I looked into Ruby and it struck me as being a truly aesthetically pleasing language, and since it espoused the "principle of least surprise" -- with caveats, naturally -- most Ruby code tended to be straightforward to read.
Python on the other hand, while clean -- I came from Perl -- had a fair bit of ugliness like the "self" argument in class methods, and the "dunders" in dunder methods (__init__). That said, the Python ecosystem, especially in the numerical computation space, was far ahead of Ruby's at the time, so I chose Python. It worked out really well and allowed me to accomplish what I needed to. To this day I have no regrets choosing Python.
Compared to many languages (like Haskell etc.) Python's design isn't the most aesthetically pleasing or orthogonal or clean. But being a Dutch design, at least initially, it was very pragmatic. And Python is by no means ugly on an absolute scale, on the contrary. It’s just not as beautiful as some others.
Python's main advantage is the energy behind it: to this day there are no converged alternatives to Numpy or Pandas on Ruby.
I can't help but wonder that if in an alternate universe/timeline, Travis Oliphant, Pearu Peterson, Wes McKinney and others happened to have encountered Ruby first, we might have a different de facto language for scientific computing and data science.