Many in the field struggle to comprehend how something as lucrative-sounding as data science can ever be considered dead, given that LinkedIn has called it "the most promising career" and Glassdoor has called it the "greatest job in America." "Data science is the field of study that combines domain expertise, programming skills, and knowledge of mathematics and statistics to extract meaningful insights from data," according to the definition.
Therefore, data science as a field will not become obsolete anytime soon unless and until we discover a way to not use data itself. However, many people believe that because the day-to-day tasks of a data scientist are quantitative or statistical in nature, they can be automated and there will be no need for them in the future.
The idea came from the fact that autoML models can partially automate some data scientist tasks like data cleansing, data visualization, and model building. Domain expertise is important. However, despite the fact that the tools might be able to finish the job quickly, many of them do not include "domain expertise" in the definition of a data scientist.
Domain expertise is the extensive knowledge that data scientists use to enhance their data science abilities. Therefore, even though a significant portion of the data pipeline and workflow is being automated, a data scientist is still required to translate the business problem into the appropriate format.
In addition, selecting the appropriate data science model based on industry is challenging. especially when the industries are so diverse; A video streaming platform would not benefit from a recommendation algorithm developed for the healthcare sector.
Tina, a former Meta data scientist, is of the opinion that correctly contextualizing a model is the most undervalued aspect of the job of a data scientist. She said, "There was a machine learning module that screened for content integrity and then demonetised them if they broke the rules" when discussing her time working on Instagram's integrity at Meta. In addition to obtaining the model's data, my task was to determine what constitutes a rule break.
Tina says, "The problem is that ML models can't detect "unknown unknowns." If you can't even measure something, how do you know if it even breaks the rule? There is dependably a harmony between free discourse and respectability."
“Data science is a field where only 50% of the potential has been realized,” states Dr. Vaibhav Kumar, senior director for data science at the Association of Data Scientists (ADaSci).
The field, in my opinion, still has a long way to go and needs a lot of work. Data scientists are still responsible for deciding what to do next, despite the fact that machine learning may be utilized in a number of the workflow's tasks. What do these aftereffects of the model mean? How can you tell if the model is actually performing well? What's the measurement," he asked Point.
Dr. Vaibhav stated, "In the field of data science, there will always be a need for human assistance, which machine learning alone cannot provide."
Is it then dying?
When it was suggested that AI might take the jobs of accountants and auditors away, the concern first surfaced in the accounting industry a few years ago. However, even if an AI program can do almost everything an accountant can, you still need an accountant's expertise to get tax credits, exemptions, and other benefits.
Similar to this, a data scientist might rely on autoML models to collect, visualize, and clean data so they can focus on business needs more. Additionally, given that data science is still in its infancy in numerous conventional fields like finance, healthcare, defense, and governance, the demand for data scientists will only grow in the future.
The funny thing is that a data scientist gathers the data that AutoML needs before it can even begin exploring data.
You must be logged in to post a comment.