Don't study data science as a career move; you'll waste your time!
In 2019, I reflected on the evolution of data science and its long-term career implications. My core thesis is that data science as a standalone discipline is not a sustainable career path—domain expertise and research methods are far more valuable.
Why Data Science Isn’t a Secure Career Path
Data science is a term created to connect problems with experts, but the current shortage of data scientists will fade as general-purpose tools become more advanced. When this happens, professionals will need deep expertise in either a specific domain or research methods to remain competitive. Most existing data science programs are too shallow to provide either.
From Pharmacist to Data Scientist: A Personal Journey
I transitioned from a pharmacist to a data scientist by continuously adapting to new professional challenges. My background in pharmacy, combined with self-taught data analysis skills, allowed me to apply scientific methods to real-world problems. This path required learning tools and methodologies on the job rather than through formal "data science" training.
Who Can Call Themselves a Data Scientist?
The title "data scientist" is self-assigned—there are no formal qualifications. In my current team, we have professionals with diverse backgrounds: a pharmacist (me), a physicist, an electrical engineer, a computer scientist, and two mathematicians. None of us had formal data science training; instead, we developed expertise through advanced degrees (M.A. or Ph.D.) and hands-on experience.
Most data science programs are useful for professionals with deep domain knowledge who need to learn data tools, or for managers. However, these programs often lack the depth required for rigorous scientific methodology or computer science fundamentals. Graduates may become mediocre programmers who can tweak machine learning libraries but lack the expertise to solve complex problems.
Lessons from the Past: The Rise and Fall of Bioinformatics
When I started my Ph.D. in 2001, bioinformatics was booming, with many companies and universities investing heavily in the field. Today, few professionals still identify as bioinformaticians—most have transitioned into data science, management, or other roles. This shift suggests that data science could follow a similar trajectory as tools become more automated and accessible.
As Barb Darrow noted in Fortune, advancements in tools like Tableau and emerging automation technologies are reducing the need for specialized data scientists. With over 100 graduate programs worldwide producing data scientists, supply may soon outpace demand, diminishing job security and perks.
How to Build a Sustainable Career
The best way to future-proof your career is to develop versatility and expertise. This is achieved not through crash courses but by solving hard problems, preferably under supervision—often through obtaining an advanced degree. A degree in an applied field (e.g., biology, finance, engineering) teaches you how to use data science as a tool while deepening your domain knowledge. This approach ensures you can bridge knowledge gaps and adapt to evolving demands.
I advocate for treating data science as a tool rather than a career goal. My path worked for me, but it may not be universal. Deep learning, for example, could redefine data science as a standalone scientific field. However, until then, domain expertise remains the most reliable foundation for long-term success.