Skip to content

Canvas, Data Analysis, Study Mode, and Voice

Canvas

ChatGPT is no longer only a chat box. Canvas is an editing surface for writing and coding projects that need revision, restructuring, or incremental improvement. OpenAI describes it as a separate interface for projects where plain back-and-forth chat is not enough.

For humanities users, canvas is useful for:

  • Essay drafting and grant text
  • Structured notes and lesson plans
  • Any document where you want to edit alongside the model rather than merely receive one more block of prose

At the time of writing, OpenAI says canvas is available on web, Windows, and macOS; check current documentation for mobile support.

Data analysis

For tabular material, data analysis is one of ChatGPT's more concretely useful features. When you upload data, ChatGPT can create interactive tables and charts, inspect rows and columns, and answer follow-up questions about the dataset in natural language.

For humanities researchers this is helpful for:

  • CSVs, catalogues, and survey data
  • Bibliographic exports and coding sheets
  • Quick exploratory work on tabular evidence

It is not a substitute for reproducible statistical workflow, but it is often a very good first pass.

Study mode

Study mode is aimed at learning rather than merely answering. OpenAI says it is available across ChatGPT plans and is designed to ask interactive questions, adapt to your level, and guide you step by step.

That makes it relevant not only for students but for academics learning a new method, tool, language feature, or technical concept.

Voice and audio

Voice mode matters less for most formal research tasks, but it can be useful for thinking aloud, note capture, travel, and accessibility.

OpenAI's desktop app also includes features for transcribing and summarising audio on supported paid and workspace plans, which can produce canvases from meetings or voice notes. That could be valuable for seminars, brainstorming sessions, and research administration, though accuracy still needs checking.

Image generation

ChatGPT can generate images using DALL-E and GPT-4o's native image capabilities. For humanities users, this can be useful for creating illustrations for teaching materials, poster design, visualising historical scenarios, or generating placeholder images for presentations. As with all generated content, the outputs should be clearly disclosed as AI-generated. See Disclosure & Ethics.