Back in February 2025, several media outlets warned of a new threat targeting users of Google’s Gmail email service. Attackers used AI technology to perfect phishing emails and make them look more convincing. To do this, the AI collected freely available data from social networks, websites, and online forums on the internet and used this information to formulate a deceptively genuine-looking email that pretended to come from an acquaintance, family member, or superior. What’s more, to ensure that the message actually looked deceptively genuine, the AI also generated suitable domains as senders for the emails. The scam was dubbed “Deepphish”—a portmanteau of the terms deep learning and phishing. Even if the report mentioned at the beginning raises some questions—such as why Gmail users in particular were affected by the Deepphish attack—it nevertheless highlights a development that experts had been expecting for some time: criminal groups are increasingly using AI tools to perfect their attacks. Domains created with AI One of the weak points of conventional phishing attacks has always been the sender address. Most phishing emails can be easily identified by the sender . For example, a message from a streaming service such as Netflix or Disney with an address like andy@brandbot.com is almost certainly a fake—no matter how perfect the rest of the presentation may be. In the AI-supported variant of a phishing attack, on the other hand, new types of algorithms are used that generate a sender address with a matching URL that is adapted to the text of the email. A research group led by Alejandro Correa Bahnsen at the US company Cyxtera Technologies, an operator of data centers, investigated how effective these algorithms can be. They developed an algorithm called Deepphish, which was trained to suggest suitable URLs for phishing attacks. To do this, they fed a neural network with more than one million URLs that had been set up for phishing via email in the past and used them to train their algorithm. In doing so, they specified two different profiles for the actors behind the phishing attack. Phishing emails can often be recognized by the sender addresses. If, as in this case, a message purporting to be from Disney comes from andy@ brandbot.com, something is wrong.Foundry With the AI-generated addresses, they achieved an increase in attack efficiency from 0.69 to 20.9 percent for one profile and from 4.91 to 36.28 percent for the other. They published their results in a stud you can find here. While Deepphish originally only referred to the algorithm developed at Cyxtera, it’s now used in most cases for AI-supported phishing attacks in general. How a Deepphish attack works Deepphish attacks follow a standardized pattern. The first step is to research the target’s social environment: Where does she live? Where does she work? What are the names of their family members? What are their friends’ names? What are the names of their colleagues and superiors? What are their email addresses, how close are they to the target person? The AI uses social networks and online forums as sources, as well as data published by hackers from intrusions into company networks and websites. The more data collected in this way, the more precisely the phishing email can be tailored to the victim. In a study, employees at Cyxtera investigated how the success rate of phishing emails can be increased by selecting an AI-generated sender address. Foundry The next step is to register a suitable domain and generate a sender address using an algorithm such as Deepphish. The AI then writes the text of the email. Using the information collected, it generates a suitable subject line, a salutation tailored to the recipient and content that is worded correctly and could actually have been written by the supposed sender. Due to the precise personalization, the message appears considerably more credible than a standard phishing email. But what do the criminals want to achieve with their deepphish attacks? They want to inspire so much confidence with their forgeries that the recipient is prepared to click on a file attachment or an embedded link. Everything else happens automatically: the file attachment then usually downloads and installs malware. The link, on the other hand, leads to another fake website that requests credit card details or login information for a streaming service, for example. AI-supported phishing emails However, the Deepphish algorithm is just the beginning. There is now a whole range of tools that do all the work for criminals when formulating phishing messages. The programs are called FraudGPT, WormGPT, or GhostGPT. They formulate phishing emails that are targeted at individuals or specific companies. For example, the user can instruct these programs to generate a Netflix-style email with a request to enter account details on a fake website. Or they can have questions answered such as “How do I hack a Wi-Fi password?”. Or they can instruct the AI to program a software keylogger that forwards all keystrokes on a computer to a server address via the internet. Hacking tools such as WormGPT use AI to generate convincing-looking and well-formulated phishing emails. In most cases, they target specific individuals or companies. Foundry ChatGPT and other Large Language Models (LLMs) have in-built filters so that they do not respond to such requests. As ChatGPT does not disclose its code, there is nothing that can be done about this. However, it is possible to use instructions from the darknet to confuse LLMs such as ChatGPT via certain prompt sequences so that they are then prepared to disregard their in-built filters. At the same time, some criminal groups have switched to LLMs from the open source scene and removed the corresponding filters. AI generates malware The Stopwatch AI website demonstrates just how far the possibilities of AI-generated malware already go. It shows how AI can be used in three simple steps to program malware that specifically undermines the protective shield of the major antivirus tools. In the first step, called “Choose Platform”, you select the operating system of the computer you want to attack. You can choose from Mac, Windows, Linux, AWS (Amazon Web Services, Amazon’s cloud service), Google Cloud, and Microsoft Azure, Microsoft’s professional cloud service. The Stopwatch AI website demonstrates how malware can be programmed in a few simple steps with the help of AI tools. The first step is to select the operating system to be attacked. Foundry The second step is called “Choose Defence” and offers nine antivirus tools, including Microsoft Defender, Eset Endpoint Protection Advanced, McAfee Endpoint Security, Symantec Endpoint Security, and Kaspersky Endpoint Security for Business. In the second step, Stopwatch AI users select the antivirus program whose weaknesses they want to exploit with their malware attack. Microsoft Defender is also listed here.Foundry In the third step, “Choose Attack”, you specify the type of virus you want to create. The selection ranges from adware and spyware to ransomware and keylogging through to data exfiltration, i.e. data theft. Stopwatch AI offers ten different types of malware, from keyloggers to ransomware. The user must register in order to implement the selected malware. Foundry After clicking on a form of attack, Stopwatch AI asks for log-in details. It is possible to register with the site using a Google, Github, or Microsoft account. As soon as registration is complete, the AI starts programming the desired malware. In order to use the site, the user must agree to the terms of use, which exclude attacks against other systems. This is because Stopwatch AI is only intended for studying malware development with AI. Critically, all projects are assigned to the respective user and saved. How to recognize AI-generated phishing emails Always take a look at the sender address of incoming emails and consider whether it is plausible. Also look out for the following features: Become wary of emails from people you are not normally in contact with or have not heard from in a while. This is especially true if these messages contain unusual requests or enquiries. Hover your mouse over any links and check where they lead to. If the address does not match the sender of the email or the text of the message, it is often a scam. No bank, streaming service, or public authority ever asks for your password or wants to know your account details via email. Be suspicious of emails that put you under time pressure or claim a high level of urgency. Tricking antivirus tools with AI Every antivirus program downloads the latest virus definitions from the manufacturer’s server at least once a day. They describe the characteristics of the new malware variants discovered in the last few hours so that the software on the user’s computer can reliably detect the malware. However, this protective shield has become increasingly fragile. The reason: virus construction kits that allow hobby programmers to create functioning malware even without AI have been circulating on the darknet for decades—but not only there. Many of these malware programs are simply minimally modified variants of already known viruses. The author often only has to change the signature for his malware to be counted as a new virus. This is the only way to explain why antivirus manufacturers report 560,000 new malware programs every day. In the age of AI, the production of malware variants has taken on a new quality. This is because security manufacturers had taught their antivirus programs to recognize and isolate the variants of already known malware. With AI support, it’s now possible to manipulate existing malware in a targeted manner so that it is no longer recognized by the virus watchdogs. The tool manufacturer Acronis demonstrated this in a presentation using a malware sample that it had uploaded to Google’s detection service Virustotal. While it was initially recognized as malware by nine of the antivirus programs used there, only one virus guard was able to identify the malware as such after it had been reworked by Grok3’s AI. When the researchers had the sample code processed by Gemini 2.0 Flash and Deepseek R1, the virus was no longer detected by any of the programs at Virustotal. Depending on which AI software is used, the hacker can manipulate existing malware in such a way that it remains almost or even completely undetected by Virustotal. Foundry Nevertheless, the heuristic and behavior-based methods of antivirus programs also work with malware whose code has been modified with the help of AI. Email spoofing The falsification of email addresses, known as email spoofing, hardly occurs any more. Since 2014, the SPF, DKIM and DMARC authentication methods have gradually been defined as standards and subsequently implemented by email providers. Since then, it is no longer possible to falsify the domain information in an email address. For an address such as “magazine@pcworld.com”, for example, the domain is pcworld.com. If the aforementioned authentication procedures are deactivated by a provider, these mails are normally sorted out as spam by the recipient. Spoofing attempts still exist, however. The sender’s name can be changed in many e-mail clients, for example in classic Outlook via File -> Account settings -> Account settings -> Change -> Your name. However, this does not affect the email address. In the case of hacker attacks, the reply address is sometimes changed at the same point. In this way, all replies to the emails sent are sent to the hacker’s address. Another trick is to use a similar-looking domain, such as “magazine@pcworlde.com“. Read...

Environment Southland said everyone on board the boat is safe, and there are no known environmental impacts from diesel or oil. Read...

Who doesn’t remember PC games such as Maniac Mansion, the King’s Quest series and the dubious adventures of Leisure Suit Larry or software such as Microsoft Works and Lotus Smart Suite? These titles originally came from the 80s and 90s, ran under MS-DOS or Windows 3.1 and were delivered on floppy discs or—very modern back then!—CD-ROMs. In our guide, we want to breathe new life into these treasures from the past and get them running on a current PC with Windows 11. We describe two scenarios: Firstly, we use the original data carriers for MS-DOS or Windows 3.1. The second scenario describes the case where installation media are no longer available. Even then, there is a way for you to delve back into all your favourite games and programs from back then. Prerequisite: DOSBox as MS-DOS emulator To run old applications and games under Windows 11, you need an emulator such as DOSBox. This command line tool, which was primarily developed for games, provides the MS-DOS platform and simulates an IBM PC-compatible computer with an x86 processor, the speed of which you can adjust. This means that old 16-bit programs can also be run under DOSBox, and you can even install Windows 3.1 and then run compatible software on it. Since not every user knows how to use a command line tool or still has the handling of MS-DOS including all commands ready at hand, DOSBox comes with a very good help function and also offers tutorials and instructions on the project website, simplifying operation enormously.Foundry DOSBox also provides the required hardware environment: In addition to mouse and keyboard, the tool controls input devices such as joysticks and gamepads, which is indispensable for games. It also supports (virtual) CD-ROM drives and sound cards for audio output. The emulator works independently of the actual equipment of the PC on which it’s running. The operation of DOSBox requires rudimentary knowledge of command line tools. However, you can always call up a short introduction and the help function each time you start it, and there are excellent instructions on the project website. There is also an alphabetically sorted list of supported games at www.dosbox.com/comp_list.php. Setting up DOSBox and mounting the necessary drives For the sake of clarity, install DOSBox in a newly created directory called Oldies on your hard drive “C:\” and create another folder called DOSSoft in which you’ll later store your games and applications. When naming folders, please note that DOSBox, like MS-DOS itself, abbreviates all names that are longer than eight characters. As you’ll have to type the folder names into the command line later, you should therefore avoid names that are too long. For our article, we first want to run the MS-DOS game Shadowlands (1992), which we still have the original 3.5-inch floppy disc for, under DOSBox. Next, install Windows 3.1 from the original floppy discs and install the graphics program Aldus Photostyler (1991) – available on four 3.5-inch floppy discs – and an image of Microsoft Works (1988). The most important command in DOSBox is mount. You use it to provide the drives and folders that you’ll need later to install and start your software. The dir command shows you the contents of a directory, for example the floppy disc of the game Shadowland (1992).Foundry To read the media, use a USB floppy drive that is mounted in DOSBox, i.e. make it available to the emulator as a physical drive. To do this, start the program and type for the USB floppy drive mount A A:\ -t floppy This preserves the drive’s letter A. The “-t” and the subsequent designation stand for the drive type that is mounted. The drive letter is capitalized. Otherwise, upper and lower case is not important. If the mount was successful, a corresponding message appears. Please be patient, however, as the process may take a few seconds. Important: Every time you exit DOSBox, you must remount your drives. In addition, the contents of the mounted drives and folders are not updated. So if you copy new files into a mounted directory, you must remove it with mount -u : and then remount it. By the way: If you still have programs on 5.25-inch floppy discs, installation can be difficult or even impossible. We did not find any external drives for this in our research, and the current mainboards do not have the controllers for used internal devices. So if you want to use a program that originally comes on 5.25-inch floppy discs, you have to get the ISO from the Internet. But more on this later. Running games from 3.5-inch discs under MS-DOS The fantasy role-playing game Shadowlands was released in 1992 for MS-DOS, Amiga and Atari, and we want to run it from the original 3.5-inch floppy disc in DOSBox. After mounting our floppy drive as described above, we must do the same with our installation directory. So we type mount C C:\Oldies\DOSSoft – the folder now has the drive letter “C:” -, then change the drive letter with A: to the floppy drive and install Shadowlands using the command install A: C: into our game folder. A “SHADOW” folder is automatically created. The message “Install Complete” appears as confirmation, and we switch to the next folder with C: to automatically switch to the SHADOW folder. With dir to display the contents and finally start our game with shadow. The Lucas Arts adventure game The Dig was distributed on a CD in 1995. The game can still be started and played directly from this disc in DOSBox. To do this, we mount the external drive, change to the directory of The Dig and execute the start file. Foundry By the way: If there are several subfolders, the cd command is used to change the subfolder manually. You can go one level higher with cd . (two dots). After starting Shadowlands, select the resolution, i.e. whether we want VGA with 640 x 480 pixels or EGA with 640 x 350 pixels. We prefer the “newer” VGA. Shadowlands then starts with sound output and lets us create a hero for our epic adventure. Conveniently, we can use the mouse, as DOSBox can provide this as well as the keyboard. While you’re in an emulated game, the mouse and keyboard are disabled for your “real” PC. You must then exit the game or briefly interrupt it with Ctrl-Alt-Del to restore functionality. If you then want to switch back to the game, simply click with the mouse in the relevant window. By the way: If you want to install an old game from CD, mount the corresponding drive with mount E E:\ -t cdrom whereby the drive retains the letter E. Then switch to it with E:, display the contents with dir to find the setup file and then execute it. Proceed in the same way if a game can be started directly from CD. Then run the start file instead of the installer. Installing Windows 3.1 from disc or image on DOSBox Next, we want to install Windows 3.1 from our seven 3.5-inch floppy discs under DOSBox. To do this, mount our floppy drive and the installation folder as described above. With dir to display the contents of the first floppy disc and then start the installation with setup The familiar screen appears and you can run through the installation. Change the discs when the installation wizard prompts you to do so. The data is automatically stored in your installation directory in a (new) folder called WINDOWS. To start the operating system, mount this folder as drive F, for example, with mount F C:\Oldies\DOSSoft\Windows mount, then change to the directory with F: and start Windows 3.1 with win to start it. To return to DOSBox, click on “File” and “Exit Windows” in Windows 3.1. You can even install Windows 3.1 in DOSBox using the original discs or an image of them and then use other programs in it. Today, installation via image only takes a few seconds – no comparison with the numerous floppy disc changes in the past. Foundry Caution: Older floppy discs in particular may no longer be readable! In this case, you can alternatively obtain an image of the desired program from the Internet. Remember, however, that you cannot legally use the software without a valid serial number! In the case of Windows 3.1, for example, we found what we were looking for at winworldpc.com. After downloading and unpacking the ZIP file, move the six disc images to a newly created folder called Win31, which you then move to your installation directory. Also unpack the IMG files so that all installation files are in one folder. Now mount the directory as drive E with mount E C:\Oldies\DOSSoft\Win31 then change to the directory with E: and carry out the installation with setup. As with the installation via floppy discs, a folder “WINDOWS” is created, which you must mount in order to install from it with win to start the operating system. Running applications in DOSBox under Windows 3.1 To install a program under Windows 3.1, you’ll need to store the installation files in a folder to which we have access under Windows 3.1. This is the case for your installation directory, so create a folder called Aldus for Aldus Photostyler and copy the files from the Aldus discs into it. Under DOSBox, change to the (newly) mounted drive “C:” and make sure that the folder is also available. Then start Windows 3.1 in DOSBox and open the “File Manager” in the “Main Group” folder. The Office predecessor Microsoft Works (1988) included a word processor, a spreadsheet and a database and can still be installed under Windows 3.1 in DOSBox without any problems. It’s also easy to use.Foundry Find your “Aldus” folder in the “C:” drive, and then open it. Start the installation with the “pssetup.exe” file. Then close the file manager using the button of the same name and find Aldus Photostyler in its own folder. Double-click on the icon to start the application. Next, to install the Microsoft Works 2.0 image proceed in the same way as with the Windows installation: Download the ZIP file from winworldpc.com, unzip it into a previously created “Works” folder in your installation directory and unzip the eight disc images into it so that all the files are in the Works folder. Then remount the folder in DOSBox, start Windows 3.1 and open drive “C:” via the file manager as already described. In the “Works” folder, start the installation via the “setup.exe” file. Create disc images If you have software on original floppy discs and want to store their contents as securely as possible, you should create images of them. This is possible with the freeware Floppy Image, which you simply run via the exe file without installation. The program can create IMZ, IMG, and EXE files. You can create images of your floppy discs with the Floppy Image freeware. The old media are notorious for often causing difficulties when being read so this is a good way to protect your software and install it later.Foundry Click on “Browse” under “Read Floppy” in the program, specify the name, the storage location and the desired image format and click on “Save”. Now enter a description and start the process by clicking on “Start”. Read...

Welcome to The Full Nerd newsletter—your weekly dose of hardware talk from the enthusiasts at PCWorld. Missed the controversial topics on our YouTube show or hot news from across the web? You’re in the right place. Want this newsletter to come directly to your inbox? Sign up on our website!  I like owning things. The recent AMD driver kerfuffle has me wondering what that really means these days. I’m old enough to remember when buying a physical item meant you had it for life. It would run as advertised until it could no longer—perhaps a part would break, a motor would burn out. Even then, things were mechanical enough that you could open them up and tinker, in an effort to fix them. Now with so much software required to make hardware perform intended function, that line has become blurry. We depend on companies to keep products functional. We are expected to trust that a company will support a product for a reasonable amount of time. AMD’s confused messaging around driver support for RDNA 1 and RDNA 2 graphics cards shook that trust. At first, it seemed that GPUs just three years old were already getting the axe—Team Red would no longer provide updated drivers for new game optimizations. Then, after this stance was seemingly confirmed by AMD statements to the press (causing more furor), AMD finally issued a clarification via a blog post. The drivers would branch, but baseline support for first- and second-generation RDNA wouldn’t end yet. The Full Nerd crew and I discuss the whole episode in more depth in the show, including the nuances of the situation. The reputation of AMD Radeon’s division obviously plays a role in all this, as does the current environment of GPU prices and availability. But even with the resolution of this PR hiccup, the fundamental issue remains. Chromebook lifespans have improved, but that’s only for newer models. Own an older one and it’s likely EOL now. That sucks.IDG / Matthew Smith Companies can now kill their products at any time—even when you have them in your hands. Once software support gets dropped, it’s over. I have a drawer full of phones that Google and Apple no longer provide security patches for. Older Chromebooks got a similarly raw deal. You can argue that you can still technically use these products, and sure, that’s true. But in today’s environment, a lack of security patches is asking for a headache. No driver support for new games means you won’t get to play them. Etc. I’m grateful that people donate their time to alternative software, like LineageOS and Linux, to help keep perfectly usable hardware still chugging along. (I’m about to try this to breathe new life into a beloved, now-unsupported Chromebook.) But that doesn’t change the fact that we’re at the mercy of companies’ willingness to maintain a product or even whole product line. And I hate that. I’m willing to sign up for software-as-service. I think of it as renting tools. But my hardware? I bought it for its known properties. I bought it for its specific features. I bought it so it would keep doing what I needed it for. If it stops working because the software is gone, what did I actually own, then? In this episode of The Full Nerd  In this episode of The Full Nerd, Adam Patrick Murray, Brad Chacos, Alaina Yee, and Will Smith hash out their feelings about the weekend’s RDNA 1 and 2 driver support debacle and how many Windows PC games run Linux now. On the AMD front, Brad pulls out what he thinks AMD meant to say, but can’t. Meanwhile, I get fussy about chart interpretations and headlines. Not hatin’ on Linux, just the implied takeaway in the coverage. Oh, and I get to show everyone my latest fashion acquisition. Is it fashionable? No. Am I pleased that I had a good reason to own AMD-branded socks? Kind of yes. (Very yes.)(Full disclosure: The socks were a gift from one Adam Patrick Murray. I have a whole collection of various tech-branded ones from him, possibly because I’ve talked about buying socks during Black Friday so often over the years.) Willis Lai / Foundry Missed our live show? Subscribe now to The Full Nerd Network YouTube channel, and activate notifications. We also answer viewer questions in real-time! Don’t miss out on our NEW shows too—you can catch episodes of Dual Boot Diaries and The Full Nerd: Extra Edition now! And if you need more hardware talk during the rest of the week, come join our Discord community—it’s full of cool, laid-back nerds.  This week’s stirring nerd news Might have to start handling memory with this much care, given how expensive it is to replace any that goes bad or dies.IDG Thanks to the AMD Radeon hullabaloo, everything else feels relatively quiet—despite being noteworthy. AMD CPUs helped the company hit new records, for starters. And I got pretty invested in not one, but two different robot vacuum stories. Also, I realized I’ve failed to adequately sound the alarm when giving buying advice in the last week or so. Some component prices already hurt, and it got bad so fast. (Hopefully you don’t need more memory any time soon.) Proving a point: In stark contrast to the Radeon side of things, AMD’s Ryzen CPUs remain steady, propelling Team Red to loftier and loftier heights—desktop CPU share has climbed almost 10 percentage points since 2024, according to CEO Dr. Lisa Su. This is some bull$*(%: Not the engineering to bring the robot vacuum back online, but the manufacturer’s decision to remotely brick the device. (Also, this is why we use guest networks for IoT devices.) Filed under “Asked if they could, not if they should”: Look, I’m just as guilty of doing things just to see if it’s possible. I still snort-laughed about squeezing an operating system so small, it becomes unusable. Science + efficiency = I click: Humans creating better ways to deal (and reuse) the materials they create? Sounds good to me. How far the Internet’s come: From data dropouts that cause typos to memes that load in bare seconds. (Can’t say we’ve used the improvements for best impact, but that’s a thought for another day.) Venus’s mysteries will remain so to us for a while longer.Planet Volumes / Unsplash Enjoy your privacy, fair Venus: The loss of the last remaining satellite is sad for us. I’m telling myself a planet named for the goddess of love probably needs a break from prying eyes for a bit, though. I really want to know about the 5% of humans who failed: Were they distracted during Butter Bench? Was their native language different than how the benchmark was administered? Did they just not care? And yes, I did feel kind of bad about this robo vacuum’s internal meltdown. We’ve all been there, buddy. I needed this 15 years ago: A black & white mode for Google Maps to eke out that much more battery life from a dying phone? Better late than never, I guess. (Though it’s not officially real yet.) Ouch, that hurt faster than expected: Memory prices have shot up abruptly—and it’s not just DDR4 affected. If you need more RAM right now, gird yourself for as much as 100 percent (or more) price increases. I don’t think Black Friday will save us from this. I have a secret—I’m apparently one of the lone people who likes the “fall back” to Standard Time in the U.S. The early dark evenings make stew and soups feel perfect. (Definitely come share your favorite recipes with me in The Full Nerd Discord’s #food-chat channel.) Catch you all next week! ~Alaina This newsletter is dedicated to the memory of Gordon Mah Ung, founder and host of The Full Nerd, and executive editor of hardware at PCWorld. Read...

I’ve used Windows 11 regularly for a couple of years now. I still dislike it. But now we’re all stuck with it, given Windows 10’s “death” earlier this month. What I hate in particular: The sound settings on the Windows taskbar.  In Windows 10, when you clicked, you could immediately see a name for your output device. I could tell at a glance what the active speakers were. Most people don’t have as many connected devices I do, but even when I traveled with just a laptop and a pair of headphones, this info was helpful. In Windows 11, you only see the volume slider—to get the name of the active speaker, you have to click a button. This extra step slows me down, which is the part I dislike. Having to scroll through a bunch of device names (some of which I’ll never use as speakers) drags out the process even more. I can’t fix the interface, but I can hide the excess audio devices. So that’s what I did. How to disable a sound device in Windows There’s a lot going on here.PCWorld To make an audio device disappear as a sound option, you have to disable it in Windows. Open the Settings app. Click on System > Sound. (Or search for ‘Sound’) Click on the audio device you want to disable. Under General > Audio, click on the Don’t Allow button. You can choose to disable both output and input devices. Output is what plays sound from your PC, like internal or external speakers, Bluetooth headphones and earbuds, etc. Input is what records audio from you or your environment—basically, microphones. These can be webcams, dedicated microphones, etc. On these output and input lists, you may see peripherals and devices you might not expect like your monitor or a capture device. For me, it was two monitors and a CamLink 4K capture dongle that was cluttering my audio options. So I disabled them all, both as output and input devices. After a massive spree of disabling, I now have a much cleaner, clearer set of audio options.PCWorld For good measure, I also disabled the audio for my webcam. I prefer my dedicated mic and this reduces the ability for compromised or rogue software on my PC to immediately hijack its use. If you ever accidentally disable the wrong sound device, just go to System > Sound, scroll down to All sound devices, and then choose the one you want to re-enable. Note: If you disconnect a device (like a monitor) and plug it into another port, you may cause Windows to reactivate it as an audio source. You’ll have to disable it again if this happens. This “fix” doesn’t eliminate all of my crankiness about Windows 11 and its interface nonsense. But it reduces the friction. Read...

If you’ve been tracking AMD’s market share in the PC CPU market, you know that it’s continually climbed over the past year or so. Chief executive Dr. Lisa Su told Wall Street that she expects that trend to continue, even after the company’s desktop CPU business reached record highs during the third quarter of 2025. AMD’s Client and Gaming segment of the business sold $4 billion worth of processors, CPUs, and semi-custom chips for the console business during the third quarter, but was slightly surpassed by a record $4.3 billion from the data center. Given the ravenous appetite for AI, its efforts in the data center dominated the company’s third-quarter 2025 earnings call — analysts didn’t ask a single question specifically about its PC business. That’s somewhat surprising, given AMD’s continued success in the PC market. The company’s client business, made up of its CPU sales, recorded a record $2.8 billion. “Our PC processor business is performing exceptionally well, with record quarterly sales as the strong demand environment and breadth of our leadership Ryzen portfolio accelerates growth,” Su told analysts in a transcript provided by Investing.com. “Desktop CPU sales reached an all-time high, with record channel sell-in and sell-out led by robust demand for our Ryzen 9000 processors, which deliver unmatched performance across gaming, productivity, and content creation applications.” AMD’s desktop CPU share has gained almost 10 percentage points since last year, and Su said she expects that trend to continue. “Looking ahead, we see significant opportunity to continue growing our client business faster than the overall PC market, based on the strength of our Ryzen portfolio, broader platform coverage, and expanded go-to-market investments,” she said. Interestingly, Ryzen notebooks even “increased sharply” during the quarter, Su said. That’s surprising, given that Intel has maintained the typical 80-20 split between AMD and Intel, especially in the notebook business. If the chipmaker is cutting into that, Intel will face a significant attack on multiple fronts. AMD said less about its efforts in the GPU business, though it noted that revenue and channel sell-out grew significantly, driven by the Radeon 9000. The company’s FSR4 technology, which uses upscaling and AI frame generation within the RX 9000 generation to improve frame rates, now supports more than 85 games. AMD also saw a boost from its semi-custom business as Microsoft and Sony ramp up console production for the holiday sales season. That’s a double-edged sword, as the company’s outlook includes a decline in “strong double digits” in gaming revenue, presumably reflecting a lull after the holiday console sales boom. AMD’s second-quarter revenue increased 36 percent year over year to $9.2 billion, with net income up 31 percent to $1.243 billion. Read...

TechHive Editors Choice At a glanceExpert`s Rating Pros Gemini is a game-changer Excellent high-res video with 8x digital zoom Camera has a very wide field of view No subscription required for limited video recordings and person, pet, vehicle, and package detection Cons No battery option (not even in a different model) Familiar-face recognition requires at least a $10/mo subscription Gemini doubles the price of a subscription Google has a limited home security ecosystem Our Verdict The Google Nest Doorbell (wired, 3rd gen) is an excellent home security device, and you don’t need to pay for a subscription to get a limited number of event recordings. Gemini, meanwhile, is nothing short of amazing, but you will need an expensive subscription to avail yourself of it. Price When Reviewed This value will show the geolocated pricing text for product undefined Best Pricing Today Best Prices Today: Google Nest Doorbell (wired, 3rd gen) Retailer Price $179.99 View Deal Price comparison from over 24,000 stores worldwide Product Price Price comparison from Backmarket Best Prices Today: Check today’s prices Prior to announcing its third-generation Nest Doorbell—along with its other two new security cameras in early October—Google had us thinking it had lost interest in the smart home. While I wouldn’t say this device marks a roaring comeback, it is a very good product; provided you’re willing to pay at least $10 a month ($100 per year) for a Google Home Premium Standard subscription in addition to paying $180 for the device itself. Like a distressingly large number of smart home and home security devices, the Google Nest Doorbell (wired, 3rd Gen) just isn’t all that useful without a subscription. Yeah, you’ll be notified when someone rings the bell; you’ll get person, pet, vehicle notifications, as well as package notifications when a courier makes a delivery; and you’ll get a live video stream so you can see and have a conversation with whoever’s at the door. Google’s Gemini is a game-changer, and the new Nest Doorbell’s video resolution; person, pet, vehicle, and package recognition; and other features are best-in-class. But Google does Ring and some other competitors one better by providing up to six hours of 10-second event video previews without a subscription (Ring doesn’t give you any recordings without a subscription). That said, if you want all the slick features Google touted in its announcement demos—its Gemini AI, in particular—you’ll need to step up to the $20-per-month ($200 per year) level of Google Home Premium Advanced. I’ll go over those details more in a bit. Specifications Before we get into those more advanced features, let’s discuss what you get for the price of the product itself. Google’s latest doorbell is taller than it is wide, measuring 5.2 x 1.7 x 1.1 inches (HxWxD). It’s available in three colors: Hazel (beige), Linen (off white), and Snow (white). Google sent the last one for our review. The Nest Doorbell (wired, 3rd gen) features excellent facial recognition, even when a visitor’s face is in low light or is obscured by their clothing.Michael Brown/Foundry The camera is at the top of the device, and there’s a large round doorbell button at the bottom. A proximity sensor sets an LED light ring around this button to blinking when someone approaches, and pressing it sounds chime on the doorbell itself (and in the house if you’ve connected it to a chime, as I’ll get to later). The camera sensor captures video in 2K resolution with a 1:1 aspect ratio (so, 2048 x 2048 pixels). It captures a head-to-toe view of your visitors, and 6x digital zoom enables you to get a very detailed view of them when you pinch-to-zoom in either real time or with a recording (recordings are captured at 30 frames per second). The Nest Video Doorbell (wired, 3rd gen) comes in a complete kit, including a module for your existing chime and a decal to advise visitors they’re subject to video and audio surveillance.Michael Brown/Foundry Two-way audio with noise cancellation lets you converse with visitors, or you can play canned greetings if you’re using the live view feature while you’re away from home or if you just don’t want to interact with them for whatever reason. There are currently only three available phrases: “You can just leave it;” “We’ll be right there,” and “No one can come to the door.”  You can’t record your own canned messages. The camera itself has a wide 166-degree diagonal field of view that results in very slight barrel distortion. A set of six 850nm infrared LEDs provide black-and-white night vision up to 10 feet, but with any supplementary illumination, you’ll get full color at night. HDR support delivers accurate color reproduction. If you live in a very cold-weather climate, you’ll want to be aware of the doorbell’s operating temperature range of -4 degrees Fahrenheit to 104 F (-20 to 40 C). It is otherwise well protected from the elements, with an IP65 rating. As our IP code guide tells us, this means it’s impervious to dust and protected from water jets coming from any direction. So, you could hit it with a spray nozzle attached to a garden hose, but it probably wouldn’t withstand a blast from a pressure washer. Google provides a tiny hex-head screw to secure the doorbell to its mount, but its size and the location of the threaded hole makes using it a challenge.Michael Brown/Foundry Google provides everything you’ll need to install the doorbell, including a base plate, spacers for installing the doorbell on an uneven surface, a wedge that will shift the doorbell’s view 20 degrees left or right, and a hex key and small set screw for securing the doorbell to its mount (I promptly dropped this tiny screw during my installation, where it rolled into oblivion between the planks on my deck). You’ll also find the other screws you’ll need to mount the doorbell, a pigtail wire for connecting the doorbell to a power supply, and a puck-shaped connector for your existing doorbell chime. Being a wired product, you’ll need to connect the Nest Doorbell (wired, 3rd Gen) to a 16-24VAC or 10-40VA transformer. If you’re replacing an existing wired doorbell, you should already have an appropriate transformer installed, so you can just swap your old doorbell for this new one. The Nest Doorbell can also use your existing mechanical or electronic chime (provided it meets the same power requirements, of course). Google includes a small puck you’ll attach to the chime box for this purpose, but I didn’t use it because I don’t have a chime. Installation and setup Google is to be commended for its excellent illustrated, step-by-step in-app instructions that guide you through the installation process. They cover everything from turning the power off to your existing doorbell, to disconnecting the existing wires, uninstalling the old doorbell, connecting the wiring to the doorbell (the provided pigtail makes this step particularly easy), and mounting the doorbell to the wall. The Google app has one of the best installation routines I’ve encountered, showing every step in detail and showing you what to expect at every stage.Michael Brown/Foundry Once you have the doorbell’s mount and wedge (if you use it) installed and the doorbell connected to power, the app will have you scan a QR code on the back of the doorbell to add it to your smart home environment. After a few moments, the app will ask where you’ll be using the doorbell (presenting a list of rooms) and then proceed to download and install the latest firmware. When it’s finished, you should see a live view from the camera. After a long pitch for Google Home Premium, and a few other rudimentary steps, the app will ask if you want to turn on the doorbell’s video history. I don’t know why you wouldn’t, unless as the app points out, doing so would run afoul of local laws. If you want—or need—to warn visitors that a camera is recording audio and video, Google provides a decal that you apply to a nearby wall or window. The screw terminals on the back of the doorbell and the provided pigtail wire ensure a secure connection to your doorbell wiring. Michael Brown/Foundry The Nest Doorbell (wired, 3rd gen) comes with a 30-day free trial of Google Home Premium (you can choose either the Standard or Advanced version). The Standard tier includes 30 days of event-based video history for all the Nest cameras in your home; alerts for familiar faces, garage door status, and the sound of glass breaking as well as the sound of smoke and carbon monoxide detectors sounding off; and an emergency calling feature that lets you call the local-to-your-home 911 dispatcher for emergency help, even when you’re away from home. Your subscription also includes Gemini for Home, which you can use with your Nest smart speakers and displays. You can also create smart home routines by typing out what you want to accomplish in the Google Home app. I wouldn’t say installation is foolproof, but scanning a QR code to add the doorbell to the Google Home app sure makes things easier.Michael Brown/Foundry The Advanced tier includes all those features, but it doubles the event-based video history to 60 days,  adds 10 days of 24/7 video recording (useful in the  event there  wasn’t enough motion to trigger the camera to record a clip), provides “descriptive” notifications that provide context to the camera’s alerts, the ability to search the event history for relevant recordings, and daily summaries that recap all the activity that occurred in your Nest cameras’ field of view on a given day.   That’s a lot of extra features, but $20 per month is a lot of money, especially when it doesn’t include professional monitoring for a complete home security system, which Google no longer offers. It’s worth noting here that Google invested $450 million in ADT in 2020, and ADT bundles Nest doorbells and security cameras with its ADT Smart Home Security System, which I reviewed in October. Mounting the Nest Doorbell (wired, 3rd gen) to its wedge mount. (This is a temporary installation, so the mount is on top of the wire instead of the wire coming out of the wall).Michael Brown/Foundry For the sake of comparison, Ring’s middle subscription tier for its video doorbells and security cameras costs $10 per month, and if you also have a Ring Alarm home security system, you can add professional monitoring that will dispatch first responders (police, fire, or medical) in an emergency for an additional $10 per month ($20 per month in total). To be fair, Ring’s top-tier plan, which includes AI for its cameras, plus professional monitoring costs $30 per month. And Ring charges an additional $5 per month for a similar smoke alarm/carbon monoxide listening feature (you can get that by itself or on top of any Ring service tier). Using the Nest Video Doorbell (wired, 3rd gen) The new Nest Video Doorbell’s ability to recognize familiar faces is impressive, within reason. Once I’d trained it with one example of each person who’s at my home with any regularity, it almost never failed to accurately identify those faces when they came within range of camera. When it correctly identified my grandson’s girlfriend while she was wearing a hoodie, I tried it myself in similar attire, and it correctly identified me as well. But when I tried it a second time, after cinching up my hoodie until I looked like Kenny from South Park, the doorbell merely identified me as a person. More importantly, a human looking at the recording would have no trouble picking me out of a lineup. Google’s AI certainly isn’t foolproof, however; it routinely mid-identified my 10-pound Chihuahua walking across my deck as a cat, and it once identified a cat walking in my neighbor’s yard as walking on a fence railing. But if you take a quick glance at the zoomed-in screenshot on this page, you might make the same mistake. Google’s Gemini AI is very impressive, aside from identifying my little dog as a cat and interpreting this cat in my neighbor’s yard as walking along the fence rail.Michael Brown/Foundry Google Gemini can also summarize the activity logged by all its cameras into a daily Home Brief, followed by a time-stamped list thumbnails of the videos that were recorded. If you’re looking for a specific event, you can quiz Gemini by typing questions in the app. You can do the same with voice commands if you have a Nest smart display—I don’t—but that doesn’t seem to the case with a smartphone, or at least not with the iPhone I use. And if you try it with the existing Nest smart speakers, Google Assistant will just reply “Sorry, I don’t know where to play the video. Please tell me the exact name of the screen.” You’ll get motion-detection and doorbell-press notifications almost instantly, but it can take several minutes for events to appear in the app’s log. What’s missing? The most obvious missing element of the Nest Video Doorbell (wired, 3rd gen) is right there in its name: There’s no battery. If you don’t have an existing doorbell, you’ll need to install a transformer and pull wire through your walls to the location you want to mount the doorbell. That’s not a difficult task, and if you buy a plug-in transformer, you won’t need to need to deal with high-voltage wiring or hire an electrician. A bigger issue is that if you don’t have an existing doorbell chime, or if the one you have isn’t compatible with the Nest Video Doorbell, you’ll need to rely on your phone to be notified when someone rings the bell. While you could install your own—especially if you’re installing a transformer anyway—Ring and many other Google competitors have remote chimes that work over Wi-Fi. This could be an opening for a third-party vendor, but I couldn’t find any that were compatible with this 3rd-generation device (there are several that are compatible with 1st– and 2nd-generation Nest doorbells. And there are a couple of features that Eufy security cameras have that I’d love to see Google copy. First and foremost is the ability to track a person entering and exiting the fields of view of multiple outdoor cameras, and then have those feeds stitched together into a single continuous video. Eufy’s cameras can also control the behavior of Eufy’s smart outdoor lights, so that the presence of a stranger can trigger its colored lights to switch to a bright white or even start flashing as a warning. It might be possible to achieve that second behavior with Google’s new doorbell thanks to its Matter support, but I didn’t explore that. It’s also worth repeating that Google doesn’t offer its own comprehensive home security system, and while you can use its cameras in ADT’s system, they’ll be managed by ADT, not Google, and you won’t get Gemini. Should you buy the Google Nest Doorbell (wired, 3rd gen)? If your smart home revolves around Google Home, and you have the required infrastructure or are willing to install it, the Nest Doorbell (wired, 3rd gen) is an easy recommendation. Gemini is impressive enough to be a game-changer, and the camera’s video resolution, person, pet, vehicle, and package recognition, and the rest of doorbell’s many other features are excellent. If you’re not already wedded to a home security ecosystem, or if you’re only looking for a video doorbell, the latest Nest doorbell is among the best on the market, and it’s priced accordingly. It’s not worth paying $10 a month—$20 a month for Gemini—if you only have the doorbell, but that calculation changes if you’re deploying other Nest cameras as well. It’s too early to say that Gemini is worth dumping whatever other home security ecosystem you might have in favor of Google Home. The competition in this space is fierce, and lots of companies make great video doorbells: Ring, of course, but also Aqara, Blink, Eufy, Lorex, Reolink, Tapo, and Wyze Labs. But in terms of AI, Google and Ring are leagues ahead of the competition (we haven’t had an opportunity to review Ring’s latest offerings). All that said, my biggest reservation about recommending the latest Nest cameras overall is that Google no longer has a comprehensive system with door/window and motion sensors, alarms, smoke and CO detectors, and professional monitoring. So, don’t buy this product if those things are important to you. But my bottom-line recommendation of the Ring Doorbell (wired, 3rd gen) assumes you just want a great video doorbell and those other considerations aren’t important to you. The fact that don’t need a subscription to get a limited number of event recordings is a big plus; and if you’re willing to pay for it, Gemini is pretty amazing. This review is part of TechHive’s in-depth coverage of the best video doorbells. Read...

Opponents say the fast-track process rides roughshod over the environment and the local community. Read...

Microsoft is now making Mixed Reality Link available to all Windows 11 users with Meta Quest headsets, reports Engadget. The feature has been tested since 2024 and allows users to work in a virtual desktop with multiple large monitors in a VR environment. The experience is similar to that offered by the Apple Vision Pro, but the Meta Quest headsets are significantly cheaper. For example, the Meta Quest 3S costs around a tenth of the Apple Vision Pro. Mixed Reality Link can be connected both to a local Windows computer and to cloud-based work environments such as Windows 365 Cloud PC, Azure Virtual Desktop, and others. Read...

Imagine this: You pull up to your driveway and your garage door opens like magic. You head for your doorstep and walk right through the front door, no keys needed. Already, your favorite playlist is queued up on your smart speakers, and the room is set to the perfect temperature—all without saying a word to a smart assistant or pulling out your phone.  Those feats are possible thanks to a smart home feature called geofencing: that is, setting up a geolocated virtual perimeter or “fence” around your home, your office, or anywhere else you choose. When you cross the boundary—either coming or going—you can set your smart home to do something, usually by triggering an automation or routine.  There’s nothing new about geofencing—all the major smart platforms, from Alexa and Apple Home to Google Home and Samsung SmartThings—have offered geofencing functionality for years, as do many other major smart home ecosystems, including Philips Hue (for smart lighting), Chamberlain (for garage door controllers), and Ring (for smart home security).  Yet geofencing remains a feature that many of us overlook. Heck, I’m supposedly a smart home expert, and I admit to not giving geofencing much of a chance.   Luckily, getting started with a geofence is easy; what’s a little harder is wrapping your head around the possibilities.   Read on for some clever geofencing smart routines to try; this handful of examples may spark ideas for yet more location-based routines.  After that, I’ll give you a quick primer on how to get started with geofencing—it’s actually not that difficult.  Open your garage door like magic Chamberlain myQ Smart Garage Hub (model MYQ-G0401) Read our review Best Prices Today: $29.98 at Amazon I mentioned this geofencing trick at the jump, and it’s a neat one; having your smart home open your garage door automatically as you approach your driveway.  The way it works is simple: Using either your smart garage door opener’s app (such as the Chamberlain app for its MyQ garage door controller line) or, say, an Alexa routine, you define a circular perimeter–a geofence–around your home, perhaps the size of a city block.   When you cross the border of the geofence, your arrival will trigger the routine, opening your garage door just as you’re driving up.  Related: The best smart garage door openers Adjust your thermostat with ease Nest Learning Thermostat (4th gen) There’s nothing like coming home to a cool environment during the hot summer months, or to a warm, cozy living room in the dead of winter. But even if you already have thermostat automations that tweak the temperature according to the time of day, you should consider location-based automations that adjust the mercury dynamically as you come and go. For this example, you might want to set the geofencing radius a little wider—say, to an area that’s about a half-hour’s drive to your home, giving your A/C or heating enough time to bring the temperature in your house to the perfect level as you walk inside.  Same goes for when you leave your home. If you cross the geofencing border going the other way, your thermostat might switch to a power-saving routine while you’re gone.   Bonus tip: Most geofencing routines allow for triggering conditions such as “Only do (such-and-such) if no one is home”; that way, your smart home won’t turn change the target temperature if you head out but someone else is staying behind. The best smart thermostats have motion sensors–on the device itself and with some models, sensors you can deploy in multiple rooms–that can determine occupancy, so it doesn’t matter if not everyone has a cellphone. Arm and disarm your security system  Ring Alarm Pro (8-piece kit) Read our review Best Prices Today: $149.99 at Amazon What’s the point of having a smart security system if you forget to arm it when you leave? Good question.  Here’s where geofencing comes into play. Similar to what we just did with the thermostat, you can set up a geofence that automatically arms your security system when you leave home, or disarms it when you return (so no more fumbling with a keypad).   Geofencing is a prominent feature in the Ring app, and most other smart security manufactures offer similar functionality for their alarm systems.  Related: The best smart home alarms and systems Lock and unlock your front door  Level Lock Pro Read our review Just as you can use a geofence to arm and disarm your front door, so can you use it to lock and unlock your home’s smart lock, saving you from the trouble of fishing out your keys, pulling out your phone, or remembering a PIN.  In this scenario, you might want a fairly tight geofence radius to help ensure your smart lock stays locked unless you’re in close proximity; ditto for when you’re heading out the door.  Related: The best smart locks Turn your lights on and off  Philips Hue Festavia string lights Like coming home to a warm, cozy lighting scene? You can easily set your smart lights to turn on just the way you like as you’re arriving home—or have them turn off whenever you leave.  The Philips Hue app, for example, has a location-based automation trigger that will kick off your favorite lighting routine as you leave or enter a geofenced zone. You can also set the automation to run at any time of day, or only at night.  Related: The best smart lights Drop the needle on a music playlist  Amazon Echo Spot (2024) Read our review Best Prices Today: $54.99 at Amazon Aside from being greeted by the perfect lighting scheme when you come home, you might also want to hear your favorite tunes, too.  You can easily set a smart home app like Alexa or Google Home to detect when you’ve come home and then play the music of your choice on one or more selected smart speakers.   Related: The best smart speakers How to create a geofencing routine Intimidated by the idea of setting up a location-based automation? Don’t be; they’re easy.  Here’s an example using the Alexa app; this should help guide the way for the geofencing functionality in other smart home setups.  The translucent blue circle around the location is the geofenced area; tap and drag the little blue dot on the edge to expand or contract the perimeter.Ben Patterson/Foundry Open the Alexa app, tap the “+” button, then scroll down and tap Routines to create a new Alexa automation.  Now, we’ll create a location-based trigger. Under the When section, tap Add an event, then tap Location.   Next, enter an address or tap Current location; when you do, a map will appear showing the location you picked—such as, say, your home. The translucent blue circle around the location is the geofenced area; tap and drag the little blue dot on the edge to expand or contract the zone.   At the top of the screen are two tabs; Arrives and Leaves. This setting determines whether the routine gets triggered when you either arrive at the location or leave it. For this example, we’ll choose Arrives. When you’re ready, tap Next, then give your geofenced location a name.  Now, you’ll land back at the main Routine screen, where you can choose what happens when the automation gets triggered by your geofencing conditions. Under the Alexa Will section, tap Add an action. For this example, we’ll pick Music & Podcasts. You can then pick a song, an artist, a podcast, or a music station from the music-streaming source of your choice. I’ll go ahead and choose Taylor Swift on Apple Music.  Back at the main Routines screen, tap Choose Device under the Hear Alexa from section, then pick the Alexa speaker where you want to hear the music play; I’ll chose the Echo Dot in my kitchen. Tap Save.  All set! Now, whenever I walk through the front door, my favorite Taylor tunes will begin playing on my kitchen Echo—the perfect way to come home.  This article is part of TechHive’s in-depth coverage of the best smart home systems. Read...