Meta has been going all in on AI, whether people want it or not, and now it’s bringing more features in that vein to Facebook. The network’s latest move is to let people use Meta AI to animate their profile photos. Because what better way to express your individuality than to use a pre-canned AI-generated animation on your own face?

Meta AI is also coming for your Facebook Stories and Memories. The network’s Restyle lets you use gen-AI to change up the aesthetic of your posts. You can once again use pre-canned stylings or give the AI assistant your own prompt.

In the company’s own words, the new tools that will create “share-worthy moments that spark meaningful interactions and conversations with friends.” I guess meaning is in the eye of the beholder. If you’re desperate to behold even more AI slop, Meta recently said its Vibes feed of exactly that content will be getting a standalone app.

The latest generation of Android flagships from Vivo, Oppo, and even Samsung include one 200MP sensor, used as the primary camera or the telephoto camera. However, the next generation of Android flagships could include three 100MP sensors.

You heard that right. According to Chinese tipster Digital Chat Station, “some” (that could be more than one) smartphone makers are “testing three 100-megapixel lenses” or cameras.

Three 100MP cameras? That’s the rumor

Although the tipster doesn’t specify the nature of these cameras, they could very well be the primary, telephoto, and ultrawide shooters. This is one of the most interesting approaches I’ve heard of lately. Here’s why.

It might sound like a 100MP primary sensor and a 100MP telephoto sensor is a downgrade from the current 200MP standard at first. But only the Oppo Find X9 Ultra has been confirmed to feature two 200MP cameras on the back.

The others, including the Vivo X300 Pro, the Find X9 Pro, and the Galaxy S25 Ultra (or even the upcoming Galaxy S26 Ultra, for that matter), include only one 200MP primary sensor, which, by the way, uses pixel binning to default to a lower resolution for increased low-light performance, saving space and capture time.

Having 100MP primary and telephoto sensors would still allow brands to capture at a lower default resolution, upscale significantly when needed, and take less storage or capture time than a 200MP sensor would.

The ultrawide camera could finally get a serious update

So far, I haven’t talked about the ultrawide sensor, because it maxes out at 50MP on the flagships. Hence, a 100MP ultrawide camera (if our interpretation of the tweet is right) would be a dramatic upgrade.

It could enable more detailed macro shots (if the sensor doubles as a macro shooter) or greater post-capture reframing potential. In addition to the rear-facing cameras, the leaker also claims that a “100-megapixel front-facing camera,” with a “small-pixel sensor,” is in the works.

Given that the Galaxy S26 series seems to be stuck with a 12MP front camera, and Chinese flagships use a 50MP sensor on their most expensive variants, a 100MP selfie shooter could deliver a noticeable upgrade.

Since it’s a small-pixel sensor, low-light photography might be an issue, but I guess smartphone makers should be able to fix it with computational trickery.

A fake 7-Zip website is distributing a trojanized installer of the popular archiving tool that turns the user’s computer into a residential proxy node.

Residential proxy networks use home user devices to route traffic with the goal of evading blocks and performing various malicious activities such as credential stuffing, phishing, and malware distribution.

The new campaign became better known after a user reported that they downloaded a malicious installer from a website impersonating the 7-Zip project while following instructions in a YouTube tutorial on building a PC system. BleepingComputer can confirm that the malicious website, 7zip[.]com, is still live.

The threat actor registered the domain 7zip[.]com (still live at the time of writing) that can easily trick users into thinking they landed on the site of the legitimate tool.

Furthermore, the attacker copied the text and mimicked the structure of the original 7-Zip website located at 7-zip.org.

The installer file was analyzed by researchers at cybersecurity company Malwarebytes, who found that it is digitally signed with a now-revoked certificate originally issued to Jozeal Network Technology Co., Limited.

The malicious copy also contains the 7-Zip program, thus providing the regular functions of the tool. However, the installer drops three malicious files:

1. Uphero.exe – service manager and update loader
2. hero.exe – main proxy payload
3. hero.dll – support library

These files are placed in the ‘C:\Windows\SysWOW64\hero\’ directory, and an auto-start Windows service running as SYSTEM is created for the two malicious executables.

Additionally, firewall rules are modified using ‘netsh’ to allow the binaries to establish inbound and outbound connections.

Eventually, the host system is profiled with Microsoft’s Windows Management Instrumentation (WMI) and Windows APIs to determine the hardware, memory, CPU, disk, and network characteristics. The collected data is then sent to ‘iplogger[.]org.’

“While initial indicators suggested backdoor‑style capabilities, further analysis revealed that the malware’s primary function is proxyware,” Malwarebytes explains about the malware’s operational goal.

“The infected host is enrolled as a residential proxy node, allowing third parties to route traffic through the victim’s IP address.”

According to the analysis, hero.exe pulls config from rotating “smshero”-themed C2 domains, then opens outbound proxy connections on non-standard ports such as 1000 and 1002. Control messages are obfuscated using a lightweight XOR key.

Malwarebytes found that the campaign is larger than the 7-Zip lure and also uses trojanized installers for HolaVPN, TikTok, WhatsApp, and Wire VPN.

The malware uses a rotating C2 infrastructure built around hero/smshero domains, with traffic going through the Cloudflare infrastructure and carried over TLS-encrypted HTTPs.

It also relies on DNS-over-HTTPS via Google’s resolver, which reduces visibility for defenders monitoring standard DNS traffic.

The malware also checks for virtualization platforms such as VMware, VirtualBox, QEMU, Parallels, as well as for debuggers, to identify when it’s being analyzed.

Malwarebytes’ investigation started after noticing research from independent security researchers who analyzed the malware and uncovered its true purpose. Researcher Luke Acha discovered the purpose of the Uphero/hero malware.

The xor-based communication protocol was reverse-engineered and decoded by s1dhy, who confirmed the proxy behavior. Digital forensics and incident response (DFIR) engineer Andrew Danis connected the fake 7-Zip installer to the larger campaign impersonating multiple software brands.

Malwarebytes lists indicators of compromise (domains, file paths, IP addresses) and host-related data observed during their analysis.

Users are recommended to avoid following URLs from YouTube videos or promoted search results, and instead bookmark the download portal domains for the software they use often.

Modern IT infrastructure moves faster than manual workflows can handle.

In this new Tines guide, learn how your team can reduce hidden manual delays, improve reliability through automated response, and build and scale intelligent workflows on top of tools you already use.

Get the guide

There’s been a lot of pushback on electronic door handles lately, as multiple carmakers — especially Tesla — have been accused of making manual door releases too hard to find and access during an emergency. Rivian is one of the companies that reportedly decided to change this on its upcoming R2 SUV, and a spate of first-look videos released Tuesday finally give us a look at what the company has changed.

First off, the front doors open from the inside in the same way as in the existing R1 vehicles. There is an electronic button that opens the door, and there’s a manual door-release latch tucked into the front part of the interior handle.

The rear doors also have an electronic button, as well as a change to the rear manual release.

On R1 vehicles, passengers have to first pull a panel off the door to access a “release cord” that operates the manual latch. On the new R2 SUV, Rivian moved this release cord to that same front-of-the-handle position as the front seat manual releases — though it’s still tucked behind a piece of plastic that must be popped out, making it slightly harder to access than the front door manual releases.

The R2 SUV isn’t set to go into production for another few months, so the company has not put out proper instructions on how to access this release. But here’s an image from a new video published by JerryRigEverything’s Zack Nelson:

He doesn’t pull out the actual cord, but it’s the best illustration I’ve seen so far of what passengers will need to do if they are in an R2 that has lost power for whatever reason, limiting the vehicle’s electronic door release.

The manual release is still behind a piece of plastic, and it’s not the most obvious or accessible way to open a door from the inside. But it’s at least in a more logical spot than just being hidden behind a panel.

These kinds of situations don’t happen often. But when they do, it’s typically during a major crash. That means every moment can be the difference between life and death.

Rivian is not alone in reworking how hard it is to access the manual releases. The most high profile example is Tesla. Bloomberg News found at least 15 deaths in crashes where there is evidence that occupants (or rescuers) were unable to open the doors. The company has said it will change how it designs its handles in order to address the problem.

And electronic door latches can present other issues. Last year, Ford had to issue a recall to fix a power-delivery problem for the electronic latches on the Mustang Mach-E.

If you’ve been thinking about documenting your life, whether it’s the exciting or mundane stuff, this might just be your moment. This GoPro Hero 13 Black bundle includes a variety of useful accessories, as well as a full suite of interchangeable lenses. It’s currently marked down to just $550 on Amazon, a $200 discount from its usual price.

GoPro

Hero 13 Black – HB Series Lens Collection

The biggest change to this generation of GoPro action cameras is the interchangeable lens system, and this kit comes loaded with basically every lens an aspiring filmmaker could ask for. That includes the ultrawide lens mod, which boosts the field of view to 177 degrees, an anamorphic lens and filters, and a macro lens for beautiful close-up shots. While it doesn’t come with a ton of attachments, it does have a case for carrying everything, and basic adhesives to get you started.

The downside here is that GoPro is still using the same sensor and processor as previous models, for better or worse. It’s one of the highest resolution and frame rate offerings you can get, with its 27-megapixel sensor producing up to 5.3K video, and up to 120 fps, although only for five seconds at a time at the highest resolution. Unfortunately, this GoPro, like others before it, still struggles in dim lighting. That said, there are some improvements to the HDR that our reviewer Scott Gilbertson said made a big difference when it came time to start editing.

Even with an upgraded battery, the stamina can be lackluster, lasting just one or two hours depending on how warm the camera gets. Thankfully, a new pass-through charging port allows you to hook up a power bank and keep filming for much longer. There are some other quality-of-life upgrades too, like a magnetic mounting system that makes swapping from surf to sand even easier.

The Hero 13 Black is our favorite GoPro, but there’s a whole world of action cameras for your next adventure, so make sure to check out our full guide if you’re curious about other options. Otherwise, this big bundle includes everything you need to get out and start shooting for just $550.

If it feels these days as if everything in technology is about AI, that’s because it is. And nowhere is that more true than in the market for computer memory. Demand, and profitability, for the type of DRAM used to feed GPUs and other accelerators in AI data centers is so huge that it’s diverting away supply of memory for other uses and causing prices to skyrocket. According to Counterpoint Research, DRAM prices have risen 80-90 precent so far this quarter.

The largest AI hardware companies say they have secured their chips out as far as 2028, but that leaves everybody else—makers of PCs, consumer gizmos, and everything else that needs to temporarily store a billion bits—scrambling to deal with scarce supply and inflated prices.

How did the electronics industry get into this mess, and more importantly, how will it get out? IEEE Spectrum asked economists and memory experts to explain. They say today’s situation is the result of a collision between the DRAM industry’s historic boom and bust cycle and an AI hardware infrastructure build-out that’s without precedent in its scale. And, barring some major collapse in the AI sector, it will take years for new capacity and new technology to bring supply in line with demand. Prices might stay high even then.

To understand both ends of the tale, you need to know the main culprit in the supply and demand swing, high-bandwidth memory, or HBM.

What is HBM?

HBM is the DRAM industry’s attempt to short-circuit the slowing pace of Moore’s Law by using 3D chip packaging technology. Each HBM chip is made up of as many as 12 thinned-down DRAM chips called dies. Each die contains a number of vertical connections called through silicon vias (TSVs). The dies are piled atop each other and connected by arrays of microscopic solder balls aligned to the TSVs. This DRAM tower—well, at about 750 micrometers thick, it’s more of a brutalist office-block than a tower—is then stacked atop what’s called the base die, which shuttles bits between the memory dies and the processor.

This complex piece of technology is then set within a millimeter of a GPU or other AI accelerator, to which it is linked by as many as 2,048 micrometer-scale connections. HBMs are attached on two sides of the processor, and the GPU and memory are packaged together as a single unit.

The idea behind such a tight, highly-connected squeeze with the GPU is to knock down what’s called the memory wall. That’s the barrier in energy and time of bringing the terabytes per second of data needed to run large language models into the GPU. Memory bandwidth is a key limiter to how fast LLMs can run.

As a technology, HBM has been around for more than 10 years, and DRAM makers have been busy boosting its capability.

As the size of AI models has grown, so has HBM’s importance to the GPU. But that’s come at a cost. SemiAnalysis estimates that HBM generally costs three times as much as other types of memory and constitutes 50 percent or more of the cost of the packaged GPU.

Origins of the memory chip shortage

Memory and storage industry watchers agree that DRAM is a highly cyclical industry with huge booms and devastating busts. With new fabs costing US $15 billion or more, firms are extremely reluctant to expand and may only have the cash to do so during boom times, explains Thomas Coughlin, a storage and memory expert and president of Coughlin Associates. But building such a fab and getting it up and running can take 18 months or more, practically ensuring that new capacity arrives well past the initial surge in demand, flooding the market and depressing prices.

The origins of today’s cycle, says Coughlin, go all the way back to the chip supply panic surrounding the COVID-19 pandemic . To avoid supply-chain stumbles and support the rapid shift to remote work, hyperscalers—data center giants like Amazon, Google, and Microsoft—bought up huge inventories of memory and storage, boosting prices, he notes.

But then supply became more regular and data center expansion fell off in 2022, causing memory and storage prices to plummet. This recession continued into 2023, and even resulted in big memory and storage companies such as Samsung cutting production by 50 percent to try and keep prices from going below the costs of manufacturing, says Coughlin. It was a rare and fairly desperate move, because companies typically have to run plants at full capacity just to earn back their value.

After a recovery began in late 2023, “all the memory and storage companies were very wary of increasing their production capacity again,” says Coughlin. “Thus there was little or no investment in new production capacity in 2024 and through most of 2025.”

The AI data center boom

That lack of new investment is colliding headlong with a huge boost in demand from new data centers. Globally, there are nearly 2,000 new data centers either planned or under construction right now, according to Data Center Map. If they’re all built, it would represent a 20 percent jump in the global supply, which stands at around 9,000 facilities now.

If the current build-out continues at pace, McKinsey predicts companies will spend $7 trillion by 2030, with the bulk of that—$5.2 trillion—going to AI-focused data centers. Of that chunk, $3.3 billion will go toward servers, data storage, and network equipment, the firm predicts.

The biggest beneficiary so far of the AI data center boom is unquestionably GPU-maker Nvidia. Revenue for its data center business went from barely a billion in the final quarter of 2019 to $51 billion in the quarter that ended in October 2025. Over this period, its server GPUs have demanded not just more and more gigabytes of DRAM but an increasing number of DRAM chips. The recently released B300 uses eight HBM chips, each of which is a stack of 12 DRAM dies. Competitors’ use of HBM has largely mirrored Nvidia’s. AMD’s MI350 GPU, for example, also uses eight, 12-die chips.

With so much demand, an increasing fraction of the revenue for DRAM makers comes from HBM. Micron—the number three producer behind SK Hynix and Samsung—reported that HBM and other cloud-related memory went from being 17 percent of its DRAM revenue in 2023 to nearly 50 percent in 2025.

Micron predicts the total market for HBM will grow from $35 billion in 2025 to $100 billion by 2028—a figure larger than the entire DRAM market in 2024, CEO Sanjay Mehrotra told analysts in December. It’s reaching that figure two years earlier than Micron had previously expected. Across the industry, demand will outstrip supply “substantially… for the foreseeable future,” he said.

Future DRAM supply and technology

“There are two ways to address supply issues with DRAM: with innovation or with building more fabs,” explains Mina Kim, an economist with the Mkecon Insights. “As DRAM scaling has become more difficult, the industry has turned to advanced packaging… which is just using more DRAM.”

Micron, Samsung, and SK Hynix combined make up the vast majority of the memory and storage markets, and all three have new fabs and facilities in the works. However, these are unlikely to contribute meaningfully to bringing down prices.

Micron is in the process of building an HBM fab in Singapore that should be in production in 2027. And it is retooling a fab it purchased from PSMC in Taiwan that will begin production in the second half of 2027. Last month, Micron broke ground on what will be a DRAM fab complex in Onondaga County, N.Y. It will not be in full production until 2030.

Samsung plans to start producing at a new plant in Pyeongtaek, South Korea in 2028.

SK Hynix is building HBM and packaging facilities in West Lafayette, Indiana set to begin production by the end of 2028, and an HBM fab it’s building in Cheongju should be complete in 2027.

Speaking of his sense of the DRAM market, Intel CEO Lip-Bu Tan told attendees at the Cisco AI Summit last week: “There’s no relief until 2028.”

With these expansions unable to contribute for several years, other factors will be needed to increase supply. “Relief will come from a combination of incremental capacity expansions by existing DRAM leaders, yield improvements in advanced packaging, and a broader diversification of supply chains,” says Shawn DuBravac , chief economist for the Global Electronics Association (formerly the IPC). “New fabs will help at the margin, but the faster gains will come from process learning, better [DRAM] stacking efficiency, and tighter coordination between memory suppliers and AI chip designers.”

So, will prices come down once some of these new plants come on line? Don’t bet on it. “In general, economists find that prices come down much more slowly and reluctantly than they go up. DRAM today is unlikely to be an exception to this general observation, especially given the insatiable demand for compute,” says Kim.

In the meantime, technologies are in the works that could make HBM an even bigger consumer of silicon. The standard for HBM4 can accommodate 16 stacked DRAM dies, even though today’s chips only use 12 dies. Getting to 16 has a lot to do with the chip stacking technology. Conducting heat through the HBM “layer cake” of silicon, solder, and support material is a key limiter to going higher and in repositioning HBM inside the package to get even more bandwidth.

SK Hynix claims a heat conduction advantage through a manufacturing process called advanced MR-MUF (mass reflow molded underfill). Further out, an alternative chip stacking technology called hybrid bonding could help heat conduction by reducing the die-to-die vertical distance essentially to zero. In 2024, researchers at Samsung proved they could produce a 16-high stack with hybrid bonding, and they suggested that 20 dies was not out of reach.