Highlights:

  • New features allow users in a meeting room to have their own video tile with a name, such as dynamic tiles and face identification.
  • Huddles, a brand-new feature, will soon be available in Google Chat. Instead of interrupting the conversation to enter a meeting, it will seamlessly add the meeting into the chat experience.

Google, the third-largest public cloud provider, recently started its annual Cloud Next conference in San Francisco. The event was kicked off by Thomas Kurian, CEO of Google Cloud, who made several significant announcements that might help the tech giant compete with Amazon and Microsoft in the evolving cloud market.

Given that many of the most recent developments and features are powered by artificial intelligence, generative AI was the focus of the Cloud Next 2023 conference. As competition is fierce, Google is under pressure to expand its artificial intelligence products.

Here is a list of all the announcements made by Google at the event.

Enhanced Vertex AI

Google announced Vertex AI around two years ago, a unified AI platform that provides access to all of Google’s cloud services. With the help of pre-trained and customized tooling, Vertex AI enables quick building, deploying, or scaling of ML models.

The ML platform as a service (ML PaaS) solution from Google Cloud, Vertex AI Vision, came next. The word completion model powered by PaLM 2, the Embeddings API for text, and other foundation models are just a few new tools and models developers can utilize from the wide availability of generative AI services based on Vertex AI earlier this year.

Google is including TII’s Falcon and Meta’s Llama 2 at Google Next. The only cloud service provider supporting first-party, open-source, and third-party models is Google Cloud. APIs for these models will be available through a new tool called Model Garden in its Cloud Platform.

Google Workspace’s Duet AI

Google declared “Duet AI” as the name for the generative AI elements in Workspace during I/O 2023. It was only accessible at that time via invitation to trusted testers. But as of now, Google is offering all users a free Duet AI trial for Google Workspace. According to reports, Google will charge USD 30 for each user’s access to Duet.

On the Google Cloud platform, Duet AI is a development interface driven by AI that offers chat and code support for developers. Google’s productivity services, such as Gmail, Slides, Drive, Docs, and others, are covered by various generative AI tools under the Duet AI umbrella. In essence, Google is competing with Microsoft’s Copilot.

With the aid of generative AI, Google intends to make Gmail, Sheets, Docs, Slides, and Meet more useful.

At I/O, Google introduced the “Help me write” feature for Gmail, which enables users to send auto-replies to emails and customize them to suit their needs best. The “Help me organize” function in Sheets allows users to request that information be organized in the sheet with just a few basic word prompts. Similarly, Google Slides now has a “Help me visualize” box that allows users to enter instructions to receive AI-generated visuals.

Google unveiled additional Workspace AI integration with Google’s other flagship apps, like Meet and Chat, at Cloud Next. The ability to take notes in real-time is one of Google’s new AI features in Meet. Simply click “take notes for me,” and the app will record a summary and action items as the meeting occurs. You can catch up on the meeting’s events by seeing a recap provided by Google.

Duet can “attend” a meeting on your behalf due to another new Meet feature. Simply click the “attend for me” option on the meeting invitation, and Google will automatically generate some text about what you might want to discuss. The meeting participants will have access to those notes to discuss them.

Other new features allow users in a meeting room to have their own video tile with a name, such as dynamic tiles and face identification. Additionally, a feature that automatically translates captions into real language can be found when a different language is spoken. The brand-new automatic caption translation capability supports up to 18 languages.

Google is incorporating Duet AI into Google Chat as a final step. To ask inquiries about your content, get a summary of the papers in a space, or catch up on missed chats, you may now interact directly with Duet AI. A redesigned UI and a new shortcut option have made Google Chat simpler. Google is also integrating smart canvas features like smart chips into Google Chat. You can now add 50,000 people to a space in Google Chat. The modification represents a substantial increase over the prior cap of 8,000 members.

Huddles, a brand-new feature, will soon be available in Google Chat. Instead of interrupting the conversation to enter a meeting, Huddles seamlessly adds the meeting into the chat experience. By the end of the year, Huddles will be made publicly previewable. Google will soon offer Chat support for third-party applications, including well-known brands like Workday, Zoho, and Loom.

70% of Generative AI Unicorns Utilize Google Cloud

Approximately 70% of Generative AI unicorns and more than half of all financed AI startups are Google Cloud clients, according to a statement made by Google during the event. This year’s conference demonstrated how Google provides the infrastructure optimized for running AI models.