Yes, pretty much. Not just deep learning, pretty much all data-related work. And, much like Rust, it's a good question, worth asking, and asking it repeatedly helps the world arrive at a more refined answer, and improves Julia in the process.