Perhaps Google ought to name its annual builders’ convention Google A/I as an alternative of Google I/O, as a result of the corporate is all about embracing generative AI this yr. Google debuted its newest language mannequin, Pathways Language Mannequin 2 or PaLM 2, which now fully-powers Bard, Google’s AI chatbot rival to Microsoft’s ChatGPT-powered Bing chatbot.
Hsiao additionally mentioned Bard will develop into “extra visible” within the subsequent few weeks, each in responses and prompts. This implies Bard search outcomes will begin returning photographs for related queries, and likewise that folks will be capable to immediate bard “with photographs” utilizing Google Lens integration. Hsiao demonstrated this with a photograph of two canines and the immediate “write a humorous caption about these two,” so you will nonetheless must immediate Bard with phrases along with the pictures (I am undecided what I used to be considering, however prompting Bard “with photographs” initially sounded rather more thrilling in my head).
Google may even be collaborating with third-parties so as to add options to Bard, reminiscent of text-prompted picture technology because of integration with Adobe Firefly (opens in new tab). The corporate is looking these “instruments” and confirmed a graphic with a number of corporations’ logos together with Kayak, OpenTable, Tripadvisor, Spotify, and Walmart. Lastly, Google has dropped the waitlist and Bard is now obtainable in 180 international locations and has added help for prompts in Japanese and Korean.
Duet AI for Google Workspace
Google additionally demonstrated its new AI-powered toolset for Google Workspace (Google’s suite of productiveness apps, which incorporates Gmail, Docs, Slides, Sheets, and Meet), which it is now calling “Duet AI for Google Workspace” (the toolset, not Workspace). It is referred to as “Duet AI” as a result of Google imagines that you’re going to be collaborating with AI, such as you collaborate with coworkers and different people in Workspace apps proper now.
Google really first introduced the upcoming Duet AI instruments in a weblog publish again in March, however most of them — save for the writing assistant software “Assist me write” — are nonetheless unavailable to the general public. The upcoming instruments will let customers do issues like generate summaries from notes, create photographs from textual content prompts inside Google Slides, create templates from textual content prompts in Google Sheets, and create customized backgrounds in Google Meet.
Google additionally demonstrated a future ‘Sidekick’ panel that may be capable to learn your content material and “have interaction” with you utilizing contextual prompts. Google Workspace VP and GM Aparna Pappu demonstrated this characteristic with an unfinished kids’s story. Asking the Sidekick for assist (don’t be concerned, it is not going to immediate you with out solicitation) generated some questions on what may occur subsequent.
A lot of the new Workspace instruments aren’t obtainable to the general public but, and will probably be rolling out selectively over the subsequent few months. Should you’re concerned about probably getting early entry to check out these options, you possibly can join the Google Labs waitlist right here.