Ask HN: How to predict next randomly picked element of a list

The Problem

Let's say that we have a machine that spits out a number (between 1 to 15) alongside a color (either green or blue) every 10 seconds.

OK so every 10 seconds we get a number (between 0 to 15) and a color (green/blue), cool, BUT we do not have any clue how these choices are made by the machine, we can just observe the output and record it at best.

The Question

can we use ML to feed in a model with a lot of these numbers and/or colors as they are put out by the machine and after a while let the model predict the next number and/or color which is going to be generated by the machine?