With machine learning-based chatbots comes the ever-present worry of inaccuracies and unreliable information. As such, a new type of discipline within the realm of AI is being introduced: machine unlearning, which aims to make AI-generated output more reliable and safe by making the model 'forget' information instead of 'learning' new ones.
With machine learning-based chatbots comes the ever-present worry of inaccuracies and unreliable information. As such, a new type of discipline within the realm of AI is being introduced: machine unlearning, which aims to make AI-generated output more reliable and safe by making the model 'forget' information instead of 'learning' new ones.