What is the point of this talk? You don't talk much about native Clojure library for deep learning, which does not exist (Cortex is still in early phase and has not seen a release since Oct 2017). If you rely heavily on Python libraries using a wrapper, why not just use Python? When you develop a piece of software you have to think about long term maintainability. How many Clojure developers would you find compared to Python developers? You need to show us what Clojure offers in the machine learning space that is not available in other languages. One of the zens of Python is "Readability counts." Compare that to 17:39, you could as well write in Assembly.
What is the point of this talk? You don't talk much about native Clojure library for deep learning, which does not exist (Cortex is still in early phase and has not seen a release since Oct 2017). If you rely heavily on Python libraries using a wrapper, why not just use Python? When you develop a piece of software you have to think about long term maintainability. How many Clojure developers would you find compared to Python developers? You need to show us what Clojure offers in the machine learning space that is not available in other languages. One of the zens of Python is "Readability counts." Compare that to 17:39, you could as well write in Assembly.