Yuka Yu (喩舜君 高原由佳)’s Post

Want a quick overview of ALL the major announcements from Google I/O 2024? 👀 Check out this 10-minute recap video from Google, covering everything from AI innovations to the latest Pixel phones. 🤖📱 #GoogleIO #AI #Google #Pixel #GDGBrunswick

🌟 𝐆𝐨𝐨𝐠𝐥𝐞 𝐈/𝐎 𝟐𝟎𝟐𝟒: 𝐀 𝐋𝐞𝐚𝐩 𝐢𝐧𝐭𝐨 𝐀𝐈'𝐬 𝐅𝐮𝐭𝐮𝐫𝐞! 🚀👨💻 Google's annual I/O event showcased groundbreaking AI advancements that are setting the stage for a smarter, more integrated technology future. Here are the top takeaways: - 🌐 𝐏𝐫𝐨𝐣𝐞𝐜𝐭 𝐀𝐬𝐭𝐫𝐚: Google's ambitious project aims to create a universal AI agent, capable of understanding and anticipating your daily needs – essentially your personal AI sidekick. 🤖 - 🚀 𝐆𝐞𝐦𝐢𝐧𝐢 𝐔𝐩𝐝𝐚𝐭𝐞: The introduction of Gemini 1.5 Pro in Workspace Labs and its multimodal capabilities means smarter, more intuitive interactions across Google products. Whether it's summarizing emails or generating video highlights, Gemini is transforming how we interact with digital content. 📊 - 🔍 𝐄𝐧𝐡𝐚𝐧𝐜𝐞𝐝 𝐆𝐨𝐨𝐠𝐥𝐞 𝐒𝐞𝐚𝐫𝐜𝐡: With AI-generated overviews and video query capabilities, Google Search is becoming more sophisticated, providing summaries of complex topics and enabling searches based on visual content. 📹 - 👀 𝐕𝐞𝐨 – 𝐍𝐞𝐱𝐭-𝐆𝐞𝐧 𝐕𝐢𝐝𝐞𝐨 𝐌𝐨𝐝𝐞𝐥: Meet Veo, Google's most advanced generative video model yet, capable of creating detailed 1080p videos from text or image prompts, enriching the way content is produced and consumed. 🎬 - 🔧 𝐀𝐝𝐯𝐚𝐧𝐜𝐞𝐝 𝐀𝐈 𝐓𝐨𝐨𝐥𝐬: The introduction of Trillium, Google’s latest CPU, offers unmatched compute performance, supporting the vast demands of AI training and operations. 🖥️ - 📱 𝐀𝐧𝐝𝐫𝐨𝐢𝐝 𝐀𝐈 𝐈𝐧𝐭𝐞𝐠𝐫𝐚𝐭𝐢𝐨𝐧: The rollout of Gemini features in Android will make smartphones even smarter, providing context-aware assistance and enhancing user interaction through multimodal inputs. 📲 𝐃𝐢𝐯𝐞 𝐃𝐞𝐞𝐩𝐞𝐫: - 𝟏𝟎-𝐦𝐢𝐧 𝐑𝐞𝐜𝐚𝐩: https://lnkd.in/eQrB46ve - Google'𝐬 𝐂𝐄𝐎 𝐅𝐫𝐨𝐦 𝐒𝐮𝐧𝐝𝐚𝐫 𝐏𝐢𝐜𝐡𝐚𝐢: https://lnkd.in/eQiJeEak #𝐆𝐨𝐨𝐠𝐥𝐞𝐈𝐎 #𝐀𝐈 #𝐈𝐧𝐧𝐨𝐯𝐚𝐭𝐢𝐨𝐧 #𝐓𝐞𝐜𝐡𝐓𝐫𝐞𝐧𝐝𝐬 #𝐅𝐮𝐭𝐮𝐫𝐞𝐎𝐟𝐓𝐞𝐜𝐡

Out of an interest in learning what Google's BARD AI can do, I enabled it for my personal Google Workspace tenant. Only preparation needed was enabling Early Access Apps AND Core Data Access Permissions. Following this goto: Bard.Google.com and click the 'Try Bard' button (bottom right). #bardai #google

Google I/O 2024 event resume. Such great features. Project Astra Veo Generative AI Video Model Trillium CPU Axion Processor Gemini AI Google Search Gemini AI Video Search Gemini AI 'Live' Voice Chat Gemini AI Gems Android 15 AI Powered Search Circle To Search Gemini AI Context Aware Android 15 with Gemini Nano with Multimodality

AI is making history now on a daily basis. Exhibit A: The recent Google I/O conference, where they unrolled a truly groundbreaking AI system that combines the power of an AI search and iteration platform with all the Google properties you already have on your platforms like Gmail, Drive etc., AND all the properties stored on your phone (photos, notes, contacts, mic, speakers, camera). It can see, hear, and think in ways we've never seen before. Get ready for the future. Between this, and the changes already coming down the pike with Microsoft Copilot, the way we work and play will take a giant leap forward. It's up to all of us to make sure we harness the best parts of this innovation, without getting our information exploited in the process. https://lnkd.in/ggEk5YQy

Wow! What a firework of exciting News at the Google I/O yesterday: 👓1 - Project Astra This is an AI assistant for everyday life. Via a camera - for example with a mobile phone or smart glasses - the AI can capture the entire environment and answer questions about it. 👀2 - Google Search Google search, is to be completely remodeled thanks to AI. AI Overview is intended to provide directly relevant information about the topics you are looking for, without having to laboriously search them together manually. 🎥3 - AI for Video, Veo Gemini can also generate videos in the future. With the generative AI program Veo, video clips can be created based on text inputs. As an example of what the model can do, Google demonstrates several prompts and the videos. 👨⚕️4 - Gems "Gems" are personal AI features for Gemini that you can create yourself. If you want an application that helps you with work, leisure planning, sports or the organization of everyday life, you can create it individually and access it faster. 💾 5 - New Trillium TPUs Trillium features the third generation of SparseCore, which describes as “a specialized accelerator for processing ultra-large embeddings common in advanced ranking and recommendation workloads.” This, will allow Trillium TPUs to train models faster and serve them with lower latency ⚡️6 - Gemini model: 1.5 Flash. ⚡ It’s a lighter weight model compared to Pro and optimized for tasks where low latency and cost matter - like chat applications, extracting data from long documents and more And much much more News about #Gemini, #Workspace, #GooglePhotos

#Google significantly amps up its #AI game including demonstrating on-device AI on #Android devices at #GoogleIO 2024. During I/O, Google also presented an initial iteration of the much talked about universal #AIAssistant. Referred to as #ProjectAstra by Google, this real-time, #multimodal AI assistant has the ability to perceive the world, identify objects and their locations, and assist with a wide range of tasks or inquiries. #Gemini #ArtificialIntelligence #LongContext

Google I/O Project Astra Google recently announced their product launches in the event and the most intriguing one to me is Project Astra. Remember that scene in Iron Man where Tony Stark asks Jarvis to identify an object and pull up its blueprints – all with a simple voice command? Well, Google's Project Astra, unveiled at the recent I/O event, might just make that sci-fi scenario a reality. But what exactly is Project Astra, and how can it transform our daily lives? Let's break it down without the tech jargon. A super-powered Google Assistant that can not only understand your voice commands but can also see the world around you. That's the core idea behind Project Astra. It's a multimodal AI assistant, which means it can process information from your phone's camera or even future smart glasses, along with your voice and text. Here's how it can be your everyday hero: 1. Instant Object Recognition: Point your phone and Astra can use its image recognition skills to tell you what it is and how to care for it. 2. Real-time Translation on the Go: Astra can translate signs, packaging, or even conversations in real-time, making travel a breeze. 3. Contextual Assistance: Show Astra the instructions or the object you're working on, and it can provide relevant video tutorials or suggest missing tools. 4. Enhanced Shopping Experience: Astra can bridge the gap between the physical world and online shopping with just a scan. Project Astra isn't just about fancy tricks. It's about using AI to understand the context of your situation and provide truly helpful assistance. Think of it as having a super-informed friend by your side, ready to answer your questions and solve problems on the fly. For learning more - https://lnkd.in/dH79wRay

Google I/O 2024 KeyNotes: 1. Gemini 1.5 Pro: Offering 1M tokens for all users and a 2M token context window for private developers, this model is designed for general performance. 2. Gemini 1.5 Flash: Optimized for tasks requiring low latency, this model is ideal for narrow or high-frequency tasks. 3. LearnLM: This latest model family, based on Gemini, is fine-tuned for learning. Google also introduced Gemini Flash, adding a fourth model to its suite. Gemini Flash is optimized for narrower or high-frequency tasks where speed matters the most. It offers 1M token capacity at a slightly lower cost than GPT3.5. Additionally, Gemini introduced Gemini Gems, Gemini Live, and Project Astra, a live video understanding personal assistant chatbot. The Gemma Model Family expanded with Gemma 2, now up to 27B tokens, and PaliGemma, the first vision-language open model. Other launches included Veo from DeepMind, Imagen 3 for creating photorealistic images, and the Music AI Sandbox collaboration between YouTube and DeepMind. SynthID watermarking was extended to text, audio, and video, and Trillium was introduced as the codename for TPUv6. #google #generativeai #gemini https://lnkd.in/dg6hqtgD