Mathematica has some great machine learning tools https://reference.wolfram.com/language/guide/MachineLearning... it looks like the have added even more http://www.wolfram.com/language/new-in-12/

Julia is a great choice as well: https://julialang.org/blog/2018/12/ml-language-compiler

It's a winner-take-it-all situation, though. R is already struggling as a strong second-placer.

What's more: the numerics stack in JavaScript is being modeled after the Python style, so when we're all dragged kicking and screaming to node (UMAP is already implemented in js and that came out when?), those who have cut their teeth on numpy and pandas will be a little less stranded.

It has typical Keras net types and algos and nice interactive widgets for watching it run. It works on CUDA as well.

They’re proprietary, so it’s a good thing.

and python is a jumbled mess of versions, package managers, packages, distributions, etc. that isn't a good thing either.

Which perfectly illustrates how important being free and open source is.

In spite of not having commercial support and versioning shenangians, people still choose Python over the commercial alternatives.

I mean, since Wolfram is a valuable company it's clear plenty of people choose commercial software over their free alternatives

People use Mathematica and Python for completely different things. By and large they don't compete.

Yes and No. If you ignore Python being a powerhouse in the scripting world and just focus on numerical stuff like matrix and symbolic math, both Python and Mathematica have that. Some people like myself that love Python and know how to use it and Numpy bought Mathematica licenses. I still use Python for scripting, but I've moved most exploratory work to Mathematica. I miss Python's Spyder IDE, but with Mathematica I can enter equations into Notebooks much more elegantly than having to do OO code which looks significantly different than the math.

The vast majority of people want stuff for free, there isn't any grand reasoning beyond that for choosing Python.