Image by Author | Ideogram
# Introduction
When I first started my data science career in 2020, the field was booming. Everywhere you looked, companies were hiring data professionals. At that time, I built a data science portfolio and managed to land several high-paying clients.
I would write data science content, such as white papers, articles, and technical documentation — which paid between USD \$500 and \$1,000 for two days of work. I built simple machine learning models and conducted analyses using tools like Tableau and Power BI. As clients started recommending my work and leaving positive reviews, I landed more projects. I worked 5 to 6 hours each day from my couch and was completely remote.
Recently, however, I’ve changed things up.
I’ve quit a few freelance jobs for a full-time data science position — one where I go to the office every day and work double the hours. And no, it’s not because the job pays more. In fact, I made more money as a freelance data scientist than I do now.
So why did I switch from a comfortable, high-paying freelance job to a full-time position that pays less?
Read on and you’ll find out the three top concerns that led me to taking this action.
# 1. Building Technical Skills
When I worked for myself, I realized I’d hit a plateau in learning technical skills. I was working more like a machine, generating repetitive outcomes for the same freelance clients. This meant that I not only worked less, but my technical knowledge had reached a standstill.
A reality check came when I attended a reputable tech conference and networked with other data professionals. I realized I hadn’t kept up with much of the technology they discussed. These data professionals were building AI agents and retrieval-augmented generation (RAG) systems, while I was refreshing the same dashboard for the hundredth time and writing white papers on Python for data science.
Don’t get me wrong — a data scientist’s value is in the outcomes they drive, and in many cases, fancy tools like large language models (LLMs) are akin to using a sledgehammer to crack a nut. However, I lacked basic knowledge of tools that were at the forefront of tech companies, and that scared me. I’ve witnessed firsthand how complacency and the unwillingness to adapt to new tools has rendered tech employees obsolete.
# 2. Being Paid to Learn
At my current full-time job, there are training courses led by AI experts that teach you to integrate LLMs into your data science workflows. Regular hackathons with teams like data and software engineering allow you to gain skill sets that go beyond your scope of work. There are peer-led tutorial sessions almost every week where other team members walk you through a problem they solved and show you how to build a similar project. This saves a ton of time and teaches you far more than most online courses.
A full-time job is the one place where you learn on somebody else’s dime, instead of having to enroll yourself in a $1,000 bootcamp.
When I focused solely on freelance work, two things happened:
- Firstly, I wasn’t incentivized to learn new things unless a client had a problem that required me to upskill.
- If I did have to learn something new, I typically paid for an online course.
And if I got stuck or didn’t understand something, I didn’t have anyone around who could help me grasp the concept.
3. AI-Proofing My Career
This might be controversial to some, but the biggest reason I got a full-time data science job is because I believe it will help secure my career from AI. And while this might sound counterintuitive, hear me out.
With my freelance job, here’s what I learned:
- How to use my existing skills to solve the client’s problem
- Gathering client requirements and using them to solve a specific technical issue
However, with a full-time job at a large tech company, my scope now involves:
- Gathering a business requirement and working with teams like product, design, and engineering to turn it into a data problem
- Making key product decisions
- Understanding how the company’s data warehouse works and using it to build data pipelines
- Building relationships with stakeholders and peers
With freelance work, you typically solve a targeted technical problem for the company — such as building a dashboard and refreshing it every quarter, or creating a machine learning model for a specific use case. The requirements are clearly specified, and you just need to focus on execution with your technical skills.
However, AI is democratizing technical skills.
It allows people who don’t know how to code to build applications. People who don’t know SQL can easily write a query and create a comprehensive dashboard. As AI continues to democratize technical skills, the value of data science freelancers will likely decline. The pay will decrease, and the space will become more competitive.
Conversely, a corporate role is multifaceted. It requires far more collaboration, domain expertise, critical thinking, and understanding of the business. As you climb the data science corporate ladder and reach higher positions within the company, you’ll become more difficult to replace (even as AI models get better). Also, you can transition to roles like business analyst or product manager and even negotiate higher salaries. To put it simply, there are many ways to move forward in a corporate role. You can oversee data solutions and drive business value in ways that don’t overlap with AI’s capabilities.
On the other hand, working a freelance job where the only value you bring is your technical skill puts you in a vulnerable position.
For that reason, I have decided to prioritize long-term career safety over short-term income. I chose a lower-paying full-time job over freelance data science roles to build a set of skills that will keep me relevant in the next decade, regardless of how AI impacts the technical side of the profession.
Summary
To summarize, I quit my comfortable, high-paying freelance roles to take a much more demanding full-time data science job. And I did it for the following reasons:
- To learn technical skills at a faster pace
- To climb the corporate ladder and prioritize long-term financial stability over short-term income
- To secure my career from AI by gaining experience and learning skills that cannot be replaced (such as business and product knowledge, stakeholder management, and critical thinking)
YMMV, however, so I encourage you to do your own research. Drop a comment below if you feel you have valuable insight for others.
 
Natassha Selvaraj is a self-taught data scientist with a passion for writing. Natassha writes on everything data science-related, a true master of all data topics. You can connect with her on LinkedIn or check out her YouTube channel.