Exécution de notre outil de synthèse sur un nouvel article
C'est la première fois que cet article est demandé et nos outils de synthèse d'IA n'ont jamais été exécutés dessus. Nous pouvons exécuter nos outils maintenant si vous cliquez sur le bouton "Exécuter" sur la page mais assurez-vous d'abord que c'est le bon article.
Why We Don't Have AGI Yet
Peter Voss, Mladjan Jovanovic
Résumé : The original vision of AI was re-articulated in 2002 via the term 'Artificial General Intelligence' or AGI. This vision is to build 'Thinking Machines' - computer systems that can learn, reason, and solve problems similar to the way humans do. This is in stark contrast to the 'Narrow AI' approach practiced by almost everyone in the field over the many decades. While several large-scale efforts have nominally been working on AGI (most notably DeepMind), the field of pure focused AGI development has not been well funded or promoted. This is surprising given the fantastic value that true AGI can bestow on humanity. In addition to the dearth of effort in this field, there are also several theoretical and methodical missteps that are hampering progress. We highlight why purely statistical approaches are unlikely to lead to AGI, and identify several crucial cognitive abilities required to achieve human-like adaptability and autonomous learning. We conclude with a survey of socio-technical factors that have undoubtedly slowed progress towards AGI.