IA en las empresas, por Félix villar

"The incorporation of AI into companies is unstoppable, and human resources will not be an exception."

En la era de la transformación digital, donde la IA emerge como una fuerza disruptiva en múltiples sectores, es fundamental abordar su integración con cautela y consideración, especialmente en ámbitos tan sensibles como Recursos Humanos. El desconocimiento de las posibilidades de la IA en la gestión de talento o la falta de certeza legal en un ámbito tan delicado pueden convertirse en obstáculos para las empresas que desean incorporar algoritmos.

Madrid, March 27th. We talked with  Félix Villarpartner of IN2 and member of the Advisory Board of AI+EqualThis computer engineer assures that the incorporation of AI in all areas of companies is "unstoppable". This reinforces the need to adopt a multidisciplinary vision that facilitates a deep reflection on the use and objectives of AI tools, as well as on the reputational and legal implications of their implementation. It is also crucial to maintain a focus on diversity and inclusion as fundamental pillars to support the ethical application of AI in HR management.

AI+Equal.- How would you describe the current impact of Artificial Intelligence in the field of human resources?

Félix Villar: "Predictive" Artificial Intelligence has been applied for years in the productive areas of large companies, however, in the field of HR departments, and as we see in the meetings we have with their managers, this type of investment is not usually deployed with the same degree of intensity. We understand that the difficulty of justifying a clear ROI, and the complexity of having the data (in many cases sensitive) necessary to be able to tackle People Analytics and AI projects, are the reasons for this lower degree of penetration. Other barriers to entry are the lack of knowledge of the possibilities of AI and legal uncertainty in an area as sensitive as HR. In any case, the incorporation of AI in all areas of companies is unstoppable, and HR will be no exception.

The importance of data quality

IA+Igual.- What are HR departments missing to be really efficient in implementing AI tools?

F.V.- First, they must know the possibilities offered by AI applied to HR. Secondly, they must understand that without quality data and well-structured information it is very difficult, not only to implement Artificial Intelligence, but also any People Analytics process. A third relevant point is to reflect on the use and objectives we pursue by implementing AI tools, and the reputational and legal repercussions of their application.

IA+Igual.- So, what do you think will be the level of implementation of AI in the HR function in five years' time?

F.V.- I am sure that the implementation of AI in all organizational processes is unstoppable, and it will also be unstoppable in the area of human resources. Perhaps the pace will be somewhat different compared to other areas, but it will still be essential to compete in terms of efficiency and productivity. The experience we are gaining in the IA+Igual project reaffirms this opinion. We are observing, for example, that companies where HR management is part of the "core" of the company are more advanced and interested in implementing AI solutions.

IA+Igual.- What are the most important challenges companies face when implementing AI solutions in this area?

F.V.- There are many challenges, but perhaps one of the most important is to understand that AI tools are just that, tools that complement us and help us make informed decisions. However, the ultimate decision must always be supervised by a human being. We cannot transfer the weight of our decisions to a tool, even if it is an AI tool, and even less so in an area as sensitive as the workplace.

Awareness, training and certification

IA+Igual.- As a member of the Advisory Board, how do you think the IA+Igual project can contribute to address possible biases and discriminations in AI algorithms used in HR?

F.V.- I think the project will help enormously in three different aspects. First, by sensitizing HR professionals on everything related to bias and discrimination. Second, by training them in the basic principles of AI, its uses and how we can unconsciously transfer bias and discrimination to these tools. And finally, by providing them with a document (the AI+Equal White Paper) that will allow them to have a manual of good practices. With all this information, professionals in the field will have more confidence to approach an AI tool implementation project with a 360º vision that minimizes bias, promotes fairness and eliminates reputational risk.

IA+Igual.- What role does diversity and inclusion play in the design and implementation of AI systems?

F.V.- An essential role. Ensuring diversity and inclusion must be one of the pillars on which the application of AI in HR is based. Precisely, one of the paradigms offered by this technology is to help manage diversity. Until now, although companies are made up of diverse people, people management has remained uniform, the same management model, with nuances, but the same for everyone. However, the analysis capacity of AI and its speed will make it possible to establish customized management models, much more in line with this diversity. The risk is to lose common sense. As an example, we have recently seen the case of Google's AI (Gemini). When the model was trained by making certain good-natured adjustments, the result was that, when asked to generate images of German soldiers from 1943, it displayed black soldiers or oriental women in German uniforms, for the sake of inclusion. The AI is not as intelligent as we think and was not able, in this case, to contextualize the complexity of the human species.

IA+Igual.- Where does the complexity of the audit process of an algorithm lie?

F.V.- I would highlight two aspects. To have the data with which the algorithm has been trained, and which have allowed us to create the model. Or, failing that, sufficient information about them to be able to create a set of synthetic data that will allow us to evaluate it. And the need to have a multidisciplinary vision, essential for a 360º evaluation of the algorithm, including all relevant aspects of its "life": origin, motivation, objectives and use, convenience, ethics, equity, legislation, biases, explainability, life cycle; and all technical and quantitative aspects related to a BIAS audit. This multidisciplinary vision therefore requires a large team to ensure this 360° vision.

IA+Igual.- What impact do you think the success of the AI+Equal project could have on the future of AI applied to human resources?

F.V.- We hope to generate knowledge that will help human resources professionals to make AI an efficient co-pilot of their function. To create a certification framework that can guarantee that the tools they use are reliable and respond to their objectives of equity, diversity and equal opportunities. And above all, help drive the transformative role they have within the organization.

Leave a Comment

Your email address will not be published. Required fields are marked *

Scroll al inicio