Hello World, Take Two!

Hello again, World! I started this blog almost 4 years ago to write about math, data and software engineering. I failed dreadfully. Today I’m reviving this space to share data science and machine learning thoughts, learnings, projects and whimsy.

To give a sense of the directions that we may take on this blog, here are some things - none of which are completely original or novel thoughts by any means - that I believe to be true and am particularly interested in exploring further.

  • Human + Machine > Human OR Machine.
  • In a world where Machine Learning / Artificial Intelligence algorithms are becoming commoditized, understanding the underlying math / theory is important.
  • In industry, data should be a product not a service.
  • Deep Learning has and will lead to incredible applications of shallow AI but is not sufficient for general Artificial Intellgence applications.
  • For companies building data products, ML/AI infrastructure is more important than ML/AI algorithms (see bullet about commoditization above).
  • ML/AI development will move beyond the Python ecosystem within the next 5 years.
  • Math rules everything around us.
  • Imposter syndrome is real.
  • There is a non-trivial percent chance that our reality is powered by a Reinforcement Learning algorithm optimized for attention.
  • The NBA is the greatest sports league.

The main goal of this blog is to use it as a space to share my knowledge, thoughts and learnings with the world in a way that is simple enough to prove to myself that I understand completely.

Written on October 13, 2019