Early impressions of the new Siri AI (AI standing for Apple Intelligence) are quite positive after the release of the developer beta. Here are a few of our favorite stories on it:

• I tried Siri AI, and so far it actually works – Allison Johnson, The Verge
• Hands on with iOS 27’s new Siri AI – MacRumors
• Apple’s new Siri AI knows when to shut up – Jay Peters, The Verge

Stay on the lookout for our content once the Public Betas are out in July.

Also highly recommend checking out this interview with Apple’s SVP of Marketing, Greg Joswiak and SVP of Software, Craig Federighi. It gives a bit more background on how Apple is trying to differentiate Siri AI from other products:

Design refinements to Apple’s Liquid Glass UI, allowing users to choose how clear or frosted buttons and bars look. Custom EQ for AirPods. Smoother network transitions between using WiFi and Cellular, so your device won’t stay on a bad WiFi connection. Use Agentic AI to fix all your passwords with a single tap. Magic Cue-like features: bringing helpful context into messages, and on your screen during a phone call (like a flight confirmation number). You can see all the new features on Apple’s website.

Apple’s big swing during this year’s keynote was an all-new Siri, now more conversational with better speech dictation and Visual Intelligence built into the Camera app, though it’s hard not to notice that Google Gemini got there first, especially with its use of Personal Context data. Two genuinely clever new additions: you can now vibe-code your own Siri Shortcuts in plain language, and Safari can build custom extensions just from a description. Plus, Safari can notify you when a website is updated. Apple also delivered a much-needed overhaul of parental controls and under-the-hood enhancements, making it feel a bit like a Snow Leopard year. Everything was centered on performance upgrades: with apps launching up to 30% faster, AirDrop 80% faster, and iOS 27 support stretching all the way back to the iPhone 11.

As we discussed in our review of the Meta Ray-Ban Gen 2, a major downside of these glasses is the build-up of creep around this product due to user misuse. A new write-up from Wired details products marketed to Meta Ray-Ban users that block the recording light, allowing them to supposedly record people without their knowledge. Even though some of these products fail to bypass Meta’s recording guardrails on the glasses, the fact that they are being marketed to millions of people via TikTok will likely further erode trust in the products and the users who wear them.

For months, fans of the popular VR community have loathed the changes Meta made to the platform, chronicled in an excellent write-up by Victoria Song, The Verge. Now, Supernatural has a chance to become what it once was: an independent company that can just focus on catering to its community of users, not burdened by the changing whims of Mark Zuckerberg. The new app is expected to launch this fall, and you can read more about the change here.

As reported by The Verge, users can expect many more tech products to have user-replaceable batteries starting next year, when new EU rules take effect. These include headphones, e-readers, portable game consoles, laptops, and more, though smartphones and tablets are the big two that have exemptions because they’re already covered by another EU law, which doesn’t require user-replaceable batteries for devices that are waterproof, and their battery will last at least three years with minimal capacity loss. You can read more details in the excellent write-up Dominic Preston did for The Verge.

Now, anyone can make accessories for it. Hopefully, this means there will be a greater variety of accessories for Google’s new device by the time the holiday shopping period comes around. And yes, we’re working on a long-term review of the Air, due out in November.

We’ve completely overhauled search on our site, so you can now search for posts, products, quick posts and site pages, all in one place. Plus, we completely overhauled the search algorithm so it should be more relevant, and we’ve added new sort options as well. And when you click on a Quick Post in search, it will now open in its very own page! Read the full update notes in our Site Changelog.

Google is rolling out a new update to Google Home that will allow users to design automations based on visual insights. So, for example, when your doorbell camera notices a package has been delivered, you can have a message broadcast on all of your Google speakers: “A package has just been delivered.” You can also ask for a Home Brief of everything going on in your home that cameras have noticed, or locate a pet via cameras.

Boy, was there a lot of new stuff! From Gemini Spark, Antigravity 2.0, a ton of new stuff coming to Gemini on Mobile, and a teaser of Google’s new Gemini-powered glasses, here’s a recap of the biggest announcements:

From new Auto Forms using personal context data, to a new Rambler feature (kinda similar to Google’s AI Edge Eloquent app), a new Pause/Play feature that makes it harder to get into addicting apps, to a huge update to Android Auto and a BRAND NEW TYPE OF GOOGLE DEVICE (Googlebooks!), trust us, it’s worth just watching the full video:

As reported by Tom Warren from The Verge, Microsoft is finally addressing one of our main critques of the Windows experience, as outlined in past reviews.

You could already remix your photos into AI-slop-looking photos, use AI-infused search, and now you can try on clothes via Google Photos, which Google recently announced in a new blog post. It’ll take all the clothes you wear in your photos and create a virtual wardrobe for you, but I can’t shake the feeling we’re one step away from Google advertising clothes it thinks we’d like to wear.