@mhoye Lots of great answers already in the thread. I think that, in 2024, there needs to be at least one session on "ML/AI In Practice"...

* how is training a model different than writing an algorithm
* how to think about data, about training, about inference
* how to improve the quality of a classifier, a predictor, a transformer
* what people with ML/AI backgrounds expect to get from software engineers, and what you should expect from them
* how to talk about safety, bias, and error