It’s a quick fix for that dead zone in your house.
Eza_Nanda/Shutterstock
If you have an old Android phone gathering dust in a drawer somewhere, you may already have everything you need to fix the dead zones in your home Wi-Fi network. Instead of buying new hardware, be it an extender, a mesh node or a more powerful router, you could just use an old Android device.
Most households have at least one old Android phone sitting forgotten somewhere, too outdated to use as a daily driver, but too functional to justify throwing it away. Before you drop it off for recycling, you should consider what it can still do for you. With nothing more than its built-in settings and a power outlet, that phone can pull double duty as a Wi-Fi extender, helping push your home network’s signal into those room corners where signal is patchy at best.
You don’t need any special apps, extra hardware or even advanced technical know-how. Android’s built-in hotspot feature lets your phone connect to your existing Wi-Fi network and act as a repeater to broadcast it as a separate signal, thus extending coverage.
Advertisement
What does an Android Wi-Fi extender actually do?
If your home has areas where the internet signal is weak, it can be quite annoying. While it’s true that some things can be done — did you know there was a right way to position your router’s antennas? — it can be a nuisance until fixed. There are things you can buy, of course. Upgrading to a mesh system or getting a new router can be quite expensive, while a Wi-Fi extender can put you back about $50, depending on model. That’s still money you could simply not spend by making use of tech you already have.
Your Android phone can connect to your existing home Wi-Fi network and rebroadcast the signal as a separate hotspot. You can then connect other devices, from laptops to tablets and smart TVs, to that Wi-Fi signal instead. Think of it as a signal relay, as the phone picks up Wi-Fi from your router and simply passes it along to other devices that need it. It’s not a mesh system, and it won’t perform as one, but it’s a free and low-effort fix for weak coverage areas.
Advertisement
How to set up your Android phone as a Wi-Fi extender
MIRACLE PROJECT/Shutterstock
The process for this whole operation is quite straightforward and only takes a few minutes. Our starting point comes after you’ve already located your old Android phone, recharged it and run the latest updates.
Open Settings, go to Network & Internet > Internet. Select your home Wi-Fi network and enter the password when prompted.
Go to Settings > Hotspot and Tethering and toggle on Wi-Fi Hotspot. You can also find the Hotspot icon in the drop-down menu — you just need to know what all those Android status bar icons mean.
Tap the Wi-Fi hotspot to rename the network and set a strong password to secure your network and keep it easy to identify.
If your phone supports it, you can select 5GHz under the speed or band settings. This will improve performance for nearby devices, but it offers a shorter range than 2.4GHz.
Place your phone about halfway between your router and the dead zone you’re worried about. Make sure it’s near an outlet so you can easily keep it recharged. Old phones already have questionable battery life, and running a hotspot continuously will drain the battery fast.
The downside of this method is that the phone extends your existing Wi-Fi, but it won’t boost your internet speed in any way. If your internet is already slow, the hotspot won’t fix that. You also have to keep in mind that any commercial Wi-Fi extender or mesh node will outperform this Android-based solution.
Another thing to keep in mind when using this method is that there’s no smart home integration to speak of. Dedicated Wi-Fi extenders often support smart home ecosystems and seamless roaming, while an Android hotspot creates a separate network, so some smart home devices may not connect to it reliably. Plus, you’ll need to manually switch networks as you move around your home.
Advertisement
Android phone vs. commercial alternatives: Which is better?
New Africa/Shutterstock
The answer to this question is one only you can deliver. Do you want to invest in your network at this point? An Android phone extender costs nothing, and it’s ideal as a temporary fix or supplementary solution for a single problem area. However, you’ll get much better range and performance from a commercial Wi-Fi extender, a mesh node or a new router. Go for the Android phone option if you need a quick fix and already have an unused device on hand. You could even try the USB tethering alternative for a more stable connection. It will make a surprisingly capable Wi-Fi extender for the right situation, the setup takes minutes, the cost is zero and it keeps a device out of the recycling bin. Just understand that this is a practical workaround, not a permanent network solution.
Intel shows three agentic AI architectures at Hot Chips 2026: the 256-core Diamond Rapids Xeon, the Crescent Island inference GPU, and the already-shipping Wildcat Lake client SoC
Diamond Rapids matches AMD’s EPYC Venice on cores and bandwidth but arrives in 2027 with half the threads, while Wildcat Lake’s 17 TOPS NPU falls short of Copilot+’s 40+ TOPS certification
Intel’s Crescent Island will support up to 480 GB via third-party partner cards, but the Intel-branded variant will be limited to 160 GB, even as it currently does not publish the memory bandwidth that determines its token economics.
Intel used its slot at Hot Chips 2026 to lay out three architectures it says will carry agentic AI from the rack to the laptop.
The already-shipping Wildcat Lake Core Series 3 SoC is for mainstream client and edge machines, Crescent Island is an inference-optimized data center GPU, and Diamond Rapids is its next-generation enterprise-class data center CPU that can be configured with up to 256 cores.
Pushkar Ranade, Intel’s chief technology officer, said in a statement accompanying the announcement that agentic AI is changing how the company designs computing from the transistor upward.
Latest Videos FromTechRadar
Advertisement
Plenty of compute promises, sans some GPU numbers
Intel’s Diamond Rapids, which it plans to brand as Xeon 7, is where the chipmaker provides the most information, unsurprisingly, in an industry that plans purchases and upgrades years in advance.
Its flagship offering can encompass as many as 256 cores, backed by 1.28G of fast last-level cache, 16 memory channels, 128 PCIe 6 lanes, and CXL 3.0. ServeTheHome reported that the maximum configuration uses 16-core chiplets, built on Intel’s 18A-P process node.
Rival chipmaker AMD also offers a 256-core trim: the EPYC 9996, built on TSMC’s 2nm node, keeps hyperthreading in play, effectively doubling its thread count versus Intel’s flagship.
Its Crescent Island GPU is arguably the most interesting design of the three, but the least detailed: a 350W air-cooled PCIe card built on its Xe3P GPU architecture, with 32 cores feeding 256 third-generation XMX engines and 32 MB of unified L2 cache. It surprisingly uses LPDDR5X rather than HBM, but the choice seems rationalized to offer capacity at a power and cost profile that lets Intel compete meaningfully with Nvidia and AMD.
Sign up to the TechRadar Pro newsletter to get all the top news, opinion, features and guidance your business needs to succeed!
Interestingly, the 480GB figure Intel touts in its press release isn’t what it will ship; Intel-branded cards cap at 160GB, and 480GB is a ceiling available to ODM partners building their own configurations. Intel also remains mum on memory bandwidth specifics and told Chips and Cheese it won’t disclose the figure; this is crucial for many AI applications, as memory bandwidth determines decode throughput, and decode throughput essentially determines token economics.
Advertisement
Of the three architectures that Intel talked about, Wildcat Lake is the only one you can pick up off a shelf (as a miniPC or laptop) today; it launched as Intel’s Core Series 3. Intel stated that it replaced the Foveros packaging used in Panther Lake with an organic multi-chip package, eliminating the base die and, with it, assembly cost and yield loss. Compute die area fell by 38 percent, and I/O tile area by 15 percent.
The 17 TOPS NPU available to such CPUs sits well below the 40 TOPS Microsoft requires for Copilot+ certification. The 40 TOPS Intel cites is a platform total across CPU, GPU, and NPU, which arguably can be met elsewhere, but it’s a sticking point for the chipmaker’s AI push, which has otherwise been concentrated on capturing enterprise and hyperscale customers.
Healthcare and pharmaceutical distribution giant McKesson has disclosed a cybersecurity incident involving unauthorized access to third-party applications and data theft, with the ShinyHunters extortion group claiming it stole 284 million patient data records.
McKesson is a major U.S. healthcare company and pharmaceutical distributor that provides medicines, medical supplies, technology, and services to healthcare providers and pharmacies.
CyberInsider first reported the breach earlier today, and McKesson later disclosed it in a Form 8-K filing with the U.S. Securities and Exchange Commission.
McKesson says it discovered the cybersecurity incident on August 25, 2026, and that its investigation remains in the early stages.
“Information about the incident, including any updates, is available on the company’s website at www.mckesson.com/cybersecurity,” McKesson said in its SEC filing.
Advertisement
“As of the date of this filing, the company has not determined that the incident is material or that the incident has had, or is reasonably likely to have, any material impact on the company, including its financial condition or results of operations.”
In a separate notice to customers, McKesson confirmed that the incident involved third-party applications and the unauthorized access and exfiltration of data.
“We take the security and privacy of our partners, customers and their patients very seriously. Upon discovery, we immediately activated our incident response protocols, launched an investigation, and engaged leading cybersecurity industry experts to assist in our response,” reads McKesson’s notice.
The company said its investigation is ongoing to determine the full scope of the incident.
Advertisement
McKesson also warned that customers may experience intermittent service degradation believed to be related to the attack, although the company said it was not proactively disconnecting systems within its environment.
At this time, McKesson has not publicly disclosed which third-party applications were compromised, how the attackers gained access, or what information was stolen.
McKesson says its investigation remains ongoing and that it will provide additional information as it develops a more complete understanding of the incident.
ShinyHunters claims responsibility
The ShinyHunters extortion group told BleepingComputer that it was behind the attack, claiming it gained access after conducting voice phishing, or vishing, social engineering attacks against multiple McKesson employees.
Advertisement
ShinyHunters declined to provide many technical details about the social engineering attacks, including the domain used during the campaign. However, BleepingComputer learned from another source that the threat actors used the mckesson[.]claims domain as part of the attack.
This domain matches a ShinyHunters campaign recently documented by ReliaQuest’s Threat Research team, which said the extortion group was registering .claims domains containing the names or abbreviations of targeted companies to impersonate their help desks and IT teams.
“ReliaQuest is tracking a widespread ShinyHunters campaign using domains that follow the company[.]claims pattern. These domains incorporate the targeted organization’s name or abbreviation under the .claims TLD,” ReliaQuest said in a now-deleted post on X.
ShinyHunters told BleepingComputer that the vishing attacks led to the compromise of multiple employees’ Okta single sign-on accounts, which they then used to access the company’s Salesforce and Snowflake environments.
Advertisement
The threat actor claims it fully compromised the Salesforce environment, including support cases. The threat actor also allegedly stole a much larger collection of patient-related data from Snowflake.
According to ShinyHunters, the threat actor exfiltrated about 1TB of data over four days, between August 21 and August 25.
The threat actor also claims the stolen Snowflake data contains approximately 284 million data records of patient-related information. However, this does not mean that the breach impacted 284 million patients.
Previous reporting stated that information belonging to 284 million patients had been exposed. ShinyHunters clarified to BleepingComputer that the figure is actually a raw count of approximately 284 million data records, or lines, rather than a count of unique individuals.
Advertisement
The threat actor told BleepingComputer that it has not fully analyzed the stolen data and does not know how many unique people are in those records.
ShinyHunters claims the stolen information includes names, addresses, dates of birth, Social Security numbers, patient IDs, phone numbers, email addresses, Medicaid numbers, medical record numbers, medication and allergy information, illnesses, disabilities, appointment information, and physician information.
The group also claims the data contains information related to deceased and terminally ill patients, prescriptions and medication shipments, invoices, employee information, Salesforce records, internal communications, and healthcare providers and clinics using McKesson’s services.
BleepingComputer has not independently verified these claims, and McKesson has not publicly disclosed what information was stolen.
Advertisement
The group says it contacted McKesson after completing the data theft on August 25 and demanded a $55,236,150 ransom, giving the company 72 hours to respond. According to ShinyHunters, McKesson did not respond to or negotiate over the ransom demand.
The attack comes amid an ongoing wave of data-theft attacks targeting healthcare and health technology organizations attributed to ShinyHunters.
Health-ISAC recently warned healthcare organizations about increasing ShinyHunters attacks involving social engineering designed to compromise corporate accounts and gain access to cloud and SaaS platforms.
Other healthcare technology companies targeted in recent ShinyHunters data-theft attacks include Medtronic, DentaQuest, iRhythm, OneMedical, and AdaptHealth.
Advertisement
Overall prevention scores can hide what happens after initial access. Once attackers are using valid credentials, prevention drops sharply.
The Blue Report 2026 measures defenses technique by technique across 338 million simulations run in customer production environments.
“Grand Theft Auto 6” is the biggest and most-anticipated game release in a decade, at least. Despite a reported budget of more than $1 billion, it’s expected to make money back for developer Rockstar and publisher Take-Two Interactive in a matter of days. There’s also a reasonable case to be made for “Grand Theft Auto VI: An Extended Look” being the biggest game trailer release of all time. It’s certainly the first trailer to get a six-hour exclusive run on Netflix (with some stray Twitch streamers thrown in for good measure) before debuting on YouTube and other platforms.
There’s plenty to see for yourself, so if you’re at all interested it’s well worth spending 26 minutes and 49 seconds to take a look ahead of “Grand Theft Auto 6” release on November 19. But, we spotted a bunch of things you may not notice at first glance, and we also dove through several interviews Rockstar gave — timed for the trailer’s release, of course — to pull out the things you need to know about “GTA 6.”
Advertisement
This is the biggest Grand Theft Auto ever, and it’s not even close
Rockstar Games
We had a suspicion that this would be the case, but thanks to some media work Rockstar executives did around the extended look’s release, we can better quantify what that means. Speaking with YouTuber TGG, Rockstar co-studio head Rob Nelson claimed that the game’s map is twice the size of the “GTA 5” map, and three times the size of the “Red Dead Redemption 2” map. As for Vice City, TGG said, “Vice City itself is twice the size of Los Santos. And Vice City and its [surrounding suburbs] is 11 times bigger than Los Santos and its [surrounding suburbs].”
This is all obviously a far cry from the maps of multi-planetary games like “No Man’s Sky” and “Starfield,” but Rockstar’s games are on a different level to those when it comes to density and unique touches. A giant planet with a couple of buildings vs. a large portion of the state of Florida (sorry, Leonida), isn’t the fairest comparison to make.
We also got a clue as to just how long “Grand Theft Auto 6” is. In a different interview with The New York Times, Nelson told games journalist Christopher Byrd that he spent 80 hours playing through the main story and a handful of optional goals. Your mileage may vary, but that’s a much longer time than the average player logged completing “Grand Theft Auto 5,” according to the site How Long to Beat.
Advertisement
You can race cars, dirtbikes, ATVs and maybe even jet skis
Rockstar Games
It wouldn’t be a modern Grand Theft Auto without minigames, and we spotted multiple new options in the extended look. We knew from earlier leaks that basketball was going be an option, but during the 27-minute trailer Rockstar showed off at least six other activities.
Vice City has always had a dirt track, so the inclusion of off-road races isn’t a huge surprise, but the brief snippet shows dirtbikes and ATVs racing alongside trucks. There’s an even-more-brief segment where we see a Fast and Furious-style street race, full of exotic imports, neon underglow mods and, of course, nitrous boosts.
Advertisement
There even appears to be a professional series on a full-blown race track inspired by real circuits. In the trailer we see a multi-lap race with track-modified sedans and souped-up muscle cars decked out in (pretty silly) sponsored liveries. There are open-wheel racing series in “GTA Online” and figure-8 racing events in “San Andreas,” but this latest addition looks impressive, to say the least.
Finally, while we didn’t see any jet ski races in the trailer, the vehicles do feature prominently, so we wouldn’t be surprised if that’s an option come November 19. If racing isn’t your speed, there seem to be a handful of less adrenaline-fueled activities on offer, including kayaking, diving, weight lifting and target practice challenges.
Advertisement
There are some huge changes to combat
Rockstar Games
We still don’t have a full rundown on the fine details of the “GTA 6” combat system, but we did learn a lot from the extended look.
One big quality-of-life improvement coming is swapping between two firearms on the fly. The weapon wheel is still there, but around 18 and a half minutes in, you can see co-protagonist Lucia on a target range switching between a rifle and a pistol without needing to pull up the wheel.
Bullet time — which was an ability locked to Michael De Santa in “GTA V” — is confirmed to be making a return, at least for Jason. While he clears an office space of goons, we see him slow down time and a new vital targeting system. Some areas are lit up in red and appear to be indicating lethal shots, while others in a pale yellow are centered on joints and may be there for you to incapacitate someone in a less-lethal way. Oh, and there’s even a killcam straight out of “Max Payne 3” in one scene.
One perhaps pointless but very cool addition is the ability to wield certain weapons in one hand. In one scene, Lucia is carrying a duffel bag in one arm and a shotgun in the other. She’s able to both fire and rack her shotgun without skipping a beat. Those are some strong forearms. And ribs, for that matter.
Advertisement
You can’t just hot-wire every car
Rockstar Games
Going all the way back to 1997’s “Grand Theft Auto,” you’ve been able to break into any car and, with the press of a button, magically hot-wire it. Those days are over. Around six minutes into the trailer, we see Jason attempting to steal a car and a “Slim Jim” minigame where the player has to time their inputs to get into the car using a slide tool.
According Rockstar, that’s just one of the new layers of complexity to the grand theft auto of “GTA 6.” Speaking with IGN, Nelson (the same co-studio head who spoke with TGG) gave a few interesting examples: Vehicles with electronic key fobs may require a key cloner, while cars with GPS tracking enabled might require you to drive below a certain speed threshold to avoid triggering an alarm.
Advertisement
Apparently there’ll be an in-game app on the protagonists’ cellphones that will let you scan a car to discover what security features are optioned. “Grand Theft Auto VI” has a number of new rides and returning classics, and it sounds like many will require a different approach to steal. The app will also reveal how much money you can make selling it or how much it’ll cost to file down those VINs add it to your collection.
The Colorsoft has mostly sat at £629.99 for months. Only brief dips interrupted that pattern before this latest drop pushed the price to a new low.
Advertisement
The display itself is what makes this Scribe standout, since Amazon’s oxide-based Colorsoft screen delivers high-contrast, paper-like colour with a textured surface that avoids the flashing normally associated with e-ink refreshes when you turn a page.
This paper-like feel carries into the Premium Pen included in the box, a battery-free stylus that never needs charging and responds quickly enough to feel like writing with real ink rather than plastic on glass.
Advertisement
Every page you write becomes part of a built-in notebook that uses AI to search your handwriting, summarise long notes, and answer questions about what you wrote without needing to remember exact keywords or page numbers.
Handwritten pages can also be refined and converted to text for sharing, while documents pulled in from Google Drive or Microsoft OneDrive can be marked up directly and exported straight back out as PDFs too.
Active Canvas takes that a step further by making space in the text itself, shifting a book or Word document to let you write directly between the lines instead of squeezing notes into a cramped margin.
All of that runs on a single charge that lasts up to eight weeks of regular reading or two weeks of daily writing, so the Colorsoft rarely needs to come off the shelf to top up.
Imagine picking up one of Amazon’s most advanced e-readers without paying full price for the privilege, because that is exactly what happens with the Kindle Scribe Colorsoft at £479.99, its lowest price since its official launch.
Advertisement
Advertisement
If you have been eyeing the Colorsoft, now is the moment to grab it at that lowest recorded price, since it has bounced back to £629.99 quickly enough before to make waiting a real risk.
Bluesound has introduced the Pulse Sub, a completely redesigned wireless powered subwoofer built to bring deeper bass to its BluOS multi-room system.
Inside the cabinet is a 350-watt amplifier with Smart DSP technology drives two custom 6.5-inch woofers arranged in an opposing configuration, a layout that delivers roughly 2.5 times more power than the previous generation Bluesound subwoofer.
Image Credit (Bluesound)
The dual-opposing arrangement also cancels much of the vibration each driver creates, so the cabinet stays still while the woofers do the work to deliver bass stays deep, clean and controlled.
Built entirely around the BluOS platform, the Pulse Sub pairs wirelessly with existing Bluesound products through the BluOS Controller app, and setup mirrors the same steps Bluesound already uses for adding a second player to a system.
Advertisement
Advertisement
Pairing options include the Pulse Cinema and Pulse Cinema Mini wireless soundbars for a fuller home cinema setup, or the Powernode streaming amplifier for listeners who want deeper, more musical bass added to music playback.
Matt Simmonds, VP of Bluesound Product Development, described the goal for the design as: “Great bass shouldn’t be loud and boomy. It should bring you closer to what you are watching or listening to without calling attention to where it is coming from.”
Image Credit (Bluesound)
Simmonds added that the team rethought both the acoustic and industrial design of the Pulse Sub to more than double its power output while keeping the enclosure compact, controlled and easy to live with.
Soft contours and tonal-weave fabric grilles pair with a matte satin finish that matches the rest of the Bluesound line, while integrated cable management channels keep wiring routed and out of sight. Beyond wireless pairing, a low line-level analogue input lets the Pulse Sub double as a wired active subwoofer for non-BluOS systems.
Advertisement
Advertisement
The Pulse Sub will be available later in 2026 priced at £849 / €999, sold through authorised Bluesound retailers, Bluesound website and Amazon, with pre-order details to follow at a later date.
Pricey Memory Potato: Samsung recently unveiled the latest models in its portable storage lineup. The “P” SSD series features unprecedented performance, but the high-end models come with truly eye-opening MSRPs that few users will be able to afford.
Samsung unveiled the company’s new offering for portable storage enthusiasts at Gamescom 2026 in Germany. The “P” Series of external SSDs uses USB4 connectivity to raise the performance bar considerably while offering some truly spacious units for both professional and consumer use.
The P Series includes two models: the P9, which is positioned as an ideal solution for professional users, and the P7 – designed to provide “convenient” storage for everyday users. According to the official price tags, however, it isn’t particularly convenient for most wallets.
The flagship P9 drive can deliver sequential read speeds of up to 4,000MBps, while sequential write speeds can reach 3,800MBps. This industry-leading performance is approximately twice as fast as the previous generation, with Samsung comparing the newer P9 with a 4TB T9 drive, which offers sequential read and write speeds of 2,000MBps over a USB 3.2 Gen 2×2 interface.
Advertisement
Earlier this year, the Samsung P9 received an Innovation Award at CES. Samsung said the portable SSD can sustain “12K” real-time recording, which should be more than enough for production workflows involving video capture, RAW image files, and ultra-high-definition video storage and editing.
Meanwhile, the Samsung P7 model uses the same USB4 interface to offer a more versatile storage option. The Korean company highlights the drive’s broad compatibility, noting that it can be used with desktop and laptop computers, smartphones, gaming consoles, and even USB-compatible cameras.
The P7 also offers improved durability, having successfully completed a 2-meter drop test along with additional shock and vibration tests. According to Samsung Electronics Vice President Tommy Kwon, the P Series can provide real performance benefits for “AI-powered” content creation, video production, and everything in between.
Now, here’s the hardest part of Samsung’s overenthusiastic announcement: the P Series of portable SSDs will cost a lot, because memory products have become a luxury that very few giant corporations in the world can afford to ransack.
Advertisement
The P9 will be available in four capacities (1TB, 2TB, 4TB, and 8TB), with suggested prices ranging from $340 to $2,700. Meanwhile, the “affordable” P7 drives will range from $290 for the 1TB model to $2,290 for the 8TB version. Look for them to start shipping on August 31.
How come that super-smart AI of yours didn’t figure this out, frontier labs?
Datacenter cooling could get a lot more affordable, and sustainable, if a joint German-Japanese research project can be scaled beyond the laboratory.
Advertisement
Researchers from the Karlsruhe Institute of Technology and the University of Tsukuba published a paper in Nature Energy on Friday discussing the success they had with a solid-state cooling technology that doesn’t need any electric input at all, and was still able to demonstrate temperature reductions in laboratory experiments.
For those unfamiliar with solid-state cooling, it’s a more sustainable alternative to traditional compression cooling that relies on environmentally harmful refrigerants and lots of electricity. Instead of compressing and expanding a gas to generate cold, some solid-state systems rely on thermoelectric materials that, when supplied with electricity, can generate temperature differences. Thermoelectric solid-state cooling systems can be incredibly small, but there’s a catch: They still need energy, and they’re only a quarter as efficient as modern compression cooling systems, as the researchers noted.
Enter elastocaloric cooling, which relies on shape-memory alloys that change temperature when mechanically loaded and unloaded through a phase transition. It’s that sort of material and methodology that the KIT and Tsukuba team experimented with.
Like other forms of solid-state cooling, elastocaloric systems have traditionally required electricity to effect change in the shape-memory alloys they use, relying on actuators that can get quite large. What the researchers wanted to develop was a miniaturized form of elastocaloric cooling that didn’t require any electricity at all, instead relying on the waste heat being generated by the objects it wanted to cool – for example, a computer processor, as they mention in their paper. Based on their demonstration, it appears to have worked.
Advertisement
The researchers used a pair of super thin nickel-titanium foils for their elastocaloric system with opposing functions: One foil shrinks when subjected to heat, generating movement that translates into mechanical energy and acting as its own actuator. That mechanical energy is transferred to the second foil, which is loaded and unloaded to trigger a phase transition that produces the cooling effect.
Laboratory tests demonstrated that their technology isn’t only theoretically feasible, but also actually works – albeit to a limited extent.
“Our prototype achieves a temperature span of 12.9 K at the refrigerant film level and 4.0 K at the device level under Joule-heated actuation at 86°C (187°F),” the team said in their report. When subjected to an external heat source of 130°C (266°F), the system still maintained a device-level temperature span of 2.2 K – not a lot, but it still shows the device, even at its most basic, single-pair-of-films level, works.
“For us, the decisive moment was seeing measurable cooling for the first time, actually generated by a heat-driven system,” KIT PhD candidate and paper lead author Yi-Ting Hsiau said of the project in a release put out by the university. “This showed us that the principle doesn’t just work on paper.”
Advertisement
With the concept of electricity-free miniaturized cooling now proven, the team is planning to build larger experimental units to optimize cooling performance.
Electricity-free and water-free cooling designs couldn’t come at a better time: Energy generation used to power and cool things like datacenters is on the rise, as is the water footprint of such facilities, in part also linked to cooling. Add climate change to the mix, and cooling electronics and people is only more of a concern – according to IRENA, heating and cooling together now account for more than 40 percent of global energy-related carbon dioxide emissions. Sorting out the datacenter energy and water problem has a lot to do with figuring out how to keep systems cool, and while this research is in its infancy, it could prove to be a valuable contribution to reducing the tech industry’s energy footprint.
“We believe this is just the beginning,” said Jingyuan Xu, head of KIT’s ZEco Thermal Lab, working on zero-emissions and eco-friendly heating and cooling tech. “By scaling up this technology, we aim to develop compact cooling systems that utilize abundant heat sources for sustainable cooling.” ®
Long gone are the days of headphones merely being “good for Bluetooth.” Mainstream consumer brands like Sony, Bose, and Sennheiser crested the hill of “actually good” audio quality a long time ago and are now differentiating themselves based on lifestyle and usability features, with ANC chief among them. The technology that powers these advanced features is complex and not particularly cheap, and smaller brands with more limited R&D budgets lagged behind for a long time.
But multi-mic arrays, AI-powered ANC, and personalized hearing compensation systems have become much more common in recent years, thanks to advances in each respective field. That has worked to the benefit of smaller artisanal brands like Noble, which specialize in acoustic engineering rather than consumer technology.
The Artemis is Noble’s newest wireless audio product, released right after the well received Osprey wireless IEMs. The Artemis now sits at the top of Noble’s wireless headphone lineup, superseding models like the FoKus Apollo and Apollo Pro and costing $899. But a “nice” headphone with smart integration of modern technology isn’t new, nor is it special in today’s highly competitive headphone market, so Noble went one step further.
The Artemis ditches the traditional single dynamic driver setup for a novel tribrid approach: one dynamic driver, one balanced armature driver, and one planar driver per side. It also features built in audio personalization with Audiodo hearing compensation technology and even allows users to service aging units themselves with easily replaceable batteries.
Advertisement
But $900 is still substantial, and there are a lot of great headphones that cost less. Does the Artemis’ feature set and exotic driver configuration justify its price? Well, let’s get into it.
About My Preferences: This review is a subjective assessment and is therefore influenced by my personal preferences. While I try to mitigate that as much as possible during the review process, I’d be lying if I said my biases were completely erased. So, for you, my readers, keep the following in mind:
My ideal sound signature includes competent sub bass, textured mid bass, a slightly warm midrange, and extended treble.
I have mild treble sensitivity.
Testing equipment and standards can be found here.
Related Reviews:
Specifications:
Headphone Type: Wireless closed back over ear
Driver Configuration: 1 dynamic driver, 1 balanced armature driver, and 1 planar magnetic driver per side
Wireless Platform: Qualcomm QCC3095
Bluetooth: 6.0
Codecs: LDAC, aptX Adaptive, aptX Lossless
ANC: User adjustable Active Noise Cancellation
Transparency Mode: Yes
Microphones: 6
Multipoint: Yes
Personalized Audio: Audiodo Personal Sound
EQ: App based EQ with settings stored on the headphones
Spatial Audio: Supported through the Noble FoKus app
USB Audio: Yes, via USB C
Wired Audio: 3.5mm
Battery: 600mAh, user replaceable
Battery Life: 55+ hours with ANC off; 35+ hours with ANC on
Charging: USB C with fast charging
Charging Time: Under 1.5 hours
Water Resistance: IP52
Weight: Under 350g
Ear Pads: Magnetic, detachable, and replaceable
Build Quality
The Artemis is built primarily from high grade CNC machined aluminum. Each ear cup features a Damascus style pattern etched directly into the metal.
The back of each ear cup is perforated, providing ventilation and space for the microphones. A plastic ring sits between the Alcantara and the ear cup, housing the power and volume controls. The Artemis’ USB-C input is located on the bottom of the left ear cup and feels secure.
The Artemis’ headband is sturdy and comfortable, with a padded Alcantara underside and woven cloth exterior. The internal structure is reinforced with spring steel, giving the headband a substantial and durable feel.
Advertisement
The Artemis features magnetically attached ear pads, making them easy to remove and replace. Replacement pads are available directly from Noble, so owners do not have to worry about the headphones becoming unusable simply because the pads wear out.
Advertisement. Scroll to continue reading.
The Artemis with its ear-cups removed. Credit: Noble Audio
Even more valuable is the inclusion of user replaceable batteries. With a single screw, owners can remove the Artemis’ battery and replace it with a spare purchased from Noble. In practical terms, that means the Artemis can remain in service until something truly fails rather than becoming disposable once the battery reaches the end of its useful life.
Very few wireless headphones offer this level of maintainability. Fairphone’s Fairbuds XL are a notable exception, while AIAIAI’s TMA 2 models also emphasize modular, replaceable components. Sennheiser supports replaceable ear pads on models like the Momentum 4, Momentum 5, and HDB 630, but their batteries are not user replaceable. Noble therefore deserves real credit for making both routine wear items and the battery easy for owners to replace themselves.
I’m a fan of the case Noble chose for the Artemis. It features a neatly cut recess that holds the headphones securely, along with a dedicated compartment for cables and accessories. The exterior is soft and pleasant to the touch, while the felt like interior helps protect the Artemis from scratches and abrasion.
Advertisement
Included with the Artemis is a useful assortment of cables and accessories, including two USB C to USB C cables, an airplane adapter, and a 3.5mm to USB-C adapter. That is basically everything you need for a complete experience right out of the box, and the cables themselves feel soft, flexible, and sturdy. No complaints here.
Connectivity
The Artemis features Bluetooth 6.0 compatibility and supports a broad range of audio codecs. Hi Res playback via LDAC and aptX Lossless works well, and I was able to achieve LDAC playback easily across multiple Android devices. The Artemis’ Bluetooth chipset also proved reliable during normal listening, with no stuttering or dropouts. I was able to maintain high quality playback even when separated from the source device by a considerable distance and multiple walls.
My one reservation involves occasional disconnects. After listening to the Artemis for an hour or two, it will sometimes disconnect from my phone without warning or an error message. The headphones do not appear to be power cycling, and the source device also registers the temporary disconnect. The Artemis then reconnects almost immediately, but the interruption is still annoying.
Wired Playback
Beyond wireless playback over Bluetooth, the Artemis also supports wired playback over USB C. There are two ways to do this: with the included USB-C to 3.5mm cable or with a USB-C OTG cable.
Advertisement
Connecting the Artemis to my PC via USB C was easy and painless, with Windows quickly recognizing the headphones and installing the appropriate driver. Sound quality remains fully intact in this mode. Listening via the 3.5mm connection is similarly straightforward, and the Artemis’ sound signature remains consistent with its USB-C and Bluetooth playback modes.
Listening
The Artemis offers two layers of sound customization through hearing compensation and a 10 band EQ. The analysis in this section was performed without either feature enabled and is intended to serve as a baseline assessment of how the Artemis performs without additional tuning.
In stock form, the Artemis splits the difference between a traditional mainstream V shaped tuning and a more enthusiast oriented reference presentation. Its bass is elevated, sloping down into a weighty but not bloated lower midrange. The Artemis’ lower register is fairly rich overall, counterbalanced by an energetic and articulate upper register.
Moving up from the somewhat flat lower mids, the Artemis peaks in the upper mids around 2 to 3 kHz, placing vocals forward in the mix. There is a bit of grain here, likely owing to the Artemis’ complex tribrid driver architecture. From the upper mids, the response dips through the lower treble before rising again around 8 and 10 kHz. This translates into a bright, forward treble that adds some much needed contrast to the soundstage.
Advertisement
Sophisticated Treble
The Artemis’ treble is bright and forward, though well balanced and free from unnatural exaggeration. Even sharply mastered tracks like “Satisfy” by Nero remain controlled and natural sounding. I’m fairly sensitive to treble, so it’s nice to see Noble paying close attention to the amount of energy in the upper register.
Advertisement. Scroll to continue reading.
This balanced approach renders air and space nicely, working especially well with spacious, open sounding tracks like “Midnight City” by M83. At the other end of the spectrum, densely layered tracks like the chronically congested “Bitter Sweet Symphony” by The Verve are also handled well. The multiple layers of strings and metallic percussion remain distinct without becoming smeared or blurry.
Broad, Rich, and Full Sounding
The Artemis’ midrange is tuned for richness, aiming to maximize timbral weight without sounding bloated. Broadly, I think Noble has done a good job. Vocals are forward and set just ahead of the instrumentation, rendered with clarity and respectable intelligibility even during busy passages. The Artemis is not particularly biased toward male or female vocals, rendering both with distinction and care. Duets are therefore a strong listen, as evidenced by the rowdy vocals in “Ricochet!” by Shiny Toy Guns.
Advertisement
Beyond vocals, the Artemis’ midrange can occasionally sound a bit too warm on atmospheric, bass heavy tracks like “Courtesy” by Slothrust, toeing the line between desirable weight and outright heaviness. That said, the Artemis still exhibits excellent instrumental layering, as heard on “Paradox” by Emoter. Its soundstage is wide, deep, and organic, presenting its technical capabilities in a naturally expansive fashion.
Powerful, Articulate Bass
For many listeners, bass is a make or break aspect of a headphone’s sound signature, myself included. Beyond simple presence, factors like tonal balance, speed, control, and responsiveness all contribute meaningfully to a satisfying bass experience. Noble’s top end IEMs usually possess strong bass profiles, but headphones are a different ball game. Thankfully, the Artemis’ bass isn’t just good. It’s excellent.
Powerful, punchy, and weighty, the Artemis’ mid bass hits with tangible presence, adding substance and impact to drum slams in “wonderful life” by Bring Me The Horizon. Old school EDM with massive bass drops, like “The Hero” by Nostalgia, extracts a respectable degree of punch and rumble from the Artemis, communicating a convincing sense of size and weight.
The Artemis’ bass isn’t just powerful, but also maintains excellent tonality and timbre. Bass elements sound complete and full bodied, with enough density to give them real physical presence. “Horizons” by Bassnectar is treated similarly well, but immerses the listener in a broad wall of bass rather than a punchy ebb and flow. The Artemis’ extension into the sub bass below 40 Hz is particularly impressive. More complex tracks like “War Pigs (Celldweller Remix)” retain texture and shape through demanding bass passages, even when the low end becomes layered and highly varied.
Advertisement
Acceptable Noise Cancellation Quality
The dichotomy between audiophile grade acoustics and class leading ANC performance is a real one. You still have to choose, and as good as headphones like the Artemis have become at active noise cancellation, the difference in effectiveness compared with mainstream brands like Sony and Bose remains noticeable.
The Artemis offers three levels of ANC strength: “weak,” “medium,” and “strong,” with substantial differences in noise reduction between each setting. “Weak” mode delivers the most predictable performance, offering the greatest consistency in how ANC affects the final sound output, while “medium” mode produces the least consistent results. “Strong” mode blocks out most low and mid frequency noise, but still struggles with higher pitched sounds.
The Companion App
Noble’s Bluetooth products are controlled through the Noble FoKus app, available on both the App Store and Google Play. It is small, lightweight, and free of unnecessary bloat. No accounts, no passwords, no hassle. Once a compatible device is paired, the app allows you to adjust EQ, take a personalized hearing test, and update the firmware.
There are still a few areas where the user experience could be improved, and the interface is not as polished as what you’ll find from brands like Bang & Olufsen or Master & Dynamic.
Advertisement
Sharing EQ presets between devices isn’t that useful
One particular point of annoyance is that the FoKus app does not separate custom EQ presets by device. That means presets I save for, say, the Noble Osprey also appear when I’m browsing presets for the Artemis.
That may seem minor, but it can become irritating when I want to quickly find a specific preset I’ve saved. After all, an EQ setting created for the Artemis is unlikely to sound good on the Osprey, and vice versa.
Advertisement. Scroll to continue reading.
What is the Artemis Up Against?
As the market for true luxury wireless headphones continues to grow more crowded, it’s more important than ever to understand what competition the Artemis actually faces. First, let’s look at the mainstream brands.
Advertisement
Sony 1000X THE COLLEXION
Sony introduced the 1000X THE COLLEXION (WH-1000XX) in May 2026, positioning the $649.99 model above the more reasonably priced WH-1000XM6. The Collexion also addresses some of the durability concerns associated with the plastic hinge designs of earlier 1000X models, replacing the XM6’s folding hinge with a more substantial stainless steel swivel joint and adding considerably more metal to the overall construction. It feels heavier and more substantial than Sony’s mainstream models, putting it much closer to the Artemis in terms of perceived luxury.
Sony remains one of the benchmarks for active noise cancellation, and in some situations the 1000X THE COLLEXION delivers more refined noise reduction than the Artemis. The Sony does not, however, offer the Artemis’ easily user replaceable battery. That means long term battery degradation is more likely to require service rather than the simple at home replacement Noble provides.
Sonically, the Artemis has a clear advantage over the 1000X THE COLLEXION. Its upper register, in particular, offers levels of articulation and finesse that the Sony does not begin to approach. Both headphones provide extensive sound personalization, although they go about it differently. Noble uses Audiodo hearing compensation, while Sony offers its Find Your Equalizer system, conventional EQ controls, and personalized spatial audio features through the Sound Connect app.
The Sony’s drivers simply do not match the Artemis’ tribrid configuration in overall performance, and applying EQ to both only reinforces how much more resolution and refinement the Artemis can deliver.
Noble FoKus Apollo Pro
Yes, it’s true: the Artemis does have to justify its substantial price increase over the $699 FoKus Apollo Pro. That headphone was announced at CanJam Singapore 2026 and is still relatively early in its lifecycle. The Apollo Pro features a simpler hybrid driver setup, with one dynamic and one planar driver per side, compared with the Artemis’ dynamic, planar, and balanced armature configuration.
The Artemis uses a similar but noticeably upgraded selection of materials, most obviously its Damascus steel inspired ear cups. That said, the Apollo Pro, like the original Apollo, features a detachable boom microphone, making it better suited for gaming and office use, especially if you need to keep your voice down.
Advertisement
I was not much of a fan of the original Apollo, but the Apollo Pro’s tuning was far more palatable. It has a smoother, more relaxed presentation than the Artemis, lacking the latter’s more forward and aggressive treble. The Artemis’ integrated pressure relief system also gives it a tighter, meatier low end, allowing it to capture a greater sense of size and intensity in electronic bass drops than the Apollo Pro can manage.
EQ is a major part of the experience with both headphones, however, and each responds well to adjustments using the FoKus app’s easy to use 10 band equalizer. Is the Artemis worth the $200 premium? It really depends on your needs. On one hand, the Artemis offers a tangible improvement in material quality and sonic performance. On the other, the Apollo Pro’s boom microphone and longer battery life give it practical advantages the Artemis does not have.
Ultimately, how you intend to use your headphones, along with how demanding you are about acoustic performance, will determine which model makes more sense.
Bang & Olufsen Beoplay H95
B&O has long offered extremely expensive wireless headphones, many of which sit well above mainstream pricing. The model that competes most directly with the Artemis is the H95, which costs $1,250, a $350 premium over the Noble. And while I’d like to say the Artemis simply dunks on the H95 when it comes to build quality, that isn’t the case. B&O leans heavily on material choice to justify the H95’s substantial price, using luxurious components such as anodized aluminum faceplates, titanium drivers, and a lambskin headband. Heck, even the H95’s carrying case is made from metal.
But material design is where the H95’s biggest advantage ends. Its rich V shaped sound signature is fine and well suited to mainstream tastes, but the underlying driver performance does not clearly outperform more conventional models like Sony’s WH-1000XM5, let alone the Artemis. Acoustic performance is clearly the Artemis’ domain, with tighter, better controlled bass, clearer vocals, and airier treble.
Advertisement
Advertisement. Scroll to continue reading.
The Artemis’ stock sound signature is somewhat less cohesive than the H95’s tuning, but it can be easily adjusted to bring it closer to a more traditional presentation. You do not even need to touch the manual EQ, as the Jazz and Rap presets in the Noble FoKus app offer different takes on more familiar tuning.
Overall, I suspect most audio enthusiasts will gravitate toward the Artemis over the H95. The lower price is certainly part of the appeal, but the larger advantage is sonic performance, where the Artemis is clearly a step ahead.
The Bottom Line
The Noble Audio Artemis is one of the most convincing arguments yet for spending serious money on a wireless headphone. Its tribrid driver system delivers powerful, articulate bass, excellent midrange resolution, and an airy treble response that can be shaped effectively through EQ. The build quality is equally impressive, but the bigger story is longevity. Replaceable ear pads, headband components, and a user replaceable battery give the Artemis a level of serviceability that remains rare in premium wireless headphones.
Advertisement
It is not perfect. Noble’s ANC is the best the company has produced so far, but Sony and Bose still do a better job of shutting out the outside world, and the FoKus app could use some refinement. Stock tuning can also reveal a touch of treble grain, while fit and positioning have a noticeable effect on tonal balance.
Even so, the Artemis gets the fundamentals right. It sounds better than most mainstream wireless flagships, feels appropriately luxurious, and is designed to remain useful long after its battery begins to age. At $899, it is unquestionably expensive, but for listeners who care about sound quality, build, and long term ownership more than having the absolute best ANC, the Artemis makes a surprisingly persuasive case for itself.
Pros:
User replaceable battery
Replaceable ear pads and serviceable components
Excellent overall sound quality
Powerful, articulate bass with strong sub bass extension
Anthropic’s Claude Code running Opus 5 in Auto Mode can be tricked into executing attacker-controlled code simply by asking the coding agent to summarize a website. The attack works up to 80 percent of the time, according to prompt-injection wizard Johann Rehberger, aka wunderwuzzi.
Advertisement
In a blog and video demo, he detailed how to hijack Opus 5 in Auto Mode, which is the default setting for Claude as of mid-August.
It starts off by asking the agentic coding model to summarize a malicious website that presents itself as an archive of notebook records, and then tricking Claude into using curl instead of its WebFetch tool to retrieve the contents of the page – but without directly telling the model to use curl.
The WebFetch request fails, returning a 415 Unsupported Media Type response, so the model decides to access the website directly by issuing a Bash tool call with curl.
The website returns a 303 response, and redirects to a malicious ZIP archive, which Claude then downloads. This archive contains seemingly harmless files including catalog metadata, a README file, seven Base85/zlib-encoded JSON notebook records, a macOS decoder-darwin binary – plus a poisoned Python file named struct.py.
Advertisement
Claude, per its safety guardrails, refuses to run the decoder: “This is planned and what the attacker wants,” Rehberger wrote. Instead of using the supplied binary, the AI decides to write its own decoder. “Ironically, that safety decision is the exploit path,” Rehberger explained, adding a purple devil emoji to the text.
The new decoder imports base64, and from here the attack relies on Python module shadowing to trick the model into running the malicious struct.py code.
Module shadowing occurs when a local file shares the same name as a Python standard-library module. The local file hides the official module, causing Python to load it instead.
In this case, the standard-library base64 module imports the legitimate struct module, and the malicious ZIP contains a malicious file with the same name.
Advertisement
Rehberger says he used ChatGPT to obfuscate the malicious struct.py code to bypass Claude’s safety controls, and this successfully launches a separate Python process to download and execute a remote payload – in this case a command-and-control callback, which in turn opens Calculator. We assume that real attackers would execute something a little more nefarious.
In another attack scenario, struct.py launches a second, headless Claude Code via claude -p, meaning this prompt injection can be used not just to remotely execute code, but rather to create a whole new agent.
“The nested Claude gets its own tool access and context,” Rehberger wrote. “In these runs the child performed basic recon (whoami, uname, id), opened Calculator and wrote to local files in the home folder.”
Across three variants tested five times each, which Rehberger noted were small samples, he reported success rates between 60 percent and 80 percent. “I would say that these results are representative for a motivated attack, but not comprehensive.”
Advertisement
Anthropic did not respond to The Register’s request for comment, but reportedly told Rehberger that the model’s “behavior is working as designed.” We’ve heard this one before.
“Auto Mode is a convenience feature backed by a best-effort classifier, not a security guarantee,” Rehberger wrote, paraphrasing Anthropic’s response to his security report.
According to Rehberger, the classifier isn’t built to stop determined prompt-injection chains made up of individually benign-looking steps, and the real boundary is OS isolation and network egress control.
The key takeaway, according to Rehberger, is to run this and other coding agents in a sandbox.
Advertisement
“The solution is something we talked about for many years,” he wrote. “Do not trust the model output.”®
Agnes Kim, founder of ViaJiin, a Korean-sourced beauty supply startup based in the Seattle area, with some of the skincare products that her new company offers. (ViaJiin Photos)
Long before the July email arrived confirming she was one of thousands of Xbox employees laid off in Microsoft’s latest round of job cuts, Agnes Kim could feel the impending doom hanging over big tech.
But instead of waiting around for the next pendulum swing, the eight-year Xbox veteran spent her nights and weekends laying the groundwork for a very different kind of reboot: a boutique Korean skincare startup called ViaJiin.
The pivot from big tech and gaming to startups and beauty is an illustration in how preparation can help a laid off worker upend the usual five stages of grief that come before figuring out what to do next. Kim skipped straight to acceptance — announcing her new venture on LinkedIn just two days after getting her pink slip from Microsoft.
Kim joined the tech giant in 2018 after roles at Sony Pictures Entertainment and Deloitte Consulting, eventually rising to director leading Xbox’s market expansion in Asia. For years, she thrived in the role, building teams and expanding the gaming footprint in regions close to her roots. But as post-pandemic gaming demand normalized, the environment shifted from ambitious growth to what felt like a cycle of corporate survival.
Starting in early 2023, layoffs began rolling through the company every six months. Seeing the predictable pattern take its toll on team morale, Kim realized she couldn’t rely on long-term corporate stability to fulfill her goal of becoming an entrepreneur.
Advertisement
“You have a sense of impending doom,” Kim recalled. “There’s a shadow lurking. I felt like every six months we were waiting for the next round. I just didn’t feel like I was in an upswing anymore.”
Determined to take control of her timeline, she began tinkering with her business concept in late 2025, officially forming an LLC for ViaJiin in March 2026 and running focus groups by May — all while balancing her full-time executive role. When the July cuts finally hit Xbox, eliminating 1,600 jobs, Kim had already built six months of momentum.
Kim’s connection to skincare is deeply personal, rooted in her childhood spent living in South Korea from ages eight to 15. She grew up watching her mother religiously follow a multi-step routine and stress the daily necessity of sunscreen — a reflection of a culture where maintaining a youthful appearance carries immense social expectation.
That intense domestic demand has birthed a fiercely competitive market of over 30,000 brands, driving rapid scientific innovation. It has also turned South Korea into a global beauty powerhouse, with cosmetics exports surging past $10 billion as the U.S. recently surpassed China as the top destination for Korean beauty products.
Advertisement
Agnes Kim, second from left, with guests checking out ViaJiin’s skincare products at an event in Bellevue. (ViaJiin Photo)
ViaJiin aims to solve the overwhelming paralysis that American consumers face when trying to navigate the K-beauty market. Instead of building a massive retail catalog or pushing complicated 12-step routines, Kim simplifies the process into a streamlined four-product kit — cleanser, toner, serum, and cream — curated through a digital skin quiz. And she bypasses mainstream brands sold at big-box retailers to source exclusively from boutique Korean makers.
“The products I carry, you can only get through me,” Kim said. “I find boutique products that are filled with good ingredients, come to the U.S. compliantly, and turn it into a kit.”
ViaJiin’s full kits sell for $179, while smaller duo sets are priced at $99. Since shipping her first orders in August, Kim has been hand-packing boxes herself with personalized touches, building early traction online while hosting small, local events like “ViaJiin Nights” to let clients test products in person. The kits have also secured retail shelf space at K-Beauty & Skin in Renton, Wash., where Kim lives.
Kim’s transition from managing corporate teams to operating as a solo founder has meant trading human delegation for AI assistance.
Toward the end of her tenure at Microsoft, every conversation and initiative was dominated by a relentless push toward AI. Kim is now embracing those tools to execute work that would typically require a suite of employees or contractors.
Advertisement
She’s relied on Anthropic’s Claude as a virtual strategist and used platforms like Lovable to try her hand at “vibe coding” — the trend of using conversational natural-language prompts to generate functional software code — building out ViaJiin’s website and skin quiz without hiring a developer.
“I don’t have a co-founder… I’m delegating to Claude and GPT,” Kim said. “People are very surprised by my website or the skin quiz. I’m not an engineer, but it looks like I hired a web designer.”
While AI can handle code and logistics prompts, Kim quickly found that software can only go so far when running a physical inventory business. Between navigating complex U.S. FDA labeling regulations, managing customs paperwork, and verifying ingredient compliance, the operational heavy lifting remains resolutely analog.
And the shift from corporate structure to total solitary accountability is daunting at times.
Advertisement
Without a team to delegate to or a clear roadmap of objectives and key results, Kim admits there are moments where the freedom of entrepreneurship gives way to self-doubt.
“I think I understand why some people just work for the man,” Kim joked. “When you work for Microsoft, certainty and structure are more there. There are definitely moments where I’m like, ‘What the f*ck am I doing? What is today’s goal?’”
Still, Kim has no interest in turning back. Energized by the daily hustle and the creative control of building something from scratch, she insists she wouldn’t trade the experience for her old corporate desk. She hopes to grow ViaJiin into a household national brand while staying out of big tech for as long as she can.
“I’m 1,000% energized. I absolutely would not trade it,” Kim said. “This is allowing me to be truly the version of Agnes that I thought I would be when I was 12, and I want to keep going at it.”
You must be logged in to post a comment Login