OpenAI introduced a new way to interact with ChatGPT on Thursday: an interface it calls “canvas.” The product opens a separate window, next to the normal chat window, with a workspace for writing and coding projects. Users can generate text or code directly on the canvas, then highlight sections of the work to have the model edited. Canvas is in beta for ChatGPT Plus and Teams users on Thursday, and Enterprise and Edu users next week.
Various consumer AI providers are converging around editable workspaces as a practical way to use generative AI. ChatGPT’s new interface offers similar features to Anthropic Artifacts, launched in June, and viral coding partner Cursor. OpenAI is racing to offer competitive pricing, and launch new capabilities of its own in ChatGPT, as a means to grow its paid user base.
AI chatbots today cannot complete large projects from a single prompt, but they can often create a good starting point. Editable workspaces, such as the canvas, allow users to fix parts of an AI chatbot’s output that are wrong, without having to scan their prompt and generate a new line of code.
“This is just a more natural interface for collaboration with ChatGPT,” said OpenAI product manager Daniel Levine in a demo with TechCrunch.
In our demo, Levine will select “GPT-4o with canvas” from the ChatGPT drop-down window. However, OpenAI says that the canvas windows will jump as soon as ChatGPT detects a separate workspace could be useful, for example for longer outputs or complex coding tasks. You can also type “use canvas” to automatically open a project window.
Levine showed TechCrunch how ChatGPT’s new features could help you write an email. Users can send to ChatGPT to generate an email, which will then appear in the canvas window. Then users can change a slider to adjust the length of the writing to be shorter or longer. You can also highlight specific phrases, and ask ChatGPT to make changes like “make this sound friendlier”, or add emojis. Users can also ask ChatGPT to rewrite all mail as in another language.
The characteristics for the coding cloth are a little different. Levine instructed ChatGPT to create a web server API in Python, which he generated in the canvas window. By pressing an “add comments” button, ChatGPT will add online documentation to explain the code in plain English. Also, if you highlight a section of code that ChatGPT has created, you can ask the chatbot to explain it to you, or ask questions about it. ChatGPT is also getting a new “code review” button, which will suggest specific edits for the code in the window, generated or written by users, for them to approve, edit themselves, or reject. If you press approve, ChatGPT will take a stab at fixing the bugs itself.
Once the web is out of beta, OpenAI says it plans to offer the feature to free users as well.
Themes
AI, ChatGPT, OpenAI, TC