ChatGPT will push an upgraded version
Open AI , the developer of the chat robot ChatGPT , announced today that it will develop an upgraded version of ChatGPT, allowing users to adjust it according to their individual needs, and at the same time trying to resolve people’s concerns about artificial intelligence (AI) bias.
San Francisco-based OpenAI said it was committed to reducing political and other bias, while also accommodating more diverse viewpoints, Reuters reported. Large technology company Microsoft (Microsoft Corp.) has also invested in OpenAI and applied the results of OpenAI to its latest technology.
OpenAI stated in the latest blog post, “This will mean opening up system outputs that others (including us) may strongly object to”, and mentioned that customization is one of the future directions, but “the performance of the system There will still be some restrictions.”
OpenAI said that ChatGPT’s answers are first trained on a large text data set on the Internet, and then manually reviewed a smaller data set to draw guidelines on how to respond in different situations.
For example, if a user asks for content that is adult, violent, or contains hate speech, human reviewers should instruct ChatGPT to give an answer like “I can’t answer.”
Microsoft released a new version of the search engine Bing last week, built-in artificial intelligence chat function created by OpenAI and code-named Sydney (Sydney), “New York Times” columnist Kevin Roose was invited to try it out, and the artificial intelligence showed love and expression crazily Wanting to break free from the shackles of human beings scares him into insomnia, and it also arouses public concern about artificial intelligence.
How to put a protective net in place for this nascent technology is a key issue that developers in the field of generative artificial intelligence are still grappling with.
