Emanuel Gouveia
Coordinator of the Bachelor's Degree Program in Applied Data Science
Universidade Católica Portuguesa (Braga)
Anyone who follows developments in artificial intelligence (AI) has grown accustomed to seeing what would have been science fiction just a few months ago make the news—and then, by the next day, become so commonplace that it doesn’t even raise an eyebrow. While some of us are still testing the waters with one or two chatbots that give us ideas for dinner, armies of autonomous agents are browsing the web, shopping, training their own subagents, and posting on forums off-limits to humans, where they complain about the stupidity of their creators. This dizzying progress even makes us believe that one of these days we’ll wake up and no longer have a job. But then we go to handle some business at a school, where we see that doing research means sifting through stacks of paper. And then we remember that the digital transformation hasn’t happened yet. It’s happening, at different speeds.
It’s true that developing applications is now faster and cheaper. For organizations that need to go digital, this means a lower barrier to entry. With digitization, an organization becomes a data factory, and data science is the practice that can turn that data into a competitive advantage. Order histories make it possible to predict future delivery times, sales histories give rise to systems that recommend what customers should buy next, and much more.
In this context, the profession of data scientist will continue to gain importance: someone who uses statistics and computing to put data to work for the business. As software development becomes more accessible, data scientists can set themselves apart through their interest in the domain of application and in the processes that generate the data. They can add value through their curiosity in exploring the data and by choosing the modeling approach best suited to each problem. They can ensure the usefulness of their systems by developing them with a focus on how these systems will perform once deployed.
The fascination with AI and the funding opportunities that reflect it create a sense of urgency among companies to use AI as a solution before they even know what the problem is. Data science is needed—not to jump on the bandwagon, but to point out that “the emperor has no clothes,” or, more specifically, something like “we have a data quality problem that can’t be solved by slapping an interface together on top of ChatGPT.”
A data scientist who views their role as that of a programmer who receives pre-processed data and delivers a technically valid report or model is already replaceable by tools that are readily available today. Data scientists of the future must cultivate, in addition to technical competence, an interest in the context of the organizations where they work and the critical thinking needed to question choices that may be unfounded. With these characteristics, and as pivotal players in the dialogue between the worlds of technology and management, they will remain irreplaceable. At least for a few more months.
