Google’s Post

Amazing start to #GoogleIO! A few highlights: - We’ve further improved Gemini 1.5 Pro by increasing its “long-context understanding” - this means the model can now follow increasingly complex & nuanced instructions, making it even more useful & helpful for people. https://lnkd.in/emqUsBK7   - We’ve created a series of tools to support the creative process across media, including video and images. Our video model Veo can generate high-quality videos across a huge range of visual styles, and our image model Imagen 3 can create images with stunning detail and realistic lighting. https://lnkd.in/eQNrP_yt  - We unveiled Project Astra, our effort to build a universal AI agent. Building on our Gemini model, we’ve developed experimental agents that can understand the context you’re in and respond quickly in conversation, making the pace and quality of interaction feel much more natural. This technology is still in early development - but the capabilities already feel amazing to me. Some of these capabilities will come to Google products like the Gemini app later this year. In the meantime, you can watch a video of our prototype - the video has two parts, each captured in real time: https://lnkd.in/esYGKwmQ - We announced an expansion of SynthID, our generative AI watermarking tool. SynthID will now be available for text in the Gemini app and web experience. It works by embedding imperceptible watermarks directly into the text-generation process. SynthID isn’t a silver bullet for identifying AI-generated content, but is an important building block for developing more reliable AI identification tools and can help millions of people make informed decisions about how they interact with AI-generated content. https://lnkd.in/e5rMtPYJ - We introduced LearnLM, a new family of models based on Gemini, designed to make learning more engaging, more personal, and more useful. LearnLM will appear in existing services like YouTube - where it will allow users to ask clarifying questions, get helpful explanations, or take a quiz on what you’ve been learning. https://lnkd.in/eGNEt3-Y - We shared our approach and commitment to delivering responsible AI, including how we’re improving our models and protecting against misuse with things like AI-Assisted Red Teaming, and how we’re building AI to benefit society by helping solve some of the biggest problems facing humanity - for example, we created AlphaFold 3 to help transform our understanding of the biological world and drug discovery. https://lnkd.in/ewFmj3X3 Congratulations and thank you to all of our teams for the great work to advance AI, creativity, safety & responsibility, and education! Everything covered in today's #GoogleIO keynote is here: https://lnkd.in/eP_YH7YR