In the midst of the boom that enthralls the technological giants, with artificial intelligence becoming the new totem of innovation, perhaps the words of the CEO of LinkedIn, Ryan Roslansky, have gone unnoticed by ordinary mortals: “The best jobs will no longer belong to those who have the most prestigious titles, but to those who are adaptable, forward-thinking and willing to learn and adopt new technological tools.”
A message as encouraging as it is disturbing. According to the company’s own data, around 70% of the offers published on its job portal already require knowledge in artificial intelligence (AI). But what do you have to do to become one of those most in-demand profiles? How difficult is it to become an expert in artificial intelligence? And, more importantly, is there a path that does not require changing sectors completely?
The first question that arises as soon as the question is raised is: where do you start studying? The starting point for those new to AI is usually data analysis and computer skills related to information processing, which allow computers to learn from data and make decisions. No less important are the basic notions of mathematics, especially statistics and linear algebra, which constitute the basis of a high percentage of artificial intelligence algorithms.
In the second step of the pyramid, according to Enrique Serrano, president of the Artificial Intelligence and Big Data Commission of AMETIC, it is required to have minimal knowledge in programming language, placing special emphasis on Python, along with libraries such as NumPy and Pandas for data manipulation and analysis, and frameworks such as TensorFlow or PyTorch, which are sets of ready-made tools that make it easy to build and train artificial intelligence models without having to program everything from scratch. “If someone is an expert and does not know how to use NumPy, Pandas… well, it is strange. In the end, in a project of this nature, only 15-20% is AI, 80% is data preparation,” he asserts.
Once the essential knowledge is acquired, it is complemented by two other areas: machine learning and deep learning. In simple words, the machine learning allows machines to learn from data to make predictions or classifications, while the deep learning uses artificial neural networks, which resemble the functioning of neurons, to detect complex patterns in large amounts of data. This makes possible applications such as image recognition, language processing or virtual assistants.
Serrano adds that using tools such as GitHub, a platform to manage and collaborate on code projects, or publishing work on Kaggle, a community to practice with databases and participate in data science competitions, shows that the professional can apply these technologies in real projects and use AI practically.
With all of this, the profile of what could be considered an AI expert is drawn: a common path between recent graduates and professionals trained in computer engineering, computing or related branches. For José Varela Ferrío, head of AI and digitalization at UGT, this technical profile—often identified as data scientist or data engineer (data engineer)—corresponds to those who have, in their words, “control over the main AI tools on the market.” However, there is a second route that is more accessible for those workers who want to make a 180-degree turn in their career path and that does not require such a high technical level: specializing in artificial intelligence within their own sector.
The path of AI beyond the technical level
The market also does not need all professionals to be AI engineers or scientists. Currently, two clearly differentiated paths coexist to enter this field, and each one offers different strengths and limitations. The first profile is that of the technical expert, someone with extensive knowledge of models, algorithms and deployments. In theory, you should be able to apply AI in almost any sector. But this breadth has an obvious limit: without understanding the specific business – law, medicine or logistics – your ability to well define the requirements or interpret the real impact of a model is lower. “He has a lot of general knowledge, but he will have more difficulties understanding what the business requirements are,” says Varela. Companies are not looking for more complex models, but rather more useful solutions, and therefore business knowledge is becoming increasingly important.
On the other side of the coin is the professional who masters a specific sector – for example, a lawyer capable of automating the review of contracts – and decides to complement his training with a master’s degree or artificial intelligence module. In this case, the strength is in hybridization: it will not be an expert in AI, but someone capable of introducing it meaningfully within its competence. The disadvantage is that this knowledge is less transferable to other sectors. At this point, Varela emphasizes that “using generative AI (ChatGPT or Gemini) has nothing to do with knowing how to deploy an AI system within a company, with all its technical, ethical and operational implications.”
“Before, perhaps you wanted to have an expert in artificial intelligence who knew something about marketing. Now what you are looking for is a marketing expert who knows something about AI. Therefore, it will be much more profitable for you to specialize in your sector with a touch of AI, than not to specialize in AI with all the technical knowledge that it requires,” indicates Rubén Nicolás Sans, director of the Higher School of Engineering, Science and Technology at UNIE University.
Human capabilities, keys
It is clear that coding, debugging data or facing the headache of generating predictive models with which a chatbot Answering doubts and questions from users is essential to consider yourself an AI specialist. But they are not the only requirements. Away from programming, there are other skills that are just as relevant, according to the experts consulted by FiveDays. And if one had to be highlighted above the rest, it would be analytical capacity. “In such a changing environment, the safest path is not hyperspecialization, but rather developing capabilities that remain stable over time: understanding the fundamentals of technology, learning quickly, exercising judgment, imagining new uses and leading transformations within organizations,” explains Boris Walbaum, founder of Forward College, a higher education institution that offers bachelor’s degrees from the University of London.
That is to say, technique is not everything. Walbaum highlights the importance of human skills as competencies that allow professionals to remain relevant, beyond mastering a specific tool. “What really makes the difference is the ability to understand what you are trying to solve, analyze a situation, question assumptions and work with others to turn an idea into something functional,” he maintains.
Added to this is communicative ability. For Varela, a profile can be technically brilliant, but if it is not able to explain the purpose of the project, it will end up overshadowed: “If in a meeting you are not able to explain what you have done, why you have done it and for what purpose you present the project, you end up having a problem. On the other hand, there are people with less technical productivity who stand out because they communicate very well what they do and why they do it.”
Finally, in the opinion of several experts, ethics in the use of artificial intelligence is another aspect to take into account. “We observe many specialists who know how to design AI to make decisions about layoffs or public aid and they have no idea about fundamental rights. It is terrible,” warns Varela. Along the same lines, Sans remembers that it is essential to be aware of how to use this new technology, when to resort to it and how to use its tools appropriately, always questioning the answers they offer.
The ideal path
Once the duality that emerges in the sector is clear – between the purely technical profile of an engineer and the professional who applies AI in their area – and the capabilities that companies demand, one last question arises: what is the ideal path to becoming an expert? The answer is, to say the least, diffuse and not without nuances. However, there is one point in common: training, whether through a university degree, a business course or a master’s degree, always with the aim of becoming familiar with the architecture of AI. And even better is to combine it with a specialization in a specific area.
And once formed, it is important to understand that AI does not completely replace a job, but rather enhances it. Automating a process is fine, but the key is to understand what each process does, why it is done that way, and how it fits into the context of a job. “You have to learn the basics, because it is very difficult to carry out an accounting analysis if you do not know the minimum accounting standards. Thinking that AI only consists of learning about AI is living with your back turned to the world,” says Varela.
As Walbaum warns, amid the AI obsession, aiming to become an absolute expert is the wrong goal: “A more realistic path is to build solid conceptual foundations, cultivate imagination, learn to work with AI tools, and spend enough time in real organizations to understand how change happens.” Because at the end of the day, technology will change, so will the models, and what seems essential today will be an accessory tomorrow. The only thing that remains is the human capacity to interpret, decide and make sense. And perhaps the market will end up asking for more humanity and fewer machines.
For more updates, visit our homepage: NewsTimesWire
