Machine Learning as a Metaphor for Human

Machine Learning as a Metaphor for Human Learning I’m interested in using machine learning as a metaphor for human learning, as a way of trying to name the mistakes made by behaviorist assumptions. We want to provide feedback, but the error signal propagates through layers of culture, identity, role, and prior experience, so that what you think something means is not what it means at levels of internal representation, and you are probably reinforcing the wrong things. 

If our goal really is to develop powerful, situated knowledge and skills (and we should question this—at what level? embedded within which cultural assumptions), we need to think about how to provide feedback indirectly. This is where progressive pedagogy comes in: tools like personally-meaningful projects and metacognition