Data Science in 10 years

17th May 2019

“I very often get the question: “What’s going to change in the next 10 years?” And that is a very interesting question; it’s a very common one. I almost never get the question: “What’s not going to change in the next 10 years?”

– Jeff Bezos, Founder of Amazon

It’s said that the only constant in business is change. And although that’s true tactically, underlying all these changes are principles that don’t change. This fact is intimately known by Amazon founder Jeff Bezos, who focuses on things that don’t change rather than the things that do. Bezos is betting that in 500 years customers will still want low prices, faster delivery, and broader selection. 

He uses this knowledge to guide invest in technology, like drones (faster delivery), new fulfillment centers (low prices) and broader selection (kindle).

Like Bezos, data scientists are investing their time and money in an ever-changing field. And while most are worried about the new coding language or library it may be useful to think about things that will be useful 100 years from now.

Here are 3 proposals:

1. Transactional data
Whether it’s Ancient Greece or Google, we have been keeping track of how people consume for ages. Having intimate knowledge of possible directions, recommendations and recurring trends in transactional data is something that will not go away soon.

An example is knowing at least 5 different analysis that you can build off transaction data: customer segmentation, marketing attribution, purchase prediction, recommendation system, and community detection.

 2. Communicating a clear takeaway
Explaining the path behind a model does little good if you can’t explain why a person should care. Being able to put a technical method in a business context allows others to understand what you are doing and invest in your idea.

There will always be a time when a small amount of people understand something that few understand. There is tremendous value when these few are able to explain to the majority about why they should care.

ex: A random forest model becomes a tool to know if A/B test data is bias. A regression model is automating the pricing of apartments.

3. Speed
People always want something faster and quicker. Anticipate that need by building common requests before they are asked. 

ex: Prepare business cases for transactional data. Build libraries to do common data cleaning. And even create your own data science textbook to quickly reference information when someone ask you about it.

The world in some ways is in constant change, and changing alongside it is essential.

But if you look closely, there are still timeless patterns that emerge, stick to these to reap long term rewards.

Read also: Bringing important development practice into Data Science

Tagged with:

Fernando Hidalgo

A Data Scientist that has worked with Discovery Communications and Prismoji. Top Data Science writer in Quora with posts that have received over 370K views. He is passionate about building machine learning products and democratizing data science education.

To give you the best possible experience, this site uses cookies. Continuing to use means you agree on our use of cookies.