AI researchers have fought for nearly 35 years as to whether artificial neural networks (ANNs) could ever be a plausible model of human cognition.
A team from New York University has created a new AI model that mimics human’s ability to generalize language learning, fine-tuned to human reactions.
To make the neural net human-like, the researchers trained it to reproduce the patterns of errors they observed in humans’ test results, matching words with colors, for example.
Much like a child gets practice when learning their native language, the new model improves its skills through a series of repetitive learning tasks, reducing the volume of data required to train today’s ANNs.
MORE