Big data is probably an overused term these days, and what is sometimes described as big is usually more like medium data. Big data is when you have so much information that it is difficult to process, or even store, on a single machine. Traditional approaches often break down with big data, as they are not adequate for the huge automatically-acquired datasets that are common today. For example, the amount of data constantly collected by IoT sensors or by our interactions with online services can be vast.
One caveat of big data and ML is that, although they are good at finding correlations in large sets of data points, you cannot use them to find causations. You also need to be mindful of data privacy concerns and be ...