Users will now be able to customize the feature based on individual needs, such as training the chatbot to specialize on a specific topic.
It’s an announcement marathon for Open AI. And after the new Sora video generation model, and the $200 subscription to ChatGPT Pro, comes Canvas. The interface is accessible through the OpenAI website and ChatGPT desktop application for Windows, the company described it as a “Google Docs with AI superpowers“. Canvas had actually already been pre-announced in October, but now new functions have been integrated.
The interface is an editing tool for writing and coding that sits alongside the user’s ChatGPT chatbot conversation. The new feature allows you to edit responses and “collaborate” with ChatGPT. Not only with Canvas will it be possible to “obtain feedback and corrections in real time”, explains the company. Users will then be able to make changes based on suggestions from ChatGPT. “Please note that Canvas is only available on GPT-4o. If you reach your GPT-4o limit, you will have to wait for it to reset continue to use Canvas“.
The new features of Canvas: how to use the interface
There will be new features compared to the beta version presented in October. For example, the interface will be able to handle the Python programming language. Not only that, users will also be able to customize the feature based on individual needs, such as training the chatbot to specialize on a specific topic.
As OpenAI explains in a note on the site: “With Canvas now you can highlight specific sections to indicate exactly what you want ChatGPT to focus on. Like a copy editor or a code reviewer, can provide online feedback and suggestions with the entire project in mind.” To activate the feature, simply select the option Compose a picture from the menu and then click on View tools.
OpenAI’s first announcements: the $200 subscription and Sora
The first of OpenAI’s 12 planned announcements was a $200 monthly subscription for its flagship chatbot. Customers who purchase it will have unlimited access GPT-4, Advanced Voiceand also an exclusive o1 version that uses advanced calculations for provide more accurate answersyou will also be able to handle longer and more complex conversations. In fact, model 01 was trained “with reinforcement learning to perform complex reasoning. o1 thinks before answering: it can produce a long internal chain of thoughts“, the company explained.
Open AI also introduced Sora, perhaps the product most awaited by users. It is a stable diffusion model that generates a video starting from a text. “The AI can generate complex scenes with multiple characters, specific types of movement, and accurate subject and background details. The model has a deep understanding of language, which allows it to accurately interpret instructions.” Sora at the moment though it is not available in Europe. OpenAI, in fact, will be able to distribute the new model when it is available in compliance with community regulations on privacy and artificial intelligence.
