NemoClaw Knowledge Wiki

AI round up by Your Own Workflow channel

4 min read

https://www.youtube.com/watch?v=mYWyq-yBJnk The video provides a comprehensive overview of the latest artificial intelligence (AI) product updates across Google’s ecosystem, including Google Workspace (specifically Google Vids), Gemini, Google Search, NotebookLM, and Google AI Studio. It focuses on new features, their functionalities, and practical applications for users. Here’s a breakdown of the updates:

  • Google Photos AI Features: Image to Video: Users can now turn photos in their gallery into short video clips. This feature is free and accessible by selecting a photo, choosing “Subtle movements” or “I’m feeling lucky,” and generating a 6-second clip. [0:31, 1:00, 1:21] Remix Photos: Users can remix their photos into various styles like 3D character, cartoon, or sketch, similar to capabilities found in tools like ChatGPT but integrated directly into Google Photos. This involves selecting a style, uploading a photo, and hitting “generate.” [0:40, 1:47] Both of these photo transformation features are integrated into a new “Create” tab in Google Photos, acting as a hub for AI projects, and are currently rolling out in the US for Android and iOS users. [2:10, 2:23]

  • Google Search Online Shopping Features: Virtual Try-on: Previously a limited experiment in Search Labs, this feature is now fully launching in the US. Users can tap a “Try it on” icon on product listings or Google Images, upload a full-length photo of themselves, and see how the clothing item looks on them. [0:41, 2:31, 2:41] Price Alerts: Google Search will now track price drops for desired items. Users can open a listing, click “Track price,” set their budget, and choose preferred colors and sizes to receive notifications when the price matches. This is designed to help users save money, especially during holiday shopping. [0:44, 3:02]

  • Google Drive Enhancements with Gemini: Video Interaction: Google Drive now allows users to interact with videos in a side panel. Users can upload their videos (or use Google Meet recordings) and use Gemini to get insights, summaries, and answers to questions. [0:24, 3:31] Automated Action Item Generation: Gemini can summarize video content, list action items, show key topics, or provide a meeting agenda. Users can also input specific prompts, such as “Generate a list of post-meeting action items sorted by participant,” which can then be exported to a shareable dashboard for team collaboration and progress tracking. [0:28, 3:45, 4:15] Performance and Bug Tracking Dashboard: Users can paste performance issues, bugs, and errors discussed in a video into Gemini Canvas and ask it to visualize them interactively, allowing for easy filtering and tracking of improvements. [4:49, 5:05]

  • Google Opal (Experimental Tool): Opal allows users to build, edit, and share mini-AI apps using natural language and visual editing, without needing to code. Users can create custom productivity AI apps with features like video and audio generation. [0:17, 5:24] Users can access Opal at opal.withgoogle.com (a VPN might be needed outside the US) and either create an app from scratch or remix existing templates from the gallery (e.g., Blog Post Writer, Book Recs, Fashion Stylist, Learning with YouTube, Video Marketer). [5:42, 5:58] The video demonstrates using the “Learning with YouTube” app to analyze a video, extract a transcript, generate a quiz, and provide a report. It also shows how to “remix” the app to add new features, like requesting a preferred language for the output report. [6:01, 6:40]

  • Gemini 2.5 Flash-Lite for Developers: This is a new, generally available model in the API, described as the cheapest model in the Gemini 2.5 family. It enables developers to build powerful apps at a fraction of the cost, with input costing about 0.40 per 1 million tokens. [7:59, 8:14] A demonstration shows how to use Gemini 2.5 Flash-Lite in AI Studio to build a prototype app that translates casual phrases into legal language, highlighting its speed and cost-effectiveness for large text inputs. [8:34, 9:24]

The presenter concludes by emphasizing the game-changing potential of tools like Opal, suggesting that developers will soon be able to deploy apps with minimal coding knowledge. [10:04]

Graph View