Efforts made to keep robot speech responsible

Artificial intelligence innovations such as OpenAI’s GPT-3 open new frontiers in language processing and generation but also enhance the risk of its use to “spread chaos, scale harm and watch the world burn.” Open AI is so concerned that it is offering a free content filter to identify profane language but also hate speech used against certain groups and people. (TechCrunch, June 1, 2022, by Natasha Lomas)

OpenAI withheld release of GPT-3 to developers until it created best practices and content guidelines. The company also prohibits projects that allow “others to knowingly generate content with political, sexually explicit, violent, harmful, and hateful themes.” (ITPro, November 19, 2021, by Connor Jones)