This white paper defines hybrid thinking as the cognitive process in which a human subject and one or several AI agents simultaneously elaborate the same chain of reasoning. It is distinguished from the instrumental use of AI (query, occasional assistance) and is positioned as its own category with specific risks and benefits. Three clinically documented risks —cognitive dependence, atrophy of core skills, and silent transfer of bias— and three measurable benefits —amplification of the range of addressable problems, iteration speed, assisted discovery— are analysed. A pedagogy of hybrid thinking articulated in six competencies is proposed, along with a set of cognitive sovereignty metrics enabling individuals and organisations to assess their degree of dependence and preserve structural autonomy. Implications for Latin American educational systems are discussed.
Chris Meniw (Sat,) studied this question.