ChatGPT itself also can remember things about users that it might probably use later —sometimes it does this robotically, or you’ll be able to ask it to recollect something. Those “long-term” memories affect the way it responds to chats. Search doesn’t have this yet—a brand new web search starts from scratch— but it surely should get this capability within the “next couple of quarters,” says Fry. When it does, OpenAI says it’s going to allow it to deliver way more personalized results based on what it knows.
“Those could be persistent memories, like ‘I’m a vegetarian,’ or it could be contextual, like ‘I’m going to Recent York in the following few days,’” says Fry. “Should you say ‘I’m going to Recent York in 4 days,’ it might probably keep in mind that fact and the nuance of that time,” he adds.
To assist develop ChatGPT’s web search, OpenAI says it leveraged its partnerships with news organizations akin to Reuters, the , , the , Axel Springer, Condé Nast, and Nonetheless, its results include information not only from these publishers, but some other source online that doesn’t actively block its search crawler.
It’s a positive development that ChatGPT will now give you the option to retrieve information from these reputable online sources and generate answers based on them, says Suzan Verberne, a professor of natural-language processing at Leiden University, who has studied information retrieval. It also allows users to ask follow-up questions.
