What to study to work in AI: new careers and key skills

With the rise of artificial intelligence, specializations in this field are undergoing a significant transformation. Previously, those interested in receiving training in this field had to come from traditional STEM fields (science, technology, engineering and mathematics). However, in addition, there are also educational programs that cover the specific skills that are most in demand.

In fact, there are more than 120,000 professionals in Europe who are highly qualified in skills related to artificial intelligence. This figure surpasses that of the U.S., according to the Atomico report, 'State of European tech 2023'. "The advance of AI affects nearly all companies, regardless of the sector or industry to which they belong, when it comes to the way they engage with their customers. Today’s AI can break down some of the barriers in accessing information and knowledge, and also brings new AI-based products and businesses," explained Emilio Parrado-Hernández, professor at Carlos III University of Madrid.

Skills and abilities to work in AI

An AI professional needs a combination of technical and analytical skills in different fields, including ethics and collaboration. Josep Amorós, Head of BBVA’s Data University, pointed to some of these key skills:

• Technical skills: it is necessary to have in-depth knowledge of machine learning algorithms, master deep learning techniques, understand natural language processing and have the ability to work with large data sets.
• Programming: master languages like Python, and use tools like TensorFlow, PyTorch and other machine learning libraries.
• Data analysis: knowing how to clean, process and analyze data using tools such as SQL, Hadoop and Spark.
• Development of generative solutions: understanding and applying prompting techniques (the practice of giving instructions to AI so it can generate new content), building chatbots and using generative AI for creative and productive tasks.
• Regulation, ethics and responsibility: being familiar with the regulation that affects AI, understanding the biases in AI models, valuing the importance of transparency and making ethical decisions.

In general, these skills not only apply to artificial intelligence. They are also critical in another field: big data. "In the last decade, we witnessed the roll-out of big data and a significant number of companies started to exploit large volumes of data through machine learning models. This ‘modus operandi’ has tremendous potential, but requires fairly qualified personnel to interact with the technology,” Parrado added.

Therefore, new profiles related to AI and the analysis of large databases are in demand by companies, such as a machine learning engineer (personnel in charge of researching and designing models), a prompt engineer (expert who trains AI language models to generate specific responses) or a computational linguist (who adjusts the way artificial intelligence functions based on natural language processing).

What to study to work in AI

Identifying what to study to work in AI depends on each professional’s specific goal, given that it spans a wide range of disciplines and specializations. “It is such a broad field that there are multidisciplinary teams in which the combination of computer science, mathematics and statistics is just as important as mastering the field of application. For this reason, more than talking about specializing in AI in general, it is important to think about training adapted to each field,”  said Jesús Fernando López, Director of the Instituto de Ciencia de los Datos e Inteligencia Artificial (Institute of Data Science and Artificial Intelligence) at the University of Navarra.

Therefore, due to the multidisciplinary nature of AI work, Josep Amorós recommends training that incorporates both the traditional fundamentals of AI and the new trends in generative AI:

• Degree in Computer Science, Mathematics, Data Science, Physics and Engineering (STEM degrees). To master aspects related to programming, algorithms and data structures, which are essential to the development of AI.
• Master in Artificial Intelligence or Data Science. To specialize in machine learning, natural language processing, big data analysis, and other critical areas of AI.
• Certifications and specialized courses. Staying up to date with courses in generative AI, deep learning and advanced data modeling and analysis  techniques is critical. In fact, most prestigious universities now offer training to study artificial intelligence, including specific courses, degrees and postgraduate specializations, such as masters and doctoral programs.  Massive Open Online Courses (MOOC) are also popular on platforms such as Platzi, Coursera and Udemy. In addition, institutions like Harvard University offer artificial intelligence courses that are adapted to different areas of specialization, such as AI Essentials for Business or for computer science students, and an introduction to programming artificial intelligence with the language, Python.
• Continuous training and professional development: attending workshops, seminars and conferences on AI, collaborating on practical projects to apply theoretical knowledge in real contexts. For example, the Grow with Google platform offers several courses on AI, which go from basic concepts on machine learning for beginners to more advanced notions designed for developers of apps and engineers. Furthermore, the company has a Generative AI Learning Center.

In addition to formal training to work in AI, Emilio Parrado-Hernández, of the Carlos III University of Madrid, puts the focus on another methodology: 'learning-by-doing'. "This discipline is very experimental. There are almost always several ways to solve a problem and it is not always clear what the best option is, so we need to try,” Parrado said. “It is also very complicated to try to introduce programs with subjects and classes on all of the novelties and tools that are constantly coming out. This reinforces the fact that self-learning is fundamental.”

Sometimes, it is the companies themselves that develop training strategies in this direction. This is the case of BBVA, which signed an agreement with OpenAI to accelerate processes and promote innovation in the bank, including training for employees. “With the advancement of generative AI, we have introduced new modular programs designed for all employees - not only the technical experts, thus democratizing access to AI knowledge,” Josep Amorós added. Therefore, studying artificial intelligence is not just to prepare for the future, but to meet a need for today’s companies.