In her book, The Worlds I See: Curiosity, Exploration, and Discovery at the Dawn of AI, ImageNet creator Fei-Fei Li continuously references the North Star as her moral signpost, guiding her on a human-centric, ethical trail as she traversed uncharted frontiers in the emergence of modern AI. Her work has been hailed as a source of inspiration and a call for continued scientific discovery to improve the human condition.
In this presentation, we focus on false stars that shimmer enticingly in the distance, tempting us with distractions and illusions, as well those steadfast constellations like Polaris (Fei-Fei Li’s North Star) that guide us with unwavering light towards developing and deploying ethical AI. In this regard, we will commence with establishing a mental image and conceptual framework of ethical AI in the context of talent assessment. This discussion will take as its centrepiece AI’s coming of age within the framework of the Assessment Centre (AC) method. Thereupon, we will investigate the manifestation and sources of biases, systemic inequalities, and practices that may harm or adversely affect AC participants. We will highlight risks and warning signs that, if not heeded, may lead to shortcuts, false insights, and ultimately compromise validity. This will be juxtaposed with guidelines for attaining fairness, transparency, and inclusivity, creating a fair and equitable talent assessment process that fosters genuine leadership potential.
The philosophical/theoretical component of this presentation will be grounded by practical insights and research findings gathered in developing AI solutions within the AC framework – particularly AI-administered and -scored video interview and simulation exercises.
We will conclude our session by proposing a taxonomy for evaluating AI-driven assessment methodologies against the Polaris or True North of ethical practice.
Jaco holds a Master’s degree in Industrial Psychology and is a registered psychologist with the HPCSA. He is a previous ACSG committee member and served on the committees of the Society for Industrial & Organisational Psychology (SIOPSA)’s Pretoria Branch and People Assessments in Industry (PAI) Interest Group.
The greater part of his professional career centred on talent assessment within the frameworks of role fit, succession, personal growth, and team/group effectiveness. Through the course of this 30+ year career, he accumulated considerable experience in back-office (strategic product and process development) and client-facing (needs analyses, collaborative solutioning, and professional practice) roles. This included developing, deploying, and managing assessment centre technologies in general leadership/business and technical/functional contexts.
Jaco’s current role at Evalex Talent Solutions focuses on product development and enhancement, particularly in the AI domain. He is enthralled by exploring new horizons in product development, particularly at the interface of behavioural science, artificial intelligence, and business acceleration.
He is a regular presenter of research papers at the ACSG.
%202.jpg)
