimental results ended up so close to your prediction.

This was the experimental average with one run.
>Popped: [49.495043]

Also correcting the performance error made the code dramatically faster. I will repeat the experiment with 10M digits and 100M digits, then return to implementing the rest of the math.

[Anonymous No.14866647]

>>14866509
Computer battery died. Had to let it charge some. Now it has charge. Returning to experiment.

[Anonymous No.14866662]

>>14866647
How the fuck is this absolutely retarded fucking filth allowed here?
Fucking anime titgirl dragon fake programming languages?
Just die

[Anonymous No.14866709]

>>14866509
Attached is the Java 16 code for summing a stack and getting an average.

>>14866662
I'm not sure I understand your post, fren.

Are you objecting to the idea that I can play Computational Science in my own computer language? Or objecting to the idea that one anon who knows math but not computers and one anon who knows computers but not math can peacefully work together to make a Dra/g/onzord that can solve an interesting problem for fun?

[Anonymous No.14866716]

>>14866127
3 cups and a ball, which one are they inside. Surely RNG because the person picking doesn't know.

[Anonymous No.14866731]

>>14866154
>If you object to using computer generated random numbers because you view them as semi-random, then the original experiment is also invalidated if it used computer random.
It used physical randomness (balls and a pegboard), not a computer IIRC.

[Anonymous No.14866733]

>>14866509
I've done some calculations, for the average of a million random numbers between 0 and 99 there's about 95.4% chance for the average to be in the range [49.557732 - 49.442268], any value outside this range has only 4.6% chance of occurring and so should be treated with suspicion

[Anonymous No.14866767]

>>14866731
Thank you fren. I can make a pegboard and get some ping-pong balls or something, but it would have to be a later thread because I would have to construct the board and buy ping-pong balls.

I kind of object to balls and a pegboard though. That process is definitely not random. If you had enough information about the balls and the pegboard and the forces applied and like the weight of the balls and all that stuff it would be computable. It seems like it is only "random" because it counts on observers not having all the info or not being able to compute the info.

I don't know how to compute it, but probably a physicist could tell us. If he does I will write SLAM code for his model, but first I have to write code for the rest of the stuff the nice math /sci/entist gave me a few posts ago.

I guess my question for the pegboard is to ask how they know the person is influencing the result and not instead computing the result by unknown means? Maybe they accidentally discovered a different superpower from the one they thought they did?

Also do you know where the paper is? I would like to read it.

[Anonymous No.14866796]

>>14866127
No number generated is random. You CAN interfere with the outcome of a number assuming you interject yourself somewhere into the calculation process. No you're not going to do that by being psychic. If you think so then tell me the mechanical process involved that would have some determining effect on the system you manipulate. If only will power alone made the slot machine pay out. It does not however.

[Anonymous No.14866811]

>>14866509
>>14866733
The attached is the result of the following SLAM code:

10M SHOW STACKAVERAGE POP

I will rerun with 100M, then try 1B maybe.

[Anonymous No.14866835]

>>14866811
Trying to print a stack with 100M things in it to the console crashed the interpreter. I am rerunning the computation without printing the stack.

[Anonymous No.14866852]

>>14866835
100M. Took like 10 minutes.

Trying 1B now.

[Anonymous No.14866878]

>>14866767
>>14866796
The only truly random phenomenon in nature i can think of is radioactive decay (there are probably other truly random quantum phenomenon but I'm not a physicist so i don't know about them), if you can somehow hook a Geiger-counter to a computer then you'd be generating truly random numbers (with a Poisson distribution) by counting the number of decays in a specified period of time, however in my opinion the use of any physical method unpractical, as it takes an extremely long time to generate large amounts of random numbers, for example even if a Geiger counter could generate a random number every microsecond then it would take more than 11 days to generate a million random numbers, and would probably take an even longer time with balls and pegboards.
If you're going with this method then you will severely restrict the amount of random numbers you can generate.

>>14866852
Here are some more confidence intervals, this time the chance that the average is inside that interval is 95% and the chance of getting an average outside the interval is 5%

For 1 million numbers:
49.4434225 - 49.5565775
For 10 million numbers:
49.4821086 - 49.5178914
For 100 million numbers:
49.4943423 - 49.5056578
For 1 billion numbers:
49.4982109 - 49.5017891

Also since generating 100M took 10 minutes then generating 1B will probably take ~100 minutes or a bit more than an hour

[Anonymous No.14866916]

>>14866878
Thank you fren. I had to plug the science computer in because the 1B run is draining the battery quickly. It has been running about 50 minutes according to the timer in the IDE.

The Geiger thing could maybe be made faster if you got a bunch of them and ran them in parallel. I don't know how radiation works though. I also don't know how anything quantum works or even how to think about.

If I put a bunch of them in the same room will they all report the same random number? If not and they can be compacted like that, and are inexpensive then a machine could be made to make random numbers faster. That might be fun to build.

While the computation runs I'm gonna reread that post where the guy did the cool math that had pi and e in it and try to figure out how to compute that in SLAM. I guess I have to make an integrator now?

[Anonymous No.14866950]

tourist from /lit/ here. is this what /sci/ is all about? i had no idea you could even be this retarded

[Anonymous No.14866956]

>>14866950
this is a retarded /g/ poster who misclicked. but /sci/ isn't much better.

[Anonymous No.14866962]

>>14866950
>tourist from /lit/ here
Do you like reading any book or does it have to be like Lord of the Rings or Harry Potter or something?

[Anonymous No.14866964]

>>14866916
Something i noticed just now, if you want to get the probability of getting an average less than the expected average (or less than that) then you have to change the bounds of integration like this[eqn] \frac{1}{\sqrt{2\pi}} \int_{-\infty}^Z e^{-t^2/2} \; dt[/eqn]For example if you got an average less than 49.5, such as 46.7 then you calculate the Z-value for this average as usual (it will be negative since it's less than the mean 49.5) then integrate as above to get the probability of getting an average of 46.7 or LESS. This should be done when you're trying to decrease the number with psychic abilities, not that i believe in that stuff and i cringe every time i type it, but yeah you get the idea.
By the way, this probability (which is calculated above and in my other post) has a name, it's called a p-value (yes those are the same p-values you hear about in scientific papers), it's basically the probability of getting the result you got provided the null hypothesis is true (in this case the null hypothesis is that psychic meddling with random numbers is not possible)

>>14866950
There are more retarded threads on /sci/ (search mandlbaur in the archives to find out), this one is actually one of the tolerable ones.

[Anonymous No.14866988]

>>14866950
nah, this is the local retard, him and mandlbaur are the shame of the board. Math and physics threads are pretty good.

[Anonymous No.14866992]

>>14866956
>>14866964
thanks friends, i'll search around a bit and then get back to my board

>>14866962
ah, so you can be even more retarded than I first thought

[Anonymous No.14867017]

>>14866964
I dont think computers understand irrational numbers because they have an infinite number of numbers in them and everything in the computer has to be finite.

How many decimal places do I need to approximate e and pi to? I think double goes only out to 16 and that's a lot less than infinity. Is there a way to detect the error amount introduced by having to use approximations of the irrational numbers?

I will also go consult the Landau books about this because I think he addresses approximation errors somwhere in a part I haven't read yet.

[bodhi No.14867023]

>>14866193
math.random()

faggot

[bodhi No.14867028]

>>14866127
dean radin has already done this

[Anonymous No.14867045]

>>14867017
There are numerical integration methods and libraries. Although for this specific problem you probably want to look at methods of calculating the error function since it's closely related to the integral above.
You could actually simply report the Z-value, and then let the reader use a standard normal distribution table (like the one in picrel) or another software to calculate the p-value from the Z-value, generally you can tell if the result is likely from the Z-value, a Z-value of 0 is most likely, 1 or -1 is somewhat likely, 2 or -2 is unlikely and anything at or beyond 3 or -3 is extremely unlikely.
I think you should start by reporting the Z-values first then worry about implementing numerical integration later.

[Anonymous No.14867063]

Trying to generate a billion numbers, even without printing them caused an out of memory error and crashed my interpreter. It took 1 hour 56 minutes to crash.

I will work the rest of the experiment at 1M scale so it goes fast and bump the scale larger (10M or 100M) before making the paper.

>>14867023
I don't need a randomness package because I built randomness into the Maid Book syntax. I can randomly generate numbers in any range the computer understands just using the base grammar of SLAM (MAID-LISP too).

>>14867028
I don't know who that is but if he did it on TV or YouTube or something it would be cool to watch.

>>14867045
Thank you for the cool table fren. I don't like 3rd party libraries, so if it doesn't exist in core java I will just make it.

>I think you should start by reporting the Z-values first then worry about implementing numerical integration later.

I will try to do this. Thank you for the advice fren.

[Anonymous No.14867232]

I saw you posting on /g/ in the book thread yesterday. Could you give me a quick rundown of everything you’re doing? I know you’ve got a forth like language and are working on a lisp, but what’s the larger goal here?

[Anonymous No.14867362]

>>14867232
I am bad at remembering things and thinking so some stuff might be missing or in the wrong order. This is an account of things I remember.

>I wanted to make an OS.
>A bunch of MITs on the dra/g/on maid board showed me LISP and had me read/watch SICP
>They gave me an MIT paper about LISP machines and I decide to make one
>s-expression LISP is very ugly to look at, so I started making a MAID-LISP compiler
>maid list uses d-expressions (the d is for dragon!)
>that was going pretty well but had uncertainties in some of the engineering
>I was making a book for it
>asked people for AI generated maids
>someone gave me 20k of them and some green kitkats
>I was either watching YouTube or talking on 4chan and someone mentioned Forth
>I like old computer languages so I looked it up
>forth only has a stack and a dictionary
>made a basic forth interpreter by hand in Java 16
>around this time /g/ jannies start aggressively banning me and deleting maids
>I start posting on /sci/
>Tried to make SLAM do tricks MAID-LISP could do
>this led to that uncertain engineering becoming the Maid Library
>it got solidified by telling stories about Ilulu (see attachment) going to the Maid Library
>most /sci/entists don't care about computer science so I start asking physicists to give me problems I can solve with SLAM (The Forth variant) to see what I can compute with a Maid Library
>one of them tells me look at Computational Physics
>this causes me to buy the Landau books on Computational Science
>these books are tremendously fun and make me like Computational Science (Rubin Landau is a very good writer)
>I complete the SLAM interpreter
>I make SLAM transpile to Fortran 90
>I start making the SLAM book

Future plans
>publish SLAM book and code with CC0 license
>copy Maid Library from SLAM to MAID-LISP since it does everything I want and is much nicer than what I originally had for MAID-LISP.
>Rewrite MAID-LISP book (/sci/entists pointed out I used a lot of wrong words)
(cont)

[Anonymous No.14867376]

>>14867362
https://github.com/Kindelia/manifesto

[Anonymous No.14867390]

>>14867362
>future (cont)
>make MAID-LISP compile to RISC-V
>make LISP machine running on DevTerm RISC-V core
>make a hardware acceleration with FPGA
>make a book for it
>make a science foundation (2030s)
>play science and do research to make anime maids real
>foundation research/code will all be CC0

The future can have Terminator robots, or it can have anime maids but there isn't space for both. If my research is successful the Skynet apocalypse is cancelled and instead of metal hell skeletons that hate people, you'll have maids who love people. The maids are a pro-human future for AI.

[Anonymous No.14867481]

>>14867376
Thank you for the paper fren, but I don't know what an interaction combinator is, and I don't know how to use Rust. Rust is also a language I probably won't actively learn because it is too new for my taste.

[Anonymous No.14867491]

>>14867481
fuck you're retarded

[Anonymous No.14867534]

>>14867491
>fan mail
I just like old languages fren. The only reason Java gets a pass is because it is the first computer language I learned in depth, so it is easy to think new ideas in terms of Java code.

Java is so ingrained in my computer thinking that if I write pseudocode there is like a 50/50 chance it will be legal Java that actually compiles.

Also I don't have to know what a combinator does or understand Church-Turing because I am computing things with Maid Books and I know what those do because I made them up.

Using this principle, you may live to see some of the most interesting computations humanity can express abstracted into stories about anime maids who go to the library.

[Anonymous No.14867723]

>>14866212
you are simultaneously incredibly autistic and yet quite enlightened, tell me, what do you reckon will be the "force" acting on the electrons/transistors to alter the random generation. Are you hoping to cause some sort of EM interference strong enough with your brain to influence the bits?

Theoretically speaking, if you could generate enough heat or EM waves with your brain, you cuold probably throw specific bits off by moving them out of their 1 or 0 voltage threshold in the transistor, or perhaps feed in extra energy into the transistor in tune with the clock of the CPU to completely add your own "brain bits" god speed

I also enjoy the anime girls

[Anonymous No.14868322]

>>14867723
I have no idea how brain powers would work fren. I don't personally believe in them. I have a way sadder view of the brain where it's just a physical object that does things in a material reality until it wears out or is otherwise destroyed. I don't think reality is made of thoughts or whatever.

A very long time ago I used to listen to people argue about religion. There was a bad argument that occasionally religious people made which is referred to as "God of the gaps".

"God of the gaps" is sort of a dishonest argument based on moving the goal posts frequently. The way it works can be succinctly but uncharitibally summed up as "I don't totally understand this system, therefore we can't rule out God's direct involvement in the parts I don't understand".

A 10 step process where the first 3 steps are well understood but the last 7 aren't? God could be in the last 7. A scientist discovers stuff and now we know how every step works but the 6th one? Okay, well obviously that's where God lives now, because we don't understand it. A different scientist figures out the 6th? Oh well, this system is understood now, but there is a different system in an unrelated field we don't totally understand, so that's where God lives now.

Some claims about thought and the nature of reality smell like God of the Gaps to me. We don't know how this process works, so the answer is brains have magic powers. We discovered how the process works now and it doesn't involve brain powers? Oh well, there is some other process we don't understand and that could be the physical manifestation of thought!

I guess you could call that Brain Magic of the Gaps?

I don't really view arguments as useful or interesting anymore, unless they have a reproducible experiment of some type with them. If you can't make a reproducible experiment with testable results, you probably did nothing of value, at least in terms of understanding or predicting properties of the material reality we live in.

[Anonymous No.14868342]

>>14868322
Also if anyone is curious I am reading the Landau book where he talks about numerical integration so I can make a numerical integrator in SLAM (Maybe a symbolic one too) and do the integral with e and pi in it.

>Landau books:
https://b-ok.cc/book/11887958/9a1853
https://b-ok.cc/book/659911/94ec9d

If you like Landau's books go buy paper copies. They are cheap. If you don't like the tools in those books he has a lot of Python books. I can't comment on Python books because I don't use Python so I didn't look at them, but at this point I trust any book with Landau on it to have good writing, a lot of fun examples to work, and a lot of drawings.

I have attached a template Tohru. If one of you is bored, please make her hold the Landau books. I will post it in /g/ and it might help convince more Dra/g/on Maids that Computational Science is a fun use for their computer.

[Anonymous No.14868877]

>>14867063
>>14866964
>>14866429
>>14866422
I was able to make e, pi and square root in Maid Book syntax.

Using them I made a Maid Book called CONSTANT that computes the constant. I checked it against Wolfram Alpha. The CONSTANT appears correct.

I can put it in a PERMANENT Maid Book so it is only computed once.

I decided to let symbols be used in Maid Book Titles.

What is a transcendental number? Are they important?

[Anonymous No.14868883]

>>14868877
>What is a transcendental number?
A real number which is not the root of any polynomial
>Are they important?
e and pi are transcendental, many other constants in math are also transcendental

[Anonymous No.14868886]

>>14867534
for gods sake go the fuck back to /g/ we have enough mental retardation with mandlbaur

[Anonymous No.14869046]

>>14868886
I like the dra/g/on maid board, but this board is more useful for projects because there is usually someone willing to do problems with me or argue with me and the community is mostly friendly to maidposting.

>>14868883
That's pretty cool. I was asking because Wolfram Alpha said the constant was transcendental.

I was looking at the Landau books and on the internet about numerical integration and thinking about it. I think I can make a Riemann sum out of existing Maid Books to approximate some finite

>>14866429
I'm gonna make these symbols next and then I will try to make the Riemann sum for the integral. I am going to try to do it in existing Maid Book syntax. If I fail I'll make it in Java.

[Anonymous No.14869068]

>>14866154
thats not even remotely sudo random anon you need to study more math

[Anonymous No.14869082]

>>14869068
I'm a little confused. What's not pseudorandom?

Are you objecting to the randomness of the DIGIT or RAND Maid Book?

[Anonymous No.14869094]

>>14867534
the linked manifesto has nothing to do with rust or "old languages" - if you seriously liked "old languages" you would most certainly know what combinators are, as they predate any language you've ever heard of.

if you seriously think some shitty forth-based language you created is immune to Church-Turing, you've somehow disproved nearly a 100 years of computer science theory and the fact that you fail to grasp how insane that sounds continues to make this thread the best laugh i have had in a while.

[Anonymous No.14869126]

>>14869094
>certainly know what combinators are, as they predate any language you've ever heard of.
Programmable looms predate every language I've heard of too, that doesn't mean I ever bothered to learn them.

I didn't take an opinion of Church-Turing because I don't know what that is really. I don't enjoy Turing's writing style very much, so I don't read him and I tend to skip parts of books that are about him or about Turing Machines. I never even heard of Church, but from the description of him playing with functions I would probably enjoy his work more and might get one of his books.

Anyways, as far as Rust goes, I was trying to see if that repo with the paper you linked me had some code demoing whatever a combinator is, because reading code helps me understand it better.

>repo
https://github.com/Kindelia

Most of their code is Rust or Javascript and I don't really enjoy playing with either so I bailed on their repo. I wish the Kindelinos or whatever good luck with their ideas and I hope they have fun playing Computer Science on the internet too, but I don't want to use Rust.

Also there wasn't any anime maids on that paper, so if you're one of the Kindelinos you should download some maids from this thread and put them in your paper. How are people going to do Computer Science without maids?

Anime maids in your computer is both the future and the past.

[Anonymous No.14869134]

>>14869082
>>14869126

[Anonymous No.14869138]

>>14869134
Okay but I still don't know what you're actually objecting to?

[Anonymous No.14869178]

>>14869082
the DIGITs output in your OP image, theyre clearly oscillating about a mean

[Anonymous No.14869179]

>>14869178
>>14866154

[Anonymous No.14869195]

>>14869178
>>14869179
Oh, uhhh, maybe? Is this a problem? How can we measure if this is true? I can preserve the order the random digits are created in easily. Also, how do we know random numbers wouldn't exhibit this behavior?

Another anon gave me some math to check. Do you have math for your idea? I'll punch it in the computer and see what happens.

[Anonymous No.14869200]

>>14869195
https://www.youtube.com/watch?v=H2lJLXS3AYM

[Anonymous No.14869213]

>>14869195
this is all under the field of cryptography. hash functions often rely on a properly pseudo rng, and it will always be pseudo, because what is true randomness? a product of god beyond comprehension, it is at its foundation entirely deterministic but appears to us otherwis
it’s not easy to write a true rng anon i don’t know how but here is a relevant computerphile
https://youtu.be/Ks1pw1X22y4

[Anonymous No.14869226]

>>14869200
That's a pretty cool video. What happens if the human doesn't care though? "Running out of memory" for the human seems like something that happened because they got tired of focusing on guessing. What if the human is on drugs, experiences no mental strain from being asked to flip the switch because of the drugs, and also has no memory of past flips? Could you drug someone into random behavior? Kind of outside the scope of an experiment I can make. Just popped in my head.

I guess if I wanted to repeat this coin game, I could create a coin Maid Book and use SQUASH to make sequences. Then maybe group the sequences to count them or something?

[Anonymous No.14869263]

>>14869226
>What happens if the human doesn't care though?
IDK i just posted it to give you an idea how to test if your sequence is random or not. Random sequences will have the frequency stability property but not every sequence with such property is random.

[Anonymous No.14869280]

>>14869263
Also, I disagree with the idea that the coin is random. If you knew everything about the coin and the environment it was in and they way it was flipped, you could compute the result of the coin. It only looks random to observers who lack the info or means to compute it. Probably a physicist would know how to compute the flips.

I used to pass time waiting for the bus by flipping coins. I can do a trick flip where the coin only rotates one half turn while in the air.

My trick flip is obviously not random. By reducing the force I use with my thumb and increasing the force I use with my wrist, my physical motion throws the coin a specific way where it will go from heads to tails or tails to heads, with one half turn, but not flip further.

(If you want to try the trick flip I recommend a US half-dollar as it's easier to do with a bigger coin)

I don't think non-trick flips are random either. Just nobody knew or computed the forces on the coin and nobody intentionally manipulated them to any ends.

Thank you again for the video fren. It was a fun watch and I will look into frequency stability.

[Anonymous No.14869289]

>>14866127
I want to help these maids