OpenAI, responsible for creating and improving the ChatGPT chatbot, has introduced new features for individual subscribers of its service. The main innovations are the work and analysis of downloaded files in different formats. The neural network also received multimodal support.
As noted above, the update includes the ability to download and process files. Once a file is passed to ChatGPT, the chatbot takes a few moments to process it and can then perform actions such as summarizing the data, answering questions, or creating data visualizations based on prompts.
This applies not only to text documents, but also to graphic images. An example published on the Threads network was in which a user uploaded a photo of a capybara and asked ChatGPT to create a Pixar-style image based on it using DALL-E 3 technology.
Another important feature is multimodal support. Now users do not need to manually select operating modes from the GPT-4 drop-down list. The chatbot, based on the context of communication with the user, independently guesses the user’s needs. This greatly simplifies working with AI and makes using the service more intuitive.
The functionality of the individual Plus subscription is now similar to the Enterprise corporate plan, which previously had more advanced features.