On Data Science and AI
As a data scientist at an artificial intelligence startup one wonders how the two compute together? The lines are fuzzy and they will not any clearer by erecting a nice Venn diagram to resolve the issue but I hope my share of thoughts from the trenches will interest the reader.
When Google economist Hal Varian famously declared Data Scientist the sexiest job of the 21st century the year was 2009. His company had yet to hire a small army of deep learning researchers who where still pondering the next big idea in obscure Canadian bars. One thing is clear: things are moving very fast.
Data and scientific methods to make sense of it is the stuff of the revolution behind both AI and data science. This is the common thread between the two. I see them converging to a problem from two different directions:
Data science on the one hand has a high level approach. It uses data to understand and affect people, systems and environments at high granularity. It often plays out at the level of business KPIs, machine failures or even social policy.
AI on the other hand is taking more bite-sized chunks of data developing a capability to understand and affect its environment completely autonomously. It’s bottom up with breakthroughs at the level of image recognition or machine translation - operating, for example, at the resolution of individual pixels.
Will the two approaches ever meet in the middle, and who will be faster? It is only natural to see the two moving closer together over time. The data science singularity where hard AI algorithms are in charge of everything is still a long way off.
In our team at Gyana we leverage AI technology as a very potent means to better understand our high level problems at a deeper level. The success in abstraction of AI algorithms from low resolution data to meaningful information directly translates into valuable insights that help answer data science questions. It’s also very powerful in bringing decision makers closer to the data and insights. In that sense we hope to take the human by his hand. Empowered by AI, everyone will be a veritable data scientist soon!
