In the technical sense what happened was in order to increase how much data you can stuff in a flash cell, each flash cell had to be able to store a varying amount of charge in discrete steps. The numbers are purely for example but let's say for single-level cells (SLC), a cell was considered a "1" if it had a voltage of over 1.2V. For multi-level cells (MLC), you could now store 00, 01, 10, and 11, which could correspond to say 0V, 0.4V, 0.8V, and 1.2V. Then we go to triple-level cell (TLC), now we can store 000, 001, 010, 011, 100, 101, 110, and 111, which would correspond to 0V, 0.15V, 0.3V, 0.45V, 0.6V, 0.75V... etc up to 1.2V.
Here's a problem, flash wears out by not being able to flush out all of the charge carriers in the flash cell. So each time you do an erase, some charge carriers remain. Eventually this builds up until the cell can no longer reliably hold a certain lower level of voltage. That's how flash "degrades." By trying to stuff more discrete levels of voltage in a cell, this lowers the amount of charge carriers needed to make the cell unreliable.
So why do this? Because you, the consumer, wanted more storage capacity. The process node flash memory cells are at is similar to processor nodes, so we're pretty much stuffing as many transistors as we physically can. The only way to get more capacity is to allow for more discrete levels of voltage to be used.