An expert in artificial intelligence has warned individuals against placing excessive trust in the popular chatbot “ChatGPT” and discussing their life secrets, work, or political preferences with it. Mike Woldridge, a professor of artificial intelligence at the University of Oxford, stated that sharing secrets and personal data with “ChatGPT” would be “utterly unwise because anything revealed to this robot can be used in training future versions of it,” as reported by the British newspaper “The Guardian.”
He added, “Users should also not expect balanced answers to their questions and inquiries because technology will tell you what you want to hear.”
The artificial intelligence expert pointed out that many humans expect AI systems to show empathy and answer their questions with logic and consciousness, emphasizing that “artificial intelligence has no consciousness or empathy.”
He continued, “Everyone should know that anything they tell ChatGPT will be used to train future versions of it, and they will not be able to delete any information they have disclosed.”
A spokesperson for OpenAI, the developer of ChatGPT, revealed last April, “We have worked to stop saving the chat… consequently, conversations will not be used when chat logs are disabled for training and improvement of our models.”
OpenAI had previously stated that around 100 million people worldwide use the popular robot every week.
Leave a Reply