"Learning To Forget – A Weapon in the Arsenal Against Harmful AI"

Researchers are bringing attention to a real issue that accompanies Artificial Intelligence (AI) technology, which is teaching it how to forget. According to scientists, 'Machine Unlearning' is an essential weapon against AI risks. They are helping to develop new methods for making AI models known as Deep Neural Networks (DNNs) forget data that poses a risk to society. The issue is that re-training AI programs to 'forget' data is expensive and time-consuming. Modern DNNs, such as those based on Large Language Models (LLMs) (e.g., ChatGPT and Bard), require a lot of resources to be trained and can take weeks or months. This article continues to discuss Machine Unlearning as a weapon against harmful AI.

The University of Warwick reports "Learning To Forget – A Weapon in the Arsenal Against Harmful AI"

Submitted by grigby1