A Notebook feature has been added to Gemini. [Photo: Google]
[Digital Today reporter Jinju Hong (홍진주)] Google has introduced a ‘Notebook’ feature to its artificial intelligence (AI) service Gemini and made it available for free to all users. It stores conversations, files and user instructions in one space and reflects them in later responses, an attempt to shift beyond simple Q&A to a ‘memory-based work tool’.
The core of the feature is continuity of context. Users no longer need to repeat the same explanation each time and can continue work based on previously accumulated information and conversations. Gemini therefore focuses not on one-off answers but on producing results that can be carried out in practice.
On April 21, IT outlet TechRadar introduced five ways to use Gemini’s Notebook feature effectively.
First is bringing recurring tasks into one place and automating them. If frequent work such as schedules, to-do lists and subscription management is gathered in the Notebook, a request such as “Please organise this week’s plan” can pull everything together. The key is that users do not have to enter it again each time.
Second is using it as a tool to turn records into plans. After accumulating notes or ideas, users can ask it to organise them into an actionable plan, structuring scattered information into practical steps. This directly helps boost productivity because it produces results closer to what needs to be done than a typical one-off answer.
Third is building preference data to improve recommendation accuracy. If users store records of food, spending and content use in the Notebook, Gemini provides personalised recommendations based on them. Its value lies in reflecting past choices rather than making simple random suggestions.
Fourth is setting a response style in advance. If users enter instructions such as ‘Be concise’ or ‘Use a conversational tone’ at the top of the Notebook, they are automatically applied to all subsequent answers. This reduces repetitive prompt input and helps keep outputs consistent.
Fifth is separating notebooks by purpose. Users can create separate notebooks for daily work, personal use or specific projects, and ask questions on the assumption of each space’s context to get more accurate answers. Dividing use is more efficient than putting all information into a single notebook.
This update focuses more on a ‘change in the way memory works’ than on improving AI performance. Gemini is evolving into a structure that accumulates user records and continues work based on them, rather than starting over each time. As a result, future AI competitiveness is likely to depend less on how many functions it offers than on how naturally it connects users’ actual work flows.