Rendered at 14:42:10 GMT+0000 (Coordinated Universal Time) with Cloudflare Workers.
costco 21 hours ago [-]
This is a nice spin on the Aaronson oracle:
> In a class I taught at Berkeley, I did an experiment where I wrote a simple little program that would let people type either “f” or “d” and would predict which key they were going to push next. It’s actually very easy to write a program that will make the right prediction about 70% of the time. Most people don’t really know how to type randomly. They’ll have too many alternations and so on. There will be all sorts of patterns, so you just have to build some sort of probabilistic model. Even a very crude one will do well. I couldn’t even beat my own program, knowing exactly how it worked. I challenged people to try this and the program was getting between 70% and 80% prediction rates. Then, we found one student that the program predicted exactly 50% of the time. We asked him what his secret was and he responded that he “just used his free will.”
vunderba 16 hours ago [-]
I'd never heard of the Aaronson oracle so thanks for the link!
I actually built this on a whim after watching that episode of TNG where Data locks everyone out of the Enterprise.
There's another version where a stats professor divides people into two groups. One group flips a coin 50 times and writes on a whiteboard the results. The other comes up with 50 results.
They leave the room and when they come back immediately know which is which. Specifically because of the over-alternation that yours mentioned. A string of HHHHHH is exactly as likely as HTHTHT, but people don't intuit that.
ShinyLeftPad 10 hours ago [-]
> A string of HHHHHH is exactly as likely as HTHTHT, but people don't intuit that.
this contradicts the comment you're replying to. the premise they are having is that HTHTHT is "less random" than HHHHHH.
dwattttt 8 hours ago [-]
The strings are equally likely if they're being purely randomly selected, but not if they're being selected by people with a poor understanding of random. In the latter case, people are more likely to select HTHTHT.
ShinyLeftPad 28 minutes ago [-]
I wonder what thought process led you to believe that even though the strings are equally likely and random people who don't choose the one you think is more random have somehow inferior understanding of randomness
Must have been looking at some external random source
codechicago277 54 minutes ago [-]
I used the digits of pi, transposing and changing the order, and got 14% on the original game. Professional poker players will use external sources of randomness to make sure they’re perfectly balancing their positions and not exploitable.
I don't get the fuss? 163 presses in and the machine scores 48%. You just ... randomly press the buttons? Should I play for even longer?
shawabawa3 23 minutes ago [-]
yeah i did 329 presses and got 55%. Doesn't seem that hard i wasn't even being very random, i just sort of zoned out and mashed them
cubefox 8 hours ago [-]
> (...) Then, we found one student that the program predicted exactly 50% of the time. We asked him what his secret was and he responded that he “just used his free will.”
- Scott Aaronson, Quantum Computing Since Democritus.
isoprophlex 8 hours ago [-]
Just use your free will, bro!
TeMPOraL 15 hours ago [-]
I got 64% at 50 steps in, then dropped down to 53%.
The secret? I distracted myself with the website text and kept pressing the buttons quickly without thinking about it.
kzalesak 13 hours ago [-]
I have managed to get 38-48% the times I tried. I believe I might have been influenced by the article I just read.
It would be interesting to see if by giving people advice on statics it would produce, on average, a measurably different outcome.
oli5679 11 hours ago [-]
A simple way to score 50%ish is to take a sequence of text and press the left button if even length, right button if odd.
This is what I also use for rock, paper scissors (mod 3). You can also use second hand of a watch for random samples, e.g splitting 0-30 vs. 31-60.
rlayton2 5 hours ago [-]
Similar to me, when I have to choose randomly I look for a number anywhere around. If it's odd, one choice, even is the other.
My strategy for rock paper scissors is to always go rock, and tell everyone that's my strategy. It's more fun for me seeing how people deal with that information than the win would be. Not very random though
Y_Y 3 hours ago [-]
I think Zipf's law means you're biased towards odd in that case (unless you allow yourself to choose empty strings)
dmurray 2 hours ago [-]
I don't think Zipf's law gives you this. It mostly says that common words tend to be short, but it doesn't tell you what proportion of the text will be common words. For example, in English writing two-letter words are found much more than one-letter words.
However, I think you're correct anyway. The right distribution to look at is something like the orange series in the bottom graph on [0]. Eyeballing it, it does look like 3/5/7/9/11-letter words appear more in aggregate than those of 2/4/6/8/10 (49% Vs 44.5%), and longer words are a rounding error.
Does someone know whether people saw the "predictions" in the actual test? Because saying random numbers blindly is not what I was doing. After my next number was predicted correctly, I changed my "strategy" to increase randomness (be less predictable).
So I don't know if this result is comparable to the one from the study.
My results were:
The house called 10 of your 87 scored numbers: 11%.
Blind guessing would average about 9.7 of them (11%).
Only 5 of the 206 people in a psychology study were more random than you. The house found you extremely hard to read.
vunderba 8 minutes ago [-]
You’re right. If you notice the instructions in the game, it encourages you to go as fast as you can. I keep the numbers up there just to make sure the voice recognition is working properly and to make the game seem more engaging.
But yeah, I agree: I think this does influence the results, since it makes it easier to be more random because you can essentially try to play in an adversarial fashion by adjusting your results according to feedback from the predictor itself.
In the original study, that wasn’t the case. They had a three-by-three grid of numbers and a metronome as well, which meant that they couldn’t slow down or speed up and had to go at a very specific rate with zero feedback until the very end.
Maybe I should add the equivalent of a "Hard" mode which makes the game more faithful to the original study.
This isn't very interesting, but it scratched an itch I had to see what would happen.
The house's record
The house called 18 of your 190 scored numbers: 9%.
Blind guessing would average about 21.1 of them (11%).
Only 1 of the 206 people in a psychology study was more random than you. The house found you extremely hard to read.
I really want to know who was more random than a RNG
newsy-combi 9 hours ago [-]
Getting 0% isn't the goal though. You want to sit right at 11.11%, the blind guessing rate. If the prediction algo always predicts you wrong, it actually always "predicts" a number you won't guess. That too is a feat, reducing the pool of numbers you're going to choose from, therefore making you more predictable.
Plasmoid2000ad 2 hours ago [-]
It looks like it's a fair implementation of guessing. Giving it digits of pi gives a very high (low) score - most times. To get a lower score I think you would have to be adversarial and/or get a lucky run. Based on the house doing better/worse on some runs I guess it not deterministic.
tyre 19 hours ago [-]
It might not be that someone is more random than an RNG, but that the distribution of numbers in your 190 vs. theirs appears more "random". It's a limited sample size.
If yours happened to contain consecutive digits or common ones (ending in zero, significant numbers like 911 or 42 or 69), that would imply less randomness even if it's actually the result of "real" RNG.
cortesoft 11 hours ago [-]
They were probably also a RNG… one of them has to win, just on chance.
FuckButtons 11 hours ago [-]
It got 8% of mine.
i_am_a_squirrel 12 hours ago [-]
i actually was, it said no one was more random than me! guess i just got lucky.
tldr; "im just like totally so random xD"
excannuk 3 hours ago [-]
[dead]
shric 17 hours ago [-]
I know pi to 100 decimal places and so that was pretty easy to cheat it
patwolf 3 hours ago [-]
I only have a few digits of pi memorized, so I had better luck using digits from fibonacci's sequence. I can derive them in my head, albeit slowly as the numbers get large.
blahgeek 12 hours ago [-]
You should prove that digits of pi are randomly distributed :)
iterance 8 hours ago [-]
That is easy. I entered all of the digits into this website and it determined it was random.
pavel_lishin 21 hours ago [-]
Is anyone sufficiently paranoid that their first thought was, "Are you going to train a model on my voice?"
(Yes, I saw the "type the numbers" option.)
panzagl 20 hours ago [-]
No, I thought of the penguin of doom...
d3Xt3r 19 hours ago [-]
Holds up spork
_diyar 21 hours ago [-]
Why would they when there are thousands of hours of radio and podcasts do train on?
233mhz 5 hours ago [-]
It could be the first step of a very long social engineering process, who knows.
scottyah 18 hours ago [-]
Maybe they are not training to make a TTS, but training on accents to make STT.
stronglikedan 20 hours ago [-]
And all that GOOG-411 data!
ShinyLeftPad 10 hours ago [-]
because they need more random
andai 17 hours ago [-]
"Hey its me ur grandson"
AnthonBerg 20 hours ago [-]
Ah, you see they are doing something far more random…
progforlyfe 21 hours ago [-]
Yes, my mind went directly here
19 hours ago [-]
plank 1 hours ago [-]
Typing, so more random.
Got a nice 9%, more random then most...
And yes, did e.g. type 5-5-5 repeated.
plank 1 hours ago [-]
Realize I said it wrong:
11%, 9 correct guesses (expectation if random would have been 9.3), three people more random in some other experiment.
tibbar 11 hours ago [-]
Many years ago, as a side project to help me apply to tech companies, I made this little site [0] where the premise is trying to pick the number that you think other people won't pick very often. I still think it's pretty neat.
Also, in part because of this project, I got to go from being unemployed in my hometown in the South to working in SF on some pretty neat projects. It's a good memory.
The pattern of holes is really interesting there. Everyone avoiding nice round numbers, so a bunch of those are very low frequency - except weirdly at the very ends of the range
vunderba 11 hours ago [-]
Nice~ The website made me recall an old mentalist force trick we learned as kids where you ask somebody to "quickly think of a two-digit odd number between 1 and 50 where each digit is a different number".
Most people will say: 100101 (binary to avoid spoilers)
frutiger 3 hours ago [-]
How are you sealing the predictions?
vunderba 15 minutes ago [-]
[flagged]
Trusteando 19 hours ago [-]
What about the following strategy? You first think of a random integer n1 from 1 to 9, then you think of another one n2, perhaps one that you have not used very often, and finally say (n1+n2) mod 9 + 1. I think this strategy could make your selection more random . Perhaps someone could try it and report the results.
e1ghtSpace 5 hours ago [-]
I found that saying two digit numbers like 82, 63, 45 etc was more random... maybe throw a couple double digits being the same in there like 77.
kelseyfrog 12 hours ago [-]
How would making your input deterministic make it more random?
Trusteando 5 hours ago [-]
Imagine that instead of two numbers you select a million numbers, and apply a zip algorithm to that sequence to obtain a file then you map the first byte of that file using (first byte) mod 9 + 1, intuitively there must not exist any correlation between consecutive tries of this strategy. Looking for this example I realized that instead of using n1+n2 you should use a non commutative operation to improve the strategy, but when using just two numbers the addition suffices.
dozerly 12 hours ago [-]
Well clearly there’s a modulo operation in there and nobody knows how they work
cortesoft 11 hours ago [-]
Depends on the algorithm they use to find patterns.
pcthrowaway 8 hours ago [-]
You're mixing up unpredictable and random
cortesoft 10 minutes ago [-]
Fair, but so did the person I replied to. For the purposes of what we are trying to do, unpredictable is what matters.
riffraff 9 hours ago [-]
> The house called 3 of your 27 scored numbers: 11%.
> Blind guessing would average about 3.0 of them (11%).
Admittedly I gave up before the 30 threshold but I think I did well enough at "beating the house"
Edit: tried again, 14/118, which is 12%. Maybe it's because I'm typing, and that's easier to do that speaking up.
8 hours ago [-]
malfist 18 hours ago [-]
This is pretty interesting. I've found it's fun to throw off the algorithm. Use your warmup to repeatedly guess the same number, and then don't use it for a while.
nico 5 hours ago [-]
Very cool, I really like that you have a typed and voice version
ehwa37 11 hours ago [-]
The house called 9 of your 117 scored numbers: 8%.
Blind guessing would average about 13.0 of them (11%).
None of the 206 people in a psychology study were more random than you. The house found you extremely hard to read.
Nice!
tchalla 15 hours ago [-]
> The house called 1 of your 32 scored numbers: 3%. Blind guessing would average about 3.6 of them (11%). Only 1 of the 206 people in a psychology study was more random than you. The house found you extremely hard to read.
christophilus 16 hours ago [-]
> Only 7 of the 206 people in a psychology study were more random than you. The house found you extremely hard to read.
I started just lagging the guesses of the machine, and it never seemed to catch on.
cromka 15 hours ago [-]
Didn't work for me, it started guessing them with 100% rate after just few of my replied guesses.
christophilus 15 hours ago [-]
I did mix it up. So, 3 or 4 trailing guesses and then some random clicks and then trailing is what I did.
rootsu 9 hours ago [-]
I just entered same number 5-6 times and it never caught it up.
alwa 21 hours ago [-]
I found it interesting when I felt my focus shift from the number-generating task to noticing (and responding to) the house’s guesses.
As soon as I noticed it repeating its guess, I had a good long run of adversarially avoiding that number… it’s kind of fun being on the machine side of the slot-machine-type racket!
maidaman 20 hours ago [-]
The first time I did badly. But the second time I got really far by deciding on a number instinctively and then picking something next to it. Maybe we instinctively pick similar numbers, but if we fight our instincts we end up with something truly random.
wren6991 2 hours ago [-]
I was able to beat this quite easily by just saying the first 100 digits of pi out loud. The house guessed 1/10, no better than random. Maybe you should add a predictor for "well-known mathematical constants".
exabrial 12 hours ago [-]
I'm reminded of the Dilbert Comic with Accounting's Random Generator.
I tried it. Several 9s later, it figured me out.
akshitgaur2005 20 hours ago [-]
Alternating and advancing from both ends seems to work, it calls n-1, and n+1 correctly if the previous guess was n, but got no calls in a sequence like - 1, 8, 3, 7, 4, 7, 1, 7, 3, 8, etc
DylanMerigaud 9 hours ago [-]
Too many alternations is exactly how people give themselves away.
The house called 19 of your 190 scored numbers and only 2 people were more random than me.... Hmmm
kps 20 hours ago [-]
I guess I've skewed the totals a little, but the site agrees that the digits of π are random.
vorticalbox 20 hours ago [-]
I got to 2/30 then it just guessed 7 from 30-50.
Did I break it?
vunderba 14 hours ago [-]
Yup! Most of the way I put the system together was by having it build several n-gram Markov models of your numerical sequences on the fly.
But we also have a couple of predictors that aren’t markov chains and are based on simpler heuristics. Without seeing your record, I'm guessing that the "numerical avoider" predictor got overweighted and just kept trying to guess the same number.
It should be fixed now!
yesitcan 10 hours ago [-]
When I see the world “sealed”, I immediately know Claude generated the site copy.
coolThingsFirst 11 hours ago [-]
Curious to know what model they use for speech recognition?
Not familiar with the sota if someone can chime in.
vunderba 11 hours ago [-]
OP here - I went with Vosk because it's pretty trivial to lock input down to a set vocabulary (in this case numbers) and from my testing works relatively well on most browsers. (webkit, firefox, etc)
I just pretended I'm playing piano to a megamix of tracks which popped into my head and did 'well'. Only 2 people more random. Typing though, think the site said that's "easier to win" on. Cba to speak 200 numbers out loud :/
Thanks, btw, actually enjoyed that. :)
vunderba 16 hours ago [-]
Haha, that’s great. Yeah, I totally get what you mean about speaking out loud.
You only really need to vocalize about ~40 numbers total, but you can actually get a similar level of RNG by "subvocalizing the numbers" before you press them on the keypad.
The main issue is that using the keypad kind of lends itself to just pressing it randomly without really thinking about the actual numbers themselves.
nightpool 19 hours ago [-]
"You spread your picks more evenly than a random draw would. Your most-used number made up 23% of your picks, where a random draw's busiest number gets about 23%."
??? huh?
vunderba 14 hours ago [-]
That's an explanatory bug on my part.
I have a little "mad-lib templatized" set of sentences for reporting on various stats, and that one was supposed to be sort of related to the law of small numbers. For example, shorter runs mean there is a larger probability of getting lopsided distributions (different from pure 11%).
It's supposed to demonstrate that even though you used a single number a disproportionate number of times, it's not necessarily statistically significant since shorter RNG runs can produce the same effects.
> In a class I taught at Berkeley, I did an experiment where I wrote a simple little program that would let people type either “f” or “d” and would predict which key they were going to push next. It’s actually very easy to write a program that will make the right prediction about 70% of the time. Most people don’t really know how to type randomly. They’ll have too many alternations and so on. There will be all sorts of patterns, so you just have to build some sort of probabilistic model. Even a very crude one will do well. I couldn’t even beat my own program, knowing exactly how it worked. I challenged people to try this and the program was getting between 70% and 80% prediction rates. Then, we found one student that the program predicted exactly 50% of the time. We asked him what his secret was and he responded that he “just used his free will.”
I actually built this on a whim after watching that episode of TNG where Data locks everyone out of the Enterprise.
https://youtube.com/watch?v=rERApU26PcA
https://youtu.be/bl5TUw7sUBs
And it inspired me to build a little game based on it. See if AI can read you mind: https://jeffyclassify.com/blog/mind-reader/
They leave the room and when they come back immediately know which is which. Specifically because of the over-alternation that yours mentioned. A string of HHHHHH is exactly as likely as HTHTHT, but people don't intuit that.
this contradicts the comment you're replying to. the premise they are having is that HTHTHT is "less random" than HHHHHH.
(archived now): https://web.archive.org/web/20260625105410/https://brainfrog...
https://aaronsonoracle.com/
- Scott Aaronson, Quantum Computing Since Democritus.
The secret? I distracted myself with the website text and kept pressing the buttons quickly without thinking about it.
This is what I also use for rock, paper scissors (mod 3). You can also use second hand of a watch for random samples, e.g splitting 0-30 vs. 31-60.
My strategy for rock paper scissors is to always go rock, and tell everyone that's my strategy. It's more fun for me seeing how people deal with that information than the win would be. Not very random though
However, I think you're correct anyway. The right distribution to look at is something like the orange series in the bottom graph on [0]. Eyeballing it, it does look like 3/5/7/9/11-letter words appear more in aggregate than those of 2/4/6/8/10 (49% Vs 44.5%), and longer words are a rounding error.
[0] https://medium.com/@davidxu90/relative-frequencies-of-words-...
So I don't know if this result is comparable to the one from the study.
My results were:
The house called 10 of your 87 scored numbers: 11%.
Blind guessing would average about 9.7 of them (11%).
Only 5 of the 206 people in a psychology study were more random than you. The house found you extremely hard to read.
But yeah, I agree: I think this does influence the results, since it makes it easier to be more random because you can essentially try to play in an adversarial fashion by adjusting your results according to feedback from the predictor itself.
In the original study, that wasn’t the case. They had a three-by-three grid of numbers and a metronome as well, which meant that they couldn’t slow down or speed up and had to go at a very specific rate with zero feedback until the very end.
Maybe I should add the equivalent of a "Hard" mode which makes the game more faithful to the original study.
"Over here we have our random number generator."
https://web.archive.org/web/20230305130159im_/https://assets...
If yours happened to contain consecutive digits or common ones (ending in zero, significant numbers like 911 or 42 or 69), that would imply less randomness even if it's actually the result of "real" RNG.
tldr; "im just like totally so random xD"
(Yes, I saw the "type the numbers" option.)
Also, in part because of this project, I got to go from being unemployed in my hometown in the South to working in SF on some pretty neat projects. It's a good memory.
[0] https://www.humanrandomness.com/
Most people will say: 100101 (binary to avoid spoilers)
> Blind guessing would average about 3.0 of them (11%).
Admittedly I gave up before the 30 threshold but I think I did well enough at "beating the house"
Edit: tried again, 14/118, which is 12%. Maybe it's because I'm typing, and that's easier to do that speaking up.
Nice!
I started just lagging the guesses of the machine, and it never seemed to catch on.
As soon as I noticed it repeating its guess, I had a good long run of adversarially avoiding that number… it’s kind of fun being on the machine side of the slot-machine-type racket!
I tried it. Several 9s later, it figured me out.
I always thought I was random, but now it's conclusive!
https://imgur.com/a/0azWegW
Did I break it?
But we also have a couple of predictors that aren’t markov chains and are based on simpler heuristics. Without seeing your record, I'm guessing that the "numerical avoider" predictor got overweighted and just kept trying to guess the same number.
It should be fixed now!
Not familiar with the sota if someone can chime in.
https://github.com/alphacep/vosk-api
Thanks, btw, actually enjoyed that. :)
You only really need to vocalize about ~40 numbers total, but you can actually get a similar level of RNG by "subvocalizing the numbers" before you press them on the keypad.
The main issue is that using the keypad kind of lends itself to just pressing it randomly without really thinking about the actual numbers themselves.
??? huh?
I have a little "mad-lib templatized" set of sentences for reporting on various stats, and that one was supposed to be sort of related to the law of small numbers. For example, shorter runs mean there is a larger probability of getting lopsided distributions (different from pure 11%).
It's supposed to demonstrate that even though you used a single number a disproportionate number of times, it's not necessarily statistically significant since shorter RNG runs can produce the same effects.