Join top executives in San Francisco on July 11-12, to hear how leaders are integrating and optimizing AI investments for success. Learn More
Want to be a data scientist in 2023? If so, you’re not alone. But rapidly shifting economic conditions and recent massive layoffs at companies like Meta may have many of the nearly 106,000 data scientists in the U.S., and those looking to enter the field — one in which the average salary is $100,274 per year — wondering what the coming year will bring. What skills will be most in demand? What is a data scientist’s typical day really like? What are the biggest industry trends?
Daliana Liu, senior data scientist at machine learning company Predibase and podcast host of The Data Scientist Show, likes to ask, and answer, those very questions. In fact, she started her podcast — which now boasts 55 episodes featuring interviews with data scientists from companies including Meta, AirBnB, Nvidia and Google — because she felt data science needed more dialogue around the trends, skills and lessons learned, directly from the voices of real professionals working in the sector.
After previously working as a senior data scientist and senior machine learning instructor for Amazon Web Services (AWS), Liu said she knows what it’s really like as a professional in the field.
“I can share some advice I didn’t know when I got started,” she said, adding that she sometimes felt alone on her career path. Data science, she explained, can feel siloed at times, especially with remote work.
Join us in San Francisco on July 11-12, where top executives will share how they have integrated and optimized AI investments for success and avoided common pitfalls.
“I felt there’s a gap between what I learned in school, and what I actually do, and I also feel very insecure sometimes,” she said. “I didn’t know a lot of other data scientists who worked in the industry, so I wished I could have a community and talk to them.”
No one mold for a data science role
Essentially, said Liu, a data scientist takes something raw and translates it into something meaningful. The power of data science, she explained, is making sense of the past to make a recommendation for the future.
“A data scientist is basically someone who solves a business problem with data,” she explained. “I created a meme with Sherlock Holmes looking at different pieces of evidence, except we have hundreds, thousands, millions of more [pieces of] evidence than Sherlock Holmes — and you have to find a statistical framework or machine learning solution to answer a question.”
What sometimes complicate the outside view of data science are the many paths professionals take to enter it and the niche skills they develop along the way. For example, Anaconda’s 2022 State of Data Science report found that 20% of students who hope to enter the data science profession say one of the biggest barriers to entry is the lack of clarity around what experience is actually required. And, those already working in the field report that their responsibilities are all over the map — system administration, actual data science or engineering, cloud engineering, research or even education.
Liu says this was her experience too, and many data scientists she has interviewed and worked with have said the same thing: There simply isn’t one mold for fitting into a data science role — and you don’t necessarily need to have a tech background.
“A lot of people I’ve interviewed have come from a non-tech background,” she said. “They’re just very interested in getting insights from data.”
And there are different types of data scientists, Liu emphasized. There are the generalists, who have a foundational toolbox around statistics, machine learning models and forecasting. And there are data scientists who are more specialized, working with product teams and helping the business run experiments or make decisions.
3 major misconceptions about data scientists
Throughout her own career and from her podcast talks, Liu has observed three major misconceptions about the profession:
1. Everyone thinks you’re a math genius.
“People think you have to know a lot of math, or have a Ph.D., said Liu. But actually, she explained, thanks to tools like Python or different data science packages, you don’t need to calculate everything. That said, “you do need to understand the foundation, and I believe everyone can learn that.”
Liu added that she doesn’t think she’s a math “genius.” In fact, “I struggled a lot in my undergrad degree,” she said. Overall, she added, no one is “cut out” to be a data scientist. “I don’t think I was ‘cut out’ to be a data scientist, I’ve failed,” she said. “Everybody has struggled and they’re still trying to figure things out. We’re all still trying to go to Google or StackOverflow to find answers.”
2. Data science is like magic.
“People say what we do is kind of magic, but in reality, what we do a lot of times is simply just spend time with the data,” Liu explained. “Some people call it ‘become one with the data’ — you want to start with something simple and build on top of data so you can understand how your solutions work.”
And, she added, sometimes keeping things simple and uncomplicated is the best way to do data science. “The simple solution sometimes works better,” she said. “I’d rather hire someone with good foundational skills, then have someone always talk about those advanced skills but don’t really know what they’re talking about.”
3. Intense technical problem-solving is the only way to communicate.
Data science isn’t just about technical skills, Liu reiterated. Often, it’s about soft skills like empathy and understanding.
“Besides looking at and really understanding the data and building models, we also talk to product managers in the business,” Liu said. “You need to have empathy for your stakeholders because eventually, your data science or insights are changing people’s behavior, or changing business aspects. You need to educate people and explain things.”
What will data science jobs look like in 2023?
With uncertainties about a pending recession and more layoffs, there are many questions about the future of the data science profession. But Liu says there are key technical skills and personal traits that will hold firm even in turbulent times.
Those include a focus on providing ROI to solve business problems; the ability to interpret models and their findings clearly for stakeholders; and prioritizing empathy for the end-users while solving the problems.
“You need to think like a business owner, even for machine learning,” said Liu. “You [might] have a lot of very technical skills [and] understand the models, but you also need to just think because you want to solve a business problem.”
She also expects diversity across gender and race to continue to increase in the field, and says she has noticed it happening already.
Even though statistics may be daunting — Anaconda’s report notes that in 2022, the data science profession is still 76% male, 23% female and 2% non-binary — Liu knows this is going to change.
“Don’t wait [to see more] people who look like you to do what needs to be done,” she said. “Maybe you don’t see a lot of people who look like you, but maybe that’s more motivation for you to become one and then be the representation, so other people can see you and feel inspired.”
Liu’s biggest piece of advice really has nothing to do with data science at all: “Find a balance between finding value for the business and also having a fulfilled, balanced life for yourself.”
VentureBeat's mission is to be a digital town square for technical decision-makers to gain knowledge about transformative enterprise technology and transact. Discover our Briefings.