News
OpenAI servers experienced mass downtime yesterday, causing chaos among its most loyal users for well over 10 hours.
For six hours straight, I sat at my desk live-blogging the fiasco here on TechRadar, trying to give as many updates as possible to an outage that felt, for many, as if they had lost a piece of themselves.
You see, I write about consumer AI, highlighting all the best ways to use AI tools like ChatGPT, Gemini, and Apple Intelligence, yet outside of work, these incredibly impressive platforms have yet to truly make an impact on my life.
As someone who’s constantly surrounded by AI news, whether that’s the launch of new Large Language Models or the latest all-encompassing artificial intelligence hardware, the last thing I want to do outside of work is use AI. The thing is, the more AI develops at this rapid pace, the more impossible it becomes to turn a blind eye to the abilities that it unlocks.
In the creative world, you’ll stumble across more AI skeptics than people who shout from the rooftops about how great it is. And that’s understandable, there’s a fear of how AI will impact the jobs of journalists like me, and there’s also a disdain for the sanitized world it’s creating via AI-slop or robotically-written copy.
But the same skepticism often overlooks the positives of this ever-evolving technology that gives humans new ways to work, collect their thoughts, and create.
After six hours of live blogging and thousands of readers reaching out with their worries surrounding the ChatGPT server chaos, as well as discussing what they use the chatbot for, I’ve come away with a completely new perspective on AI.
Yes, there are scary elements; the unknown is always scary, but there are people who are truly benefiting from AI, and some in ways that had never even crossed my mind.
More than a chatbotAn hour into live blogging the ChatGPT outage, and I was getting bored of repeating, “It’s still down” in multiple different ways. That was when I had an idea: if so many people were reading the article, they must care enough to share their own reasons for doing so.
Within minutes of asking readers for their opinions on the ChatGPT outage, my inbox was inundated with people from around the globe telling me how hard it was to cope without access to their trusty OpenAI-powered chatbot.
From Canada to New Zealand, Malaysia to the Netherlands, ChatGPT users shared their worries and explained why AI means so much to them.
Some relied on ChatGPT to study, finding it almost impossible to get homework done without access to the chatbot. Others used ChatGPT to help them with online dating, discussing conversations from apps like Tinder or Hinge to ensure the perfect match. And a lot of people reached out to say that they spent hours a day speaking with ChatGPT, filling a void, getting help with rationalizing thoughts, and even helping them to sleep at night.
One reader wrote me a long email, which they prefaced by saying, “I haven’t written an email without AI in months, so I’m sorry if what I’m trying to say is a bit all over the place.”
Those of us who don’t interact with AI on a regular basis have a basic understanding of what it can do, often simplifying its ability down to answering questions (often wrongly), searching the web, creating images, or writing like a robot.
But that’s such an unfair assessment of AI and the way that people use it in the real world. From using ChatGPT to help with coding, allowing people who have never been able to build a program an opportunity to do so, to giving those who can’t afford a professional outlet for their thoughts a place to speak, ChatGPT is more capable than many want to accept.
(Image credit: Shutterstock/Rokas Tenys)ChatGPT and other AI tools are giving people all around the world access to something that, when used correctly, can completely change their lives, whether that’s by unlocking their productivity or by bringing them comfort.
There’s a deeply rooted fear of AI in the world, and rightfully so. After all, we hear on a regular basis how artificial intelligence will replace us in our jobs, take away human creativity, and mark the beginning of the robot uprising.
But would we collectively accept it more if those fears were answered? If the billionaires at the top were to focus on highlighting how AI will improve the lives of the billions of people struggling to cope in this hectic world?
AI should be viewed as the key to unlocking human creativity, freeing up our time, and letting us do less of the mundane and more of enjoying our short time on this planet. Instead, the AI renaissance feels like a way to make us work harder, not smarter, and with that comes an intense amount of skepticism.
After seeing just how much ChatGPT has impacted the lives of so many, I can’t help but feel like AI not only deserves less criticism, but it deserves more of an understanding. It’s not all black and white, AI has its flaws, of course it does, but it’s also providing real practical help to millions of people like nothing I've seen before.
You might also likeApple Vision Pro is unquestionably one of the most powerful pieces of consumer hardware Apple has ever built, but the pricey gadget is still struggling to connect with consumers. And that's a shame because the generational-leaping visionOS 26 adds even more eye-popping features to the $3,500 headset, which I think you'd struggle to find with any other mixed reality gear.
Apple unveiled the latest Vision Pro platform this week as part of its wide-ranging WWDC 2025 keynote, which also introduced a year-OS naming system. For some platforms like iOS, the leap from, say, 18 to 26 wasn't huge, but for the toddler visionOS 2, it was instantly thrust into adulthood and rechristened visionOS 26.
This is not a reimaging of visionOS, and that's probably because its glassiness has been amply spread across all other Apple platforms in the form of Liquid Glass. It is, though, a deepening of its core attributes, especially around spatial computing and imagery.
I had a chance to get an early hands-on experience with the platform, which is notable because Vision Pro owners will not be seeing a visionOS 26 Public beta. Which means that while iPhone, iPad, Apple Watch, and Apple TV owners are test-driving OS 26 platform updates on their favorite hardware, Vision Pro owners will have a longer wait, perhaps not seeing these enhancements until the fall. In the interim, developers will, of course, have access for testing.
Since much of the Vision Pro visionOS 26 interface has not changed from the current public OS, I'll focus on the most interesting and impactful updates.
See "me"(Image credit: Apple)During the keynote, Apple showed off how visionOS 26 Personas radically moves the state of the art forward by visually comparing a current Persona with a new one. A Vision Pro Persona is a virtual, live, 3D rendering of your head that tracks your movements, facial expressions, and voice. It can be used for communicating with other people wearing the headgear, and it's useful for calls and group activities.
Apple has been gradually improving Personas, but visionOS 26 is a noticeable leap, and in more ways than one.
You still capture your Persona using the front-facing 3D camera system. I removed my eyeglasses and held the headset in front of my face. The system still guides you, but now the process seems more precise. I followed the audio guidance and looked slowly up, down, left, and right. I smiled and raised my eyebrows. I could see a version of my face faintly on the Vision Pro front display. It's still a bit creepy.
(Image credit: Future)I then put the headset back on and waited less than a minute for it to generate my new Persona. What I saw both distressed and blew me away.
I was distressed because I hate how I look without my glasses. I was blown away because it looked almost exactly like me, almost entirely removing the disturbing "uncanny valley" look of the previous iterations. If you ever wonder what it would be like to talk to yourself (aside from staring at a mirror and having a twin), this is it.
There was a bit of stiffness and, yes, it fixed my teeth even though part of my setup process included a big smile.
It was easy enough to fix the glasses. The Personas interface lets you choose glasses, and now the selection is far wider and with more shades. I quickly found something that looked almost just like mine.
With that, I had my digital doppelganger that tracked my expressions and voice. I turned my head from side to side and was impressed to see just how far the illusion went.
Facing the wall(Image credit: Apple)One of the most intriguing moments of the WWDC Keynote was when they demonstrated visionOS 26's new widget capabilities.
Widgets are a familiar feature on iPhones, iPads, and Macs, and, to an extent, they work similarly on Vision Pro, but the spatial environment takes or at least puts them in new and unexpected places.
In my visionOS 26 demo experience, I turned toward a blank wall and then used the new widget setup to pin a clock widget to the wall. It looked like an actual clock hanging on the wall, and with a flip of one setting, I made it look like it was inset into the wall. It looked real.
On another wall, I found a music widget with Lady Gaga on it. As I stepped closer, a play button appeared in the virtual poster. Naturally, I played a little Abracadabra.
Another wall had multiple widgets, including one that looked like a window to Mount Fiji; it was actually an immersive photo. I instinctively moved forward to "look out" the window. As the vista spread out before me, the Vision Pro warned me I was getting too close to an object (the wall).
I like Widgets, but temper the excitement with the realization that it's unlikely I will be walking from room to room while wearing Vision Pro. On the other hand, it would be nice to virtually redecorate my home office.
An extra dimension(Image credit: Apple)The key to Vision Pro's utility is making its spatial capabilities useful across all aspects of information and interaction.
visionOS 26 does that for the Web with spatial browsing, which basically can turn any page into a floating wall of text and spatially-enhanced photos called Spatial Scenes.
(Image credit: Apple)visionOS 26 handles the last bit on the fly, and it's tied to what the platform can do for any 2D photo. It uses AI to create computational depth out of information it can glean from your flat image. It'll work with virtually any photo from any source, with the only limitation being the source image's original resolution. If the resolution is too low, it won't work.
I marveled at how, when staring at one of these converted photos, you could see detail behind a subject or, say, an outcropping of rock that was not captured in the original image but is inexplicably there.
It's such a cool effect, and I'm sure Vision Pro owners will want to show friends how they can turn almost all their photos into stereoscopic images.
Space timeI love Vision Pro's excellent mixed reality capabilities, but there's nothing quite like the fully immersive experience. One of the best examples of that is the environments that you enable by rotating the crown until the real world is replaced by a 360-degree environment.
visionOS 26 adds what may be the best environment yet: a view of Jupiter from one of its moons, Amalthea. It's beautiful, but the best part of the new environment is the control that lets you scroll back and forth through time to watch sunrises and sunsets, the planet's rotation, and Jupiter's dramatic storms.
This is a place I'd like to hang out.
Of course, this is still a developer's beta and subject to significant change before the final version arrives later this year. It's also another great showcase for a powerful mixed reality headset that many consumers have yet to try. Perhaps visionOS 26 will be the game changer.
You might also like- E2 SSDs aim to balance storage performance capacity and efficiency
- New form factor fits rising demand for warm tier data storage
- High density flash could reduce reliance on hard drives long term
As workloads shift and cold data heats up under AI and analytics demands, the traditional split between high-speed SSDs and cost-effective hard drives is no longer serving every use case.
A new SSD form factor known as E2 is being developed to tackle the growing gap in enterprise data storage. Potentially delivering up to 1PB of QLC flash per drive, they could become the middle-ground option the industry needs.
StorageReview claims the E2 form factor is being designed with support from key players including Micron, Meta, and Pure Storage through the Storage Networking Industry Association and Open Compute Project.
Solid speeds, but not cutting-edgeE2 SSDs targets “warm” data - information that’s accessed often enough to burden hard drives but which doesn’t justify the cost of performance flash.
Physically, E2 SSDs measure 200mm x 76mm x 9.5mm. They use the same EDSFF connector found in E1 and E3 drives, but are optimized for high-capacity, dense deployments.
A standard 2U server could host up to 40 E2 drives, translating into 40PB of flash in a single chassis. StorageReview says these drives will connect over PCIe 6.0 using four lanes and may consume up to 80W per unit, although most are expected to draw far less.
Performance will reach 8-10MB/s per terabyte, or up to 10,000MB/s for a 1PB model. That’s faster than hard drives but not in the same class as top-end enterprise SSDs. E2’s priorities will instead be capacity, efficiency, and cost control.
Pure Storage showed off a 300TB E2 prototype in May 2025 featuring DRAM caches, capacitors for power loss protection, and a flash controller suited for this scale. While current servers aren't yet ready for this form factor, new systems are expected to follow.
It’s fair to say E2 won't replace hard drives overnight, but it does signal a shift. As the spec moves toward finalization this summer, vendors are already rethinking how large-scale flash can fit into modern infrastructure.
You might also like- These are the fastest SSDs you can buy right now
- And these are the largest SSDs and hard drives on the market
- Here's why 100TB+ SSDs will play a huge role in ultra LLMs in the near future
- Apple has just released the world's first 'haptic' trailer for its F1 movie
- The trailer vibrates your phone in time with action sequences
- The F1 movie pulls into theaters internationally from June 25
I thought I'd seen every movie trailer gimmick by now, but Apple has just produced a novel one for its incoming F1 movie – a 'haptic' trailer that vibrates your iPhone in time with the on-screen action.
If you have an iPhone (Android fans are sadly excluded from the rumble party) head to the haptic trailer for F1: The Movieto open it in the Apple TV app. You'll then be treated to two minutes of vibrations that's probably also a taste of what it's like to being a celebrity in the middle of a social media storm.
The trailer's 'haptic' experience was actually better than I was expecting. I assumed it would be a simple, one-dimensional rumble that fired up during race sequences, but it's a little more nuanced than that.
To start with, you feel the light vibration of a driver's seat belt being fastened, before the vibrations ramp up for the driving and crash sequences. There's even a light tap to accompany Brad Pitt's character Sonny Hayes moodily bouncing balls against a wall as he ponders coming out of retirement for one last sports movie trope.
Sure, it isn't exactly an IMAX experience for your phone, but if ever there was a movie designed for a haptic movie trailer, it's Apple's F1 movie...
One last Pitt stopApple's F1 movie was also the star of its recent WWDC 2025 event, with the livestream opening with Craig Federighi (Apple's Senior Vice President of Software Engineering) donning a helmet before doing a lap around the roof of its Apple Park building.
There's currently no date for the movie to stream on Apple TV+, with the focus currently on its imminent theater premiere. It officially opens internationally on June 27, but there are some special, one-off screenings in IMAX theaters on June 23 (in North America) and June 25 (internationally) for keen fans who signed up on the movie's official website.
The trailers so far suggest that F1 is going to effectively be Top Gun: Maverick set on a race track – and with both movies sharing the same director (Joseph Kosinski) and screenplay writer (Ehren Kruger), that seems like a pretty safe bet. F1 World Champion Lewis Hamilton was also involved to help amp up the realism.
If the haptic-powered trailer has whetted your appetite, check out our interview with Damson Idris who also stars in F1 and gave us a behind-the-scenes look at what the movie was like to film. Hint; they used specialized tracking cars to help nail the demanding takes flawlessly.
You might also like- ‘I had the craziest spin of my life’: F1 movie star on becoming a race car driver for the upcoming Apple and Warner Bros film
- Apple and Warner Bros. will take you behind-the-scenes of a racing cockpit in this new thrilling F1 movie trailer
- F1 and the Grand Prix invade Apple Maps – but where's Brad Pitt?
If you’re a sports fan like me, you may have had some complaints in the past about your TV when trying to watch sports. Whether it’s reflections while watching a game in the afternoon or blurring during fast motion, something always seems to need tweaking.
Another issue: a TV that appears dim, with a flat-looking image, particularly for field sports such as football and rugby.
Even the best TVs can struggle with sport, but thankfully, there’s a TV tech that’s ideal for sports fans: mini-LED.
Mini-LED: perfect for sports fansMini-LED TVs are not only becoming increasingly popular but also more affordable. This tech delivers an improved picture over standard LED by using backlights with smaller LEDs (hence the mini part).
By miniaturizing the LEDs, a higher number can be used, which results in increased brightness. It also allows for a higher number of local dimming zones in the backlight, which helps to boost contrast and improve black uniformity.
Mini-LED TVs can hit significantly higher brightness levels than other TV panel types, with 2,500 - 4,000 nitspeaks possible in flagship models. But for sports fans, it’s fullscreen brightness – the level of brightness that the TV can sustain over its entire screen area – that matters most, and once again, mini-LED TVs here regularly beat other panel types, including the best OLED TVs.
To provide an example of that from our TV testing, we regularly measure fullscreen brightness levels of between 580 - 800 nits on the best mini-LED TVs. But even the brightest OLED TV we’ve tested, the LG G5, topped out at 331 nits in our fullscreen measurement.
I’ve picked three models below that are examples of the best mini-LED TVs for sports.
1. Samsung QN90F(Image credit: Future)The Samsung QN90F is the perfect TV for sports. Not only does it deliver exceptionally high brightness levels – 2,086 nits peak and 667 nits fullscreen in Filmmaker Mode – but it has a Glare-Free screen (first introduced in the Samsung S95D OLED) that effectively eliminates reflections, making it perfect for afternoon sports watching.
The QN90F also delivers the superb motion handling that's essential for fast-paced sports. Even for movies, we found we could get smooth motion, with no sign of the dreaded ‘soap opera effect’, by setting both Blur Reduction and Judder Reduction to 3.
The QN90F delivers vibrant colors, strong contrast and realistic textures for a brilliant picture. And when viewing from an off-center seat, there’s little sign of the backlight blooming that results in contrast fade, meaning it’s great for watching in large groups.
The QN90F is a premium-priced TV, with the 65-inch model we tested priced at $2,499.99 / £2,499 / AU$3,499, but if you’re a sports fanatic, it’s worth the investment. Plus, you can expect prices to drop at some point in the near future.
2. Amazon Fire TV Omni Mini-LED (Image credit: Future)When I first began testing the Amazon Fire TV Omni Mini-LED, I didn’t anticipate it would be such a good TV for sports. But in its preset Sports mode with Smoothness (Judder Reduction) set to 4 and Clarity (Blur Reduction) set to 10, sports looked impressively smooth. Color was also surprisingly accurate in that mode, which is unusual as I’ve found the Sports mode makes colors look oversaturated and garish on most TVs.
Something unique about the Omni Mini-LED is that it’s nearly ready out of the box for sports. In contrast, I found when testing competing models such as the Hisense U6N and Hisense U7N that more setup was required to get sports looking right.
The Amazon Omni mini-LED is a significantly more affordable TV than the Samsung QN90F, with its 65-inch model often discounted down to $949.99 / £949.99. It may not have the same level of sports prowess as the Samsung QN90F, but it’s great for the money.
3. TCL QM7K / TCL C7K Image 1 of 2TCL QM7K - US (slide 1) & TCL C7K - UK (slide 2) (Image credit: Future)Image 2 of 2(Image credit: Future)This entry is a hybrid as the TCL model name (and specs) will vary depending on which side of the pond you’re on. Either way, it’s the mid-range model in TCL’s 2025 mini-LED lineup.
Both of these TVs deliver exceptional brightness at a mid-range price, with the TCL QM7K and TCL C7K hitting 2,350 nits and 2,784 nits HDR peak brightness, respectively. More importantly, they hit 640 nits and 678 nits HDR fullscreen brightness, respectively – very good numbers for watching sports in bright rooms.
These TVs require some motion setup. Since I'm based in the UK, I tested the C7K, and I found that I needed to tweak the Sports or Standard picture mode by setting Blur Reduction to 3 and Judder Reduction to 6. I also needed to lower the color setting in Sports, as it was oversaturated in its default settings.
Once this was completed, the C7K was a solid TV for sports. It isn’t quite as effective as the two models above, but it is still a very good mid-range option overall. If the QM7K is anything like its UK counterpart, then the story for that model will be the same.
Again, for the 65-inch models of these two sets, you’re looking at paying $999 / £1,099. That’s a similar price to the Amazon Omni Mini-LED, which has the best motion of the two, but with the TCL, you’re getting that extra hit of brightness.
You might also like- I compared a new cheap mini-LED TV to a mid-range model, and here’s why upgrading is more than worth a small price increase
- Sony Bravia 9 review - an ultra-premium mini-LED TV with performance to match
- How far mini-LED TV tech has evolved in 9 years and the 'dark side' of Netflix – what I learned on my visit with TCL
- Binarly spotted a legitimate utility, trusted on most modern systems utilizing UEFI firmware, carrying a flaw
- The flaw allowed threat actors to deploy bootkit malware
- Microsoft patched it the June 2025 Patch Tuesday cumulative update
Microsoft has fixed a Secure Boot vulnerability that allowed threat actors to turn off security solutions and install bootkit malware on most PCs.
Security researchers Binarly recently discovered a legitimate BIOS update utility, signed with Microsoft’s UEFI CA 2011 certificate. This root certificate, used in the Unified Extensible Firmware Interface (UEFI) Secure Boot process, plays a central role in verifying the authenticity and integrity of bootloaders, operating systems, and other low-level software before a system boots.
According to the researchers, the utility is trusted on most modern systems utilizing UEFI firmware - but the problem stems from the fact it reads a user-writable NVRAM variable without proper validation, meaning an attacker with admin access to an operating system can modify the variable and write arbitrary data to memory locations during the UEFI boot process.
Microsoft finds 13 extra modulesBinarly managed to use this vulnerability to disable Secure Boot and allow any unsigned UEFI modules to run. In other words, they were able to disable security features and install bootkit malware that cannot be removed even if the hard drive is replaced.
The vulnerable module had been circulating in the wild since 2022, and was uploaded to VirusTotal in 2024 before being reported to Microsoft in late February 2025.
Microsoft recently released the June edition of Patch Tuesday, its cumulative update addressing different, recently-discovered, vulnerabilities - among which was the arbitrary write vulnerability in Microsoft signed UEFI firmware, which is now tracked as CVE-2025-3052. It was assigned a severity score of 8.2/10 (high).
The company also determined that the vulnerability affected 14 modules in total, now fixing all of them.
"During the triage process, Microsoft determined that the issue did not affect just a single module as initially believed, but actually 14 different modules," Binarly said. "For this reason, the updated dbx released during the Patch Tuesday on June 10, 2025 contains 14 new hashes."
Via BleepingComputer
You might also like- The first UEFI bootkit malware for Linux has been detected, so users beware
- Take a look at our guide to the best authenticator app
- We've rounded up the best password managers
- NordPass and NordStellar reviewed terabytes of data
- The analysis uncovered poor password practices in the healthcare industry
- Organizations are lacking staff training and strong policies
Hygiene in hospitals and clinics is essential, but cyber-hygiene - despite being equally important - is constantly being neglected, experts have warned.
A report from NordPass and NordStellar has claimed weak password practices are “dangerously common” in the healthcare industry.
Based on a review of 2.5TB of data extracted from various publicly available sources (including the dark web), the two organizations found that different medical institutions, including private clinics and hospital networks, all rely on “predictable, recycled, or default passwords” to protect critical systems. As a result, sensitive patient data, and possibly their health, is placed at immense risk.
Carelessness“When the systems protecting patient data are guarded by passwords like ‘123456’ or ‘P@ssw0rd,’ that’s a critical failure in cybersecurity hygiene. In a sector where both privacy and uptime are vital, this kind of carelessness can have real consequences,” said Karolis Arbaciauskas, head of business product at NordPass.
The report also lists the most frequently used passwords identified in the healthcare sector. If you’re using any of these (or a variant), make sure to change them for something tougher to crack:
- fabrizio19
- 123456
- Melu3@12345
- @Vow2017
- Mercury9.Venus8
- password
- Marty1508!
- Carlton@1988
- 12345678
- @Vowcomm2018
- papa
- 12345
- Durson@123
- P@ssw0rd
- Simetrica
- Raffin2209!
- Asspain28#
- Smith
- neuro
- default
The teams warn passwords that reflect personal names, simple number patterns, or default configurations, are all prime targets for brute-force and dictionary attacks, in which cybercriminals automate the process, and try out countless combinations until they break in.
To make matters even worse - one break-in is more than enough to wreak havoc, as lateral movement can compromise entire networks, expose sensitive data, and result in different malware and ransomware infections.
The report stresses that healthcare institutions “lack clear password management policies or staff training,” which is why they are recommended to enforce strong password policies, eliminate the use of default or role-specific passwords, use a business-grade password manager, train the staff, and introduce 2FA wherever possible.
You might also like- Major data breach at healthcare giant Yale Health affects 5.5 million people - here's what we know
- Take a look at our guide to the best authenticator app
- We've rounded up the best password managers