Kinda. The 'group' refers to a group scale and/or conformal transformations. In the context of Deep Learning, the 'scale' transform is akin to adding layers.
"it's easy to imagine cases where a great programmer might invent things worth 100x or even 1000x an average programmer's salary" ... yet they get paid maybe 1.5x
Ruby is like Disco. It was fun for a while, but now it is old news.
I was forced to learn Ruby while at Aardvark, as well as the Rails and the Agile process.
It was true today as it was then; Ruby, Rails, and Agile are absolutely useless for real data science and machine learning, and this is what companies need to move forward.
Indeed, to bring machine learning tech to the Aardvark team, I had to fight tooth and nail against the Rails consultants and the Agile process guys.
Ruby is a useful gateway language for learning other modern scripting languages such as python, lua, and scala
I agree that Ruby is now legacy technology.
Let me be clear; I use it all the time. It is useful to build stable projects based on solid ideas, and I use it for dev ops, for crawling the web, and for making simple DSLs. Chef is very good (compared to, say, Ansible) , and the Bundler deployment system is quite nice. It is easy to build web endpoints in Sinatra, and the Nokogiri gem is a pleasure to work with compared to, say, java JSoup.
But Ruby is not moving the needle forward. It does not even have a decent NumPy-like module, which makes it completely useless for modern back end work in data science, machine learning, and the real things companies need to generate revenue
Rails is a good example of being overspecialized. They are so good at web programing circa 2007 that the future passed them by.
Python. Lua. Scala. Closure. These are the future.