Pseudo-random Vs. Truly-random.

Question

Why is it so important to have truly random generators? Is it not enough to have pseudo-random generators?

BTW. Do we have any true-random generators implemented yet?

Answer 1

It's usually not important unless you need it to be secure. Pseudo random can produce repeatable patterns, and can even be known what numbers it'll output if you know the starting conditions. So if you're doing random dice numbers for monsters in a video game, it's most of the time fine to have psudeo random. But if you're generating encryption then perhaps you'll need a good randomization method.

Yes "true" random generators do exist. There are a few methods (I'm going off of memory here for examples, so they might not be completely correct) like taking in white noise through a microphone, or using quantum qbits. Though if you think the universe is deterministic, then it can be argued those aren't true random either. So if you're trying to secure yourself against universe spanning entities resimulating the entire universe to find your random number, I think you're out of luck.

Answer 2

Kirk Schafer : Thank you very much. it was really informative🙏🌹

Answer 3

[only for exploration, not competing for answer]

TL;DR: Seeds aren't as helpful as they appear.

Prediction:
PRNG's generate their next outputs by performing reversible operations on previously-generated values. This is how they are deterministic + why we pick random-ish initial seeds. Some algorithms do this 'worse' than others, like .NET:
https://code.sololearn.com/c2UZ8pn0oZPz/
No matter the initial seed, I need only 56 outputs--then 100% sync'd--because later seeds ARE prior public outputs.

Reversibility, even when seeds hidden:
Although javascript hides its seeds, they're still reversible because the algos 'leak' some seed bits per output; with proper analysis the next number is reliably known after 2-3 priors.

Vetting:
Naive averages of PRNG's output should hover ~50%:
https://code.sololearn.com/WA2TQBOHm0hz/
...but more sophisticated analyses may reveal repetitions, diagonalization, etc.

If you'd like me to link something using more of Random.org's API, please let me know.

Answer 4

have a look at https://random.org
they have also an API
https://code.sololearn.com/W2GRa6bqfo1z/?ref=app

the code is incomplete and a bit broken but the part retreiving the random numbers works

Answer 5

what is NOT random? (veritasium)
https://youtu.be/sMb00lz-IfE

Answer 6

To go more "truly" random, seed your random generator with things like the time, the mouse position, your ping to a server, processor information, the current screen pixels and so on.

Answer 7

Search perlin noise for true random number generation.