Question, answer. Question, answer: AI apps such as ChatGPT and Google Gemini have so far been operated in the chat format. Google is now significantly expanding this comparatively rigid operating concept. — Photo: Andrej Sokolow/dpa
SAN JOSE: Google has rolled out an expansion of its Gemini AI platform aimed at moving users away from the rigid Q&A chat format and improving how we interact with AI-generated results.
The new feature, called "Canvas," provides an interactive workspace within Gemini, allowing users to edit and optimise AI-generated content directly.
Until now, AI systems like ChatGPT have entirely relied on text prompts, but Canvas enhances this concept by enabling real-time, interactive modifications to the prompt and generated output.
If, for example, you add the words "for my 12-year-old daughter" to the heading "Plan for learning the periodic table of chemical elements", the AI recognises this in real time as a change in the specifications and rephrases the following text in a child-friendly way.
Not only can you tweak the original prompt but you can also highlight paragraphs and tell the AI to make changes to tone and text length, for example.
Beyond text editing, Canvas also supports programming tasks, such as writing and debugging code, as well as explaining existing code. This positions Gemini as a versatile tool for both content creation and technical applications.
Initially launched in English, the feature will soon support other languages, Google says.
Google's move comes amid fierce competition in the AI space, particularly with OpenAI, Microsoft and emerging players from China like DeepSeek. – dpa
