A must for machine learners and educationalists alike.
In an age when the digital world is dominated by machine learning, it is timely that Dehaene has written about how humans go about learning. It turns out that machine learning is slightly simpler than the learning as done by the human brain. For example, newborn babies appear to start life with pre-existing knowledge circuits. These enable the baby to grasp quickly concepts such as the physical nature of objects, numbers, probabilities, faces and language. This potentially puts Dehaene at odds with most supporters of artificial neural networks, who see no reason for going beyond learning by associating and connecting data. Educationalists get a steer too. Discovery-based teaching gets a hell of a bashing.
Link to book: https://www.waterstones.com/book/how-we-learn/stanislas-dehaene/9780241366462
You may also like to browse other neuroscience books: https://www.thesentientrobot.com/category/neuroscience/neuroscience-books/
