Google debuts a slew of generative AI features as fight with rivals OpenAI, Microsoft, and Apple heats up
Google (GOOG, GOOGL) unveiled a slew of generative AI products at its Google I/O developer conference on Tuesday, including its Gemini Live assistant, updates for its Android and Workspaces platforms, and an overhaul of its flagship Search product.
The announcements are part of Google’s broader effort to reestablish itself as Silicon Valley’s AI leader after it was caught off guard by Microsoft’s (MSFT) partnership with OpenAI in 2022.
The new products also come a day after OpenAI unveiled its GPT-4o AI model and days before rivals Microsoft and Apple (AAPL) debut their own major generative AI updates.
The biggest announcement is that Google is updating its Search platform to include generative AI results at the top of the screen. The feature, which was originally available as part of a preview in the company’s Search Labs, is now coming to all US users and rolling out globally in the coming future.
Another one of the biggest announcements was a new, personalized AI assistant called Gemini Live. The feature, powered by the company’s Gemini 1.5 Pro model, will allow you to speak directly to the assistant on your mobile device, providing you with responses in a number of natural-sounding voices.
On the Android front, Google says it’s expanding the assistant to interact with apps like Google Messages and Gmail, allowing users, for example, to drag and drop generated images from Gemini into texts and emails. Users can also use the app to ask specific questions about YouTube videos, and it will scour the video’s captions to find an answer.
Google is also updating its on-device version of Gemini, called Gemini Nano, with new multimodal features. Multimodal means that a generative AI model can work across mediums including text, images, and video.
Google’s vice president of engineering for Gemini experiences, Amar Subramanya, told Yahoo Finance he’s been using Gemini Live for brainstorming, bouncing ideas off of the assistant. Subramanya said early testers have also been using Gemini Live for scenarios including translation.
In the coming months, Google says Gemini Live will be able to access a user’s cameras, allowing the assistant to interact with the real world and things it sees. OpenAI has also shown off these capabilities in demos of its GPT-4o.
In one example, Subramanya said he asked the assistant to find a recipe for a pineapple upside-down cake for 15 people and add the ingredients to his Keep shopping list. The assistant found a recipe for eight people, scaled the proportions from eight to 15 people, and put the items Subramanya needed to put it together into his shopping list.
Gemini Nano, meanwhile, permeates different parts of Android.
For instance, Gemini will now provide audio descriptions of on-screen images in the TalkBack accessibility software, which Google says will help address the problem of unlabeled images people with vision problems encounter online and in messages from friends and family.
Gemini Nano can also detect if you’re speaking to a potential scammer on the phone.
The company says it does this by picking up “conversation patterns commonly associated with scams.” Google says all of the processing for scam detection happens on your phone, and conversations aren’t uploaded to the web.
As for Google’s Workspace, the company says it is bringing its more powerful Gemini 1.5 Pro model to its side panel in Workspace apps.
The option allows you to continue working in an app like Docs or Gmail and use Gemini at the same time in a smaller side panel. Google says Gemini 1.5 Pro is capable of referencing larger data sets, so if you’re using it in Gmail it can summarize information from a greater number of emails.
Gemini for Workspace in the Gmail mobile app, meanwhile, can summarize email threads, making them easier to read on a smaller screen; provide contextual smart replies based on the contents of an email; and search for specific information in a long-buried email.
Google has just a few days before its biggest rivals launch their own AI products.
Microsoft is likely to announce updates to its AI-powered Copilot for Microsoft 365 productivity suite at its Build conference, which kicks off May 20, while Apple will debut a new generative AI-powered version of its Siri voice assistant at its WWDC event on June 10, according to the New York Times.
Now, we’ll wait to see if they’ll steal Google’s thunder. Again.
Email Daniel Howley at dhowley@yahoofinance.com. Follow him on Twitter at @DanielHowley.
Click here for the latest technology news that will impact the stock market.
Read the latest financial and business news from Yahoo Finance
