AI heavy lifting isn't just model training. There's about a million data pipelines and processes before the training data gets loaded into a PyTorch tensor.
also done in native code
Ehhh... if you're lucky. I've seen (and maybe even written) plenty of we-didn't-have-time-to-write-this-properly-with-dataframes Python data munging code, banged out once and then deployed to production. I'll take performance gains there.
From personal experience, no. I ended up writing rust bindings to call from python that turned minutes of loading into seconds.
That is the exact value proposition of Mojo.
Pythonesque (sane) syntax, great AOTC language features, great performance, memory safety, excellent Python interoperability. What’s not to like?