Local Memo: A Roundup of the Big Announcements from Google I/O

Damian Rollison

Damian Rollison

Share

Local Memo: A Roundup of the Big Announcements from Google I/O

Local Memo: A Roundup of the Big Announcements from Google I/O

  In this week’s update, learn about all the important announcements from Google I/O; ChatGPT’s new update supporting web browsing and plugins; Meta’s multisensory AI model; Twitter’s new CEO; the launch of related topics in desktop search; and Google’s support for multiple languages in local listings.   

A Roundup of the Big Announcements from Google I/O

  Last week’s Google I/O conference was heavily focused on the theme of AI, with AI feature announcements for products such as Google Photos and Workspace as well as new offerings like Project Tailwind, an AI-powered notebook tool that helps you conduct research on uploaded documents as an aid to writing and studying, and Codey, a generative AI model designed to help users write code.    Among the most heavily anticipated announcements was that of an AI-powered search interface. At the conference, Google unveiled what it calls the Search Generative Experience, or SGE. This new search interface, according to Google, “helps you quickly find and make sense of information. As you search, you can get the gist of a topic with AI-powered overviews, pointers to explore more, and ways to naturally follow up.”   The SGE interface will add a chatbot component to the top of the search results screen called the AI Snapshot, which will offer a broad array of options that combine the power of generative AI with Google’s ability to deliver relevant web content. The result is something like featured snippets on steroids, and some are already predicting that content publishers will have their very existence threatened now that Google can deliver answers to just about any query directly in search. That said, in demonstrating the new search experience Google deliberately showcased attribution to third-party sources such as TikTok (see the video below). Users can sign up to get early access to SGE via Google Labs.    The company also announced several updates to the Bard chatbot, which will begin to feature Google Knowledge Graph results, Google Maps content, and support for Google Lens, including the ability to add captions to uploaded photos. Google has removed the Bard waitlist, opening up the tool to users in 180 countries with plans to support 40 languages. Bard will also soon be able to export content to Workspace apps like Docs and Sheets, and will have access to third-party plugins like Zillow, Uber Eats, and Kayak.    Bard is now utilizing Google’s PaLM2 language model, also announced at I/O. PaLM2 improves upon the former PaLM model, first launched in 2022, with improved multilingual, logic, and coding abilities. PaLM2 is now powering features in 25 Google products, including Bard, Workspace, and Cloud.   Among many other new products and features I don’t have the space to cover here, Google also announced the launch of Perspectives, a new component of the SERP that helps users “find and explore diverse perspectives” in addition to standard search results. Perspectives will feature content from “long- and short-form videos, images and written posts that people have shared on discussion boards, Q&A sites and social media platforms,” accessible through a new Perspectives search filter.   Finally, Google announced that this summer, it will add a new “Immersive View for Routes” to Google Maps. This will allow users to preview a route in 3D before starting a trip, providing predictions for drivers, cyclists, and walkers about traffic, bike lanes, weather, and other important details.   

   

ChatGPT Rolls Out Plugins for Plus Users

  OpenAI has announced the latest version of ChatGPT, which will include beta support for web browsing as well as the integration of over 70 third-party plugins, a far greater number than previous announcements indicated. ChatGPT Plus users will have the option to enable these features through a new “Beta features” option in their user settings. ChatGPT is supposedly able to evoke the correct plugins as needed, for instance if you ask for information about any recent news not included in its 2021-and-earlier dataset. As for third-party plugins, users have the option of enabling up to three plugins in a given chat session.   

An image showing which third-party plugins you can use with ChatGPT

 

The ChatGPT Plugin Store, courtesy Search Engine Journal / OpenAI

 

Meta Open Sources New Multisensory AI Model

  Meta has released to the open source community a new AI model called ImageBind that appears to take multimodality to a new level, integrating inputs from text, audio, images, and video as well as movement and temperature. There are no practical applications for the model as yet, but it points to a future type of virtual reality experience that would be much more immersive than current AI models. Meta’s announcement suggests that other modalities, such as “touch, speech, smell, and brain fRMI signals,” could be added to the model in future. The decision to open source ImageBind is a controversial one, with rivals like OpenAI maintaining that open sourcing harms innovation and gives an advantage to malicious actors.   

An image sharing examples of multi-sensory AI including audio, images, depth, and text

   

Examples of multisensory AI, courtesy Meta

 

Twitter Will Have a New CEO

  Elon Musk has named Linda Yaccarino as Twitter’s new CEO. Yaccarino was previously the chairman of global advertising and partnerships at NBCUniversal Media, where she has worked for more than a decade. Twitter also has a new corporate name: X Corp, echoing Musk’s vision to turn the platform into “X, the everything app.” In Musk’s tweet announcing the news, he states that Yaccarino will “focus primarily on business operations” while he remains in charge of “product design & new technology.” After the announcement, Yaccarino’s follower count on Twitter quickly jumped from 7,187 to a count (as of this writing) of 387,900.  

Google Brings Related Topics to Desktop Search

  Last December, Google added related topic filters to mobile search. Now the same feature is launching on desktop. The filters allow users to drill down on topics specific to the intent of a query, such as “Vegetarian,” “For two,” and “Family” in response to a query for dinner ideas. Though the feature is now rolling out to all desktop users, it has been in a testing phase for several months now, so many users may already have seen the new filters. Filters are shown in the order Google deems to be most useful for a given query, but users can select “All Filters” to access others that may not be shown prominently.   

An image showcasing Google's related topics for the search "dinner ideas"

 

Search filters on desktop, courtesy Search Engine Journal

 

Google Will Display Business Addresses in Multiple Languages

  Google is able to display and accept suggestions for business names and addresses in multiple languages, according to findings from Stefan Somborac. For instance, in Quebec, Google will display addresses in both French and English. When a user visits the “Suggest an edit” interface for a given listing, he or she is given the option to update address information in one or more languages, depending on the language settings that user has configured. Darren Shaw confirmed that if the default language settings are changed in search, Google will offer different language options in “Suggest an edit” for both business name and address. 

Subscribe to Local Memo!

 

Signup to receive Local memo updates and the latest on localized marketing, delivered weekly to your inbox.