Here’s What You Need to Know About This Week’s Google Announcements

Image via Google

Google announced a lot of AI stuff this week at their Cloud Next 2024 event in Las Vegas. Among announcements about open source tools and new chips, there were a few important bits of AI news that should be on your radar.

Google Vids

Image via Google

Google is introducing a video creation tool, Google Vids, to its Workspace suite. It allows for in-browser video editing, writing, and production. Think of it as an additional tool alongside Sheets and Docs. Basically the tool makes videos automatically using your footage, stock content, and text. You can create a variety of video content, including product pitches, training materials, and team videos. Starting with a simple prompt, users can incorporate files from Google Drive or utilize Google’s stock content. The AI then crafts a storyboard for the video, streamlining the video creation process.

Gemini 1.5 Pro

Gemini 1.5 Pro, Google’s most advanced generative AI model, is now accessible in public preview on Vertex AI, their AI development platform aimed at enterprises. Gemini 1.5 Pro expands the amount of context it can process 128,000 to 1 million tokens. The means you can add approximately 700,000 words or 30,000 lines of code to your prompts, significantly surpassing the input limitations of competing models like Anthropic’s Claude 3 and OpenAI’s GPT-4 Turbo. As we noted before, the model lets you paste entire books and ask for interesting quotes, debug long sections of code, and offer more oodles context for chat responses. It’s probably the most powerful model on the market right now, at least until the other folks catch up.

Imagen 2

Google is expanding access to Imagen 2, the latest version of its AI model capable of creating and editing images from text prompts. Launched at Google’s I/O conference in May, Imagen 2 was developed with technology from Google DeepMind and represents a significant upgrade in image quality over its predecessor. This new model introduces the ability to generate text and logos within images. Imagen 2 aligns with competitors like OpenAI’s DALL-E 3 and Amazon’s Titan Image Generator in text and logo generation but stands out by supporting multiple languages, including Chinese, Hindi, Japanese, Korean, Portuguese, English, and Spanish. It also offers logo overlay capabilities on existing images.

Vertex AI Agent Builder

In a move that scratches a huge itch for every industry, Google has announced its AI Agent Builder, a tool that lets enterprises build agents that can search your website or internal data to find answers.

The tool integrates with the previously released Search and Conversation tools. It leverages the advanced Gemini large language models and incorporates both RAG APIs and vector search techniques. The goal is to offer an agent or chatbot that won’t hallucinate or give your customers free flights when they get confused.

Ready to start using AI like a pro?


Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.