Author: Matthew Burruss
Publisher: Matthew Burruss Blog
Publication Year: 2022
Summary: The following article discusses how OpenAI’s GPT-3 has the potential to provide the world with truly transformative technology. Engaging in meaningful conversations about ways that this technology could be harmful is the most important hurdle. The model has been found to be biased towards certain gender, races, and religion. While the male descriptors seemed to have a wide spectrum that covered a variety of adjectives, many of which were positive, the female descriptors focused largely on external, appearance-related factors. The model also shows other biased associations of sentiment with religion, owing to the fact that it was trained on the text people have written across the internet. The solution to this lies in more deeply exploring the biases and ethical considerations of the model.