The popular chatbot ChatGPT, the prime example of emerging generative artificial intelligence (AI) technology, will soon be able to memorize information and data about its users, its developer company OpenAI announced on Tuesday.
The function is initially being tested in a small group. For example, the software could remember that you have a daughter who likes jellyfish – or in which format you prefer to receive summaries of meetings at work, OpenAI explained in a memo on its website.
If you then ask ChatGPT to design a birthday card for your child, for example, the picture could include a jellyfish wearing a party hat.
For ChatGPT to memorize information about users in the future, they must ask the chatbot to do so the first time the function is executed. You can then also query what the software knows about you – and delete all or individual details.
Custom interactions alongside the memory feature are intended to make the chatbot more helpful, OpenAI said. Temporary chats are available for conversations without personalization. The information from these chats is not used for further training of the software.
ChatGPT is an AI chatbot that triggered much of the artificial intelligence hype a year ago, after being launched in November 2022.
AI chatbots such as ChatGPT are trained with huge amounts of information and can formulate texts at the linguistic level of a human.
The principle behind this is that they estimate word for word how a sentence should continue. However, it can still give completely wrong answers, even if it only receives factually correct inputs.
