Subscribe to DSC Newsletter

The Data Scientist Will Be Replaced By Tools | Forbes

Do you agree with this? I don't, I think this Forbes article is using a provocative title to get you to read it. While assembler programmers in the seventies were eventually replaced by compilers and programming language interpreters, I believe that real statisticians and data scientists can't fully be replaced by machines or software. When they are, it results in software misuse and faulty analyses.

But data scientists do produce tools to fully automate a number of tasks, such as fraud detection, automated car driving, automated weather forecasts, and automated bidding strategies. Still, these tools require maintenance and permanent machine learning - something difficult to fully automate. Data modeling and data architecture (including identifying the right data sets and right fields) are also difficult to automate for new problems where templates don't exist. And production of automated tools itself is (sometimes, not always) difficult to outsource to a robot. That's an area where data scientists (which have many skills including engineering, computer science and business management) are very useful. 

Along the same lines, do you believe that the following jobs can be replaced by tools:

  • Lawyers (yes, only up to some extent)
  • Doctors (yes, only up to some extent - could be partially replaced by crowd sourcing, and surgeons partially replaced by robots)
  • President of United States (could a robot do a worse job?)

Anyway, here's the articleThe Data Scientist Will Be Replaced By Tools

We’ve barely started to use the term “data scientist” and the demise of this new profession is already predicted? Well, it’s not one more “rise of the machines” prophecy but instead the provocative title of a proposed panel for the upcoming SXSW.

The organizer of the panel, Scott Hendrickson of Gnip, has provided a useful run-down of some of the arguments for and against the possible disappearance of data scientists. Supporting the proposition are the current scarcity of data science talent and a slew of startups providing “data science as a service.” As an example of the opposition to the “democratization of algorithms,” Hendrickson quotes Cathy (Mathbabe) O’Neil who wroterecently that “if your model fails, you want to be able to figure out why it failed. The only way to do that is to know how it works to begin with. Even if it worked in a given situation, when you train on slightly different data you might run into something that throws it for a loop, and you’d better be able to figure out what that is.” In other words, machines will never have the deep understanding of the tools of data science that is required to practice data science.

Views: 3862


You need to be a member of Data Science Central to add comments!

Join Data Science Central

Comment by Bob Vanderheyden on February 28, 2016 at 4:01pm

15+ years ago, I was told that statisticians would be obsolete in 5 years, because software would replace them.  I have more job security today, than I had back then. As long a Data Scientists evolve, to tackle more advance problems, they will still be in high demand in 20 years.

Comment by Jonathan Seller on September 5, 2012 at 11:07am

I wouldn't worry about it too much; I'm a software developer and tools were supposed to replace me a long time ago. Peoples expectations of what that actually means isn't even close to the reality. 

Comment by Tom Riecken on September 4, 2012 at 11:01am

Well... there will probably be an abundance of intelligent tools for reporting and trend monitoring.. But integration, ETL, data validation, data integrity and project scoping... That can't be automated unless you've got Artificial General Intelligence, but by that point nobody will have work anymore.

Comment by Chandrasekhara S. ("C.S.") Ganti on September 4, 2012 at 4:40am

Yes, I don't agree -- like Vince Granville indicated, I have not  truly yet  earned formally a Data Scientist title - and the demise is premature --  if  you  consider dealing  with lots of Data / Data Mining is a per-requisite.

However, I  do very well appreciate all the Data / tools and Technologies as an Applied  Statistician / Operations Researcher (for a long time). Yes, we did save / or proved $$$$$ benefits based on Logistic Regression / Simulation / Modeling for large industrial and property Insurers / for Trade Associations / Federal Regulators.  Lot of Luck for for Data Scientist future.

Comment by HURT on September 4, 2012 at 2:09am

However the article raises a right question as there is a risk of bubble for each new "technology" (in short). The tool providers also aim at democratizing their tools. Actually, some should be able to help tracking the causes of their failures, gaps..., not even considering machine learning.

More over, automatization of new tasks is an axis of the tool supply, for example dedicated to targeting marketing campaigns.

Designing new uses is a field where data scientists should bring value.


I published another very light commentary about one month ago. It's in French but some automatic translators can succeed, no ?

Comment by Mary Kardel on September 3, 2012 at 3:12pm

Data scientist has become a generic title. Is market research the same as cancer research? No, both use statistical tools, but for different jobs. Use the right tool for the right job. If you are not a carpenter, have you ever tried to build a dog house or bird house from scratch? You must know what you are doing to choose the right tools and know how to use them. In my opinion, data scientists will no longer be needed when the pocket calculator replaces mathematicians.

© 2019   Data Science Central ®   Powered by

Badges  |  Report an Issue  |  Privacy Policy  |  Terms of Service