ChatGPT Tests New Feature: Automatically Selects Relevant Tools Based on Context

2023-10-30

OpenAI is rolling out new beta features to ChatGPT Plus subscribers. According to reports from subscribed users, this update includes the ability to upload and process files, as well as support for multimodal capabilities. Essentially, users no longer need to select options like "Browse via Bing" from the GPT-4 dropdown menu; the system will guess the user's needs based on context. These new features bring the office functionalities from OpenAI's ChatGPT Enterprise plan into the standalone personal chatbot subscription. After uploading a file to ChatGPT, it takes some time for the bot to process the file before it can be used. The chatbot can then perform operations such as summarizing data, answering questions, or generating data visualizations based on prompts. This chatbot is not limited to text files alone. On Threads, a user posted a screenshot of a conversation that included an uploaded image of a squirrel and requested ChatGPT to create a Pixar-style image of it using DALL-E 3. They then uploaded another image of a jittery skateboard and asked to incorporate it, iterating on the concept of the first image. Somehow, ChatGPT ended up putting a hat on the squirrel in the end? Please note that the HTML tags have been retained, while the style and class attributes have been removed.