What does the term "Garbage In$,$ Garbage Out" $($GIGO$)$ mean in the context of machine learning?

Machine learning models can function effectively regardless of the input data quality.

Garbage data is automatically filtered out by machine learning algorithms during processing.

Machine learning models are capable of correcting errors in the input data automatically.

The quality of output from a machine learning model is dependent on the quality of the input data; poor data quality will result in poor outcomes.