NotebookLM had a busy last week as the company rolled out five new AI features to the platform.
Google also started rolling out the agentic AI shopping features in AI Mode
Photo Credit: Google
Google's teams working on artificial intelligence (AI) products had a busy time last week. The company rolled out a large number of features and improvements across several platforms ahead of the holiday season. The biggest batch of upgrades came from NotebookLM, the AI-powered learning platform, which introduced five new features, while Google Photos was second in line with four new capabilities. Apart from these, DeepMind, Gemini, and even the Messages app received updates. We have listed them all down here, in case you missed out on a few of them.
On X (formerly known as Twitter), the NotebookLM team dropped five new features in 48 hours; however, the highlight was their announcement posts. There is a bit of context to it. Among the NotebookLM loyalists, a much-requested feature is “folders” to help users organise their projects better.
After the team dropped a major new feature, custom video overview styles, many users complained that it wasn't the new folders announcement. These comments were so many that Lizzie Tao, Brand and Social Media Manager at Google wrote, “I am cracking up at the number of people who are like "WHAT???? we wanted FOLDERS" on objectively one of our coolest launches to date. i promise!!!! I am WORKING on folders!!! @NotebookLM will get folders if it's the last thing I do!!!”
After that, every single new feature announcement by NotebookLM began with a version of, “We were kidding earlier, here is what you really wanted..” tricking users into thinking Folders was about to be released. However, it was not part of the scheduled releases for the previous week.
The context aside, the company did release some interesting new updates. Apart from the abovementioned custom video overviews, which allows users to type a prompt and generate video slides in the requested style, the company also introduced chat history to let users revisit previous sessions; image as a source; and integration of Deep Research. Flashcard and Quizzes were also rolled out globally to all mobile users.
Google Photos' team also had a busy time rolling out as many as four new features for users. Nano Banana image editing AI model has now been integrated into the app, letting users edit their images by just typing text prompts.
The recently added Create tab in the app is now getting a new AI templates feature. Powered by Nano Banana, these templates let users create images based on popular edits. Users can, for instance, select the photoshoot template and immediately transform their images in that style. The team also added a new Ask button to the images in the Google Photos library. The button opens up a text box and lets users start a conversation about the content of the image, find related images, or even describe the edits.
Finally, the Help me edit feature was expanded to iOS users in the US, allowing them to make edits to images with either text or voice-based instructions. However, it is not supported by the Nano Banana model.
Nano Banana was also expanded to Google Messages via a new feature dubbed Remix. The feature allows app users to reimagine any image without leaving the interface. To use the feature, users can long-press a photo in the chat. It will show the Remix button, tapping which will let users reimagine the image.
Gemini also announced two new updates for its app. First, the Veo 3.1 model now allows users to upload multiple reference images alongside the video prompt. The AI will ensure all of the videos are used to generate the final video.
Additionally, new shopping features were also added to the app. Users in the US can now ask the AI chatbot queries about shopping and see inspirations, product listings, comparison charts and pricing across multiple e-commerce platforms directly within the chat.
The same feature is also available in AI Mode in Search, and it also gets two new agentic capabilities. First is agentic checkout which lets users set a target price for a product they're looking to buy, and the AI agent continues to monitor the price in the background. Whenever it reaches the user's price point, the agent will automatically make the purchase, with permissions.
Apart from this, the company also started rolling out the Let Google Call feature. Using agentic capability, AI Mode can contact local stores to check stock availability, booking a reservation, pricing details, and more. After retrieving the information, it is presented to the user. The feature discloses that it is an automated call and merchants have the option to opt out of the service.
DeepMind was also not behind when it comes to AI announcements. The company's AI division released a research preview of the Scalable Instructable Multiworld Agent (SIMA) 2. It is an AI agent, powered by Gemini, which can play video games autonomously or following the user's instructions. The main purpose of the model is to test how AI behaves while experiencing novel worlds, and navigates across them. The end goal of the project is to use AI as the brain for general-purpose robots.
Get your daily dose of tech news, reviews, and insights, in under 80 characters on Gadgets 360 Turbo. Connect with fellow tech lovers on our Forum. Follow us on X, Facebook, WhatsApp, Threads and Google News for instant updates. Catch all the action on our YouTube channel.