In computer science, this would be somewhat like a project-based learning environment with greater autonomy and personalization where students develop their knowledge, and in turn prove their knowledge, by creating artifacts. Projects that target attempt to solve a real-life problem or that which is required by the industry would be the ideal cases. What I am advocating for is a more practice-based educational system whose aim is to nurture communities of interest.
This would imply letting go of the behaviorist approach where we try to control the environment through broad syllabus and hope that the information sticks (and transforms to knowledge). Each student has a different interest, pace, and style for learning and we need to cater to that. In schools, this is not supported at all – syllabus/standards are so vast that sometimes I am amazed that I went through all that! In college too, we have broad syllabus with limited opportunities to implement the knowledge and as Father Sarducci mentions in the video above, hardly any of it is learnt in the end. We can mitigate the problem by being flexible in the syllabus, focusing on active use of knowledge through practical projects, removing performance-based exams (summative assessments), using formative assessments with priority to student’s engagement in their topic of interest, and supporting involvement in community of interest.