OpenAI filed a countersuit against Elon Musk on Wednesday, saying on X that “Elon’s nonstop actions against us are just bad-faith tactics to slow down OpenAI and seize control of the leading AI innovations for his personal benefit.”
In the lawsuit, OpenAI’s lawyers argue that “Musk’s continued attacks on OpenAI, culminating most recently in the fake takeover bid designed to disrupt OpenAI’s future, must cease. Musk should be enjoined from further unlawful and unfair action, and held responsible for the damage he has already caused.”
Musk, who was part of the initial founding team at OpenAI, initially sued last spring, saying he wanted to force the company to “return to its mission to develop AGI for the benefit of humanity” instead of pursuing profits. (The Verge’s editor-in-chief, Nilay Patel, found Musk’s legal case against OpenAI “hilariously bad.”)
Earlier this year, Musk also offered $97.4 billion to buy OpenAI, saying in a statement that “it’s time for OpenAI to return to the open-source, safety-focused force for good it once was.” OpenAI’s board of directors unanimously rejected the offer, which today’s filing called a “sham bid.”
Disclosure: The Verge’s parent company, Vox Media, partners with OpenAI.
Meta’s latest whistleblower, Sarah Wynn-Williams, got a warm reception on Capitol Hill Wednesday, as the Careless People author who the company has fought to silence described the company’s chief executive as someone willing to shapeshift into whatever gets him closest to power.
The message was one that lawmakers on the Senate Judiciary subcommittee on crime and counterterrorism were very open to. Their responses underscore that amid CEO Mark Zuckerberg’s latest pivot in cozying up to the right, his perception in Washington has not yet totally changed, even as he reportedly lobbies President Donald Trump to drop the government’s antitrust case against the company.
“He’s recently tried a reinvention in which he is now a great advocate of free speech, after being an advocate of censorship in China and in this country for years,” subcommittee Chair Josh Hawley (R-MO) said, pointing to longtime conservative allegations that Meta has suppressed things like vaccine skepticism and the Hunter Biden laptop story. “Now that’s all wiped away. Now he’s on Joe Rogan and says that he is Mr. Free Speech, he is Mr. MAGA, he’s a whole new man, and his company, they’re a whole new company. Do you buy this latest reinvention of Mark Zuckerberg?”
“If he is such a fan of freedom of speech, why is he trying to silence me?” Wynn-Williams asked in response. Meta convinced an arbitrator to order her to stop making disparaging statements and halt further publishing and promotion of the book, which details Meta’s alleged dealings with the Chinese government and claims of sexual harassment from a top executive. Meta spokesperson Andy Stone has called Careless People “defamatory,” but the book’s publisher said it would “continue to support and promote it.”
“We don’t know what the next costume’s going to be, but it will be something different”
Wynn-Williams also told Hawley that Zuckerberg “is a man who wears many different costumes. When I was there, he wanted the president of China to name his first child, he was learning Mandarin, he was censoring to his heart’s content. Now his new costume is MMA fighting or free speech. We don’t know what the next costume’s going to be, but it will be something different. It’s whatever gets him closest to power.”
At the hearing, Wynn-Williams testified that during her time at the company between 2011 and 2017, Meta and Zuckerberg were willing to “undermine American national security” in service of currying favor with the Chinese government. She accused Meta of working on “censorship tools” that the Chinese government could use to silence critics and provided the Chinese Communist Party American user data.
In a statement, Meta spokesperson Ryan Daniels called Wynn-Williams’ testimony “divorced from reality and riddled with false claims. While Mark Zuckerberg himself was public about our interest in offering our services in China and details were widely reported beginning over a decade ago, the fact is this: we do not operate our services in China today.”
Sen. Amy Klobuchar (D-MN) said she found it “ironic” that China was a focus of the hearing, given that when she tried to pass a tech antitrust bill, “one of the things that kept being thrown in my face and in those of others that work on this, is that ‘you’re actually going to destroy us and then China will dominate,’” she said. “Your book actually reveals the extent to which Facebook was willing to put growth over the US national interest to gain favor with the Chinese Communist Party.”
Lawmakers dared Zuckerberg to testify before their committee himself to clear up their issues with her statements. “Stop trying to silence her, stop trying to gag her, stop trying to hide behind your lawyers and millions of dollars in legal fees you’re trying to impose on her,” said Hawley. “Come to this committee, take the oath, sit there, let us question you, and give the American people the truth. We will be waiting for you.”
Wynn-Williams told the subcommittee her testimony “may be the last time I’m allowed to speak” given the legal restrictions. “It’s not going to be the last time you’re allowed to speak if we have anything to do with it,” Sen. Richard Blumenthal (D-CT) said. “What I would say to Mark Zuckerberg is, stop gagging Ms. Wynn-Williams, let her speak the truth, and you come here and tell us your version of the truth, if you have the guts to do it.”
Adobe is building AI agents for Photoshop and Premiere Pro that can suggest ways to edit your photos or videos and then carry out the tasks for you, according to a blog written today by Ely Greenfield, Adobe’s CTO of digital media.
Adobe Photoshop’s agentic AI, or what the company calls its “creative agent,” will be presented in a new floating Actions panel that will recommend context-aware edits after analyzing your photo. For instance, it will be able to suggest removing people standing in the background or creating a greater depth of field by blurring everything behind the subject. All you need to do is click the suggestion and it will be carried out automatically.
Long-time Photoshop users are used to manually manipulating photos by tediously masking people and objects and then creating layers so changes can be made to only certain parts of the image. Adobe has already added AI features that let you extend and fill photos across a larger canvas, or delete unwanted objects or people from the background using Distraction Removal.
Adobe’s vision is that Photoshop users will be able to prompt agents with natural language, making it easier to learn the steps needed to perform a task (although the agent will still be able to do it for you). And you can continue prompting the agent to make more changes, or manually make adjustments in the layers. In one example video, someone asks the agent to clean up an image and add a text box behind a person, and the agent then lists out steps including: remove background people, auto brighten, remove distracting objects, create “subject” layer, create text layer, and organize layers.
For Premiere Pro, Adobe will build on the new Media Intelligence feature introduced last week, which analyzes videos for objects and composition so you can find the footage you need. A future agent will let you direct the agent to make a rough video cut.
“While AI can’t replace human creative inspiration, with your input it can make some educated guesses to help you get your project off the ground,” Greenfield wrote in the blog. “It can also help you learn how to perform complex tasks with a few simple keystrokes, helping you grow as an editor.” Premiere Pro’s creative agent will eventually help editors refine shot choices, adjust color, mix audio, and more. Adobe also just launched Generative Extend, which uses AI to add seconds to your clips to help fit a transition.
Adobe will introduce the technology behind the first AI agent, which will be for Photoshop, at its Max event in London on April 24th.
Aqara’s Matter controllers, including the Camera Hub G5 Pro, Doorbell Camera Hub G410, Hub M3, and Hub M100 pictured, can now support over 50 Matter device types.
The company is also bringing advanced bridging to all its Matter controllers, allowing you to integrate Aqara’s many advanced automation features into any Matter-compatible smart home platform via scenes and triggers. This includes Apple Home, Amazon Alexa, and Google Home. The new update also lets you bridge Matter devices that are not yet supported by a platform into those ecosystems.
The new Matter device types that can now connect to Aqara Matter controllers include:
Robotic vacuums
Dimmable plugs
Smoke and CO alarms
Air quality sensors
Leak, freeze, and rain sensors
Curtains and blinds
Fans and air purifiers
Air conditioners and heat pumps
Water valves and pumps
Pressure sensors
Solar panels and battery storage
Electric vehicle chargers
Washing machines and dryers
Fridges and freezers
Cooktops, hoods, and kitchen surfaces
Ovens and stoves
Water heaters
Video players and speakers
This is in addition to sensors, thermostats, switches, and lights, which were already supported. Aqara says the update will roll out with Aqara Home version 5.1.9 and Matter Controller firmware version 4.3.5 later this month.
Aqara seems committed to making its platform open and interoperable, stating that it believes interoperability is essential to move the smart home industry forward. “By extending Matter support, we aim to offer users greater flexibility and ensure that Aqara products remain useful across a wide range of platforms,” said Eugene You, CEO of Aqara.
Aqara has long worked with all the major smart home platforms, including Apple Home, Amazon Alexa, Google Home, and more recently SmartThings, and it’s now positioning itself as a central part of the Matter smart home.
Aqara hubs work with its line of Zigbee-based smart home products; several are also Thread border routers. The company says they have been optimized to support Thread network sharing, meaning they can join existing Thread networks managed by Apple HomePods, Amazon Echos, or Google Nest hubs, for example. However, compatibility will depend on those platforms’ implementation of Thread. This should help provide a stronger network for Thread smart home devices.
“By extending Matter support, we aim to offer users greater flexibility and ensure that Aqara products remain useful across a wide range of platforms.”
Advanced bridging, which was introduced with the M3 hub earlier this year, will come to all current and upcoming Aqara Matter controllers, even the entry-level Hub M100. This brings two benefits. First, you can bridge Matter devices into platforms that don’t yet support that device type, and second, you can use Aqara’s advanced automation features and scenes in any Matter-compatible platform.
The latter ability means that you can use Aqara’s G4 Video Doorbell’s facial recognition feature as a trigger in an Apple Home automation that turns your lights on when you come home. The facial recognition trigger is added to Apple Home as a virtual “occupancy sensor,” according to Aqara.
Other advanced features that can be translated as either “virtual occupancy sensors” or “virtual plugs” include package detection from a camera’s on-device AI, fall detection from Aqara FP2 mm-wave sensor, and automations triggered by a device’s energy usage. Fingerprint sensing, gesture control, and presence sensing can all also translate, according to Aqara.
This advanced bridging feature is not native to the Matter spec, but is something a hub manufacturer needs to develop. Home Assistant has similar capabilities that allow you to import its scenes, scripts, functions, and other non-Matter-supported entities to connected platforms.
Aqara says advanced bridging is essentially a stopgap solution that allows users to access its full features in any Matter platform until (or if) Matter gains support for these more advanced features.
But it appears that, tariffs or no, the new Framework Laptop 12 won’t exactly be cheap.
Framework has just opened Laptop 12 pre-orders in the UK, Europe, and Canada starting at €569/£499/$719 CAD. Hours after we published this story, Framework announced it will open US preorders starting at $549 tomorrow, April 10th.
That’s hundreds less than Framework’s other DIY laptops, but a good bit higher than today’s entry-level Windows laptops and Chromebooks even before you consider that price doesn’t include storage, memory, ports, or an operating system. If you want those preinstalled, the machines start at €849, £749, and $1,079 CAD respectively, which is getting closer to what Framework charges for its Laptop 13.
But with the Framework Laptop 12, you’re also paying for a 2-in-1 backflipping touchscreen rather than the latest chips. At £499, it starts with a six-core 13th Gen Intel Core i3-1315U from 2023, with a 10-core i5-1334U available for £150 more. The £749 prebuilt model starts with the same i3 chip, 8GB of RAM and 512GB of storage, and you’ll pay £200 more to upgrade to the i5 and 16GB of memory.
While I’m mostly versed in US pricing, my UK colleague Jess Weatherbed assures me that similarly spec’d laptops can be had for anywhere between £400 and £600, and I can see more powerful thin-and-light machines available at the prices that Framework’s asking today. Again, repairability and modularity aren’t free.
(We’ll add US comparisons tomorrow once Framework reveals full US pricing; as of evening April 9th, it’s only revealed the US starting price.)
Framework isn’t yet selling the color-matched stylus with removable battery it showed off as a companion to this Laptop 12; that will go on sale “soon” and you’ll have to buy it separately. But you can also use any USI 2.0 or MPP 2.0-compatible stylus, the company says, with support for both Windows and Linux.
Speaking of colors, it appears that the prebuilt Laptop 12 will only be available in black for now. You’ll have to choose the mostly preassembled “DIY” version to get cute colors like the mint and pink model.
The first batches of DIY and prebuilt models should ship in June and July, respectively.
Update, April 9th: Added that Framework will put the US version on preorder tomorrow starting at $549.
The day before several major tech leaders appeared before Congress, begging for ways to get more energy for the nascent American artificial intelligence industry, Donald Trump signed an executive order offering a solution: increased coal production.
As part of a series of executive orders released Tuesday designed to promote the rapid growth of the coal industry â opening federal lands for mining, designating coal as a critical mineral, and using his emergency authorization powers to relax environmental regulations on coal â Trump signed one explicitly aimed towards powering energy-hungry AI data centers using Americaâs âbeautiful clean coal resourcesâ, as Trump described it. The order directs the Commerce, Energy, and Interior Departments to conduct studies determining âwhere coal-powered infrastructure is available and suitable for supporting AI data centers,â as well as whether it would be economically feasible.Â
âYou know, we need to do the AI, all of this new technology thatâs coming on line,â Trump said on Tuesday during a signing ceremony for all four executive orders. âWe need more than double the energy, the electricity, that we currently have.â
Thanks to Framework, one of the most transparent computer companies out there, we now have a sterling example of just how ridiculous President Donald Trump’s tariffs can make life in the year 2025.
Today, the company made all of its computers more expensive. Then, it almost immediately reversed that to bring its computers back to their original prices,in real time, all because Trump posted to social media that he was instituting a tariff pause less than a day after his new tariffs began. Then, Framework clarified that some price increases will still apply on China-made components.
Here’s the timeline of events:
At 12:01AM ET Wednesday, Trump’s new tariffs went into effect, including an incredible 104 percent tariff on China and a 32 percent tariff on Taiwan.
At 1:18PM ET, just over 12 hours after the new tariffs were officially in place, Trump announced a 90-day pause on tariffs for most countries but an increased 125 percent rate for China.
At 1:48PM ET, Framework announced that nope, actually, it’s bringing prices back to normal.
At 3:43PM ET, Framework announced that some prices will still reflect increased tariffs on China.
Things are still a bit convoluted, but here’s the general lay of the land: since Trump is still imposing 10 percent tariffs on Taiwan, Framework will continue a pause on selling some of its cheapest laptop configurations “where we’re unable to absorb the remaining 10% tariff.” Additionally, some of Framework’s made-in-China modules now cost more, like $15 for a USB-C port rather than $9, or $49 for an Ethernet card rather than $39 — and Framework says systems that ship with those modules will see those same price increases.
A 10 percent price hike wasn’t the only change Framework made in response to Trump’s tariffs. This very morning, Framework delayed opening preorders for its new entry-level Framework Laptop 12 in the United States even as it put it on sale in other territories. (Here’s my new story about that machine.)
You can read more about Framework’s thought process on the tariffs in its likely-soon-to-be-revised blog post published earlier today, where it initially promised to absorb part of the increased cost. While it builds systems in Taiwan, Framework says that many of its modular components are still manufactured in China where Trump’s tariffs are strongest, and it’s possible that prices for those will continue to go up.
Here is that whole original blog post for posterity, as written by Framework CEO Nirav Patel:
As a result of the new tariffs that were announced last week, we have a series of unfortunate price and availability adjustments we need to make for US customers. We manufacture most of our products in Taiwan, for which we now face a new 32% import tariff into the US. We’re absorbing part of this cost temporarily, and we are increasing prices on in-stock laptops and new system pre-orders by approximately 10% for US customers. For our lowest-priced configurations, where we can’t afford to absorb the tariffs, we’re currently pausing sales to the US. We’re also delaying the pre-order launch of Framework Laptop 12 in the US. We’re continuing to monitor changes to tariffs, and we will make additional adjustments if needed. For non-US customers, there is currently no tariff impact, and we’re keeping the same pricing and availability of our products.
First, I want to acknowledge that this sucks, for you, for us, and for our mission to remake Consumer Electronics. We will get through it. This isn’t the first challenge we’ve faced, and it won’t be the last one. We’ll navigate through and keep focused on delivering great products and fulfilling this mission. We’re going to remain open and transparent throughout and try to bring clarity to a messy situation wherever we can.
Next, let’s go into more detail on the specific changes we’re making. Our Framework Laptops, Mainboards, Framework Desktop, and a subset of our modules are made in Taiwan, which means they are now impacted by a 32% import tariff to the US. Those are the products we’re increasing pricing on by 10% in the US, and we may need to increase this further if tariffs persist. Many of our modules are currently manufactured in China, where we face between 104% and 129% (!!!) tariffs. On those modules, we’re also absorbing part of the tariff and increasing pricing, depending on the category. Our Western Digital storage is manufactured in Malaysia, which now faces a 24% tariff. We’ve also increased storage pricing by up to 10%, but recommend that you purchase it elsewhere for your DIY Edition if possible, like directly from the Western Digital website.
These changes are going into effect now in the US for both in-stock orders and for new pre-orders. We’re enacting tariff absorption along with price increases as a temporary measure while we track the evolving situation around tariffs. In the event tariffs are removed, we’ll reset pre-orders that haven’t yet been fulfilled back to their previous prices. If tariffs persist or increase, we’ll likely need to increase US prices further. For existing US pre-orders of Framework Laptop 13 (AMD Ryzen AI 300 Series) and Framework Desktop, we’re still determining how to handle the tariff impact. In the event we need to adjust pricing, we will ask for your confirmation on the new price before finalizing your pre-order. We’ve also temporarily removed the functionality to edit configurations of existing US pre-orders, but you’ll still be able to edit your order at pre-order finalization. As always, pre-orders remain fully refundable. We’ll keep you updated on US pre-order timing for Framework Laptop 12. We’re reserving some manufacturing capacity for US orders to be able to ship alongside orders from other countries.
For US orders of parts and modules that ship from our New Jersey warehouse, we’ve temporarily paused ordering while we implement changes that let us decouple pricing between laptop configuration items and items in the Framework Marketplace. When we open ordering again, we’ll continue to sell items that are already in inventory in the US at the original price, while updating pricing for each item when we import new inventory. Canadian orders that ship from our US warehouse will also remain at the original CAD price at the moment, but we may need to make future price adjustments on items that are made in China, on which we face some tariff impact.
Now, let’s get into more detail about how the tariffs work. We’re keeping this apolitical and sharing more about how this works operationally. When goods are imported into the US, tariffs are assessed based on the country of origin and the HTS (Harmonized Tariff Schedule) code that the goods are classified as. We ship all of our products DDP (Delivered Duty Paid), meaning we pay tariffs as goods clear customs and include the cost inside of our product pricing. The tariff is calculated against the value of the product at import, meaning our cost as the importer, rather than the final price we charge for the product. The country of origin is defined as the last location in which “substantial transformation” occurs. For computers, US Customs has specifically defined substantial transformation as the location at which the main circuit board is assembled.
When starting Framework five years ago, we anticipated increased trade challenges between the US and China, and we chose to build most of our manufacturing and logistics footprint in Taiwan. At the time, imports from China to the US were subject to 7.5-25% Section 301 tariffs, with exceptions that included laptops, but not laptop parts. Because we’ve assembled our Mainboards in Taiwan since 2022 (we also do final system assembly and make magnesium parts and some aluminum parts there), the country of origin for our laptops is Taiwan. While this has largely shielded us from earlier rounds of tariffs on imports from China, all countries with a meaningful electronics manufacturing infrastructure are impacted by the current round. We are actively investigating paths to perform Mainboard assembly in the US, but our current manufacturing partners do not have necessary infrastructure in place. We were also already in the process of moving some module production from China to lower-tariff regions like Thailand and Indonesia. Migrating manufacturing partners or setting up new manufacturing infrastructure is a theoretical long term solution, but is not something we can execute ahead of tariffs coming into place this week.
This brings us back to the price and availability adjustments. Our products are built around longevity, and our business is too. We’ve built excellent Supply Chain and Logistics teams to be ready for this kind of disruption, and they are exploring every possible option to get back to normal US fulfillment. We’ll share changes and solutions as we come up with them, and we’ll remain transparent throughout. Thanks for continuing to follow along the journey.
We’ll let you know if Framework changes its plans further.
Update, 4:31PM ET: Added that Framework will continue to charge more for made-in-China modules, though it will no longer do a global 10 percent price hike on all US systems.
Were you clamoring for two AI-generated podcast hosts to talk through the key points of your document before you send it to your boss? No? Well, you’re getting it anyway. Google has announced new Gemini features coming to Workspace apps, including NotebookLM-style AI podcasts integrated right into Google Docs. There’s a wide spread of new features in this announcement, from an editor that can make suggestions in your documents to a tool that will help make sense of your spreadsheets, all with Gemini at the center.
Maybe the most uncanny of them all is the podcast feature, which exists in a couple of places already. It grabbed a lot of attention when Google offered it as part of its NotebookLM research tool, and it’s also available in the Gemini app when you upload a file. This update, which is coming to Workspace accounts “in the coming weeks,” will bring audio overviews right into Google Docs, along with the option to have AI read your article out loud. Having your writing read back to you is genuinely useful, but I used Gemini to generate one of these “podcasts” based on an article I wrote and I’m not in a hurry to do it again.
Another feature coming to Docs is a prompt to “Help me refine.” Rather than just doing the writing for you, it will leave comments with suggestions about how you can tighten up an existing draft. I’m familiar with this concept as an editor, and they’re hella useful. If you don’t have access to, you know, a person editor, an AI version might not be a terrible idea. This one will be available “later this quarter.”
Further out, Google has some lofty but vague promises for Sheets. Right now, you can ask Gemini for help with specific questions and tasks in Sheets, like generating visuals. A feature called “Help me analyze” offers a broader scope to help you make sense of your spreadsheet. In theory, it will help you identify trends and “guidance to get you started,” kind of like an on-demand data analyst. It sounds like we’ll have to wait a bit to try this one out — it’s coming “later this year.”
President Donald Trumpâs latest round of tariffs â including a now 125 percent levy on Chinese imports â will hit the smart home industry hard.
Many smart home device makers are already struggling, thanks, in part, to Trumpâs first round of tariffs. Increased competition from budget smart home companies largely based in China has also played a part, and so has slower takeup from homeowners than the industry expected.
The sky-high tariffs on Chinese goods and components present a huge challenge for all smart home manufacturers. Most companies source products from China, and even if they moved manufacturing to other countries, they still may use Chinese-made components. The new, steeper tariffs on countries like Vietnam, where many companies relocated manufacturing, compound their problems.
TP-Link’s Tapo D225 Video Doorbell Camera comes with perks like free local recording.
One of our favoritedoorbells, TP-Link’s Tapo D225 Video Doorbell Camera, is on sale for just $79.99 ($20 off) at Amazon. That’s a buck shy of its best price.
A lot of video doorbells come with extra hidden subscription fees if you want to make the most of them, but this one doesn’t. It lets you record video to a microSD and offers smart alerts for people, packages, pets, and vehicles.
You can use it hardwired if you want 24/7 video monitoring or, if you prefer, with a battery pack. Running off batteries would allow you to place it where you might not already have existing doorbell wiring, but removes the constant video recording option.
The Tapo D225 can also call your phone when somebody presses the doorbell, even if you don’t have the app installed, which should make it harder to miss a visitor.
Aside from offering perks rivals typically charge extra for, the 2K-capable Tapo D225 has a clear camera with a wide 180-degree field of view that can show your entire porch. It’s a bit chunky, and you can’t really use your existing chime, but it’s a great buy at this price.
President Donald Trump’s tariffs have only been in effect since midnight ET, and hours later he’s already announced “a 90 day PAUSE, and a substantially lowered Reciprocal Tariff” of 10 percent for countries that haven’t retaliated against the tariffs he created.
However, Trump also said in his post on Truth Social that he will be increasing the rate for China — which has placed tariffs on US goods in the past days — to 125 percent.
After Trump’s post, the New York Times reports White House press secretary Karoline Leavitt told reporters the tariff level would be brought down to a universal 10 percent, a significant drop for many countries that produce large amounts of US-bought goods, like Vietnam.
Secretary of Commerce Howard Lutnick tweeted that “(Treasury Secretary)Scott Bessent and I sat with the President while he wrote one of the most extraordinary Truth posts of his Presidency. The world is ready to work with President Trump to fix global trade, and China has chosen the opposite direction.” Meanwhile, Bessent spoke to reporters at the White House, claiming the now-delayed tariffs were about getting “negotiating leverage” for Trump instead of capitulation, with rewards (in the form of still-higher tariffs) for countries that didn’t retaliate.
Based on the lack of respect that China has shown to the World’s Markets, I am hereby raising the Tariff charged to China by the United States of America to 125%, effective immediately. At some point, hopefully in the near future, China will realize that the days of ripping off the U.S.A., and other Countries, is no longer sustainable or acceptable. Conversely, and based on the fact that more than 75 Countries have called Representatives of the United States, including the Departments of Commerce, Treasury, and the USTR, to negotiate a solution to the subjects being discussed relative to Trade, Trade Barriers, Tariffs, Currency Manipulation, and Non Monetary Tariffs, and that these Countries have not, at my strong suggestion, retaliated in any way, shape, or form against the United States, I have authorized a 90 day PAUSE, and a substantially lowered Reciprocal Tariff during this period, of 10%, also effective immediately. Thank you for your attention to this matter!
OS 2.0 is now available for the Teenage Engineering EP-133 KO II synthesizer. | Image: Teenage Engineering
Nearly a year and a half after it first launched the EP-133 KO II synthesizer in late 2023, Teenage Engineering has released its first major software update. The OS 2.0 update, which can be downloaded to the synth through a web browser or installed using a USB cable, includes features like creating new samples from existing samples or beats, hands-free sampling while playing another instrument, and the ability to layer and play more sounds simultaneously.
You can see all of the new features in the OS 2.0 release notes, but the most significant update appears to be new resampling capabilities that let you process or add effects to existing samples to create brand new ones. The update also allows you to create new samples by capturing snippets of a beat made up of previously recorded samples. The KO II is also getting a new hands-free sampling mode letting you record the sounds from an instrument that requires two hands to play – like a piano – without requiring a third hand to press record or stop on the synth.
Teenage Engineering has also increased the number of sounds that can be layered and played simultaneously on the KO II from 12 mono and six stereo to 16 mono and 12 stereo with OS 2.0 installed. Other upgrades include a new song mode that “adds the ability to chain scenes and create longer, more structured track arrangements,” and sidechaining that “allows one sound to control the volume of another.” So the volume of a bass drum in a beat can be automatically lowered whenever the sound of a kick drum plays so it’s not drowned out.
The $299 KO II falls somewhere in between the company’s $59 Pocket Operator synthesizers and its expensive but highly-capable $1,999 OP-1 Field. It’s positioned as a sort of advanced musical toy, but musicians might now find it a more capable song-making tool with OS 2.0.
Whether you’re buying your first pair of AirPods or replacing a well-used older pair, it’s easy to understand why so many iPhone owners prefer Apple’s earbuds and headphones over other options. Apple has spent years gradually adding new features that work best — and sometimes only work at all — within the company’s walled garden. These convenient tricks include audio sharing, automatic switching between Apple’s various devices, Apple TV integration, and more.
There are four AirPods models to choose from, each with a distinct appeal and upside. You likely already know whether you’re more interested in earbuds or over-ear headphones, so that’s the first divide. In the case of headphones, your only choice is the AirPods Max. But in the earbuds category, Apple has three options: the AirPods Pro, AirPods 4 with active noise cancellation, and the plain old AirPods 4.
The best AirPods overall
If you want the earbuds that guarantee you’ll get the latest and greatest software features from Apple, the second-generation AirPods Pro should be at the top of your list. They offer the whole gamut of ecosystem exclusives. And if you’re the type of person to misplace things, you’ll be thankful that the earbuds’ case has a U1 chip for pinpoint location finding. So whenever they go missing, you can track right where they are in the room using Apple’s Find My app.
The AirPods Pro deliver noise cancellation that’s effective at tamping down the volume of your surroundings in all but the most hectic environments. Their sound quality is on another level compared to the regular AirPods — helped in no small part by the in-ear seal you get from their silicone ear tips.
In 2024, Apple introduced three major hearing health features, including a hearing test, automatic hearing protection, and the option to use the AirPods Pro as clinical-grade hearing aids if you have mild or moderate hearing loss. No other AirPods (including the AirPods Max) offer this functionality, so it definitely sets the Pros apart. And the same can be said of their volume controls; you just swipe on the stems to raise or lower the loudness of your music, which is a handy advantage that the similar-looking AirPods 4 lack.
But not everyone likes the way ear tips feel. Even with venting and the other measures Apple has taken to improve comfort and reduce that clogged-up sensation, some people just prefer the regular, one-size-fits-most AirPods. Their open design means you’re never cut off from the outside world. The AirPods Pro have an impressive transparency mode, but the AirPods 4 don’t need to pipe in your surroundings; you can always hear what’s happening around you.
Apple sells two versions of the AirPods 4. No matter which you choose, you’ll get the same sound quality, microphone performance on voice calls, and IP54 water resistance. (The latter also applies to the AirPods Pro.) The more expensive pair features active noise cancellation and a more advanced charging case. You might be surprised by how well the noise cancellation works even with an unsealed design; it reduces office ambience and annoying hums, but louder sounds and voices still cut through the ANC. As for the case, the pricier AirPods 4 include wireless charging and Find My integration.
If you’re really interested in noise-canceling earbuds, you should consider saving a bit more and stepping up to the AirPods Pro 2. They offer much better audio quality, stronger ANC, and a better experience. I’d only steer you to the noise-canceling AirPods 4 if you’re really opposed to ear tips.
I’ve gone back and forth recommending Apple’s $549 AirPods Max over the last few years. They have a standout design, but the good looks and premium feel come at the cost of weight. These headphones are considerably heavier than most competitors and can get fatiguing over extended use. But Apple has also made smart choices: the fabric ear cups on the AirPods Max are much more pleasant in warm weather than the pleather you’ll find on Bose and Sony products, and won’t get your ears sweaty.
The AirPods Max provide impressive noise cancellation, and the transparency mode is clearer and more natural-sounding than any other set of headphones on the market. As for sound, Apple’s headphones have a pleasant sound signature that works well across a wide range of genres. They’re easy to listen to with an impressive soundstage for closed-back headphones.
You lose some features — like conversation awareness and adaptive audio — that are included on the AirPods Pro; this is a side effect of the AirPods Max using Apple’s older H1 chip. But the company recently added lossless audio over USB-C, meaning these headphones offer the richest audio fidelity of the whole lineup.
We spend dozens of hours using AirPods products for our initial reviews, and that testing extends for weeks and months afterward. That’s where we assess sound quality, mic performance, and overall reliability. Where applicable, we try the active noise cancellation in a wide range of different environments and have put the hearing protection of the AirPods Pro through the paces at concerts. We also closely examine ecosystem benefits and how the AirPods and Apple’s other products work together.
WordPress.com has launched a new AI-powered site builder in early access that can construct your WordPress webpage, including fully written text, layout, generated images, and more. WordPress.com says the AI website builder can make “beautiful, functional websites in minutes.” However, it can’t yet do the heavy lifting of creating an ecommerce site or one with “complex integrations,” but the company says to stay tuned for such features.
You’ll need a WordPress.com account to start the free trial for the WordPress AI Site Builder, and while you won’t need a credit card on file to have fun with the tool, you will need to get a WordPress.com hosting plan to actually make the site usable, which starts at $18 per month (less if you pay for a whole year or more).
The builder works by inputting prompts into a chatbot, so you can tell it that you need a personal or business site, describe how you want the header to look, what kind of colors you like, etc. WordPress.com says being as specific as possible in your first prompt will get you better results, but the chatbot can ask you for information like your business name and location. It then makes you a site complete with AI-generated headers, images, and text. You can then continue prompting the chatbot to tweak anything from colors, styles, photos, and information that you want included.
I tried it out and made my own fake retro videogame store website. I got a super basic but serviceable site laid out in sections like Discover, Events, and Visit Us. It populated a collection of random AI-generated gameroom images, including one with a girl leaning on a CRT displaying an unknown Tetris-like puzzle game. It also presented a “Sega Saturday Showdown” event that you can RSVP for (coupled with an entirely unrelated image of a woman holding a box of Christmas cookies).
Once the site is built, you can transfer it to WordPress.com’s hosting service, where you can continue working on it manually or return to the AI builder. The tool only works for new WordPress instances and can’t yet be used for pre-existing sites.
The Nintendo Switch 2 Pro Controller isn’t the only new Nintendo gamepad getting two back buttons: the Joy-Con 2 Charging Grip has the additional buttons as well, according to Nintendo’s online listing for the controller.
Like the Pro Controller, the new back buttons are GL and GR that can be remapped to other buttons, the listing says. However, the Switch 2’s Charging Grip, at $34.99, is cheaper than the Pro Controller, which costs $79.99, so the grip is a more affordable way to try out those new buttons.
However, as with the original Switch’s charging grip, this new grip only charges your Joy-Con while it’s plugged in. “Attach the Joy-Con 2 controllers to the Joy-Con 2 charging grip to use as a classic-style controller – or a handy charging station with the included USB-C charging cable,” the listing says.
I wish the new charging grip had an included battery. But the ability to charge the Joy-Con while playing and use the new back buttons does make it a step up from the basic Joy-Con 2 grip included with the Switch 2. (Though the back buttons do look quite small.)
The Joy-Con 2 Charging Grip is scheduled to launch on June 5th alongside the Nintendo Switch 2. Nintendo has delayed preorders for the console in the US and Canada and hasn’t yet said when they’ll go live in those countries.
Google is trying to make it easier for users of its video AI model Veo 2 to make cinematic-looking generations and edit real footage. The new Veo 2 capabilities are available to preview via Google Cloud’s Vertex AI platform, alongside other updates to improve Google’s text-to-image generator, Imagen 3, and audio-related AI models.
New Veo 2 features include inpainting, which can automatically remove “unwanted background images, logos, or distractions from your videos” according to Google, and outpainting, which extends the frame of the original video into a different format. The latter tool will fill the new space with ai-generated video footage that blends into the original clip, similar to Adobe’s Generative Expand feature for images.
The update also lets Veo 2 users select cinematic technique presets to include alongside their text descriptions when generating footage, which can be used to help guide shot composition, camera angles, and pacing in the final results. Example presets include timelapse effects, drone-style POV, and simulating camera-panning in different directions.
A new interpolation feature has also been added that can create a video transition between two still images, filling in the beginning and end sequences with new frames.
Editing capabilities in Google’s text-to-image model Imagen 3 have also been updated to “significantly” improve automatic object removal, according to Google, providing what are supposed to be more natural results when removing distractions. Both Veo 2 and Imagen 3 are already being used by companies like L’Oreal and Kraft Heinz for marketing content production, with Kraft Heinz’s digital experience leader Justin Thomas saying the type of task that “once took us eight weeks is now only taking eight hours.”
On the audio side, Google has released its text-to-music model, Lyria, in a private preview and rolled out an “Instant Custom Voice” feature for its synthetic speech model, Chirp 3. Google says that Chirp 3 can now generate “realistic custom voices from 10 seconds of audio input,” and that a new transcription feature is launching in preview that can identify and separate individual speakers to provide clearer transcriptions for calls where multiple people are talking.
These updates are just a handful of AI-related announcements that Google made today. Gemini 2.5 Flash, the latest version of the company’s efficiency-optimized Flash model, will soon be available on Vertex AI. Google says that Gemini 2.5 Flash “automatically adjusts processing time” based on the complexity of the task to provide faster results for simple requests.
Google is also updating its enterprise-focused Agentic AI tools this week to allow AI agents to communicate with each other and perform tasks across platforms like PayPal and Salesforce. Meanwhile, a new section is being launched on Google’s Cloud Marketplace for companies to browse and purchase AI agents built by third-party Google partners.
Jess Weatherbed is a news writer and part of our international team. Jess describes herself as “mostly a generalist who covers a wide range of stories before my US and Canadian colleagues jump online.” Their main beat, however, is with the creative industries, where they cover what tech and software they are using, how creators are being displaced, and more. We saw her wonderful Twinkly Strings setup when she showed us her home workspace last year and thought they looked great.
How did you first find out about Twinkly Strings?
A family friend had some of Twinkly’s outdoor range set up in their garden, and I was looking for customizable Christmas tree lights at the time. I hate the idea of keeping them in a box for most of the year, so my goal was to have something that I could use as general decor year-round, rather than buying something that could only live on the wall. I enjoyed the Twinkly Strings so much that I ended up buying two sets of the Flex light strips soon after to work into some wall murals.
How difficult were they to install?
Installation is easy for both. It takes a couple of minutes to go through pairing them with your home Wi-Fi network, but once that’s done, they should be pretty stable. I’ve purchased cheap RGB addressable lightbulbs that unpair from my wireless network if I so much as sneeze near them, but from what I can recall, I’ve only ever set the Twinkly lights up once.
The Strings are no harder to wrap around a tree than any traditional lights I’ve used, and they sit around a room nicely with some adhesive hooks. The Flex light strip is a bit more labor-intensive, as you ideally need to drill the wall attachments in. You can use the adhesive pads that come with it if you’re in a rental, but I found that those lose their “stick” easily and detach, especially in warmer weather.
What do you like about them?
They just work. I wasn’t happy about the price, but I’m glad I made the investment given my poor experiences with cheaper RGB lighting. I have them paired with smart home routines via Alexa so, for example, they switch on at sundown every day, and then turn themselves off at midnight. I can use voice commands to change the colors, or just open the app — it’s responsive, reliable, and lets me pick from a huge variety of presets or create my own custom designs. I host a lot of themed parties and social events, so it’s nice to quickly change them to suit whatever vibe I need. Plus, they bring so much color into my living space. The decor might not be to everyone else’s taste, but it suits mine perfectly.
Is there anything about either or both that you wish were different or that you think would improve them?
The plug end of the power adapter is massive and can get in the way if I have multiple things to hook up, sometimes covering free sockets entirely on extension strips. I also wish the Strings and Flex had an option with translucent cables so they’d be a bit easier to match with surrounding environments, but I’m nitpicking. The biggest issue is that they’re not cheap, so I’d love for them to be more affordable.
Who would you recommend them to?
If you’re looking for some customizable lighting accessories, these could spare you the headaches you might experience with cheaper products. They’re a tad overkill if you just want something that can jump between white, red, green, and so on, but for bright and gorgeous lighting effects, I’m really glad I splurged. My original set of Strings has survived daily use for around five years now with no signs of stopping, and they still bring me as much joy as the day I set them up.
Older versions of some of these fonts were previously available in Adobe’s Font library, but were removed in January 2023 when the creative software giant ended support for Type 1 fonts, a system that uses the outdated PostScript file format. Now, Adobe says the new fonts are integrated into Creative Cloud at no extra cost — meaning that designers won’t need to pay Monotype to license them or download them locally onto their devices.
This update should result in “fewer frustrating missing font pop-ups” according to Adobe, which users may have previously encountered when trying to use fonts that were removed, or otherwise unavailable on Adobe Fonts or Creative Cloud apps. The font expansion also includes Times New Roman, Arial, and Proxima Nova, alongside options that support alphabets used by languages like Arabic, Chinese, Japanese, Korean, Thai, and Hindi.
Senators Chris Coons (D-DE) and Marsha Blackburn (R-TN) are again introducing their Nurture Originals, Foster Art, and Keep Entertainment Safe, or NO FAKES, Act, which standardizes rules around using making AI copies of a person’s faces, names, and voices. This time, the bill — previously introduced in 2023 and 2024 — has the backing of a major web platform: YouTube.
In a statement announcing its support, YouTube claims the act “focuses on the best way to balance protection with innovation: putting power directly in the hands of individuals to notify platforms of AI-generated likenesses they believe should come down.” It joins a list of supporters that already included SAG-AFTRA and the Recording Industry Association, in spite of opposition by civil liberties groups like the Electronic Frontier Foundation (EFF), which have criticized previous drafts as too broad.
The 2024 version of the bill said that online services (like YouTube) can’t be held liable for storing a third-party-provided “unauthorized digital replica” if it removes the material in response to claims of unauthorized use, and notifies the uploader that it has been removed. Another exception is if the service is “primarily designed” or is marketed for its ability to produce deepfakes.
During a press conference announcing the bill, Senator Coons said that part of the “2.0” refresh included addressing free speech concerns and caps for liability.
YouTube has also expressed support for the Take It Down Act that would make it a crime to publish non-consensual intimate images, even if they’re AI-generated deepfakes, and force social media sites to have processes to quickly remove these images when reported. The latter provision has been strongly opposed by civil liberties groups, and even some groups that advocate against NCII; despite this, it has been passed by the Senate and advanced out of a House committee earlier this week.
Today, YouTube is also announcing the expansion of a pilot of the “likeness management technology” it debuted last year in partnership with CAA. YouTube pitches the program as a way for celebrities and creators to detect AI copies of themselves and submit requests to have the content removed. According to YouTube, some top creators now participating in the pilot include MrBeast, Mark Rober, and Marques Brownlee, among others.
At Y Combinatorâs Little Tech Competition Summit in downtown DC last week, there was an air of optimism and something verging on camaraderie among a surprising crowd. Sen. Cory Booker (D-NJ) â barely recovered from his 25-hour speech about preserving democracy on the Senate floor a day earlier â showed up to discuss the importance of competitive tech markets. Sen. Josh Hawley (R-MO), a staunch ally of President Donald Trump and longtime Big Tech critic, struck similar notes. Lina Khan, former Federal Trade Commission chair during the Biden administration, posed for a photo next to former Trump adviser Steve Bannon, who has openly praised her work.
About half a mile away, at the American Bar Associationâs annual spring antitrust meeting, the mood was less triumphant. In the Khan era, the meeting served as a place for antitrust defense attorneys to grumble about President Joe Biden’s tough merger and antitrust policies. This year, the familiar policy conflicts were shot through with a sense of foreboding over Trump’s gutting of regulatory agencies and attacks on the legal profession itself.
Inside the airy, light-filled room, many speakers at the Little Tech Summit appeare …
