In 1974, Danish computer science pioneer and Turing award winner, Peter Naur proposed “datalogy or data science” as an alternative name for computer science.
However, when it comes to modern data science, it is said that Dr. DJ Patil and Jeff Hammerbacher, who were the heads of analytics and data at LinkedIn and Facebook respectively, coined the term ‘data science’ in 2008.
Fast forward to 2012, Harvard Business Review hailed Data Science as “the sexiest job of the 21st century” and in 2016 Glassdoor marked it as the highest paying job of the year. And this elevated the entire scenario of data science.
The Prominence of Data Science
Modern businesses are awash with data. Hands down. And in the past few years, these businesses have realized that they can outpace their competition and win by making sense of this data.
In order to solve complex business problems and elevate data-driven decision-making efforts, companies are seeking to hire the best talents.
Furthermore, as the popularity of data science continues to grow in leaps and bounds, you’ll see that it’s a no-sweat to incorporate base-level data science.
The new-age startups are also playing a major role in democratizing data science. They are enabling companies with all sorts of data and computing power they need. Meaning, they have eliminated all the hardware and software that were costing a king’s ransom.
Will Data Science Become Obsolete? This is What Data Scientists Said
But (there’s always a BUT).
As the evolution of this affair continues to progress, many are skeptical about the longevity of data science. People believe that it will become obsolete in the next few years. Worst still, data science might even be extinct.
Is this true, by any chance? Is data science really going to go out of the scene?
This is the current debate and data science professionals all around the world are trying to put their thoughts.
Recently, we got in touch with a few data scientists to know what’s their take on this affair and this is what they have said:
(We will keep adding inputs from other data scientists as well.)
In my opinion, Data Science will not become obsolete, although the way it is perceived and worked on will evolve and change with time.
In the past few years, data science boomed significantly. A lot of enthusiasts did successfully joined the tribe.
Now that research in the field has increased, the field of data science is expanding and requires more knowledge. Institutions have started taking benefit of this scenario by offering specialized courses. And this is why the DS boom might slow down a bit.
Also, data science has become a fancy name in most of the industry. I have seen people implementing ML algorithms in every possible problem even where it can be solved with simple statistics.
Going forward, heavy investments by industries on DS may fall if it is not too much required with future research. But the need for AI ML will not go away from industry and our daily lives.
Implementations like Smart home devices, search, social media, etc. will keep on growing and will be requiring new features/improvements and research on a time to time basis.
Overall I think the field will evolve and eventually those professionals will survive who have a thorough understanding of ML, Maths, and Computer Science.
Is the role of a data science professional going to become obsolete? No. Never.
Let me put it very simply.
For instance, I am sitting with my client to discuss a certain problem and its solution. I cannot ask my AI bot to sit with my clients. It’s ultimately the professional who is going to help me with insights.
But yes, there are certain day-to-day tasks of a data scientist that can be automated, which is happening, and it is basically to help data scientists get rid of some mundane activities.
Data science may evolve and might get some new elements to it, but it is never going to become obsolete. Rather, it is also going to require significant human involvement.
The sole reason Data Science or data scientists may become obsolete is AI. But in the coming years, the Data Science requirements will continue to increase till AI is smart enough.
Artificial Intelligence will be responsible for taking away a lot of responsibilities, only those who will learn to adapt and change their skills according to the changing society will survive.
For that state to come, machine learning algorithms will need to be smart enough. So, we are safe till someone develops those but that can take a couple of decades.
As per my experience, I don’t think it will become obsolete because when we talk about Machine Learning & Computer Vision, it has a lot to cover in the next 20 to 30 years.
Building AI solution is another factor because only a few companies make it to the production stage for business use cases.
Why is that happening? The major part is Data Set.
If we talk about India, most of the industries don’t have Data Pipeline which makes it a difficult job for AI engineers to build a solution.
I guess now the time is to focus on Data Collection & Data pipeline techniques. Learners are more focused on learning python, machine learning & other skills but these are secondary factors.
And I feel, due to this reason we have more than 7 lakh certified Data scientists struggling to land their first job. The priority should be on understanding business use cases and then how to convert them into Machine Learning solutions or AI solutions.
The fundamental proposition of data science includes understanding a problem statement, acquiring relevant data to exploit, and knowing the correct means to do it. While the former seems much easier a task from a technical point of view, it probably is the hardest to automate.
In theory, AI is supposed to predict the correct mathematical algorithm that would, in turn, be implemented towards a given problem statement. But, the question is, how reliable would automation be in the world of utter randomness and extreme variation? And can AI facilitated solutions bring about 100% client satisfaction?
Well, not all intelligence is artificial. Analyzing problems logically and being aware of the underlying practical issues could hardly ever be automated.
If at all, implementation of AI in certain domains of data science would only add leverage and serve as another tool in a data scientist’s toolbox rendering data science far from obsolete.
Subscribe To Our Newsletter.