Msot computers use Pseudo-random number generators to generate random numbers. These work by getting a seed (most often the current time to a ridiculous accuracy) and then use complicated algorithms to generate a sequence of numbers that look random. The key thing is that with the same seed they will give the same sequence. And since the seed is often relatively small (roughly four billion different possible seeds) that means you have roughly four billion possible sequences of numbers.

For most things these pseudo-random numbers are fine but for some purposes you want truly random. With only four billion possible seeds you could in theory given a list of a hundred generated random numbers take a pretty good guess at what the next one is using brute force (generate those four billion sequences and see which ones match the list of numbers you are given). For cryptographical things this is a disaster. For gambling websites this is a disaster.

And interesting fact is that using a standard pseudo random number generator you can shuffle a deck of cards and get approx 4,000,000,000 different orderings of the cards. A deck of cards however can have about 80,000,000,000,000,000,000,000... Actually its an 8 with 67 zeros after it. I'm not writing all of those.

This is why true randomness is a very popular concept. And somebody has done this with dice. By building a giant machine that rolls 1.3 million dice a day. This makes it pretty much as good as it gets in terms of true randomness.