Since its debut last year, ChatGPT, an AI chatbot that has attracted attention (and is partly feared) in all fields, including education, journalism, and the military industry. Tech companies like Google are the most shaken.
# I don't want confidential data to be used for ChatGPT training
Amazon Lawyers Consider ChatGPT Risks. Business Insider reports that Amazon employees have been instructed not to share confidential Amazon information, including code currently in development, with ChatGPT. It seems that concerns within the company have increased as Amazon officials witnessed ChatGPT giving a response very similar to Amazon's confidential data .
An internal memo, shared on Slack, explains how important this is to avoid the risk of confidential data being used as future training data for ChatGPT.
# Don't become the modern version of "King's ears are donkey ears"
Asking AI to write code for you, or asking you to correct your own code, is already a way to use ChatGPT. Sometimes the code that comes out is perfect, and sometimes it has holes.
However, before using the code written by AI, many people may not be aware of the risk of sharing the base. Since the other person isn't human, I feel like it's okay to talk about secrets, but... AI absorbs what it is given more and more and connects it to learning. It is impossible to predict where and how the secrets you confide in will be output.
In the future, I wonder if we need literacy about the risk of sharing secrets with AI. We are entering an era where AI is included in "Don't tell anyone".