There is a significant danger that this article will have aged terribly. You see, I asked everyone what they were playing this weekend on Thursday, rather than the usual post-lunch scramble on a Friday. You see, I took Friday off to travel to Wales to spend a long weekend with my family. Who knows what happened between my polling of the team on Thursday and Friday? Perhaps Valve surprise released Half-Life 3 and everyone is playing that instead. Maybe they all went off videogames in the interim.
I can only hope they thought to go into the CMS and update the article accordingly. Otherwise, I'll look like a right plonker.
Like many people at companies preoccupied with discovering the next "goose that lays the golden egg", Half-Life 2 and Portal writer Erik Wolpaw has been "poking around" with generative AI. He and a small team at Valve have been testing out different applications, in what Wolpaw assures us isn't a "concerted" effort at implementing the soul-regurgitating, workforce-abrading gadgetry in any particular new game.
Wolpaw's current feeling is that generative AI isn't very good at anything "creative", like cracking jokes. But he does think Large Language Models could make for entertaining NPC voice reactions in games such as Grand Theft Auto and, indeed, Wolpaw's own Left 4 Dead, because AI is marvellous at being a fawning little gopher. It is fantastic at "going along with whatever insane thing you say and kind of adjusting to the flow of that".
One of the most popular JavaScript libraries, Axios, was recently the victim of an attack that had fake, malicious versions available to roll out to developers. These malicious versions install a remote access trojan (RAT), which is, as the name implies, a kind of malware that allows an attacker to access compromised devices from a remote location.
Google has identified the attackers responsible as likely being UNC1069, "a financially motivated North Korea-nexus threat actor" that goes by CryptoCore.
They compromised the Axios maintainer's npm account, npm being a trusted online registry of JavaScript code for users to share and use. Two poisoned packages were added to the Axios npm, and these added a new dependency that installs a RAT.
Malicious code never got into the official Axios software itself, which remains safe, but instead two separate malicious versions were published from an account that usually publishes legitimate Axios versions. Given the way npm works, these compromised, fake versions were able to be pushed to some developers.
The attack was staged almost a day in advance, the two poisoning attacks were timed pretty precisely, and evidence was erased post-exploit, pointing towards a calculated rather than opportunistic attack.
As cybersecurity company StepSecurity explains: "This was not opportunistic. It was precision. The malicious dependency was staged 18 hours in advance. Three payloads were pre-built for three operating systems. Both release branches were poisoned within 39 minutes of each other.
Every artifact was designed to self-destruct. Within two seconds of npm install, the malware was already calling home to the attacker's server before npm had even finished resolving dependencies. This is among the most operationally sophisticated supply chain attacks ever documented against a top-10 npm package."
However, it's important to note that developers using Axios wouldn't have been automatically infected. The malicious versions would have been automatically installed by many projects whenever they next run an npm install command. How often this command is run depends entirely on the company—maybe every week or two, or with a new package install.
(Image credit: Chris Ratcliffe/Bloomberg via Getty Images)
Given that the malicious versions were removed within a few hours, it's likely that most developers using Axios are safe. However, BitDefender says its "telemetry confirms RAT execution attempts on customer systems, blocked by GravityZone and says "the blast radius is not theoretical."
The company recommends identifying exposure, assessing for prior compromise, and monitoring outgoing. Malwarebytes says: "If you are a developer deploying Axios, treat any machine that installed the bad versions as potentially fully compromised and rotate secrets. The attacker may have obtained repo access, signing keys, API keys, or other secrets that can be used to backdoor future releases or attack your backend and users."
Someone from a cybersecurity site and educational malware repo, VX-Underground, recently explained the severity of this on X as follows: "The impact from Axios being compromised is devastating, the fallout from this will be a massive headache. This is unironically a malware nuclear missile and will likely be studied in the future."
Everything will be a web app. That was once the assumption for the PC, even if the emergence of smartphone apps shifted the emphasis somewhat. But now Microsoft wants to put the focus back on running apps actually built for the Windows operating system.
Rudy Huyn, a Partner Architect at Microsoft, recently made an appeal on X (via Windows Latest) for software devs to form a new team that will build Windows apps. When queried over the specific nature of those apps, Huyn was very clear. "100% native", he said.
That "100% native" claim will take some proving. It implies apps that are built with WinUI with no WebView content. The latter, essentially, is a mini browser that apps can use to display web content.
Many Windows apps and indeed elements of Windows itself, including even some parts of the Settings app that seem like they were natively coded for Windows, are actually running on WebView underneath it all. Generally, that's because it's faster and easier to serve up content and application services via a browser than to code an app. Indeed, Microsoft Copilot and Microsoft 365 Copilot are WebView apps.
Likewise, even third-party devs that once coded native apps have been moving to browser-based alternatives, an obvious example being WhatsApp. That particular move is arguably a good example of why web apps can be problematic. Just for starters, running a WebView 2 app means spooling up a Chromium browser instance in the background, which can be much more memory-intensive.
Whatapps' move from native to web app in Windows is notorious for its ballooning memory footprint. (Image credit: WhatsApp)
Funnily enough, this all ties in neatly with an observation made by a former Microsoft operative. Former Windows development lead Mikhail Parakhin said he'd previously worked on a so-called 20/20 project, which aimed to reduce both the Windows install size and idle memory consumption by 20%.
Microsoft hasn't committed to those targets publicly, but it has made "lowering the baseline memory footprint for Windows" a core Aim. Shifting as much app code as possible from WebView to native would certainly help with that.
All in all, this seems like yet another indication that Microsoft has now got its priorities right with Windows. Instead of ramping up the AI slop, Microsoft appears to be focusing on the nuts and bolts of the OS itself. If that's true, it's very, very good news indeed, especially when you consider how much system memory costs these days.
Way back in 2017, Roland carved out a little niche for itself with the introduction of the Go:Mixer line. The small, portable audio interfaces are a convenient way to connect a mic and multiple musical instruments (or audio sources) to your phone for more professional public performances or on-the-go recording. At this year’s NAMM show, the company unveiled the latest in the family — the Go:Mixer Studio — and it’s the most premium version to date.
The Studio adds a display, multitrack output and onboard effects along with a far more luxurious design. At $300, there’s also a far more luxurious price tag. The Go:Mixer Pro-X was already a capable option, and competing products from Mackie and Zoom are also vying for your hard-earned musical dollars. The big question, then, is can the Studio make a case for itself at this elevated price point?
Right off the bat, in terms of usefulness, the Studio is a solid step up from the Pro-X thanks to the addition of a second XLR port. So if your band is a duo, or you simply need two microphones, each performer can now have their own. This also opens the Studio up for basic podcast situations, too. Technically, you could always connect more microphones through other inputs, but now you can do so without adapters or additional hardware like preamps. The rest of the connectivity remains similar with ¼-inch line-in and guitar ports, headset mic support, a 3.5mm aux input and USB-C for audio from your phone and connecting to the app.
Other headline upgrades include a much higher maximum sample rate of 24bit/192kHz (the Pro-X capped out at 16bit/48kHz) and there’s MIDI connectivity for the first time in the Go:Mixer series. The new effects consist of a compressor, EQ and reverb. EQ and compression are available at the channel level, allowing for a good amount of creative control over your mix, while reverb is global. There’s a decent selection of different types of reverb, too, with enough controls to configure them to your taste. I found some of them to be a bit robotic, or not very musical, but others sounded more traditional and appropriate for my vocals and synthesizers.
The Go:Mixer series was doing just fine without a display up until this point, but the benefits of having one are instantly clear. On the Pro-X, the only visual feedback for your levels was a solitary LED that indicated your audio was in the red. If you had multiple inputs, you might not even know which one was too loud. The first benefit of the Studio’s display, then, is visible VU meters. They’re not huge, and the display only shows information for three tracks at a time. This means you might have to page through a few screens to see the one you want, but it’s infinitely more useful than before.
The next obvious advantage of the screen is being able to control settings on the device via a menu. Navigation is intuitive, with the screen divided into three sections, corresponding to the three knobs just below it. The default screen, for example, shows the channels Mic 1, Mic 2 and Guitar/Bass. Turn the first knob clockwise to change the gain of Mic 1. The second knob for Mic 2 and so on. Click a knob and, where applicable, you’ll enter a sub-menu where those three dials control whatever is shown above them. This dynamic system works pretty well and took seconds before it felt natural.
The main limitation is that you can only see three of the mixer channels on screen at a time and there’s no way to manually reorder them. If you have a microphone connected and USB audio playing at the same time, you can’t see the levels or control both of those things from the same screen. You have to keep paging screens back and forth.
The good news is that Roland’s Go:Mixer Cam mobile app does offer a visual mixer that lets you see more or less every channel on screen at once and adjust levels quickly that way. It’s primarily designed for creating videos of your performance, but it doubles as a remote mixer if needed. There is one caveat with the app, though, which is that you won’t be able to use your phone as a USB audio source — say, for backing tracks — if you want to record video with the Go:Mixer Cam app. That’s something to be mindful of.
Roland 's Go:Mixer Studio has a display for the first time in the series
James Trew for Engadget
The app does have a cool feature, which could also be a lifesaver: You can change the “mix” after recording. If you record a performance, but find that your vocals are a bit low, or your synth is too high in the mix, you can adjust the levels and re-export it with better balance. You have options to export as video or audio only, so you can share one to YouTube and then a version for Soundcloud all from the same app. Small detail, but if you want to use the app and have the mixer sample rate set to something other than 48kHz, it’ll warn you that it needs to revert to 48kHz and restart the device before you can carry on.
If you prefer recording on the desktop, there’s also a GoMixer Editor app for Windows and Mac. It’s actually a much easier way to change settings and see what’s going on thanks to the extra visual real estate. The EQ section for each channel looks like a regular software EQ where you raise or lower points on a frequency chart. The compressor also has visual feedback to show when it’s active, which is lacking on the device itself. Obviously, the Studio has a mobile focus, but the desktop app has two big selling points.
First, if you prefer to set your mix levels, compression amount and so on at home, you can do that more easily with the desktop app and then save it in a memory slot. You can then quickly recall this “Scene” on the device while out at a gig. The second is that, for the first time in the series (according to me at least), the Studio is a viable mixer and audio interface for the desktop. The build quality is solid and weighty, not like the light plastic of previous models. It feels premium and this could just as well be used at home for streaming and podcasting as much as on the go. The desktop app makes it even more useful in this scenario.
In terms of what’s missing, this might be very use-case specific. I enjoy using this for electronic music production or pseudo DJ type performances. As such, I’d love to see at least one fader rather than just knobs, but this is true of every model to date. I’d also love for there to be a way to see all the channels at once on the device’s screen. I know it’d be a bit cramped and there’d be no easy way to adjust the mix at the same time, but as an overview you could drop into, it might be handy. And if we’re out here making wishes for any Studio Pro model, an SD card slot for native recording would really elevate the portability element so you wouldn’t need to connect a phone, just a power bank.
The Go:Mixer Studio has two XLR inputs
James Trew for Engadget
Roland has a few competitors in this space, most notably IK Multemida which makes a few portable interfaces. Perhaps the most similar is the iRig Pro Duo and Quattro. The Duo comes in a little cheaper than the Studio at around $235 but lacks a display and build quality. I also personally find IK Multimedia’s apps, while functional, less user friendly. Mackie has the M Caster Studio ($200) which adds Bluetooth connectivity but has fewer physical ports — that one too is a little older. Zoom’s interfaces often center on their ability to record directly onto the device, but have more of a vocal/spoken word focus. The H5 Studio ($299) has a display, built-in mic and onboard recording, but its mixer functionality and outputs for live performance are secondary features.
For musical performers, Roland continues to dominate this niche, and the Go:Mixer Studio is clearly the company’s most refined interface to date. The connectivity covers most use cases, even podcasting, and the layout of the dials makes it easy to use in live environments. The display is a welcome addition that goes a long way to making this feel both more useful and more premium. Perhaps the biggest selling point this time around is that the Studio no longer feels like an extra interface you bring with you for live gigs. It can easily be your main desktop audio interface too, making that $300 price tag suddenly feel a lot more palatable.
This article originally appeared on Engadget at https://www.engadget.com/audio/roland-gomixer-studio-review-portable-professional-and-plenty-of-polish-130000723.html?src=rss
WWDC 2026, the latest version of Apple's yearly developer conference, runs from June 8-12, and by all appearances the company has some important updates to outline. In comparison to Liquid Glass, the design material Apple introduced last year and now uses across all its operating systems, the new features the company is rumored to announce might not be aesthetic, but they could make just as big of a splash. Namely because Apple might finally be ready to show off its second stab at an overhauled version of Siri.
If you're curious to see the company's new plans for yourself, you can watch Apple's WWDC 2026 keynote live on its website, YouTube channel or the Apple Developer Bilibili channel in China. Apple will also host its Platforms State of the Union stream and individual developer workshops on its developer website if you want to learn even more details about the software updates the company will release later this year. Luckily, we do have some sense of what Apple has in store, and it looks like stability improvements and AI are the company's big focuses for the updates coming to iOS, iPadOS, macOS, watchOS, visionOS and tvOS this fall.
A Snow Leopard-esque approach to stability and performance
Apple released Mac OS X Snow Leopard in 2009, primarily as a way to clean up the performance and refine the new features the company released with Mac OS X Leopard two years prior. The decision to essentially "take a year off" to focus on making everything about the company's desktop operating system feel better was well-received, and Apple is apparently planning to have iOS 27 serve a similar role.
Bloomberg reports that Apple's upcoming update will be "focused on improving the software’s quality and underlying performance" and that the company's "engineering teams are now combing through Apple's operating systems, hunting for bloat to cut, bugs to eliminate and any opportunity to meaningfully boost performance and overall quality." Those fixes will presumably extend to the company's other operating systems, too.
Some of this effort may also be focused on cleaning up the visual changes introduced in Apple's big switch to Liquid Glass. The design overhaul has been controversial among the company's diehard fans, and Apple has already introduced tweaks in updates that arrived after the release of iOS 26 to make Liquid Glass interfaces more legible. Bloomberg reports the company could go a step further in its next updates and add a system-wide slider that will allow users to adjust the intensity of Liquid Glass (visual effects like translucency and reflectivity) they want in the interface.
The chatbot-ification of Siri
While stability and performance improvements will be a major focus of this year's updates, Apple is also rumored to be making some major changes to Siri. When the company first introduced Apple Intelligence at WWDC 2024, it promised to launch an updated version of the voice assistant that could use your personal context (like the information securely stored on your iPhone) to act across apps. Apple delayed those features in March 2025 and then announced a partnership with Google in January 2026 to use Gemini models to presumably make them possible.
Those features might finally arrive in this year's updates, but Apple is reportedly also changing how users interact with Siri by making the assistant more like a chatbot, according to Bloomberg. This would make the assistant more interactive and natural to speak to, and could open up other possibilities, like letting users direct Siri to perform two actions at the same time. Developers will reportedly also be able to integrate their own AI assistants with Siri, much like OpenAI has with ChatGPT.
New places to talk to AI
The chatbot version of Siri will be accessible in the usual ways, but also reportedly through a standalone Siri app. The new app will let users prompt the assistant to take care of tasks on their device, search the web and even access news, not unlike current Gemini and ChatGPT apps. Bloomberg writes that the app will also be a way to review past conversations with Siri and receive suggestions of prompts to try with the new chatbot version of the assistant.
Users will also be able to interact with Siri inside Apple's other apps via a new feature called "Ask Siri." This may appear as an option in app menus, and allow you to ask the AI assistant questions about content in the app. It's not clear if this will be as in-depth or capable as Google's Ask Maps or Ask Photos features, but it at least seems like Apple's thinking along the same lines as its partner.
This article originally appeared on Engadget at https://www.engadget.com/big-tech/what-to-expect-from-wwdc-2026-110000086.html?src=rss
Screenshots promoting "Instagram Plus" have been spotted by users in the Philippines and Mexico in recent days. According to social media consultant Matt Navarra, a subscription to Instagram Plus includes several Story-focused features not otherwise available to Instagram users. This includes the ability to create multiple audiences for Stories posts, search the list of people who have viewed your Story, extend Stories longer than 24 hours and create "spotlight" Stories. It also mentions something called “super hearts” for reacting to Stories.
A spokesperson for Meta confirmed the test to Engadget, saying that Instagram Plus is currently available in “a few countries.” The spokesperson added that “preview” would allow people to see some of another user’s Story without “showing up as a viewer”. There’s nothing quite like paying to be sneakier on social media.
The idea, as we’ve seen so far, seems closely modeled after Snapchat+, which also offers bonus features to the app's power users. Launched in 2022, the service has now become a significant driver of non-advertising revenue for the company.
Apple has been around for fifty years. It reimagined personal computers, catalyzed the era of the smartphone, enlarged an iPhone and called it the iPad and garnered a strong position in wearable tech through its Watch series and its AirPods. It also popularized software and services like its App Store, FaceTime, iCloud, iMessages and many more.
However, Apple gives and it takes away. For the MacBook Air to exist, the disk drive had to go; ditching the home button led to edge-to-edge iPhone screens. The weight of Apple's influence has led to entire product categories following suit. Or, more typically, there's resistance, complaining and (eventually) following suit.
Two new models of Meta Ray-Ban AI glasses are on the wayand they're going to be catered to those who use prescription lenses, according to a Bloomberg report. While these are supposed to be announced next week, the report noted that these won't be a "new generation" of Meta's smart glasses.
You can already add prescription lenses to Meta Ray-Ban's AI glasses, so who cares? Well, the upcoming models will come in rectangular and rounded styles and will be sold through traditional prescription eyewear channels. It seems like a way to get Meta’s wearables in front of a new group of prospective customers.
Now available on the App Store and Google Play Store, the official White House App claims to give Americans "a direct line to the White House." According to the press release, the app provides "unfiltered, real-time upgrades straight from the source." The White House App's News tab features a carousel of about 35 articles that seem suspiciously cherry-picked, favoring the Trump administration. As Engadget’s Jackson Chen notes, in the Affordability window, the app points out year-over-year price drops for items like eggs, milk and bread, but conveniently omits the recent surge in gas prices.
In the Social tab, there's even a button to "Text President Trump," which auto-populates a new text with "Greatest President Ever!"
Bluesky is the latest social media platform to throw its hat into the AI chatbot ring. Bluesky, but specifically its chief innovation officer Jay Graber and her new Exploration team, built a new AI assistant called Attie that's designed to help users create custom feeds. Graber called Attie an "agentic social app" that's built on its its open-source framework called the AT Protocol.
To use Attie, users can punch in prompts in natural language to generate social feeds without having to know how to code. On the Attie website, examples include prompts like, "Show me electronic music and experimental sound from people in my network" or "Builders working on agent infrastructure and open protocol design."
Attie
"It feels more like having a conversation than configuring software," Graber described Attie in a blog post. "You describe the sort of posts you want to see, and the coding agent builds the feed you described."
Graber added that Attie is a separate app from Bluesky and users don't have to use the new AI assistant if they don't want to. However, since Attie and Bluesky were built on the same framework, it could mean there will be some cross-app implementation between the two or any other app built on the AT Protocol. Attie is currently available on an invite-only closed beta, but anyone interested can sign up for the waitlist on its website in the meantime.
This article originally appeared on Engadget at https://www.engadget.com/ai/blueskys-next-product-is-an-ai-assistant-that-helps-build-custom-social-media-feeds-163140902.html?src=rss
President Donald Trump may have a tendency to put his name on everything, but his administration decided to go with the more authoritative The White House App for his latest pursuit. Now available on the App Store and Google Play store, the official White House App claims to gives Americans "a direct line to the White House."
From the press release, the app provides "unfiltered, real-time upgrades straight from the source." In more practical terms, the White House App is a one-stop shop for official communications from the administration and more. On the app, you can find press releases, livestream announcements and even a photo gallery, along with turning on notifications so you get official communications as soon as they happen.
However, it only takes a few minutes of digging through the app to question its value. The White House App's News tab features a carousel of about 35 articles that seem suspiciously cherry-picked with articles that favor the Trump administration. In the Affordability window, the app points out year-over-year prices that have dropped for things like eggs, milk and bread, but conveniently omits the recent swell in gas prices.
In the Social tab, there's a button to "Text President Trump," which auto-populates a new text with "Greatest President Ever!" before ultimately trying to get you to sign up for a marketing blast. The press release mentioned a way to "send your voice and feedback directly to the Administration" but the app's functionality doesn't seem to promote that. Most notably, there's even a way to submit a tip to Immigration and Customs Enforcement in the app's Get in Touch section.
While it's convenient to have all your Trump administration announcements in one place, the White House App is mostly just a portal that ends up opening external websites. Traditionally, official White House accounts on social media platforms are passed on during the transition of presidents. However, it's hard to say what will happen to the app after Trump leaves office, but one only has to look towards the lasting triumph of TrumpRx, Trump Mobile or even Trump University.
This article originally appeared on Engadget at https://www.engadget.com/apps/the-white-house-app-is-just-as-weird-and-unnecessary-as-youd-expect-175354004.html?src=rss
This week's Borderlands 4 patch has arrived, firing some gear balancing tweaks into the looter shooter's innards, then probably dropping a cringey catchphrase. Developers Gearbox have also revealed some of what's in store for the larger patch that's set to arrive next week, with performance fixes back on the menu alongside some more in-depth balancing that'll touch "those unintended interactions" folks have been relying on for overpowered builds.
It's all in the line with the nice rhythm the studio have settled into following an initial flurry of activity in response to the game running not goodly on PC out of the gate. Last week brought a big patch featuring the first round of vault hunter balancing tweaks, so this week sees a chaser precede the next shot.
Control: Ultimate Edition will be available on the iPhone, the iPad and the Apple Vision Pro in early 2026, its developer Remedy has announced. The developer says you can either "tap into the action with touch controls," which presumably includes hand tracking and gestures on the mixed reality headset, or use controllers to play the game. Remedy first made it available for the Apple ecosystem when it released Ultimate Edition for Mac back in February this year.
Remedy originally released Control in 2019 for the PlayStation 4, Xbox One and Windows computers. In the action-adventure game, you take on the role of Jesse Faden, the new director of the Federal Bureau of Control (FBC) who's also searching for her kidnapped brother Dylan. The game is set in the Oldest House, the headquarters of the clandestine US agency that studies and contains paranatural phenomena.
Ultimate Edition is the definitive version of the title and bundles the base game with the Foundation and AWE (Altered World Events) expansions. While both are continuations of the main game, AWE is a crossover between Control and Alan Wake, an older title by Remedy about a crime author whose wife disappears during a trip to a small mountain town. Remedy hasn't announced a specific release date or price for the game yet, but it's currently listed for $40 on the Apple Store for Mac computers.
This article originally appeared on Engadget at https://www.engadget.com/gaming/remedys-control-is-coming-to-iphone-ipad-and-vision-pro-early-next-year-120100226.html?src=rss
Sora, OpenAI's app and social network for AI-generated videos, has been downloaded over one million times, according to Sora head Bill Peebles. The app reached one million downloads in less than five days, Peebles says, "even faster than ChatGPT did." That's despite OpenAI only making the app available in North America, and its decision to require users to have an invite to actually use it.
Like TikTok, Sora offers an endless vertical feed of videos, only Sora's videos are AI-generated rather than uploaded by users. Creating a 10-second video of your own is as simple as writing a prompt to OpenAI's Sora 2 model in the app. And through the Sora's Cameo feature, you can even create videos of yourself and anyone else who's agreed to share their likeness to the service.
sora hit 1M app downloads in <5 days, even faster than chatgpt did (despite the invite flow and only targeting north america!)!
team working hard to keep up with surging growth. more features and fixes to overmoderation on the way!
The limited guardrails OpenAI has put on Sora has already led to a rash of videos featuring OpenAI's Sam Altman and content that clearly infringes on copyright. The fact that Sora can so readily create videos of recognizable characters like Pikachu raises questions about what OpenAI's model was trained on, and has unsurprisingly prompted pushback from the larger entertainment industry.
In response, the company has updated Sora to give users more control over what videos their likeness can appear in. OpenAI plans to offer similar controls to rights holders, giving them "the ability to specify how their characters can be used (including not at all)," according to Altman. It's not clear why these controls weren't available when Sora launched, but both seem like good changes.
Because of Sora's invite system, it's difficult to say if the over one million downloads the app has received translates to as many users. It's not unusual for someone to download an app and never use it. Whatever the case, OpenAI's bet on AI-generated videos seems like it might be a winning one, provided the company finds a way to actually make more money than it looses generating videos for Sora.
This article originally appeared on Engadget at https://www.engadget.com/ai/openais-tiktok-of-ai-slop-hit-one-million-downloads-faster-than-chatgpt-181216271.html?src=rss
A dedicated Borderlands 4 player has killed over 3,000 bosses and spent 150 hours uncovering the drop rates for items in the game. Gearbox Software keeps the item drop rate hidden from fans, but one Borderlands 4player has done in-depth research and discovered that the chances of gaining specific Legendary items are surprisingly low.
Essentially, you make the product, you should be forced to use it—which Graphite apparently enacts by randomly deleting employee accounts on a daily basis, a process it refers to as "onboarding roulette".
The facepalms, they are not big enough. Anyway, the whole idea is to test the onboarding process for its AI code review software, a traditionally difficult problem to solve due to the userbase being your primary test subjects. Fear not, however, as apparently dogfooding is the way forward. As co-founder and CTO Greg Foster explains:
"Our solution at Graphite has been to run a roulette script, randomly deleting one of our engineers' Graphite accounts every day at 9 a.m. We don’t just reset onboarding—we delete their account, tokens, configured filters, uploaded gifs, and more."
"Isn't that frustrating? Sure." continues Foster. "Folks on our team come to work to code new features, not to find themselves logged out and forced to recreate their accounts from scratch. We were cautious when first trying the technique, but the benefits became clear immediately."
(Image credit: Tim Clark)
I presume those benefits don't include members of your team immediately searching for a new job when they realise they were screwed with deliberately, but still. Foster seems remarkably chipper about the whole affair, continuing:
"Like most products, Graphite aims for fast, bug-free, and painless onboarding. The best way for us to ensure this is to suffer through onboarding once every day ourselves. Across our full Eng-product-design team, any individual only gets deleted once a month on average. But one teammate a day hitting a sharp edge has proven enough to find and motivate fixing issues.
"Deleting employee accounts has created dogfooding on one of our most critical and hard to test surfaces. We’ve caught tens of bugs, and created user empathy in a traditional blindspot. I’d strongly recommend other product teams consider automatically deleting employee accounts for the same benefits."
Somehow I don't see "onboarding roulette" catching on, despite the supposed benefits. I can only imagine the emails I'd send if my job deliberately kicked me out of the platform I was building for the sake of testing, although I'd guess that some choice language would be involved. And as for "dogfooding?" I'll leave the Pedigree Chum to the pooches, thanks.
Although Windows 10 machines will still function from October 15, every day thereafter poses a security risk to machines without these updates. As your device ages, it becomes ever more vulnerable to bad actors.
This is where our good pal Rufus comes in. It is an open-source program that creates bootable drives, which can be used to install the likes of Linux or Windows on a machine. Crucially, Rufus can be used to bypass hardware requirements for Windows 11 and can even be used to make offline Microsoft accounts. If you plan on upgrading to Windows 11, Rufus is how we recommend doing so.
How to use Rufus to upgrade to Windows 11
(Image credit: Microsoft / Rufus)
You will need to download two things to get Windows 11 up and running via Rufus (and have a USB drive ready). First, naturally enough, is Rufus, which you can download from the official website. You can grab a portable version of the software, one for ARM64 devices, or the standard Windows x64 one. The standard one is the one you want in most cases.
After this, you need to grab an ISO for Windows 11 via the Microsoft website (at the very bottom of the page). Pick Windows 11 (multi-edition ISO for x64 devices) in the drop-down menu, select your language and get it downloaded.
From here, boot up the Rufus application from your main device and click your new Windows ISO in boot selection. Then click Standard Windows Installation in Image Option. Hitting start at the bottom will open up the Windows User Experience tab, which allows you to customise how this boot drive operates.
You can check to remove Windows 11's strict hardware requirements or even force your install to use an offline Windows account, which Microsoft has been trying to block our ability to do for years.
Using Rufus will remove all data on the USB drive you want to install it on, so make sure to clear off anything important before turning it into your boot drive.
From here, simply plug the stick into any device you want to install Windows 11. Your device should be able to pick up the boot drive automatically, but you can also choose to install from BIOS as your system starts up, should you run into any problems. To get into the BIOS, you will need to press F1/F2/F10/Del (annoyingly, it depends from manufacturer to manufacturer) as you boot up your computer.
Why you should use Rufus
(Image credit: Microsoft)
Being able to update easily and straight from a boot drive is handy in itself, but one of Rufus' superpowers is that it can be used to get around Windows 11 hardware requirements.
Though far less of an issue now (due to the inevitable encroachment of time phasing out older unsupported hardware), when Windows 11 launched in 2021, it required your rig to have a TPM. TPM 2.0 is a hardware-specific bit of security technology, and is seldom found in CPUs pre-2016. It essentially stores important cryptographic data used for enhanced security. However, should your rig be so old as to not to pass this requirement, Rufus can be used to get around it.
Rufus also gets around the requirement to have/make an online Windows account. Simply click to remove both requirements when you create your Windows 11 boot drive, and you can upgrade quickly, easily, and for free. However, there are still reasons to update the old-fashioned way.
Why you might not want to use Rufus
(Image credit: Microsoft)
Upgrading from Windows 10 to Windows 11 (when you have the right hardware and already have a Microsoft account) can actually be done straight from the Windows update settings on your PC.
If you have no reason not to update and don't want to work around Microsoft's nonsense, you can do it without needing a USB drive or any other downloads.
If your CPU is fairly recent and you still can't seem to pass the TPM 2.0 requirements, there's a chance the crucial security setting has not been enabled on your PC. You can turn it on easy enough through your BIOS, and Microsoft has put together a good guide on how to do so on a range of motherboards.
When directly upgrading from Windows 10 to Windows 11, you can't use Rufus to bypass the account requirements set by Microsoft. However, there's another way around those, if you'd like to follow me down this dark alleyway to discuss some illicit business.
(Image credit: Future)
All you need to do is hit Ctrl + Shift + J during the Windows 11 set-up process, at the Secondary Keyboard Layout screen, but before the account sign-up section. From here, type:
WinJS.Application.restart("ms-cxh://LOCALONLY")
After this, hit enter, then close the console, and you can make an offline account. Though Microsoft will bug you to log in forevermore in your Settings.
Previously, you could just press Shift + F10 to open the command prompt during setup, then type 'oobe\bypassnro' to get around the online account requirement. However, that was disabled by Microsoft last March. My only worry here is that Microsoft is clearly going some way to stopping offline accounts, so it can't be guaranteed this fix will stay around indefinitely.
No matter which way you install Windows 11, make sure you get all the latest drivers after your update. As a chronic "I'll update it when I run into problems" person, it's always good to have a digital cleanup every now and then.
Rufus has been a bit of a hero to many PC gamers over the last few years, as it has managed to brute force its way past many of our least favourite Windows decisions.
Though you can get around the offline account problems on-device, it's hard not to look at Rufus as a beacon of hope for whatever Microsoft has cooking up next. It's a good safety plan should Microsoft continue to filter out other workarounds in trying to foist its account onto unsuspecting users.
The era of the AI PC is upon us, with Microsoft forcing Copilot on every 365 product like Word, Excel, and PowerPoint. When the time comes to swap to what I assume will be called Windows 12, I'm hoping Rufus will keep the ever-growing creep of AI at bay for a while longer.
Login
