As a Frequent NotebookLM User, Here Are 5 Features I Wish for in Gemini Notebook

Recently, Google unveiled its rebranded note-taking tool, now known as Gemini Notebook. While initially the changes seemed superficial, indications suggest major enhancements are on the horizon.

One notable new feature is “Collections,” allowing users to organize related notebooks akin to playlists. Moreover, a recent examination of the app’s APK suggests users may soon be able to include website links as sources directly within Gemini, among other potential updates.

This development sparked my curiosity regarding additional features I would hope to see in Gemini Notebook going forward. Here are five improvements that I believe would significantly enhance the user experience.

While one notebook suffices for many, complex projects may benefit from multiple notebooks to enhance organization in the early phases. For instance, a graduate student working on a thesis might manage several notebooks, each dedicated to a particular chapter. This approach helps at first, but becomes cumbersome once the direction is clear. An option to merge related notebooks would streamline the process.

Currently, Gemini Notebook does not support the merging of notebooks directly. Workarounds exist, such as utilizing third-party tools or cross-analyzing multiple notebook links, but these methods complicate what should be a straightforward process. A native merging feature would suit the AI-centric platform well, though certain challenges may explain its absence.

The primary reason behind the absence of this feature is the tool’s strict adherence to source grounding which aims to minimize AI inaccuracies. Combining multiple standalone notebooks could heighten the risk of these inaccuracies, but advancements in AI technology may soon make this more feasible.

Imagine investing significant time compiling a research project in Gemini Notebook only to need your original sources later. Ideally, you should have copies saved, but this isn’t always the case.

Right now, there’s no practical method to download previously uploaded sources like PDFs. Similar to the merging issue, workarounds do exist but aren’t ideal. The best option is utilizing Google Takeout to archive your data, allowing manual retrieval of old sources. However, a simple one-click option for individual downloads would be tremendously helpful in instances where the original materials are lost.

I appreciate how Gemini Notebook facilitates the creation of interactive quizzes and visual aids from notebook sources. However, there’s potential to extend these features further.

Picture an enhancement that converts your sources into engaging interactive formats, like competitive quizzes designed as games. I’d also love to see simple learning games reminiscent of the 90s, like the classic Oregon Trail.

This may be a tall order compared to simpler quiz games, but exciting developments are on the way.

Recently, new options labeled “App within the Studio” appeared, hinting at features like “interactive web pages,” “fun games,” or “visual tools.” The depth of these gaming features remains uncertain, but it’s a promising direction.

Imagine if Gemini Notebook integrated live-sync functionality with tools like Obsidian or GitHub, eliminating the need for manual uploads. This would allow for seamless updates to notebooks as databases change, facilitating complex inquiries that merge structured and unstructured data.

While this capability would greatly enhance Gemini Notebook, implementing it may pose challenges within the existing AI architecture and could introduce security vulnerabilities. Though thrilling, it’s understandable why this feature hasn’t been rolled out yet.

Gemini Notebook permits single sources of up to 500,000 words or around 1,000 pages. However, this limitation may pose issues for larger businesses or academic research, as it may fail to fully parse extensive documents, resulting in lost information.

Although one can upload large files, the application sometimes struggles to read them entirely, leading to incomplete data retrieval. Breaking documents into smaller sections or employing markdown can enhance readability but add to user workload.

While it’s understood that AI computing costs are a primary factor in these limits, raising thresholds for premium users would certainly be beneficial.

While many of the above features would be welcome, they are not essential for me. There may be numerous similar tools out there, but the unique experience offered by Gemini Notebook is hard to match. Admittedly, if this list had been created last year, it would have been much longer. Google has made significant strides in refining this tool, and improvements are likely to continue under the new Gemini branding.

Are there any features you think should be included that I haven’t mentioned? Please share your thoughts in the comments!