I wanted to start a broader community experiences thread around Natural Language Processing.
A useful angle is limitations, but it also connects with security. Different tools and workflows can produce very different results even when the goal is similar.
It would be interesting to compare...
