Gemini needs to speak. On a regular basis.
Google I/O 2026 kicked off this week, and to nobody’s shock, it was nearly fully about AI.
To start out with, Google is rolling out its “clever, AI-powered Search field” globally. As a substitute of simply autocompleting your textual content, it makes use of AI to anticipate your intent and allow you to formulate questions. It is possible for you to to make use of photos, video information and whole Chrome tabs as direct search inputs. Its AI Mode, now powered by the brand new Gemini 3.5 Flash, will nonetheless reside alongside it, to your follow-up questions/corrections.
Google additionally launched Gemini Spark. Operating within the cloud, Spark is a digital assistant that may autonomously monitor bank card statements for hidden subscriptions, observe updates out of your child’s faculty emails, or pull notes collectively right into a Google Doc. It may well even work together with third-party apps like OpenTable and Instacart to finish duties—although it guarantees to ask to your affirmation earlier than making any last purchases or sending emails. Google could have discovered a solution to take the enjoyable out of purchasing.
There have been glimpses of future {hardware}: Google and Samsung teased a collaboration with Mild Monster and Warby Parker and provided the primary have a look at two fashions of Android XR sensible glasses. The glasses will help you chat with Gemini, get real-time audio translation within the speaker’s voice, translate real-world textual content in your line of sight and snap images on the go.
Naturally, with all these new options, Google adjusted its AI subscription tiers, including a brand new $ 100-per-month mid-range (hah!) tier. The AI Extremely Plan will provide 5 occasions increased utilization limits than the usual $20 Professional plan, in addition to precedence entry to Google’s Antigravity coding instrument and 20TB of cloud storage.
In the meantime, down from its authentic $250 price ticket, Google’s top-tier Extremely plan options 20 occasions increased utilization limits and unique entry to Undertaking Genie — Google’s experimental analysis preview that allows you to construct interactive 3D worlds utilizing real-world Google Road View imagery.
We dig additional into just a few extra of the largest bulletins beneath.
— Mat Smith
Xreal’s Undertaking Aura smartglasses are a maximalist tackle Android XR
As a substitute of delicate, normal-looking glasses with some sensible options, Xreal’s smartglasses goal for an immersive AR expertise, with a concentrate on leisure. That method could be very a lot the identical as the corporate’s Android XR-powered Undertaking Aura, however based mostly on Karissa Bell’s time with these new smartglasses, they could provide extra than simply one other wearable display screen.
In contrast with the low-key (arguably lower-spec) audio-only smartglasses from Warby Parker and Mild Monsters, Xreal’s Undertaking Aura packs in three cameras, a 70-degree area of view and hand-gesture navigation. The primary gesture is a pinching movement that may really feel fairly acquainted to anybody who has used different AR setups, akin to Apple’s Imaginative and prescient Professional.
Talking of which, Undertaking Aura comes with a tethered puck, similar to Apple’s wearable show. It comes with a trackpad and a fingerprint sensor, though neither labored on the demo mannequin we examined. The corporate hasn’t revealed pricing particulars but (anticipate these alongside the formal launch later this yr), however they’re prone to exceed Xreal’s One Professional glasses, which value $650.
Continue reading.
Google’s Gemini Omni can generate ‘something from any enter,’ beginning with video
Gemini Omni is Google’s new gen-AI mannequin that may “create something from any enter”. Gemini Omni Flash is rolling out now to the Gemini app, Google Movement and YouTube Shorts. It is apparently the following evolution of Nano Banana and, presumably, its present video generator, Veo 3.1. It helps you to “mix photos, audio, video and textual content as enter and generate high-quality movies grounded in Gemini’s real-world data.”
Apparently, Omni additionally higher understands bodily forces akin to gravity, kinetic power and fluid dynamics, making scenes extra sensible.
Spotify is including extra AI gunk for podcasts and audiobooks
Introduced on its investor day, Spotify is increasing a function that permits customers to generate “private podcasts.” It introduced this instrument final month, with the choice to make use of AI brokers akin to OpenClaw and Claude Code to whip up artificial audio. The corporate says it is including a function that lets customers generate private podcasts immediately inside Spotify. It notes that after you enter a immediate, it’ll generate audio that attracts on elements akin to your Spotify style profile and world data. You too can feed in textual content, PDFs and hyperlinks to offer the instrument extra context for what you want to listen to about. Spotify says eligible Premium customers within the US will get entry to private podcasts subsequent month. The Roman Empire, Ultimate Fantasy VII Remake and Liverpool FC, please. In beneath 30 seconds. Go.
Continue reading.
Meta’s newest try to repeat its rivals is a Reddit clone
Meta has launched a brand new app referred to as Discussion board. Social media guide and analyst Matt Navarra noticed Discussion board within the App Retailer, the place it’s described as “a devoted area for the conversations that matter most to you,” created particularly for Fb Teams. As Navarra notes, Meta appears to be pushing the concept the app may also help customers get “actual solutions” from “actual folks,” making it sound similar to Reddit, providing an escape from the atrocious genAI search outcomes we have all been affected by lately. (Did we point out Google I/O was this week?)
Continue reading.
