In many beginner AI settings, one row represents one example. That example might be one customer, one house, one support ticket, one application, or one event.
This sounds almost too simple to emphasize, but many learners never slow down enough to make it concrete. They see a table as a blur of fields instead of as a collection of distinct cases.
Once the learner sees one row as one example, the rest of the table becomes easier to interpret. The columns describe properties of that example, and one of those columns may be the target the model is supposed to predict.
That shift matters because later training concepts rely on it. A learner who cannot describe what one row stands for will struggle when models and labels enter the picture.