OpenAI, the company that made ChatGPT, recently introduced a new feature for the chatbot that lets users upload PDF files and get them analysed. The functionality is currently in beta and available for ChatGPT Plus members.
The new feature works with various types of documents like PDFs, data files and other files. ChatGPT’s latest feature can come in handy for those who don’t want to go through lengthy documents and want to get information like summaries, understand various data points and generate graphs and charts based on the data in a file.
You can enable the ‘Advanced data analysis’ feature by heading to the ChatGPT settings and turning on the feature from the ‘beta features’ section.
A user who goes by the handle ‘luokai’ on Threads shared some screenshots of their conversation with ChatGPT, using the feature to upload an image of a capybara and asking the chatbot to generate a Pixar-styled photo from the existing image.
They then uploaded an image of a skateboard and asked the chatbot to make an image where the capybara is holding it. However, the AI-powered chatbot seems to be hallucinating as it automatically added a hat to the skateboard carrying capybara.
Apart from introducing the ability to analyse files, OpenAI also introduced multimodal support, a feature that can automatically select various modes like Browse with Bing, Code Interpreter instead of the user manually changing it. It looks like OpenAI is still rolling out some of these features, so it might take some time before they are available to all ChatGPT Plus members.
Credit: The Indian Express