Machine learning is a term that gets thrown around a lot these days, but for a long time I wasn't 100% clear on the definition. At a very basic level, we can take these two extremely common words individually and derive meaning easily. We can conclude that machine learning is, of course, when machines (like computers), learn stuff on their own without being explicitly told what to do by a programmer. Right?
While this is the basic meaning, without understanding how this occurs you could be somewhat unsettled by this prospect. You would be right to assume this puts us squarely on the road to Skynet, but we don't need to worry too much about our machine overlords just yet.
What has made machine learning so prominent today is the explosion of data that the internet has provided. With all this data, computers are continually getting new, more, and better information to analyze that can influence their decision making. Programs that can provide different results based on new information are using machine learning- but they always come back to some core rules or regulations that influence the output they generate. Thus the "learning" is just deriving new answers to the same questions once they get more information.
This is how Optimize works- as you exercise more, Optimize gets more information about your workout habits, ability, and progress. Therefore, as you progress, plans and recommendations continue to get more and more personalized.
A much better explanation of machine learning with great visualizations can be found here.