diff --git a/src/conclusionandoutlook.tex b/src/conclusionandoutlook.tex
index 20916a7..7e74142 100644
--- a/src/conclusionandoutlook.tex
+++ b/src/conclusionandoutlook.tex
@@ -9,6 +9,9 @@ However, the higher the sampling space $\mathcal{S}$ the higher the gains but th
 Another possible drawback is that reducing the uncertainty might not always be the best choice.
 If a system gets certain about samples that does not always mean this improves the accuracy, since it can simply be certain about the wrong thing. \cite{RubensRecSysHB2010}
 
+Active learning can have more influence if the model and the task is more complex and convergence is slower.
+The more decision points are required the more active learning can help.
+
 \subsection{Outlook}\label{subsec:outlook}
 
 Results might be different with a multiclass classification task and segmentation tasks.
diff --git a/src/experimentalresults.tex b/src/experimentalresults.tex
index 63fed6d..a746bb2 100644
--- a/src/experimentalresults.tex
+++ b/src/experimentalresults.tex
@@ -88,7 +88,8 @@ The combination of Dagster and Label-Studio is a good choice for building an act
 Dagster provides a clean way to build pipelines and to keep track of the data in the Web UI\@.
 Label-Studio provides a great api which can be used to update the predictions of the model from the dagster pipeline.
 
-% todo write stuff here
+Overall this option should just be chosen if the solution needs to be scalable and deployed in the cloud.
+For smaller projects a simpler solution just in an notebook or as a simple python script might be more appropriate.
 
 \subsection{Does balancing the learning samples improve performance?}\label{subsec:does-balancing-the-learning-samples-improve-performance?}