The article shows that it's difficult to calculate certain robust estimators used in statistics for handling outliers in data. These estimators, like LMS and LTS, are important for accurately estimating parameters in regression analysis. The researchers found that these estimators are hard to compute, making it challenging to get reliable results when dealing with data that contains outliers. They also discovered that a specific procedure in the R programming language has a very low chance of giving the correct answer for certain data sets. The article also discusses how to create new robust estimators to address these challenges.