Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Can you use your "free will"? Try your hand

people.ischool.berkeley.edu

Post reply on HN
Print view View original

1–10 of 279 posts

Re: Can you use your "free will"? Try your hand

#3

Post by ilaksh »

It sounds like the only way to "beat" this is to try to find some unrelated feed of noise to base the key presses on. Because when you actually do think about it deliberately, you fall into a pattern.

One way to do this without an external data source might be to just visualize (or look at) each object in turn clockwise of a room you are familiar with. Or use any other random source of data. Then come up with a rule that applies about half the time to some objects and half to the other. Or rather, come up with the rule before thinking of the room.

Maybe count the number of letters in the name of the object. If it's even, hit the f key, and if odd, hit d.

Re: Can you use your "free will"? Try your hand

#5

Post by valleyer »

fortran77 wrote: post #4

I was consistantly able to get it to be at or near 50% for long periods simply by typing English sentences in morse code, "f" for dot and "g" for dash

Nice. I don't even know Morse code, but simply pretending to code in some letters (saying "dit" and "dah" in my head) worked.

Re: Can you use your "free will"? Try your hand

#6

Post by endominus »

From the Github description, a quote: "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 can get it to about 60%. Given the algorithm used, it seems like it should be trivial to "trick" it by just cycling through novel 5-grams over and over, but to be honest I can't be bothered to check. And that would be rather contrary to the point of the exercise, if somewhat undercutting of the premise that free will is equivalent or comparable to randomness.

Re: Can you use your "free will"? Try your hand

#9

Post by von_lohengramm »

ilaksh wrote: post #3

It sounds like the only way to "beat" this is to try to find some unrelated feed of noise to base the key presses on. Because when you actually do think about it deliberately, you fall into a pattern. One way to do this without an external data source might be to just visualize (or look at) each object in turn clockwise of a room you are familiar with. Or use any other random source of data. Then come up with a rule…

Granted, I only tested this for a few minutes, but I found that if I thought about "repeat" vs "switch" instead of "d" vs "f", then I could reliably get under 50% prediction. Sure, the prediction rate was high 40's, but that at least means I won!

A lot of it comes down to avoiding auto-piloting into a 50/50 distribution. Subconsciously, "random" means approaches 50/50 since that's how it works, right? I'm sure given enough time, it'll be able to predict me with >50% accuracy, but that's entirely okay. Random doesn't have to be fairly distributed. As long as it doesn't get to 100% then I still have personhood :^)

Post reply on HN