2 ms·
Quoting von Neumann about software rngs is ignoring most of a century of development in the field. There are a plethora of high quality CPRNGs that are almost c
by adgjlsfhk1 5y ago
Quoting von Neumann about software rngs is ignoring most of a century of development in the field. There are a plethora of high quality CPRNGs that are almost certainly totally unbreakable if you mix in a few bits of hardware randomness per megabyte of random numbers.
- bee_rider 5y agoAlso this was in a paper where he was working on a PRNG, and later he says "If the [pseudorandomly generated] digits work well on one problem, they seem usually to be successful with others of the same type." So this is more of a "Yeah this is technically wrong but I did it and it worked" sort of thing.