Google’s initial ventures into the generative AI market have yielded various results. However, this hasn’t deterred the tech giant—under the umbrella of parent company Alphabet—from introducing new products, services, and functionalities.
For example, Google recently unveiled that all eligible users in the U.S. opting to utilize its experimental AI-powered digital note-taking application, NotebookLM, can do so freely. Initially accessible to a limited group via a waitlist in July, NotebookLM has now been enhanced with Google’s latest Gemini AI model, specifically the “Gemini Pro” variant, positioned as the most advanced among the three iterations released this year.
The Gemini Nano, a compact version requiring less computational resources, is tailored for smartphones operating on Google’s Android open-source platform. Conversely, a more extensive version with increased parameters (representing connections between artificial neurons) is slated for release in the upcoming year.
Operational Mechanism of NotebookLM
As reported by Wired, NotebookLM was developed in collaboration with artist Steven Johnson. This innovative tool enables users to consolidate multiple documents from their desktops or Google Drive into a unified digital workspace. Subsequently, users can interact with Google’s AI, leveraging the uploaded documents to seek answers.
According to Google Labs Editorial Director Steven Johnson and Product Manager Raiza Martin, when files are uploaded to NotebookLM, the system swiftly grasps the pertinent information, enabling it to respond to queries based on the provided resources. This statement was outlined in a Google blog post introducing the foundational version of NotebookLM.
In conjunction with the Gemini update, NotebookLM has undergone enhancements enabling it to analyze and process up to 20 documents simultaneously, each containing 200,000 words.
For instance, a student preparing for an exam could upload various study materials to NotebookLM and engage in a dialogue with the AI by posing practice questions. Similarly, a medical researcher could submit previous work or data, utilizing Google’s chatbot interface to seek insights from the AI.
The concept is groundbreaking, aiming to democratize the creation of personalized AI assistants that cater to diverse users across industries, offering tailored knowledge retrieval for individual tasks, projects, or organizations.
Initial Observations and Limitations
Regrettably, NotebookLM currently lacks the capability to analyze or browse web links, including those embedded within user-provided notes. This limitation stands out as a notable gap for a company renowned for its web crawling and indexing prowess.
To analyze external information, users must manually save and upload PDFs of web pages or copy-paste text into a Google Doc within their Google Drive for NotebookLM’s review.
During informal testing, occasional shortcomings were observed, such as overlooking crucial details like prominent figures in the U.S. Declaration of Independence, despite their presence in the source materials. Additionally, the AI occasionally struggled to extract information from provided PDFs. Nonetheless, after multiple attempts, NotebookLM eventually produced accurate responses.
Pictures of Google NotebookLM tests. VentureBeat is a payment.
