Technology

• AMD announces FSR 4’s exclusivity to RDNA 4 Radeon RX 9070 series GPUs
• It’s unclear why the RX 9060 won’t have FSR 4
• Team Red’s main rival Nvidia has made DLSS 4 backwards compatible

There were hardly any doubts over AMD conceding ground to Nvidia at CES 2025, with the unveiling of new GPUs on both sides overshadowed by AMD’s recent shift to a focus on mid-range options, but Team Red has also confirmed that FSR 4 will be exclusive to the new RDNA 4 GPUs - and I’ll be honest, that sucks.

As noted by VideoCardz, FSR 4 will only be available to Radeon RX 9070 series GPU owners, including games that already have FSR 3.1 support, such as Call of Duty: Black Ops 6. This is in stark contrast to FSR 3, which is available for all AMD users and beyond thanks to its open-source design, serving as a huge benefit for handheld gaming PCs as well.

Now, FSR 4's technology is said to be 'developed for RDNA 4' (the hardware microarchitecture of the new Radeon GPUs) which could be the main reason for this exclusivity. Despite this, it isn't exactly clear yet why the upscaling method is being omitted from other hardware, especially after Nvidia's announcement of DLSS 4 being available for all existing RTX GPUs - a step in the right direction for Nvidia on this occasion.

(Image credit: AMD) Should we be worried about AMD's position in the GPU market?

In previous articles, I’ve made my frustrations clear regarding Nvidia’s decision to make DLSS 3 and Frame Generation exclusively available for RTX 4000 series GPU owners. Team Green has now made amends with DLSS 4, which will be available for all RTX GPUs except for the Frame Generation feature, which remains exclusive to the RTX 4000 and the recently revealed RTX 5000 series (with the new-and-improved ‘Multi Frame Generation’ only on RTX 50 series).

This brings me to AMD’s decision to focus on mid-range GPUs this generation, which was a concern, to say the least - but after hearing that FSR 4 will only be available to RX 9070 series GPUs, I honestly believe Team Red could be in big trouble.

Even with a fresh focus on the midrange space, the lack of backward compatibility will hurt AMD’s gamer-friendly image - especially with Nvidia’s claim of the RTX 5070 performance being equal to the RTX 4090’s (most likely while using Multi Frame Generation and DLSS 4), priced at $549 / £539 / AU$1,109. Pricing on AMD’s new cards has yet to be revealed, but if it can’t undercut Nvidia’s new midrange champion, Team Red could be in dire straits.

If AMD’s RX 9060 doesn’t have access to FSR 4 (which at the very least is implied in the FSR 4 slide pictured above), the RTX 5070 will be a much easier choice at an affordable price. While FSR 4 could prove to be a massive leap over the previous rendition, it’s hard to see this beating Nvidia’s offerings unless the new Radeon GPUs are cheaper than the new Blackwell GPUs.

You may also like...

• Nvidia GeForce Now set to get a lot better on Steam Deck – and that's great news for playing AAA games on Valve's handheld
• AMD Ryzen 9 9950X3D and 9900X3D announced at CES 2025
• AMD announces new Radeon RX 9070 XT and RX 9070 graphics cards at CES 2025

• A new teaser for The Last of Us season 2 was revealed at CES 2025
• The hit series, which returns in April, looks like it'll directly adapt large swathes of the video game series
• Fans are already worried about the reaction that Kaitlyn Dever will get for playing Abby

A new trailer for The Last of Us season 2 has got fans waxing lyrical over every small but significant detail that makes up its near 70-second runtime.

The latest round of footage, which accompanied The Last of Us season 2's release month announcement last night (January 6) at CES 2025, is jam-packed with quick-fire clips that suggest it'll be as true to its identically titled game series as possible.

Considering how The Last of Us season 1 largely mirrored what happens in the first game in Naughty Dog's video game franchise, plus the involvement of The Last of Us creative director Neil Druckmann, that's to be expected. Nevertheless, it's pleasing to see how much of The Last of Us Part II is being adapted for the show's sophomore outing.

Ever since the sizzle reel-style trailer dropped online, fans have gushed over how authentic it looks, too. Threads that popped up on The Last of Us and R/Television sub-Reddits are full of comments from fans new and old discussing the level of detail on show. Where the latter group are concerned, the inclusion of a large, silver hook, the sound of that haunting siren, a snippet of that dance sequence involving Isabela Merced's Dina and Bella Ramsey's Ellie, and quotes pulled directly from The Last of Us Part II are call-backs to hugely significant moments in the action-adventure survival horror game. I won't spoil what these mean for anyone who hasn't played the games, but believe me when I say they're incredibly important moments.

The Last of Us TV show's newest teaser comes four months after the first unsettling trailer for season 2 confirmed that I wasn't ready to have my heart broken all over again (if you know, you know). It also arrives two months after The Last of Us season 2 was confirmed to be launching on in early 2025. Thanks to a stinger at the end of this most recent sizzle reel, we know it'll be with us sometime in April.

Why are people worried for Kaitlyn Dever, the actor who'll play Abby in The Last of Us TV show?

(Image credit: HBO)

Exciting as this new teaser is, it's also led to the resurfacing of fan worries about one of the Max series' new cast members.

Last January, Kaitlyn Dever was cast as Abby, The Last of Us Part II's deuteragonist. Now, the video game series' fanbase isn't concerned about Dever's acting abilities – quite the opposite, in fact. The No One Will Save You star is incredibly talented and, as many have pointed out since season 2's latest footage was released, she seems tailor-made to play the morally complex character in one of the best Max shows' second chapter.

Some people still can't separate the artist from the art

So, what's everyone panicking about? Some simply don't like Abby because of an absolutely devastating moment in The Last of Us Part II that she's involved in (don't worry, I won't spoil anything here). Others, meanwhile, weren't enamored with being forced to play as her alongside Ellie in that game, either, especially when some gamers found Abby to be an unlikable character. The sad reality, though, is that a small portion of the franchise's fanbase hates Abby because she's a muscular woman and subsequently feel emasculated in comparison. It's a bizarre opinion to have, not least because Abby is a fictional character, but there's a toxic minority who resent her for it.

That brings us onto Dever's casting. Per IGN, Dever was given extra security while season 2 was being filmed because of the vitriol that's continually aimed at the character she's portraying. Clearly, some people still can't separate the artist from the art – and it's for this reason that the aforementioned Reddit threads were also full of commenters rightfully expressing concern for Dever's wellbeing ahead of season 2's release. I'm certainly among that number, and I hope – perhaps naively – that the Booksmart and Rosaline alumnus is spared the inevitable discourse that'll likely come her way once season 2 begins.

The Last of Us season 2 will launch on Max (US), Sky/Now TV (UK), and Binge (Australia) in April.

You might also like

• The Last of Us season 2 will be shorter than the first – here are 3 reasons why that's exciting for the hit Max show
• The Last of Us season 2 will contain brand-new material that was cut from the original games
• Here's everything we know about The White Lotus season 3

The Samsung Galaxy Ring was unveiled to much fanfare in early 2024. Hinted at the winter Galaxy Unpacked presentation and revealed at the Mobile World Congress tech conference in Barcelona, it didn't actually launch until July last year, and consequently is less than a year old. But the world of fitness and wearable tech is a fast-moving one, and the latest rumors indicate Samsung could have a new version of the best smart ring on the market ready for launch as early as the January 22 Samsung Galaxy Unpacked event.

A full release seems unlikely given the short life span of the Samsung Galaxy Ring, but it's very possible we could see at least a glimpse of the next version very soon.

So what could the Samsung Galaxy Ring 2 bring us for 2025? The current version is brilliant and trumps the Oura Ring 4 as our top pick because it doesn't require a subscription, but that doesn't mean there isn't plenty of room for improvement. So here's everything we know so far about the Samsung Galaxy Ring 2, and everything we'd like to see.

Samsung Galaxy Ring latest rumors and leaks Latest news

Samsung tipped to unveil the Galaxy Ring 2 at its January Unpacked event.

Cut to the chase

• What is it? The next smart ring from Samsung
• When is it out? Probably mid-to-late 2025, with a teaser unveiling in the meantime
• How much will it cost? Don't expect much change on the current price of $399/£399/around AU$750 - we're also hoping Samsung continues to eschew the subscription model

Samsung Galaxy Ring: Price and release date rumors expectations

(Image credit: Zachariah Kelly / TechRadar)

• No word of price change or increase
• Tipped for unveiling at Unpacked on January 22
• Might not go on sale until later this year

When it comes to release date and pricing information, we don't have too much to go on for the Samsung Galaxy Ring 2. There are no rumors or leaks to suggest Samsung will change the current pricing structure of the Galaxy Ring with its next model.

As such, we can probably expect it to cost around the same as last year: $399/£399/ or around AU$750. One crucial aspect of the pricing was Samsung's choice not to opt for a subscription model with the Galaxy Ring. It's a vital advantage over the rival Oura Ring range, and one we definitely hope sticks around this year.

There are a few more rumblings when it comes to a release timetable. A leak in November revealed that Samsung was "planning to launch its Galaxy Ring 2 model a bit earlier than originally scheduled." Of course, this is hard to decipher when we don't know when Samsung was originally planning to launch the Ring, and sooner could mean days, weeks, months, or even years. More helpfully, a December 2024 leak suggested we would get an "introduction" to the Samsung Galaxy Ring 2 at January's Unpacked event, which has now been confirmed for January 22.

As with the unveiling of the original model, this is likely to be a glimpse ahead of a full unveiling later in the year. The Galaxy Ring was shown off at Unpacked and MWC before its full unveiling in July 2024. We'd expect the Samsung Galaxy Ring 2 to follow a similar pattern.

Samsung Galaxy Ring 2: Probable design and features

(Image credit: Zachariah Kelly / TechRadar)

• Rumored to be thinner
• Samsung has patented an adjustable band to solve sizing
• Larger sizes also a possibility
• Longer battery life rumored

Samsung is unlikely to reinvent the wheel when it comes to smart ring design, which is severely limited in its form by virtue of the tech and its purpose. However, a couple of interesting design changes could be in the works.

According to the same leak which pointed to a sooner-than-expected launch, the new Samsung Galaxy Ring 2 will be thinner than its predecessor. Thankfully, that won't come at the expense of battery life, which the leak claims will also improve. The information also points to "more features," but details are scant on what could entail.

It was revealed in October that Samsung has patented a new type of smart ring with an adjustable band that could solve the sizing issues that currently dog the smart ring market. The adjustable inner band would change size, while the outer band remains fixed. It's not a perfect solution, but as a patent, this is likely years away and an unlikely inclusion for the Samsung Galaxy Ring 2.

Another, more confused rumor, points to the prospect of new Samsung Galaxy Ring sizes in the pipeline. These were first reported in September, with X leaker Max Jambor claiming the Galaxy Ring will be available in two additional sizes, 14 and 15, "in the coming weeks." Jambor has since revised his prediction stating the new sizes will be available in January.

The Galaxy Ring will be available in two additional sizes, 14 and 15, in the coming weeks pic.twitter.com/00vP18RAKqSeptember 20, 2024

While some outlets have taken this to mean that the Galaxy Ring 2 will feature new larger sizes, Jambor doesn't specify this, and a more in-depth report from 91Mobiles indicates these are additional sizes for the current model, rather than the Galaxy Ring 2.

Samsung Galaxy Ring 2: What we want to see

Thus, we have exhausted all of the concrete information out there about the Samsung Galaxy Ring 2. But what does the second iteration need to succeed? After months of testing and use, here are some things we'd like to see.

1. A leaf out of Circular's Sizing book

The Circular Ring 2 can be sized using a smartphone camera, not a plastic sizing kit. (Image credit: Circular)

The Circular Ring might not have set the world alight, but the Circular Ring 2 unveiled at CES 2025 features a massive upgrade in smart ring tech.

It's leaving behind physical plastic sizing kits in favor of a smartphone-powered Digital Ring Sizing function, and it's a feature we'd love to see come to every smart ring out there, especially the Galaxy Ring 2.

2. More workout detection

The Oura Ring 4 can automatically track 40 different types of activity. (Image credit: Future)

While the Galaxy Ring offers walking and running workout detection, the Oura Ring 4 offers 40 automatic workout profiles based on your tastes, and even works to pick up more challenges exercise profiles like rowing.

The Galaxy Ring 2 could definitely benefit from more automatic workout detection, which is a mainstay of the best smartwatches and the rival Oura.

3. More gesture controls

(Image credit: Samsung)

Samsung's Galaxy Ring features a limited number of gesture controls, a vital advantage over the Oura Ring lineup. We'd love to see Samsung expand the offering.

Right now there are two gestures, you can dismiss alarms, or take a picture or video, by using the double pinch gesture. More options including a function to answer phone calls would do just fine, thank you, but we'd love to see a remote screen control functionality – picture Tony Stark/Iron Man scrolling through holograms with a swipe of his hand.

4. 10-day battery life

10 days might be a big ask, but the Samsung Galaxy Ring's seven-day battery life can't compete with the eight days on offer from the Oura Ring 4. Even matching that would be a bonus, but we'd love to see double figures. As the leaks we've mentioned suggest, we can expect at least some battery life improvement with the Galaxy Ring 2, so here's hoping it's a hefty jump.

You may also like

• Samsung Galaxy Ring review: The top smart ring for Samsung users
• I spent 24 hours with the Samsung Galaxy Ring, and it's scarily good so far
• The Samsung Galaxy Ring is brilliant, but it misses the mark in three key areas

Maingear has long been a custom PC build staple for a lot of gamers, whether that's prestige products like the Maingear Turbo, the Maingear MG-1, or the dozen or so other product lines the New Jersey-based PC builder has put out over the years.

Well, you can honestly put all those aside, because the newest crop of Maingear PCs are in a class all their own.

Maingear is showing off three gaming PC product lines at CES 2025, and all three of them are stunners in their own right.

First, there's the Maingear Shroud Signature Collection, which continues Maingear's collaboration with the popular video game streamer Shroud.

The second series is the Maingear Apex Rush, and quite frankly, I'll skip over the details here for a moment, because I have a lot more to say about them in a minute.

Finally, there's the Maingear Apex Force, which is an absolute beast of a machine, complete with angled chrome piping like a bonafide hotrod.

I'll go through each in turn, so read on to learn more about why Maingear's newest gaming PCs are the best I've seen at CES this year.

The Maingear Shroud Signature Collection Image 1 of 4

(Image credit: Maingear)Image 2 of 4

(Image credit: Maingear)Image 3 of 4

(Image credit: Maingear)Image 4 of 4

(Image credit: Maingear)

First up is the Maingear Shroud Signature Collection, which is a pair of PCs built in collaboration with professional streamer and former Valorant pro Shroud.

The PC, which features a UniSheet heat-formed glass panel to showcase the components inside from the front and the side, comes in either white or black, features a built-in LCD screen inside the case, a glass GPU airflow director emblazoned with Shroud's signature (get it?), and a special badge on the PC case listing the PC's serialized order number and date of assembly.

Given the PC's namesake, the components inside are equally impressive.

At launch, it will come with the AMD Ryzen 7 9800X3D CPU, the top-tier Nvidia RTX 5000-series GPU (whichever is the highest you can get when you order), the ASRock Phantom Gaming X870E Nova WiFi motherboard, 48GB T-Force Xtreem ARGB DDR5-8000, 2TB T-Force Cardea PCIe Gen 4.0 SSD, a 1200W FSP PTM Pro Gen 5 80+ Cybernetics Platinum power supply, and a Maingear 420mm AIO cooler with 25mm thick radiator and liquid metal thermal interface for the CPU.

While these specs aren't expected to change, who knows what the future will bring. The above will at least serve as a baseline spec for the collection.

Maingear Apex Force Image 1 of 4

(Image credit: Maingear)Image 2 of 4

(Image credit: Maingear)Image 3 of 4

(Image credit: Maingear)Image 4 of 4

(Image credit: Maingear)

Also introduced at CES 2025 is the Maingear Apex Force custom-built, fully liquid-cooled PC.

Starting with an enormous full tower case based on the Phantek NV9, cable management is kept discreetly hidden in a rear chamber of the case while the panoramic wrap-around tempered glass shows off the internals.

Inside, you'll find precision-engineered 10-degree-angled chrome tubing that matches the angle of the bottom case panel, which pumps liquid through a 420mm-long, 60mm thick AIO radiator fitted with three 140mm wide, 30mm thick fans to carry away heat from both the CPU and GPU.

The Apex Force is also customizable with Intel and AMD's latest CPUs, paired with the best Nvidia graphics cards available at the time of your order.

The custom cooling loop is also customizable, including the color of the coolant used, the type and color of the tubing, and the material and color of the tube fittings used throughout the build.

Maingear Apex Rush

OK, so I waited til the end here to talk about the Maingear Apex Rush Artwork editions, because these are simply too sick to not dig into in more detail.

There are three distinct PC designs, called 'Cyber Gamer', 'Good Fortune', and 'Vaporware Drive'.

Image 1 of 4

(Image credit: Maingear)Image 2 of 4

(Image credit: Maingear)Image 3 of 4

(Image credit: Maingear)Image 4 of 4

(Image credit: Maingear)

Cyber Gamer has a very distinct Cyberpunk 2077 vibe, including soft, braided cooling tubes, reminiscent of a cybernetic implant. There's a cyberpunk-styled woman on the case as well, giving the case some additional character. Any fans of Netflix's Edgerunners anime will probably already be drooling at this point.

Image 1 of 3

(Image credit: Maingear)Image 2 of 3

(Image credit: Maingear)Image 3 of 3

(Image credit: Maingear)

Good Fortune is an Asian-inspired design, with a distinctive charm and animal motif throughout, soft color pallete, and adorable koi fish, foxes, and cat characters on various parts of the case. Inside, brass-colored tubes and fittings blend in perfectly with the rest of the case's aesthetics.

Image 1 of 4

(Image credit: Maingear)Image 2 of 4

(Image credit: Maingear)Image 3 of 4

(Image credit: Maingear)Image 4 of 4

(Image credit: Maingear)

Vaporware Drive, however, is the design that triggered my 80's kid lizard brain, touching something primordial that I didn't even know still resided in my psyche.

OK, that might be overstating things, but my god, just look at this case. While all of the Apex Rush Artwork cases are full of interesting details, Vaporware Drive's back panel shows off a city skyline with a setting 80's sun (yes kids, the sun actually lost horizontal stripes when it set back then. Look it up), and there's a building with a neon sign that reads 14.37 DPS.

If you don't know, that's the damage per second of an upgraded StarCraft II zergling, which is the kind of absolutely unnecessary detail that makes me love this design so damned much.

It should also be noted that there is also an undecorated Apex Rush design, which features all of the same internals, but none of the extra flair beyond the already awesome-looking case, internals, and overall design.

Pricing and availability

At the moment, no pricing or availability has been specified for any of these PCs, though you can go to Maingear's website to learn more, customize your builds, and order directly from the company once the PCs are available.

Maingear is also offering a $100 reserve option for these PCs which will hold your place in the order and build queue,, a deposit that is entirely refundable and which will be put toward the final cost of the PC that Maingear builds.

Lenovo's Legion Go gaming handheld was one of the standout releases of 2024, and is arguably the best PC gaming handheld after the Steam Deck, but one thing has ultimately held it—and every other Windows-based handheld—back from truly dethroning the Steam Deck: Windows.

Windows 11 just isn't a great platform for a PC gaming handheld, it just isn't. Even the ability to install gaming stores and services other than Steam doesn't make up for how clunky the Windows 11 desktop is to navigate on an 8-inch screen without a keyboard and mouse (and the various touchscreen experiences have been very disappointing).

Now, Lenovo is launching a new PC gaming handheld at CES 2025, and it finally does away with the worst part of the PC gaming handheld experience. The Lenovo Legion Go S will be the only other handheld besides the Steam Deck to have a SteamOS license, giving you a Steam Deck-like experience on much more modern hardware.

There will also be a Windows version of the Legion Go S as well, in the Glacial White colorway (the SteamOS-powered Legion Go S will come in a color called Nebula Violet).

You will also have the choice of chips to power the new handhelds, either the AMD Ryzen Z1 Extreme, or the new AMD Ryzen Z2 Go, which is exclusively for the Legion Go S.

Specs for the Lenovo Legion Go S Lenovo also showed off the Legion Go 2 prototype

In addition to the Lenovo Legion Go S, Lenovo also showed off a prototype for the Legion Go 2, which continues with the innovative design of its first gaming handheld, including its detachable, Joy-Con-like controllers and built-in kickstand on the game tablet itself.

The prototype is obviously just that, a prototype, so the specs on the device should it ever make it to market aren't set in stone, but the unit Lenovo showed off this week features the new Ryzen Z2 processor and includes up to 32GB of faster 7500MT/s LPDDR5X memory, up to 1TB PCIe 4.0 SSD storage, and a gorgeous 8.8-inch 1920 x 1200p OLED display running at 144Hz, 500 nits brightness, and 97% DCI-P3 color gamut for stunning and vivid visuals.

Should this handheld ever make it to market, it's so pretty that you might even forgive it for running on Windows 11, but hope springs eternal that Lenovo's SteamOS license eventually makes it to the rest of its PC gaming handheld portfolio.

• Lenovo ThinkBook Plus Gen 6 has a 14-inch OLED that extends to 16.7-inch
• With a price tag of $3500, it will target businesses and enterprise users
• This ThinkBook is the latest innovation in form factor from Lenovo

If you want a laptop with a larger screen but don't want to sacrifice portability - and the GPD Duo, with its dual 13.3-inch OLED screens feels like overkill - the ThinkBook Plus Gen 6 Rollable from Lenovo could be exactly what you’re looking for.

Part of the PC maker's lineup of AI-powered business solutions being unveiled at CES 2025, the laptop, which was originally presented as a rather unlikely concept two years ago, features a vertically expanding rollable 14-inch OLED display capable of extending to a 16.7-inch panel with the press of a key or via hand gestures to the camera. The 120Hz OLED screen offers 400 nits of brightness and 100% DCI-P3 color accuracy, delivering sharp and vibrant visuals suitable for professional workloads.

Weighing 3.73lbs (1.69kg) and measuring 19.9mm in thickness, the laptop balances a slim design with enhanced productivity features. It includes an edge-to-edge ThinkBook keyboard with a ForcePad and comes with dual Thunderbolt 4 ports and Intel Wi-Fi 7 (5 Gig) for fast data transfers and connectivity.

ThinkBook Workspace

Powered by up to an Intel Core Ultra 7 processor (Series 2) with Intel Arc Xe2 graphics, the Copilot+ PC supports up to 32GB of LPDDR5x memory running at 8533MT/s and offers up to 1TB of PCIe Gen4 SSD storage. The 66Whr battery, paired with a 65W USB Type-C GaN Nano adapter, offers reliable power for extended use.

The Copilot+ PC comes with AI-powered tools like Lenovo AI Now and Cocreator from Paint, blending on-device intelligence with cloud capabilities. These tools provide features such as task suggestions, workflow optimization, and offline creative assistance for content generation and data analysis.

The device also introduces the ThinkBook Workspace, designed to improve multitasking with split-screen capabilities and dedicated widgets for frequently used apps. Its unique vertical screen design encourages better posture, making it particularly useful for professionals working long hours.

The ThinkBook Plus Gen 6 Rollable will be available in Q1 2025 with a starting price of $3,499.

(Image credit: Lenovo) You might also like

• These are the best business laptops around
• And these are the best laptops for graphic design you can buy right now
• The first true dual screen laptop is finally here for productivity fans

• Lenovo ThinkCentre neo Ultra Gen 2 adds Core Ultra CPU options and Thunderbolt
• But form factor remains the same, as does the rest of its specs
• The refresh comes at no extra cost as it keeps its retail price

At last year’s CES, Lenovo introduced the ThinkCentre neo Ultra, its first ultra-small form factor AI desktop for SMBs and a competitor to Apple’s Mac Studio.

That device launched in July 2024, and now at CES 2025, Lenovo has unveiled an upgraded version that truly wows.

The first generation of the ThinkCentre neo Ultra was powered by 14th Gen Intel processors, up to the i9-14900T. The upgraded version is now outfitted with Intel’s Core Ultra CPUs, promising a big leap in performance.

Thunderbolt 4 and much more

(Image credit: Lenovo)

The original neo Ultra featured a wide range of ports. On the front, it had two USB-A ports, a USB-C port, and a headphone/mic combo jack (3.5mm).

The rear panel offered four USB-A ports, two HDMI 2.1 ports, four DisplayPort 1.4a ports, and an Ethernet (GbE RJ-45) port. That meant it was capable of driving up to eight displays. The refresh doesn’t change the port arrangement but it does make a welcome switch.

Between the two rows of rear ports on the original model were two punch-out ports. The first could be configured for HDMI, DisplayPort, USB-C with DisplayPort functionality, or VGA (dust off your old CRT monitor!). The second could be configured for HDMI, DisplayPort, VGA, or 2.5GbE RJ-45. Both punch-outs are now gone, replaced by additional ventilation and a new Thunderbolt 4 port.

Aside from the new processor options and the addition of Thunderbolt 4, everything else remains the same.

The form factor hasn’t changed - it still measures 195 x 191 x 108mm (3.6L). All other specs, including four M.2 slots, an Nvidia RTX 4060 GPU, and up to 64GB of DDR5 RAM, are unchanged.

Another thing that hasn’t altered, and which we’re very pleased about, is the price. The ThinkCentre neo Ultra Gen 2 will be available in February 2025, with an expected starting price of $3,099.

If you’re happy with the specs of the original (and would prefer a VGA port instead of Thunderbolt 4 – hey, we’re not judging), Lenovo is still selling the first-gen model, and it's currently 49% off at $1,580.49.

More from TechRadar Pro

• These are the best mobile workstations around
• And these are the best business laptops available today
• ThinkCenter neo Ultra has up to 16 ports and can power an old CRT monitor

• Sony has revealed new versions of its PS5 accessories
• The Midnight Black Collection launches on February 20, 2025
• Preorders are set to begin on January 10 at PlayStation Direct

Sony has unveiled its Midnight Black Collection, a brand new colorway for a range of PS5 accessories launching next month.

As the name suggests, the Midnight Black Collection includes new, classic black versions of the PlayStation Portal remote player, DualSense Edge wireless controller, Pulse Elite wireless headset, and Pulse Explore wireless earbuds.

The entire collection is set to be released on February 20, 2025, with pre-orders beginning on January 16 at PlayStation Direct.

Sony has already provided the prices for each piece of hardware, which you can read below, and won't cost more than the original white models, aligning with the leaks reported in November.

The PS Portal, DualSense Edge, and Pulse Explore wireless earbuds will cost $200, while the Pulse Elite wireless headset will cost $149.99.

• PlayStation Portal remote player – Midnight Black: $199.99 / £199.99 / €219.99 / 34,980 YEN
• DualSense Edge wireless controller – Midnight Black: $199.99 / £199.99 / €219.99 / 34,980 YEN
• Pulse Explore wireless earbuds – Midnight Black: $199.99 / £199.99 / €219.99 / 34,980 YEN
• Pulse Elite wireless headset – Midnight Black: $149.99 / £129.99 / €149.99 / 22,980 YEN

"Each Midnight Black accessory features a unique, rich shade of black, with sleek detailing on various buttons and accents such as the PlayStation logo," Sony said in the latest PlayStation blog post.

"The Midnight Black design extends to the included charging case for Pulse Explore earbuds, charging hanger for Pulse Elite headset, and carrying case for DualSense Edge controller. The Pulse Elite wireless headset and Pulse Explore wireless earbuds will also include a felt grey carrying case."

These new accessories will complement PlayStation's Midnight Black DualSense wireless controller and PS5 console covers, which released in 2021 and 2022.

You might also like...

• The best PS5 deals for January 2025
• Best PS5 games 2025 - top PlayStation 5 titles to play right now
• Do you turn off your PS5 or put it in Rest Mode when you're done? According to Sony, the decision is a 50/50 split for all players

• Apple says your Siri data is never sold or used for marketing purposes
• The statement comes as the company settles a $95 million privacy lawsuit
• Apple settled to move on from the lawsuit and avoid further costly litigation

Apple says that data collected by Siri has never been sold to advertisers or used to build a marketing profile of its users. Instead, the company argues that its virtual assistant has been “engineered to protect user privacy from the beginning.”

The statement (via BBC News) comes as the tech giant has settled a class action lawsuit focused on its Siri assistant. The litigators had alleged that Siri had eavesdropped on Apple users without their knowledge or permission. As well as that, the lawsuit also claimed that user voice recordings were shared with advertisers.

For instance, lead plaintiff Fumiko Lopez argued that Siri recorded her and her daughter without their consent. The couple believe they were served targeted ads after discussing various products, including Nike Air Jordans.

In response, Apple said that “Siri data has never been used to build marketing profiles and it has never been sold to anyone for any purpose,” adding that collected data is only used to improve the virtual assistant.

The tech firm has settled the lawsuit for $95 million. It denied any wrongdoing and pushed back against claims that it “recorded, disclosed to third parties, or failed to delete, conversations recorded as the result of a Siri activation” without permission.

The company said it settled to “avoid additional litigation so we can move forward from concerns about third-party grading that we already addressed in 2019.”

Apple’s reputation for privacy

(Image credit: Unsplash [Omid Armin])

Apple has long built a strong reputation for user privacy. Unlike some rivals, many of its features work entirely on-device, meaning user data is not sent to Apple’s servers, while the company said that outside experts could verify the privacy controls of the Private Cloud Compute aspect of its Apple Intelligence system.

With Siri, though, some data – including audio of your requests – is sent to Apple’s servers. However, Apple’s Siri terms state that “Unless you opt in to Improve Siri and Dictation, your audio data is not stored by Apple.” In other words, your audio is processed on Apple’s servers, but the company does not keep it after the job is done.

Transcripts of your audio may be stored by Apple. However, both your audio data and transcripts are anonymized: they are assigned a random identifier and are not tied to your Apple Account. The identifier is disassociated from the data after six months, and Apple may keep the data for up to two years.

Apple’s terms lay out that Apple does not use Siri data for marketing purposes and that it is never sold to anyone else. Audio is only stored if you opt in to improve Siri but, being an opt-in process, this feature is not enabled by default.

Still, if you don’t feel comfortable with Siri collecting and processing any data at all (including transcripts), you can turn off Siri by opening the Settings app and going to Apple Intelligence & Siri > Talk & Type to Siri, tapping Off and disabling all of the toggles.

If you want to delete your Siri and dictation transcripts, go to Apple Intelligence & Siri > Siri & Dictation History in the Settings app, then select Delete Siri & Dictation History.

You might also like

• Apple called Android a ‘massive tracking device’ before the iPhone’s privacy push
• Apple opens up Private Cloud Compute to security researchers, offers bug bounties up to $1 million
• I'm a privacy expert—here are the 4 iOS 18 features I'm excited about

• Sony Afeela 1 is slated to offer high levels of autonomous driving
• Prices will start at $89,900 / £72,000 / AU$143,000
• In-vehicle entertainment is top of the agenda

Sony’s first foray into the world of electric vehicles is arguably one of the hardest working concepts of the CES show in Las Vegas.

Starting life as the Sony Vision-S, the project has since morphed into the Sony Honda Mobility (SHM) Afeela, gradually revealed over five consecutive years of CES fanfare.

Considering the Porsche Taycan can manage 320kW charge speeds, that figure already feels old fashioned.

During that time, we have been teased with infotainment details and treated to a mock-up that has gradually evolved in design. But this year at CES 2025, the company finally announced that the $89,900 all-electric model is now available to pre-order for a fully refundable reservation fee of $200, with deliveries anticipated in mid-2026.

But there’s a catch, as Sony’s tie-in with the Japanese legacy automaker is currently only available to residents of California, although it is expected to be opened up to other US states soon.

This year’s party piece saw SHM boss Yasuhide Mizuno summon the “near production ready” Afeela 1 to the CES stage using just his smartphone, while indicating that the suite of of 40 sensors (including cameras, LiDAR, radars and ultrasonic sensors) would allow for high levels of autonomous driving.

(Image credit: Sony Honda Mobility)

Inside, all occupants are treated to a huge dose of entertainment, thanks to displays at every seat, complete with Sony’s 360 Spatial Sound Technologies to offer an “immersive” in-car experience.

But Autocar states that the model is destined to ride on the same platform that will underpin the forthcoming and yet-to-be-driven Honda 0 Series, with official SHM specs clarifying that it will use a 91kWh battery pack that sends power to dual 180kW motors at the front and rear.

Total power output will be 482bhp, with the fastest charging speeds limited to just 150kW via Tesla’s North American Charging Standard (NACS) connector. Considering the Porsche Taycan can manage 320kW charge speeds, that figure already feels old fashioned.

According to SHM, the Afeela 1 will be available in a more expensive Signature trim level, which will set customers back around $102,900 (£82,000 / AU$165,000).

Despite the heady asking price and dazzling 'computer-on-wheels' technology, it is already falling some way behind the likes of Lucid where the important automotive stats are concerned. Lucid offers its 2025 Air Touring model with an EPA estimated range of 406 miles, as well as a total power output of around 620hp.

The more expensive Grand Touring model, which costs around $10,000 more than the range-topping Afeela 1, offers a staggering 512-mile EPA estimated range and 819hp, with a 0-60mph sprint time of just three seconds.

Entertainment takes centre stage

(Image credit: Sony Honda Mobility)

Clearly, Honda’s tie-up with Sony will major on the entertainment factor, with the two companies already promising the next generation of in-car technology. This comes courtesy of a powerful system on a chip (SoC) capable of 800 ToPs (Trillions of Operations Per Second), machine learning, advanced AI and Tesla-baiting levels of autonomy.

You can control various in-car functions via natural voice interactions thanks to the Afeela ‘Personal Agent’, which SHM says will converse with occupants and even suggest activities, while the Unreal gaming engine is used to make everything from maps to menus more interactive and visually arresting.

But Honda is responsible for how this car will ride and handle – and it's a manufacturer with a patchy EV history.

There’s also the fact that sound has been optimized for each individual seat, with Sony offering its noice-cancelling know-how to allow all of those riding aboard to enjoy entertainment from built-in apps, such as Spotify, TikTok and more.

At CES last year, then-boss Izumi Kawanishi even showcased how the vehicle could be driven with a PS5 controller, suggesting that Sony’s blockbuster games would make their way inside the car.

But Honda is responsible for how this car will ride and handle – and it's a manufacturer with a patchy EV history.

Granted, its Prologue model has been very well received in the US, but the all-electric e:Ny1 and the now-defunct Honda e haven’t been so lucky. In fact, the e:Ny1 was a massive disappointment when I drove it last year.

Honda said it would “rip up the rulebook” when it unveiled its plans for the 0 Series at last year’s CES – unofficially stating that it would essentially start afresh where EVs are concerned.

Slimmer batteries, lighter vehicles, better handling and improved efficiencies are all promised by this new platform, but we are yet to see if the Japanese marque can deliver.

So far, the pure EV specs offered by the Afeela 1 already feel a little long in the tooth, especially given that customers won’t receive their cars until 2026.

At which point, MG claims it will have released a vehicle with solid-state battery technology and all-electric ranges that could easily be in excess of 600 miles.

You might also like

• MG says the first cars with solid state batteries will arrive as early as next year
• The Xiaomi SU7 Ultra EV has been unleashed – and it’s already bagging speed records
• Garmin brings its radar safety tech to motorbikes for the first time, but I'm not convinced

As artificial intelligence (AI) and other emerging technologies continue to advance in scope and sophistication, pressure is mounting on IT teams to rapidly deploy these tools. That pressure is exacerbated by relentless resource constraints and the increasing struggles to retain talent.

These growing pressures have substantial implications for how IT professionals spend their day. Although an IT team’s initial priority resides in ensuring service availability and reliance, frequently, a substantial portion of their time is invested into crisis management – time that could be better used innovating.

This is why AIOps, the application of AI to IT operations, has gained widespread popularity when leveraged with generative AI. AIOps alleviates routine tasks and helps foster innovation by simplifying frequent problems, detecting anomalies, and accelerating automated responses.

Cutting through complexity

Contemporary IT teams are tasked with monitoring complex, hybrid environments, often depending on a wide range of tools. Among these, certain platforms stand out for their intuitive user experience and ability to integrate applications, heuristics, and workflows into a cohesive framework aimed at enhancing operational efficiency.

The purpose of these systems is to be as accessible as possible, even to less-skilled, first-level operations teams. By providing AI and machine-learning-driven insights to all skill levels, raw data can be transformed into actionable insights and recommendations.

Sophisticated causal AI can get to the root cause of complex problems when data sources and tools have been effectively consolidated. But generative AI takes this even further – it translates these causes into digestible summaries, providing proactive predictions and solutions. Simultaneously, generative AI can also leverage operational, service management, and DevOps to save time for IT teams.

Avoiding problems by proactively identifying them

The use of reactive, traditional monitoring tools leaves organizations vulnerable to an array of pain points. Many only notify teams of issues once they have already occurred, causing emergency troubleshooting, slower systems, and potential shutdowns. As systems grow in complexity, anticipating and addressing problems before they arise becomes crucial. Proactivity ensures business continuity, which should include change risk management impact (both scope and severity).

Ideally, organizations should be informed of issues before they affect operations rather than scrambling to mitigate impact after issues arise. This includes utilizing predictive AI that can identify capacity and resource issues, as well as potential service disruptions or declines, and implementing automated measures to resolve them.

Optimizing vast amounts of data to enhance business performance

Organizations are shifting towards advanced, enterprise-level tools equipped with machine learning capabilities – with the speed in which these systems evolve often exceeding human monitoring and management. These tools process and analyze a wealth of system data from complex IT environments, transforming this data into actionable insights, and driving automated responses. IT professionals can then automate actions based on a comprehensive understanding of their systems' operations and their impact on business objectives.

Organizations can better optimize valuable IT resources by leveraging their data analytics and automated actions. This allows them to prioritize tasks while improving value creation and innovation activities.

Considering the challenges

Due to the range and complexity of modern IT infrastructure, networks and applications, as well as the heterogeneity of the data the systems produce, machine learning models and AI are often seen as necessities for IT operations.

KPIs like failure prediction, mean-time-to-repair, and root cause analysis have become a typical primary focus for IT teams. However, due to the complexity and volume of the data that employees are dealing with, they struggle to keep up fast enough to make significant progress on these metrics. If an organization relies on manual, labor-intensive processes to meet these metrics, it is difficult to cost-effectively scale and standardize efforts.

Despite this, organizations will also face challenges when implementing AIOps technologies to automate these processes. These challenges may include:

• Data quality: it is crucial that the vast quantities of data sources used to fuel these tools are constantly monitored for biases and errors. Low quality data can cause issues ranging from flawed outputs to misuse.
• Scale and complexity: as IT operations grow and data and tools expand, there’s the added challenge of resolving and service modelling assets that span from cloud to mainframe and application to network.
• Silos: typically, IT teams are often isolated rather than being under a single umbrella, leading to data inconsistencies and a lack of standardization.
• Automation confidence: it can be difficult to get the right context for root cause isolation and recommended actions based on historical occurrences.

Laying the groundworks for AIOps

For AIOps to be successfully integrated, organizations must integrate existing tools, provide out-of-the-box advanced AI/ML, and accelerate automation. Business leaders must consider the use cases that are important to the organization and start small to prove the value. By doing this, AIOps can enhance the quality and speed of business decisions.

A strong AIOps strategy also requires cultural considerations. Organizations need to standardize processes to simplify automation, enhance governance to support new roles, and effectively address organizational change management. In practice, this means internal goals need to be aligned, teams must be equipped to embrace failure and grow from it, and cross-functional collaboration needs to be encouraged. A cultural shift towards open and consistent communication will assure that employee resources are effectively used and that everyone is working towards a common objective.

Ultimately, if approached strategically, AIOps can substantially streamline IT operations. It lays the groundwork for automation to be deeply ingrained into all IT activities, transforming organizational efficiency and innovation.

We've featured the best IT management tool.

This article was produced as part of TechRadarPro's Expert Insights channel where we feature the best and brightest minds in the technology industry today. The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/news/submit-your-story-to-techradar-pro

• The KaraPod uses condensed atmospheric water for making coffee
• It uses a UV light, heat, and replaceable filters to clean the water
• It will be available to buy soon for $249 (about £200 / AU$400)

We've seen lots of interesting ideas at CES 2025, but this one has proved one of the most controversial among the TechRadar team: a coffee maker that works like a dehumidifier and uses water condensed from the air in your home to brew espresso.

The KaraPod is a countertop version of the KaraPure 2 – a machine that condenses atmospheric water and runs it through a purification system that includes heat treatment, UV light, and replaceable filters to produce up to 2.5 gallons (10 liters) of clean drinking water per day. Manufacturer Karawater says that the filtered water has "an unbeatable pH value of 9.2+", though there's no strong evidence that drinking alkaline water has real health benefits.

Howard E LeWine MD, Chief Medical Editor at Harvard Health Publishing, says: "Because the fluid in the stomach is so acidic, once regular or alkaline water gets down to your stomach there will be little difference in the resulting stomach fluid pH. You could potentially raise the stomach fluid pH by drinking a lot of alkaline water, but it would only be temporary. Even if you drank enough alkaline water to slightly raise the pH of your blood, your kidneys would quickly go into action to rebalance your blood pH."

That said, using alkaline water to make coffee will certainly change the way your finished brew tastes. Whether it's for better or worse will be a matter of personal opinion.

(Image credit: KaraPod) Water, water everywhere

As our colleagues at Live Science explain, you shouldn't drink water from a regular dehumidifier – or use it to make coffee. The dehumidifier's various parts, including its condensing coils, aren't perfectly clean, and its storage tank could provide a nice habitat for bacteria and molds to grow. That's not normally a problem, because the water collected there isn't meant to be potable – you could use it to mop your floor or flush a toilet, or just pour it down the sink.

That's not the case with the KaraPure 2 and KaraPod, which filter and clean the water they collect, and circulate it to keep it fresh. Nevertheless, we still can't help feeling somewhat uneasy. Most of us drink treated tap water all the time, but the source of water collected in our homes is a little too personal, with none of the palatable anonymity of a municipal water supply. Steam from cooking last night's dinner? Rainwater from your wet running shoes? Vapor you breathed out overnight? Rising damp? Toilet plume?

If you can get past that mental hurdle, the KaraPod could be the coffee maker for you, and you'll never need to refill its tank. As TechHive reports, the machine uses compostable coffee pods, with 6oz, 10oz, and 12oz pour sizes. It can also dispense hot or room-temperature water. It will be available to order soon direct from Karawater for an early bird price of $249 (about £200 / AU$400).

You might also like

• The future of coffee – 3 hot trends brewing in 2025
• Tired of boring coffee makers? The new L'OR Barista was designed by an award-winning architect
• Shhh - KitchenAid's new coffee machines can brew up to 40 hot and cold drinks, and they're whisper quiet

We’re covering all of the latest CES news from the show as it happens. Stick with us for the big stories on everything from 8K TVs and foldable displays to new phones, laptops, smart home gadgets, and the latest in AI. And don’t forget to follow us on TikTok for the latest from the CES show floor.

• Suunto has unveiled its new waterproof headphones
• The Suunto Aqua feature bone conduction and an IP68 swimproof rating
• They can store 8,000 songs and use AI technology to analyze and improve your swim stroke

2025 is only seven days old, but we might already have a contender for the best waterproof headphones of 2025, the new Suunto Aqua.

Unveiled January 7 and available to order now, the Suunto Aqua are bone conduction headphones with an IP68 waterproof rating. As such, they can be used for running, cycling, and swimming to provide non-intrusive audio whatever your discipline of choice.

The best bone conduction headphones on the market have become a mainstay audio companion of triathletes, runners, cyclists, and more thanks to their light weight, ergonomic fit, and because they leave your ear canals free so you can stay alert in your surroundings.

Suunto's Aqua headphones cost $179 / £175 (AUS availability to be confirmed) and they tick all the boxes of a great bone conduction experience. Here's the full rundown.

Suunto Aqua: All the details

(Image credit: Suunto)

Priced at around the same point as the Shokz OpenSwim Pro, the Suunto Aqua headphones weigh just 35 grams and feature 32 GB of onboard storage, the latter enough for 8,000 songs in MP3 format (a crucial feature when you're in the pool devoid of Spotify or Apple Music).

Battery life is rated for 10 hours of playtime, with an extra 20 hours available from the included charging pod, enough for even the most grueling of swims, rides, or runs. They'll also charge fully in less than an hour, and offer three hours of listening time from just 10 minutes of charging.

As you might expect from waterproof headphones, they boast a stellar waterproof rating of IP68, good for submersion in up to five meters of water for two hours. Other key features include Bluetooth 5.3 and head movement control, as well as a dual microphone array and environmental noise cancellation for taking and making calls.

Behind the specs though, the Suunto Aqua boast a formidable secret weapon: AI-powered swim stroke analysis.

Suunto's Aqua headphones can identify stroke posture, head pitch angle, breath frequency, and glide time so that you can track and improve your swim stroke, adjusting your technique through the Suunto app. Not only will they keep you entertained in the water, but they come with their very own built-in AI-powered swimming coach.

On paper, the Suunto Aqua bone conduction headphones look mightily impressive, look out for our full review to see whether they can knock the OpenSwim Pro off their perch.

You might also like

• SONR wants you to swap bone conduction swimming headphones for a… head puck
• How I used waterproof headphones to power through my swims during triathlon training
• Water-resistant vs waterproof headphones: what's the difference?

(Image credit: Future)

We’re covering all of the latest CES news from the show as it happens. Stick with us for the big stories on everything from 8K TVs and foldable displays to new phones, laptops, smart home gadgets, and the latest in AI.

And don’t forget to follow us on TikTok for the latest from the CES show floor!

We’re in the midst of an AI hype cycle. On either side are extremes: from solving all problems, accelerating humanity’s demise, to being massively overhyped.

In the middle is everyone else with varying degrees of awareness of both the opportunities and risks. There is still much uncertainty, partly due to the nature of the technology itself: AI systems are complex and often not fully understood, leading to significant unknowns.

It’s human nature to understand how things work, even if it is at a conceptual level. For example, you don’t need to be a Computer Scientist to understand that web-search is akin to looking up relevant keywords within a massive database for matches. The big difference here is how AI finds patterns and comes up with answers that are not readily intuitive nor explainable to the user, even for experts. So the opaqueness of AI is undermining the trust we might place in it.

It’s why we’re seeing efforts to establish ways of improving trust (such as the growing activity within the explainable AI field). For instance, the UK government has introduced an AI assurance platform, while the European Union’s AI Act aims to ensure better conditions for developing and using the technology.

And this trust is something that needs to be established soon. AI will become so pervasive that even if we tell workforces to proceed with extreme caution, most will struggle to even make a clear distinction between what is AI and what is not AI.

The AI technology cycle

Why do I think AI will become that embedded when there are still big gaps in our understanding? For one thing, AI has been successfully used for decades in business for finding patterns or making predictions. It is only within the past few years after a research paper published in 2017 cracked a problem that gave rise to the LLMs that we see (and are discussing) today. No one can predict the future, however a lot of people are investing in it: 84% of CIOs expect to increase their funding in AI by 33% in 2025; the scale of these investments necessitates that companies look at five- or ten-year cycles.

It’s a balancing act between putting initiatives in place geared towards becoming AI-first, while trying to fix today’s problems and ensuring sustainability for the next few years. It’s a situation that most technologies now considered commonplace have been through. We go from idealization and realization to finding practical applications that solve real problems, followed by hype and excitement. Eventually, we recognize the technology’s limitations and address the challenges that emerge, and the technology becomes integrated into daily life in a way that moves beyond the hype cycles.

Where AI diverges from previous technologies is its intrinsically random and opaque nature, which is very different from traditional software. This has significant implications for all aspects of deploying AI. For instance, in security, while many current practices are applicable for securing AI, they serve more as analogies than direct solutions.

The doctor will see you now

Think of it like going to a doctor. When you walk in and say, "I don’t feel well," the doctor doesn’t reply, “Give me your DNA, and I’ll tell you everything wrong.” It doesn’t work like that because, aside from cost issues, DNA is an immensely complex system that we are still trying to understand, it can only reveal certain predispositions, and doesn’t capture environmental factors. Twins, for instance, have the same DNA but can develop different ailments.

Instead, doctors look at your family history, perform some tests, ask questions, and try different approaches to figure things out. Doctors are looking at the problem – your illness – through a socio-technical lens. The investigations into family history, your lifestyle, recent activities, and the social element: what’s going on in your life that could be contributing or the cause of the problem? The technical aspect is the fix – what medicine we have available to treat you – but the social element heavily influences it.

A truly integrated socio-technical approach

As time goes by, it is increasingly apparent that we need to apply the same logic to securing AI. Cybersecurity is recognized as a socio-technical field (most security issues are resultant of people, after all). Right now, there appears to be different framings between the social and the technical. We talk about social engineering, insider threats, educating employees on the risks of opening unfamiliar attachments. Separately, we deploy technical measures to provide security or mitigate the impact of attacks.

Where securing AI will differ is in the need to embed the social and the technical within the same techniques as opposed to viewing them as separate. We’re already seeing examples of where our expectations clash with what the AI models deliver: one recent example was Google Gemini telling someone to ‘please die’.

This highlights multiple points to consider. First, the opaque nature of AI: LLMs don't think like humans do (although it can be good at fooling us to believe otherwise), so it’s hard for us to understand how it could generate such a response based on an innocuous conversation around aging.

Second, if an LLM can output such a response through what appears to be an innocent conversation, what could happen when there is a deliberate attempt to generate malicious responses?

Finally, the Gemini incident underlines the importance of looking at the context in which AI tools are being used and how they are onboarded into an organization. This has a major social dimension, not within the system itself, but rather in how people interact with it.

Further complexity with pervasive AI

How is this any different from any other tool or technology solution? We’ve noticed a tendency for people to anthropomorphize AI to the degree that they haven’t done with any recent technology. The average user is having what they think are conversations with AI. They’re not writing code to generate a response or action, but talking like they’ve used search engines or even other people to find information or do something.

The biggest mistake we could make is assuming that, as AI models become commonplace, the level of attention devoted to their risks can drop off. Even with clear warnings, most users aren’t going to distinguish what is and what is not AI. Our focus needs to be on what informs our AI tools, the models and overlays they’re composed of, and where the weak spots are.

The challenge is significant, but so is the opportunity. With the right approaches we can ensure that AI enhances our world without compromising the trust that underpins it. The journey to secure AI isn’t just about protecting systems—it’s about shaping the future.

We've featured the best AI video editor.

This article was produced as part of TechRadarPro's Expert Insights channel where we feature the best and brightest minds in the technology industry today. The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/news/submit-your-story-to-techradar-pro

• Ultrahuman Rare smart ring unveiled at CES 2025
• It's a smart ring coated in precious metals: gold, rose gold, and platinum
• Other smart rings just use colored titanium

Health and wellness tech company Ultrahuman, maker of the Ultrahuman Ring Air smart ring, has unveiled its latest addition to its smart ring lineup at CES 2025: the Ultrahuman Rare, a line of ‘luxury’ smart rings plated with 18-karat gold and platinum.

The current top contenders for the title of best smart ring, the Oura Ring 4 and the Samsung Galaxy Ring, offer gold and silver colorways (while Oura also offers rose gold) but they’re not actually gold. Oura’s offering, for example, is “titanium with physical vapor deposition (PVD) coating in Gold, Silver, and Rose Gold colors” according to the Oura site, while the Samsung Galaxy Ring is similar, listing its material as titanium only in all colorways.

Real gold is a notoriously soft metal, but this is a 'luxury' smart ring more along the lines of jewelery, and less suited for bashing about. In a press release, Ultrahuman founder and CEO Mohit Kumar said: “With Rare, we’ve reimagined what a smart ring can be, creating a masterpiece that not only brings personal wellness but also embodies the pinnacle of luxury and craftsmanship.”

The ring comes in three colorways: Dune (with its outer shell wrapped in 18K gold), Desert Rose (18K rose gold) and Desert Snow (PT950 platinum), and will be on sale in luxury stores Printemps in Paris and Selfridges in London.

We don’t yet have pricing for all regions, but after reaching out to Ultrahuman, a representative confirmed “we’re looking at various locations in the US and the rest of the world. For instance - New York, Milan, Dubai and India.” We do know the cost in GBP:

(Image credit: Future)

We’re covering all of the latest CES news from the show as it happens. Stick with us for the big stories on everything from 8K TVs and foldable displays to new phones, laptops, smart home gadgets, and the latest in AI.

And don’t forget to follow us on TikTok for the latest from the CES show floor!

• Dune - £1,499 (around $1,900 / AU$3,000)
• Desert Rose - £1,499 (around $1,900 / AU$3,000)
• Desert Snow - £1,799 (around $2,250 / AU$3,600)

Rare users also gain a lifetime of access to Ultrahuman X, Ultrahuman’s device insurance package.

Watch the Ultrahuman Rare preview below: Analysis: Legit smart ring or PR stunt?

This sort of super-luxury stuff isn’t new to the tech-scene, whether it’s gold-plated phones or Hermes-branded versions of the best Apple Watches. There are always devices being packaged with exclusive branding or premium materials designed to add a little extra onto the price tag and generate headlines like this one.

However, given that rings occupy the jewelery space, a gold or platinum-plated smart ring makes a lot of sense. It allows people to wear them as legitimate style pieces, while enjoying the benefits that come with a smart ring – especially one as good as Ultrahuman’s. We rated the Ultrahuman Ring Air very highly in our review, scoring it four stars, and I think there could be a real market for these luxury versions.

You might also like...

• Amazfit Active 2 is a super-cheap AI fitness watch that responds to voice commands like Alexa
• Garmin unveils brand new Instinct 3 with AMOLED display and massive solar charging upgrade
• Circular Ring 2 unveiled at CES – and it's the end of plastic sizing kits for smart rings

• Nvidia reveals supercharged mini PC with Blackwell architecture
• A single Project DIGITS supercomputer can handle 200b LLMs, two can handle 405b
• It’ll cost from $3,000 for access to the GB10 Superchip, with Blackwell GPU and Grace CPU

Nvidia has unveiled Project DIGITS, a brand-new mini PC designed to get AI into the hands of more users through the company's flagship Blackwell hardware.

With the GB10 Grace Blackwell Superchip, Nvidia promises a petaflop of AI computing performance, making the device ideal for prototyping, fine-tuning and running large AI models as compute demands increase for businesses and consumers alike.

“Placing an AI supercomputer on the desks of every data scientist, AI researcher and student empowers them to engage and shape the age of AI," noted Nvidia CEO Jensen Huang.

Nvidia shoehorns Blackwell Superchip into mini PC

Based on the Grace Blackwell architecture, GB10 Superchip uses a Blackwell GPU with CUDA cores and fifth-generation Tensor Cores and a Grace CPU with 20 power-efficient cores built with the Arm architecture. Nvidia boasts of “best-in-class power efficiency, performance and connectivity.”

Project DIGITS also includes 128GB of unified DDR5X memory and up to 4TB of NVMe storage. A single supercomputer can handle up to 200-billion-parameter large language models, but two connected via NVIDIA ConnectX networking can handle up to 405-billion-parameter models.

Nvidia has set up a dedicated web page for prospective customers to sign up for more information regarding Project DIGITS, dubbed “a Grace Blackwell AI Supercomputer on your desk.” The supercomputer will be available from Nvidia and “top partners” from May, and prices will start at $3,000.

Further information regarding configurations has not yet been confirmed, and it’s unclear whether any further customization can be done beyond SSD storage upgrades.

You might also like

• Here is our guide to the best laptops for programming
• Consider the best workstations and best mobile workstations
• "100% Nvidia's fault": Jensen Huang admits Blackwell AI chips had a concerning design flaw

Onkyo preps new budget-friendly active speakers and a line of high-end streaming amplifiers.

Nothing says romance like platinum packed with sensors.

• TCL's 2025 TVs will feature Dolby Vision Filmmaker Mode
• TCL joins LG and Philips in adding the feature
• Director-approved picture when viewing movies with Dolby Vision HDR

Among the many specs accompanying the new TVs being introduced at CES 2025, one feature in particular is capturing our attention: Dolby Vision Filmmaker Mode.

This feature was introduced by LG in 2024 in the LG C4, LG G4 and LG B4 OLED TVs, and it applies the principles of the UHD Alliance’s Filmmaker Mode picture preset to the increasingly widespread Dolby Vision content on 4K Blu-ray disc and the best streaming services. Those principles are basically “do no harm” to the creator’s intent by adding motion smoothing, inaccurate color settings, excessive sharpness and other enhancements to movies displayed on the TV’s screen.

The basic version of Filmmaker Mode simplifies the process of turning off motion smoothing and other aggressive enhancements by providing that option in a single picture preset. This preset can be used for viewing programs with both standard and HDR10 high dynamic range (HDR), where it will provide a “director-approved” picture.

Many of the best TVs feature Filmmaker Mode, including models from Samsung, Panasonic, Philips, Hisense, and Vizio. Now, TCL has announced at CES that its new TVs coming in 2025 will feature a Dolby Vision Filmmaker Mode. In this, they join both LG and Philips, the latter of which implemented the feature in its new OLED TV models introduced in the second half of 2024, such as the Philips OLED809.

Why is Dolby Vision Filmmaker Mode important?

When most TVs that feature Dolby Vision HDR support show a program with Dolby Vision, the TV will default to a Dolby Vision custom mode, sometimes with a couple of presets that can be selected such as Dolby Vision Normal or Dolby Vision Dark.

A problem with these modes is that they all apply motion smoothing to the image by default, along with other picture settings that can result in less-than-accurate pictures. The Dolby Vision IQ preset, for instance, uses a light sensor in the TV to vary brightness based on your room’s ambient light conditions, and it could make pictures look overly dark or drab.

With Dolby Vision Filmmaker Mode, brightness, color temperature (the relative “warmth” or “coolness” of the TV’s color rendition), picture sharpness, and motion settings are all optimized in the same manner as regular Filmmaker Mode, resulting in director-approved picture quality. Also similar to Filmmaker Mode, the Dolby Vision version is set and forget, with no additional adjustments required.

Will more TVs be getting Dolby Vision picture mode in the future? Likely, the list of companies who already incorporate standard Filmmaker Mode in their TVs will eventually add Dolby Vision Filmmaker Mode. Samsung doesn’t support Dolby Vision HDR in its TVs at all, so in that case, the mode is a moot point.

The best ultra short throw projectors such as the Hisense PX3-Pro are also increasingly adding Dolby Vision HDR support, so the next frontier for Dolby Vision Filmmaker Mode could very well be the best projectors.

You might also like...

• Disney's 4K Blu-rays are finally getting Dolby Vision HDR
• What is Dolby Atmos? The 3D sound tech for TV and headphones explained
• The best 75-inch TVs: the top big-screen TVs we've tested

We’re covering all of the latest CES news from the show as it happens. Stick with us for the big stories on everything from 8K TVs and foldable displays to new phones, laptops, smart home gadgets, and the latest in AI.

And don’t forget to follow us on TikTok for the latest from the CES show floor!

• Nvidia will add a native GeForce Now app to the Steam Deck
• Confirmed to be 'later this year' with greater VR headset support
• Ultimate members will have access to ray tracing, a 4K resolution, and DLSS 3 on the cloud

Nvidia's GeForce Now has been available to Steam Deck users via a script available on its download page, but Team Green will be launching a native app for Valve's handheld later this year.

Announced in a blog post, will expand its support on numerous devices such as the Apple Vision Pro, Meta Quest 3, Meta Quest 3S, and notably the Steam Deck, providing access to AAA games in greater graphical fidelity and performance.

Steam Deck owners will have access to ray tracing at a 4K resolution when hooked up to a TV achieving 60 frames per second, thanks to streaming using powerful gaming PCs in the cloud - Steam Deck OLED users will also benefit from HDR10 support across multiple games, along with DLSS 3 for high and stable frame rates and Nvidia Reflex for lower latency (when using the Ultimate membership).

GeForce Now gives you access to over 2,100 games which continues to expand with new additions such as the upcoming Doom: The Dark Ages, so if certain AAA titles can't quite cut it on the Steam Deck (notably games like Cyberpunk 2077), Nvidia's streaming service is your best bet.

Wake up Samurai... (Image credit: CD Projekt Red) Is this a game changer for the Steam Deck?

I've never taken cloud gaming services seriously as I've always opted to run my games locally, along with not wanting to depend on a good internet connection, but this is a positive move from Nvidia I can't ignore. Considering the limitations of Valve's Steam Deck in terms of its processing power for games, there's only so much that can be done by users to balance great graphics and performance (compared to newer handhelds like the Asus ROG Ally X and Lenovo Legion Go).

In this case, Team Green's GeForce Now is the solution for both handheld and docked circumstances - using an Ultimate membership guarantees great performance using ray tracing at higher resolutions. While this can also be beneficial in the case of other handheld gaming PCs, this will prove to be far more impactful for the Steam Deck performance-wise (mind you, in a non-native experience), and the fact that it will be fully supported on SteamOS.

You may also like...

• GeForce Experience is dead – replaced by the Nvidia App – and good riddance
• Nvidia’s GeForce Now Day Pass rents you an RTX 4080 GPU-powered gaming PC for 24 hours
• Latest Nvidia RTX 5000 power usage rumors make me scared that my PSU will be nowhere near enough for the RTX 5080