During the event, Google’s development team marked a new era in artificial intelligence (AI) by showcasing advances in Gemini, its AI assistant, along with other AI products and features across its portfolio.
Read on to learn about five of the most important takeaways from the event.
Aside from that, the company revealed it will be offering a private preview of an even larger model of Gemini 1.5 Pro with a context window of 2 million tokens. It also introduced Gemini 1.5 Flash, a smaller, faster AI model.
In addition, Pichai introduced new AI features coming to Google services. Enhanced Gemini interactions will enable users to communicate with AI through Google Messages. The upcoming Gemini Live feature, accessible to Gemini Advanced subscribers, offers a natural conversational experience by leveraging advanced speech technology. Users can converse with Gemini, select from various voices and interrupt for clarification. Gemini Live can assist in tasks such as job interview preparation and will integrate camera capabilities later this year for environment-based discussions.
Google also launched Gems for Gemini Advanced subscribers, offering a personalized AI experience. Users can create custom Gemini versions and provide role descriptions such as “gym buddies” or “coding partners.” Gemini will then process these inputs to generate tailored Gems for users to interact with.
Furthermore, Google introduced Search with Video, which allows users to ask questions and obtain detailed solutions by “showing” Google a live video feed. During the demonstration, one of Google’s developers asked Google to help her figure out why her record player wasn’t working and to identify a specific part. Google could tell what part of the record player she was asking about (the tonearm) and provided instructions for troubleshooting.
NotebookLM will basically function as a private tutor, generating study guides, quizzes and audio overviews based on input text material and more personalized adherence to learning styles.
In collaboration with YouTube, Google has also launched Music AI Sandbox, an innovative platform that offers musicians creative support. Musician Michael Rebillet likened it to having a friend offering creative suggestions. Musicians can now compose new instrumentals from scratch, experiment with sound transfer and enhance their tracks.
Finally, there’s Veo, Google’s newest AI-powered video-generation software, which is capable of generating 1080p videos that are over a minute long.
They company also introduced a new addition to the Gemma family — PaliGemma is an open vision-language model available on multiple platforms like GitHub, Hugging Face, Kaggle and Vertex AI Model Garden.
For Pixel users, Gemini Nano with multimodality offers richer image descriptions and scam alerts during phone calls. Google is bringing Gemini Nano’s multimodal capabilities to TalkBack later this year, where it will enhance image descriptions for users who are blind or have low vision. With TalkBack, Gemini Nano will provide detailed and accurate information about images, even without an Internet connection.
With Gemini on Android, users can ask questions about videos and PDFs, isolating specific features within images using Circle to Search.
Alphabet traded flat on the day of Google I/O. It closed Wednesday (May 15) at US$173.88.
Securities Disclosure: I, Meagen Seatter, hold no direct investment interest in any company mentioned in this article.
Leave a Reply