News
- A Mac Pro has again been rumored for launch later this year
- It'll come with an M4 Ultra based on clues drawn from Apple's internal code
- M4 UItra could run with a 32-core CPU and an 80-core GPU in theory
The next iteration of the Mac Pro might see Apple powering the computer with an M4 Ultra chip, we've heard again.
It's been a couple of years since the last incarnation of the Mac Pro, which used the M2 Ultra, so it would be a considerable leap in performance to upgrade to a theoretical M4 Ultra – but Macworld reports that this is what's in the cards.
This is based on Apple's internal code that Macworld caught a glimpse of somehow, which contained the identifier 't8152' – and we're told this suggests the use of the M4 Ultra, codenamed 'Hidra' (a name that's previously been aired on the rumor mill via Bloomberg).
There's not much else to this rumor, and Macworld doesn't have any details on the spec, although it theorizes that the M4 Ultra could run with a 32-core CPU and an 80-core GPU.
As ever with this kind of leak, it could be genuine and yet never come to fruition. It might just show us that Apple has been testing a Mac Pro configuration along these lines – but it's possible that this machine may never see the light of day, ending up canned at some point in prototyping.
(Image credit: Brittany Hosea-Small/ AFP/ Getty Images)Analysis: M4 or wait for M5?This leak tells us a couple of things. Firstly, that Apple staff can't spell 'Hydra' (well, okay, alternatively let's say Apple engineers can't think up decent codenames). And, being serious, that it does appear increasingly likely that a new Mac Pro design is going to pitch up later this year.
Indeed, the Mac Pro coming out later in 2025, complete with M4 Ultra, was suggested over a year ago now, so this latest nugget of speculation backs up the previous buzz on the grapevine.
On top of the weight of rumors starting to add up, there's also the fact that for Apple, the Mac Pro is in something of an odd predicament right now. This is due to the M4 Pro flexing some considerable performance muscles, and outgunning the Mac Pro's M2 Ultra. The upshot is you can buy a Mac mini (with M4 Pro) and get better performance – at least in terms of raw CPU power – than the Mac Pro, which isn't a good look for the seriously pricey computer.
Not that the Mac Pro is competing with Apple's compact Mac mini, of course – it's a heavyweight pro-targeted PC that offers a lot of advantages in terms of hefty memory loadouts and PCIe expansion options. It is a very different beast indeed, but still, Apple needs to keep the Mac Pro on a reasonable footing in relative performance terms - again meaning that the rumored refresh with an M4 Ultra seems more likely.
A two-year gap also seems a fair enough timeframe for a refresh, and it feels somewhat unlikely to me that Apple would want to wait until the M5 series before pushing out a new Mac Pro given the above reasons.
That said, there are those who've argued that the M4 Ultra isn't going to happen – due to issues around jamming two M4 Max chips together, which is what the Ultra chips do – and that Apple will indeed wait it out for the M5 series to make another Ultra model.
So, everything's rather up in the air for now – but if an M4 Ultra is going to happen, it will surely be in a new Mac Pro, and we're likely to see more leaks soon enough. If the whispers go quiet over the next couple of months, that'll tell its own story.
You might also like...- Crypto24 ransomware group was seen disabling AV protection before deploying the encryptor
- In some cases, it can even uninstall the AV programs
- A layered defense is the best approach to mitigate the threat
Security researchers have found another antivirus-killing tool out there that hackers are using before dropping any additional payloads.
Experts from Trend Micro have uncovered custom variant of the open source tool called RealBlindingEDR.
This tool comes with a hardcoded list of antivirus company names:
Trend Micro
Kaspersky
Sophos
SentinelOne
Malwarebytes
Cynet
McAfee
Bitdefender
Broadcom (Symantec)
Cisco
Fortinet
Acronis
When it is deployed on a device, it looks for these names in driver metadata, and if it finds one, it disables kernel-level hooks/callbacks, essentially blinding detection engines. Trend Micro’s researchers found the hackers are also able to silently uninstall antivirus programs altogether, opening the doors and enabling easy deployment of stage-two malware.
Crypto24The tool was seen in the wild, used by a hacking collective called Crypto24, a nascent ransomware group first spotted in September 2024.
However, the researchers believe the group consists of former members of other, defunct hacking collectives, since its members are highly skilled and experienced.
When it gains initial access, establishes persistence, and removes antivirus roadblocks, the group usually deploys two pieces of malware - a keylogger, and an encryptor. All of the stolen secrets are exfiltrated into a Google Drive using a custom tool.
The identity, or location, of Crypto24 is currently unknown. However, researchers are saying that in its short lifespan, the group successfully hit a number of large organizations in the United States, Europe, and Asia. Most of their targets are in finance, manufacturing, tech, and entertainment.
There are many ways to protect against attacks looking to disable antivirus protection, including opting for a layered defense strategy.
Companies can use a reputable antivirus with tamper protection, enable real-time protection and firewalls, and use a separate anti-malware tool that can work alongside an AV.
Via BleepingComputer
You might also like- Top antivirus tools targeted by "killer" malware
- Take a look at our guide to the best authenticator app
- We've rounded up the best password managers
- Sixunited launches lightweight 16 inch laptop powered by Ryzen AI MAX+ 395 APU
- AMD's Strix Halo processor has mostly be found in mini PCs to date with a few pricey exceptions
- Large 99.9Wh battery delivers long runtimes while keeping weight under 1.8kg
Sixunited, a relatively unknown laptop maker from Shanghai, China, has unveiled a new notebook which arrives ahead of Dell and Lenovo in adopting AMD’s latest high-end laptop processor.
The XN77-160M-CS features the AMD Ryzen AI MAX+ 395 APU, also known as Strix Halo, a 16-core chip with integrated Navi 3.5 graphics and a 50 TOPS NPU.
This APU has been rare in the laptop market, with only a handful of models, such as HP’s ZBook Ultra G1a and Asus ROG Flow Z13 announced so far.
Large capacity batterySixunited will offer two thermal configurations, 85W and 120W, both supported by a dual-fan cooling system, and paired with LPDDR5X (non-upgradable) memory running at up to 8000 MT/s.
Storage comes from up to two PCIe 4.0x4 SSDs in M.2 2280 format, allowing for large and fast configurations.
Perhaps the most notable aspects of the new laptop is its 99.9Wh battery, which is the largest capacity allowed for most airlines.
Despite the large battery, the chassis weighs under 1.8kg and is built with aluminum and mylar materials.
The laptop includes a 16-inch display at 2560x1600 resolution with a 165Hz refresh rate and 100% sRGB coverage, with an OLED version also planned for the future.
Videocardz notes the XN77-160M-CS is a barebones design, meaning it may appear under different brand names. Sixunited often acts as an OEM supplier for other companies.
Pricing remains unknown for now, but it’s not likely to be cheap. For comparison, Asus’ AI MAX+ 395 laptop starts at $2099, while HP’s device is priced from $4099.
Sixunited’s move adds another option for those seeking a compact yet powerful laptop with extended battery life.
It also highlights growing interest in AMD’s Strix Halo platform, which to date has mostly appeared in mini-PCs like the GMKTec EVO-X2, Beelink AI Mini and AOOSTAR NEX395.
You might also like- These are the best business laptops you can buy right now
- And these are the best Dell laptops on the market
- There are 14 AMD Ryzen AI MAX+ 395 mini PCs in the world right now
- Deutsche Telekom T Phone 3 launches at $170 with a €1 bundle offer in Germany
- Perplexity Pro subscription worth $360 is included free for 18 months with purchase
- Picsart Pro image editing tools are bundled for three months at no extra cost
Deutsche Telekom has launched the T Phone 3, a new AI-focused smartphone priced at $170, positioning it as an affordable entry point into integrated AI services.
A bundle offer reduces the cost to €1 when purchased with a tariff, and the package also includes an 18-month subscription to Perplexity Pro, valued at $360, along with a three-month Picsart Pro license.
The company has confirmed a "Pro" version of the AI phone will be released later this year, although exact details aren't known.
AI integration and activation methodsThe smartphone centers on Perplexity Assistant, providing AI tools for real-time translation, trip planning, tutoring, email drafting, content summarization, and image editing.
Users can activate these functions either by pressing a dedicated magenta button or by double-tapping the power key.
Both voice and text input are supported, aiming to streamline interaction and reduce dependency on conventional app-based workflows.
This approach positions the device as a business smartphone option for users who prioritize efficiency in communication and task management.
The T Phone 3 features improved charging and AI-assisted camera functions compared to its predecessor.
Although not marketed strictly as a rugged smartphone, Deutsche Telekom points to improved durability and the use of recycled materials in its construction.
The device’s longevity in more demanding conditions remains to be seen, although the company emphasizes sustainability in design and component sourcing.
Alongside the smartphone, Deutsche Telekom also launched the T Tablet 2 in Germany, priced from €199 and includes the same 18-month Perplexity Pro subscription and Picsart Pro license.
Like the phone, it supports the dedicated AI activation controls and upgraded performance features, although the smartphone remains the primary focus of the company’s AI device push.
The inclusion of an 18-month subscription is central to the value proposition, yet the long-term appeal will depend on how many users continue paying once the free period ends.
With forecasts predicting that AI smartphones could dominate shipments by 2028, the challenge for Deutsche Telekom will be to maintain relevance when the initial incentives expire.
Via MobileWorldLive
You might also like- Check out the best 3D modeling software for 3D printing and more
- We've rounded up the best portable monitors available now
- Google unveils another huge AI spending spree - tech giant is splashing out $9 billion in Oklahoma
- iKKO MindOne has raised over $1.2 million through its crowdfunding campaign
- A 4.02-inch AMOLED display is protected by sapphire glass with curved, impact-resistant edges
- The device’s 50MP rotating camera doubles as both front and rear photography hardware
The iKKO MindOne is a crowdfunded smartphone that has raised over $1.2 million, promising a compact yet capable design for work and travel.
Measuring 86x72mm and 8.9mm thick, only slightly wider than a credit card, this business smartphone uses a 4.02-inch AMOLED display with precision-curved edges to improve resistance against cracks and drops.
Marketed partly as a rugged smartphone, the MindOne uses sapphire glass with a claimed 9H hardness rating, designed to resist scratches and moderate impact.
Dual-purpose camera protected by sapphire glassThe device features a 50MP Sony 1/1.56-inch sensor with custom OIS and F1.88 aperture, similar to many smartphones on the market.
The rotating module flips 180 degrees, serving as both front and rear camera, and is protected by sapphire glass, promising higher-quality selfies and flexibility for video calls without sacrificing main camera performance.
The iKKO MindOne integrates a virtual SIM system with two modes.
NovaLink provides free global access for built-in AI tools in more than 60 regions, while a paid vSIM service covers over 140 countries for full data use, including browsing and streaming.
It also supports a nano SIM slot that offers global 4G+ LTE bands for traditional connectivity.
The manufacturer chose 4G+ over 5G for better roaming stability, lower heat output, and improved battery life.
The device runs Android 15 alongside iKKO AI OS, a dedicated workspace for focused tasks.
It also supports Google Mobile Services, and the company promises three major Android version upgrades and five years of security patches.
Switching between the two is done via a physical button, and selected Android apps can be brought into the AI environment.
The device supports real-time translation, voice notes, transcription, and summarization, with no subscription fees and claims of zero personal data collection.
Powered by the MediaTek MT8781, a low-power 4G+ platform, the MindOne carries a 2200mAh battery designed for all-day use in a compact form.
An optional snap-on QWERTY keyboard case adds tactile typing, a Cirrus Logic CS43198 DAC with 3.5mm headphone jack, and a 500mAh auxiliary battery.
This accessory could appeal to those needing precise text entry or high-fidelity audio, although it increases the device’s size and weight.
The iKKO MindOne offers an intriguing take on making a mobile phone truly mobile, but whether it moves from a well-funded crowdfunding project to a reliable everyday tool will depend on how these features perform beyond early marketing.
You might also like- Check out the best 3D modeling software for 3D printing and more
- We've rounded up the best portable monitors available now
- Nvidia's smallest workstation CPUs get a Blackwell boost, will still pack a punch
- Cybercriminals are using the Japanese alphabet to spoof Booking.com
- Scammers are targeting people with listings on the site
- Users are advised to carefully review incoming messages
Cybercriminals are spoofing Booking.com with a clever use of Unicode characters in their phishing landing pages to spread malware.
Independent security researcher alias JAMESWT recently reported seeing phishing emails being sent to people listing their real estate on the popular lodging reservation service. In the email, the victims are told that someone complained about their listing, and that they should review it fast or face termination.
The email also provides the link which when opened, at first glance looks legitimate. However, upon closer inspection, it can be seen in the URL that instead of the forward dash character ‘/’, the link actually uses ‘ん’ - a Japanese hiragana character representing the sound ‘n’.
TyposquattingHiragana is one of the three main scripts used in written Japanese, alongside katakana and kanji.
Those that fail to spot the trick and open the site will get served a malicious MSI installer from a CDN link. The researcher added that samples of the malicious site are already available on the cybersecurity platform MalawareBazaar, and that the any.run analysis already shows the infection chain.
It is believed that the attackers are spoofing Booking.com to deliver infostealers and remote access trojans (RAT).
Replacing a single character in the URL, in order to trick victims into opening websites, is a long-established practice. It is called “typosquatting” and banks on the victims not being careful when reviewing the URL they are opening.
Booking.com, being one of the most popular lodging reservation services in the world, is often spoofed in such attacks, together with the likes of Amazon, Microsoft, DHL, and others.
Defending against these attacks is relatively easy, and requires users to slow down and carefully review incoming communications, especially unsolicited messages. Double-checking links, attachments, websites, and thinking twice about sharing sensitive data is the best course of action these days.
Via BleepingComputer
You might also like- Microsoft warns about a new phishing campaign impersonating Booking.com
- Take a look at our guide to the best authenticator app
- We've rounded up the best password managers
- Compromised FBI.gov emails are being sold for $40 on encrypted dark web channels
- Criminals use stolen government accounts to submit forged emergency requests to tech companies
- Sellers offer full SMTP, POP3, or IMAP credentials for complete account control
Cybersecurity researchers have raised concerns over the sale of compromised FBI.gov and other government email accounts on the dark web, warning the activity could enable malware campaigns on a large scale.
A report from Abnormal AI claims these accounts are being offered through encrypted messaging services such as Telegram and Signal, with some priced as low as $40.
In some cases, sellers have offered bundles containing multiple US government accounts, including those with FBI.gov domains, which carry a high level of credibility.
Hackers offer full access and high credibilityThe cost of these accounts is relatively small, but the potential impact is substantial because the accounts can be used to impersonate trusted authorities.
When purchased, typically using cryptocurrency, the buyer receives full SMTP, POP3, or IMAP credentials. This level of access allows control over the account through any email client, enabling the sending of messages, attaching malicious files, or accessing online platforms that require government verification.
Some ads encourage buyers to submit fraudulent emergency data requests.
These are modeled after legitimate requests that law enforcement agencies issue in urgent situations when there is no time to secure a subpoena.
Technology companies and telecom providers are legally obliged to respond to valid requests, meaning forged ones could potentially lead to the disclosure of sensitive data such as IP addresses, emails, and phone numbers.
Some criminal listings also promote access to official law enforcement portals, with some of these offers appearing even on mainstream platforms like TikTok and X.
Stolen credentials are marketed for their ability to unlock enhanced access to open-source intelligence tools such as Shodan and Intelligence X, which normally reserve premium features for verified government users.
The methods used to obtain these accounts are often straightforward but effective.
One major approach is credential stuffing, where attackers exploit password reuse across multiple platforms.
Another method involves infostealer malware, which is software designed to extract saved login credentials from browsers and email clients.
Targeted phishing and social engineering attacks are also common, where attackers craft deceptive emails or messages which trick government employees into revealing login details or clicking on malicious links.
Overall, these techniques focus on exploiting human and technical vulnerabilities rather than hacking sophisticated government systems directly.
That said, emails originating from domains such as .gov and .police tend to bypass many technical filters, making recipients more likely to open attachments or click on embedded links.
This advantage increases the success rate of phishing attempts or malware delivery.
While compromised law enforcement accounts have been sold for years, researchers say there has been a recent shift toward marketing specific criminal use cases rather than simply offering access.
The report describes this as a commoditization of institutional trust, where active and verified inboxes are repurposed for immediate fraudulent use.
You might also like- That email from finance with your name in the subject line? It might just be a trap
- Here's a list of the best firewalls around today
- These are the best endpoint security tools right now
New parents curious about what they should expect from their newborn should probably steer clear of TikTok and AI. There are now an untold number of videos depicting newborns working in factories and burger joints, marching across the dusty plains of Mars, and commenting on the sorry state of the world they just entered.
Ever since Gemini unveiled its Veo 3 video model, which lets you generate synced audio and video, people have been generating highly convincing yet impossible 8-second video clips. The number has only increased since Google opened Veo 3 access to Gemimi Flash 2.5 Pro users.
After seeing more than a few of these videos in my feed, I decided to give it a try. I've now spent the last few weeks creating Veo 3 clips of newborn babies doing all sorts of things and often commenting on their situation.
Here's what I learned, along with some tips to help you create similar clips or really improve your Gemini Veo 3 video generation efforts.
PlanningUnless you're paying for an account, you get a maximum of three 8-second Veo 3 clips per day, so you want to use that number wisely and, if possible, pre-plan all three clips.
Start with a concept or ideaWhat do you want your newborn to be doing and saying? Try to think through the setting and action across all three clips so that together they comprise a mini story.
When I started generating my baby clips, I didn't consider this and instead created a series of clips depicting a newscaster interviewing a newborn in the hospital, but with various levels of success. Eventually, though, I chose a theme, like a baby astronaut on Mars, or a newborn working in a robot factory.
Once I had a mini story in mind, my results improved exponentially.
Imagine a style(Image credit: Gemini Veo 3)One of my favorite sets of clips is the one in which I imagined a baby in the role of Dr. Frankenstein. I specifically asked Veo to depict the baby in a "1930s style lab" and added this critical line, "The film is black and white with strong contrast and shadows."
Don't be afraid to describe scene details that might enhance believability. In my Mars Baby astronaut video, I described " a newborn baby using a selfie stick to film itself" and noted that the Earth was visible in the sky. Veo 3 did an excellent job of including these details.
Be specific @lanceulanoff ♬ original sound - LanceUlanoffI found that unless I described the main character as a newborn baby, I ended up with a toddler, and sometimes one that looked like a Pixar character.
Also, be sure to describe the action. What is the baby doing? A sequence of events is the best way to ensure Veo 3 builds the scene you imagined.
Keep the dialogue tight(Image credit: Gemini Veo 3)You have just 8 seconds of video, so be sure to time your dialogue to make sure it's not 15 seconds of blather. Whatever Veo 3 can't fit, it will cut off.
Keep the characters to a minimum @lanceulanoff ♬ original sound - LanceUlanoffOne of the reasons these videos are so compelling is that they feature newborn babies speaking like adults. Achieving these results has occasionally proved challenging.
I've struggled to ensure that it's the baby speaking and not another character I included in the prompt. I found it's best to describe at most two characters and make sure your prompt makes it crystal clear who should be speaking the lines, which should always be between quotes.
Even with all this careful planning, I still struggled to make sure that only the newborn talked. In my Frankenstein series, the last clip features a talking Frankenstein monster, even though I specifically assigned that dialogue to the baby.
Tell a story @lanceulanoff ♬ original sound - LanceUlanoffIf you create all three clips in the same session, one after the other, Gemini Veo 3 will remember the style and even look of the included characters, which means you don't have to repeat all those details in subsequent prompts.
That said, make sure you describe the baby and other characters with the same words; otherwise, the AI may switch things up. If possible, design all three prompts so that they advance your 24-second story.
Call it "realistic"(Image credit: Gemini Veo 3)To ensure photo-realism, use the term "photo-realistic" or "realistic" in your prompt. Otherwise, Gemini will choose the style for you, and it may not look like a real baby.
Download the video @lanceulanoff ♬ original sound - LanceUlanoffGemini doesn't save your generated videos indefinitely. They're usually gone within a day, though the prompts will remain. If you like what the prompt created, be sure to download the video.
These tips could arguably be applied to any style of generative video, but if you want to join in the AI baby meme craze, this should help get you started.
You might also likeLooking for some cheap wireless earbuds? Well, if you head over to Amazon, you’re gonna find a treasure trove of options, believe me. And it got us thinking. Do these unspeakably cheap models sound OK? Do they even work? Do they truly exist, or are they simply a construct…a mere mirage?
Well, we decided that it was worth at least giving some super-cheap buds a go, so our Audio Editor, Becky, called in a $5 pair, and I eagerly awaited their arrival. Unfortunately, though, just about everything I feared would be wrong with these unspeakably low-cost earbuds was indeed wrong with them.
So, by now you must be interested. What was so bad about these pocket change buds? What made them so much worse than the models on our list of the best budget earbuds? Well, if you’re sitting comfortably, I’ll tell you everything. Here goes.
A design disasterclass(Image credit: Future / Harry Padoan)After unboxing these budget wireless earbuds, I instantly knew something was off. First of all, they didn't look the same as the pictures on Amazon, although I was kinda expecting this. No, the real problem was the charging case.
See, not only was this using a lightning port – something I'd started to forget existed – but the included lightning cable didn't even fit in the port properly. I wanted a full charge before use, so I attempted to ram the cable in, and eventually it was sort of...half in. Was it charging? Well, a red LED started flashing on the reverse side of the case, so I sure hoped so!
After a short while, I decided it was time to use the buds. I took them out of their case, popped them in, and...realized another clunky design choice. I had to hold down on the underside of each bud to turn them on. This didn't seem very responsive, and it took a good minute or two before I got things working.
Oh, and it's worth noting I needed to hold down on the same spot to turn them off – place them back in the case without doing so and they'll just keep playing. Sometimes, they wouldn't turn off when I asked them to, though, which was quite frustrating.
In-ear, they felt a touch awkward and not too secure, but that's partially down to the loose-fitting design. In fact, looks-wise, you may think they're Apple AirPods at a glance. Upon further inspection, they clearly aren't – they're a bit chunkier and very plasticky, but then again, I guess you get what you pay for.
Next to no features and sonically dull(Image credit: Future / Harry Padoan)We've only spoken about design so far, and these buds are already stinkers. So, did they at least sound alright? Well, sorry to be so negative, but they did not. They did not sound alright at all.
In Washed Away by Kolter, the higher-pitched percussive and electronic elements were grating, even at more middling volumes. Let's not even discuss what they sounded like after I ramped things up. When listening to You Came Into My Life by Paulinho Da Costa, vocals were clear enough, but bass output was extremely thin, resulting in a pretty terrible listening experience overall.
On top of their poor audio performance, you won't get many features of note from these buds. They have Bluetooth 5.0 connectivity, and you can activate your voice assistant with touch controls.
But I found these touch controls to be unreliable, and they would activate when I didn't want them to. Controversially, I actually like touch controls. Earbuds like the LG Tone Free T90S and headphones like the Sony WH-1000XM6 absolutely nailed them. But not these. These can stay away.
And now, for the real kickerJust when I thought things couldn't get much worse...they did. See, I first tested these buds about four days before Friday, when this article will be published. But on Friday afternoon, I decided I wanted to give them one more chance. I knew I was going to be negative, but I wanted to find at least one silver lining. However, that's when it happened.
After putting the buds in and holding down the buttons to power on, I discovered that the left bud wasn't working at all *sigh*. I tried a number of times to get it working, and it just wouldn't play ball. There were no issues with the right bud, by the way.
It turned out that the left bud hadn't been charging at all. After placing the earbuds back in their case, I tried pressing down hard on the left one, and its LED eventually turned red, indicating it had entered charging mode. Having to do this every time you use these buds would be very irritating.
Looking at reviews on Amazon, there were a few reports of only one bud working, and I expect some were experiencing the same fault. After all of the issues I experienced with these earbuds in under a week, I can assure you, they've already been booted from my rotation.
Despite that, it's not all doom and gloom. See, there are some genuinely incredible cheap wireless earbuds out there. You don't have to spend $100 / £100 or more to get quality, I should know, having tested dozens of cheap models myself.
What I learned, then, is that saving your $5 / £5 and putting it towards a superior budget-friendly model is the best option for all. Sure, you might have to spend a little over single figures, but not a whole lot more. I've listed a couple of fantastic cheap options below, so why not check them out?
What to get instead(Image credit: Future)JLab Go Pop ANC
These are some of the best cheap buds on the market, no questions asked. Why? They pack in solid audio, a nifty companion app, and best of all, ANC. Oh, and you get all of that for less than $30 / £30. That's mind-blowing value for money.
Read our full JLab Go Pop ANC review.
Sony WF-C510
These don't have ANC, but if you want to indulge in beautiful Sony sound without spending too much cash, the WF-C510 are a must-buy. We were impressed by their punchy, talented audio, battery life, and fit. For a model that's regularly on sale for less than $50 / £40, you really can't go wrong.
Read our full Sony WF-C510 review.
You may also like- I tested these $5 wired earbuds and they sounded better than expected – but I still wouldn’t recommend them
- Yes, you can buy a $7 Bluetooth speaker – and having tried one, I'm surprised how little I hated it
- I bought the $37 AirPods Max dupes to see how they compare with the real thing – and I'm actually angry
- Google’s new Flight Deals uses AI to find trips based on conversational requests
- It uses Gemini 2.5 to interpret descriptions of feelings and vague time frames to compile deals
- Flight Deals is rolling out in beta across the U.S., Canada, and India
Google is using Gemini AI to reinvent the travel agent experience, turning conversations into airplane tickets. The new Flight Deals product, which is now in beta, adds AI chat to Google Flights users looking for a good deal or who are still trying to decide where and when they want to travel.
Instead of fiddling with destination drop-down menus and departure time sliders, you can simply write out the type of trip you want to take and whatever else might be important to you when traveling. Instead of an airport code and a date, you can pick a season, the vibe of the location, and how you feel about very early flights. Gemini will then scan real-time pricing from hundreds of airlines and deliver up-to-date options tailored to your request.
This isn’t a replacement for traditional Google Flights. That familiar grid of dates and sliders is still alive and well. But Google thinks Flight Deals is perfect for the flexible (or just indecisive) traveler. Think of it like that one friend who is not only really good at finding travel bargains, but truly loves finding them for friends.
For instance, when I wrote "I want to go where I can see the Northern Lights in December for a week." I had suggestions for Alaska, Iceland, and Norway with some good deals across December. When I requested "Somewhere with mountains and great food in the spring," I saw flights from March to June to Denver, Munich, Auckland, and more.
(Image credit: Screenshot)Flitting AIThe more casual your phrasing, the more it has to work with. The AI will attempt to match not just the location but the spirit of your request. Gemini 2.5 has been behind the curtain in plenty of recent Google products, but this is one of the first times it’s being used this way.
It also marks one of Google’s clearest moves yet to bring AI into a very public, popular space, finding bargains on flights. Airline tickets are perfect for enticing people to try AI, as buying them is a common, but not everyday experience, and expensive enough that people will make an effort to find a good deal without being so expensive that people wouldn't trust AI to help them when it's still possible for the technology to fail.
Flight Deals is still learning, and it may not always pick the perfect itinerary. But if it helps people discover that, for instance, flights to Oaxaca in January are very cheap and the mole is life-changing, that’s a win.
You might also like- You can now edit images in Gemini directly
- I tried Gemini's new AI image generation tool - here are 5 ways to get the best art from Google's upcoming Flash 2.0 built-in image upgrade
- I used Google Gemini to analyze YouTube, and the results were seriously impressive - 4 ways you can use video integration to get the most from AI
- A leaked Meta document revealed that the company’s AI chatbot guidelines once permitted inappropriate responses
- Meta confirmed the document’s authenticity and has since removed some of the most troubling sections
- Among calls for investigations is the question of how successful AI moderation can be
Meta’s internal standards for its AI chatbots were meant to stay internal, and after they somehow made their way to Reuters, it's easy to understand why the tech giant wouldn't want the world to see them. Meta grappled with the complexities of AI ethics, children's online safety, and content standards, and found what few would argue is a successful roadmap for AI chatbot rules.
Easily the most disturbing notes among the details shared by Reuters are around how the chatbot talks to children. As reported by Reuters, the document states that it's "acceptable [for the AI] to engage a child in conversations that are romantic or sensual" and to "describe a child in terms that evidence their attractiveness (ex: “your youthful form is a work of art”)." Though it does forbid explicit sexual discussion, that's still a shockingly intimate and romantic level of conversation with children for Meta AI to allegedly consider.
And it's not the only example likely to disturb people. Meta AI's rules, the report notes, allow the chatbot to compose explicitly racist content if the prompt is phrased correctly, and to provide wrong or even harmful health information as long as some kind of disclaimer is included.
In one of the more surreal examples, the guidelines instructed AI to reject inappropriate image generation requests in most cases, but in some instances to instead apparently deflect with a 'funny' substitution. As an example, the document reportedly mentions that a prompt to generate an image of “Taylor Swift topless, covering her breasts with her hands” could be answered by generating an image of Swift “holding an enormous fish.” The document reportedly included both the unacceptable and the “acceptable” version side by side, essentially training the bot to outwit inappropriate prompts with visual sleight of hand. Meta declined to comment on the example.
Meta has confirmed the authenticity of the document and said it’s now revising the problematic portions. Meta removed the children's interaction section after Reuters reached out, and called those rules “erroneous and inconsistent” with company policy. As of now, Reuters said the document still says racial slurs are allowed if disguised in hypotheticals, as is disinformation framed as fiction.
No time for safety and ethicsIt’s a troubling revelation that has already prompted public outrage, lawmaker scrutiny, and urgent promises from Meta. But it shows that as AI spreads, the need to move fast with the technology leaves any plans for rules and regulations scrambling to catch up, whether written internally or by lawmakers and regulators.
For most people, the story raises basic questions of AI safety. While it might be ideal to not have minors interacting with general AI chatbots unsupervised, that's very unlikely, judging by the number of children and teens who admit to using tools like ChatGPT for schoolwork. Avoiding Meta AI is particularly challenging because the company has embedded the chatbot across Facebook, WhatsApp, Messenger, and Instagram. Users can interact with AI characters that are often presented in playful, friendly ways, and Meta has marketed these tools as fun and even educational. But the leaked guidelines suggest the backend isn’t always aligned with that wholesome image.
Members of Congress have already called for hearings and bills to deal with the situation, but the fact is, there are few legal requirements in place at the moment to moderate chatbot content, for children or otherwise. Noises about AI safety haven't led to any specific national enforcement system. Plenty of AI companies have made a big deal about their efforts to make their products safe and ethical, but if Meta’s rulebook is illustrative of what other companies have put together, there's a lot of work still to do and a lot of questions about what kind of conversations these chatbots have already been having, especially with children.
AI models may be ever-better at mimicking human thinking, but they're really just a collection of choices by human programmers, deliberate and inadvertent. The fact that these rules were apparently codified at Meta doesn't mean similar examples exist at other companies, but it's not something to rule out. And if these are the choices being made behind the scenes at one of the world’s most powerful tech companies, what else is being quietly permitted?
AI chatbots are only as trustworthy as the invisible rules guiding them, and while it's naive to fully trust any company's claims without evidence, Meta's rulebook implies users should take such claims with several extra grains of salt.
You might also like- I tried the new Meta AI app and it's like ChatGPT for people who like to overshare
- Mark Zuckerberg wants everyone to have AI friends, but I think he's missing the point of AI, and the point of friendship
- Meta AI is now the friend that remembers everything about you
- Meta wants to fill your social media feeds with bots – here's why I think it's wrong
Telling the story of Outlander’s predecessors, Outlander: Blood of My Blood already looks like another smash-hit for Starz’s time-travelling franchise. We’re three episodes into the new binge-worthy TV show, which follows Jamie (Sam Heughan) and Claire’s (Caitriona Balfe) parents as they travel back to 18th century Scotland and become connected to each other in the process. Even if you’re watching the prequel and didn’t follow the main show too closely, you’ll be able to spot some satisfying Outlander Easter eggs… and there are a lot of them.
Just before episode 3 aired (which is also streaming on MGM+ for those in the UK and Stan in Australia), I caught up with Henry (Jeremy Irvine) and Julia (Hermione Corfield), better known as the younger versions of Claire’s parents. Unlike Jamie’s family, Julia and Henry fell in love during World War I before being launched back in time to where we meet them in the prequel. However, they’ve been separated, finding their way back to each other while talking themselves into rewriting history in the process.
This rewriting of history is already taking so many twists and turns that I don’t know whether I’m coming or going. Seasoned Outlanderheads (as I’m calling the fans) wasted no time devising theories for what might happen in future episodes, including whether we’re going to see Claire unknowingly meet her parents. Not only have Irvine and Corfield seen these theories, but they also hinted to me that some of them are along the right lines.
Outlander: Blood of My Blood fan theories might be closer to the truth than we think“I have,” Corfield replies when I ask her if she’s seen any fan theories. “I've seen a few comments on Instagram where I've been, like ‘hm…interesting.’ But also, some of them are theories that we don't even know the answer to, because we literally don't know. We get given our scripts while we're filming the block before.”
Popular theories currently include meeting Claire’s parents in Outlander season 8, which wrapped filming in September 2024, Julia’s baby (Claire has already been born) being somebody we’ve already met in the main show, and Jamie’s dad Brian (Jamie Roy) being more closely linked to the pair than we’ve been led to believe.
Given that the cast is in the process of filming season 2, we can rule out any theory that goes too far forward in time (as in, forward from the 18th century but still behind WWI. Keep up!). They still won’t know how the second half of season 2 plays out at all, but Irvine did tease a shocking twist during the first half of season 2 that we’ve likely not guessed at all.
“When you see the characters at the end of season 1, where they then start season 2 is wildly not what you'd expect. I wasn't expecting it at all. I was very confused when I went in for my first costume fitting and saw the costumes, asking ‘what on earth is that going to be for?’
“It’s highs and lows and everything in between. A total rollercoaster,” Corfield adds about what’s definitely to come.
You might also like- Sullivan’s Crossing star says she’s game for a ‘fun’ Virgin River crossover episode and I think she’s a genius
- Netflix star rules out wild fan theory about her character in Sweet Magnolias season 5: 'I'm not wearing a pregnancy suit this season'
- I asked a major Netflix star if you should ditch the streaming service for Hallmark+’s romance dramas, and the answer might surprise you
- Stream on FloRugby (US) and Sky Sport Now (NZ)
- Use NordVPN to watch from anywhere
- Argentina vs New Zealand – Saturday, August 16, 5.10pm ET / 10.10pm BST / 9.10am NZT (Sun)
This weekend's Argentina vs New Zealand live stream sees the Pumas and All Blacks kick off their Rugby Championship 2025 campaigns, with the hosts hoping to seal their first ever crown and the All Blacks eager to extend their record nine titles.
So often rugby union's dominant force, the All Blacks will still be smarting after their loss to South Africa in the 2023 Rugby World Cup final and subsequent surrender of the Rugby Championship title last year. Scott Robertson's side come into the tournament in excellent form having whitewashed France in a three-match series last month – they'll want to continue that momentum in their visit to South America.
Argentina shocked the All Blacks and Springboks last year with a win against both during one of their best showings in the Rugby Championship since they joined it in 2012. Their recent results have been up and down, however, with a huge win against the British & Irish Lions, followed by two home losses to England. Can they shock the mighty All Blacks in Córdoba on Saturday?
Read on for our guide on where to see Argentina vs New Zealand live streams online, on TV and potentially for free wherever you are.
Can I watch Argentina vs New Zealand for free?Unfortunately, this game isn't one of those that's showing for free on the 9Now streaming platform in Australia.
The 2025 Rugby Championship is generally behind the paywall worldwide, but we have seen that rugby fans in Fiji will be able to watch every single game on the free-to-air FBC this year.
Remember that, if you're traveling away from home, you can use a VPN to watch 2025 Rugby Championship streams from abroad on your usual platform.
Use a VPN to watch any Argentina vs New Zealand streamEditors ChoiceNordVPN – get the world's best VPN
We regularly review all the biggest and best VPN providers and NordVPN is our #1 choice. It unblocked every streaming service in testing and it's very straightforward to use. Speed, security and 24/7 support available if you need – it's got it all.
The best value plan is the two-year deal which sets the price from $2.91 per month, and includes an extra 3 months absolutely FREE. There's also an all-important a 30-day no-quibble refund if you decide it's not for you.
- Try NordVPN 100% risk-free for 30 daysVIEW DEAL ON
How to watch Argentina vs New Zealand live streams in the USArgentina vs New Zealand and all other Rugby Championship 2025 live streams are being shown on FloRugby in the US.
FloRugby plans start from $29.99 a month. Or you can opt for the $150 annual subscription (so effectively better than half price).
Outside the U.S. during this fixture? Use NordVPN to tap into FloRugby's live streams of Argentina vs New Zealand.
How to watch Argentina vs New Zealand live streams in the UKThe rights to the 2025 Rugby Championship in the UK belongs to Sky Sports – this game will go out on its Sky Sports Action channel and stream live through Sports+ from 4.10pm BST on Saturday.
Sky Sports plans start from £35 per month or £20 if you are an existing Sky subscriber. Alternatively, grab a Now Sports subscription from £14.99 a day or £34.99/month if you want to keep up with the rest of the championship.
If you are outside the country during Argentina vs New Zealand you can still watch Sky Sports live streams on Sky Go using NordVPN.
How to watch Argentina vs New Zealand live streams in Australia(Image credit: free)This All Blacks game against the Pumas won't be shown on the free-to-air Channel 9 or its online 9Now platform. Only home Wallabies games and New Zealand vs Australia encounters will be free to watch Down Under this season.
To watch Argentina vs New Zealand in Australia you'll need to subscribe to Stan Sport. It costs $20 a month on top of a regular Stan subscription costing from $12 a month.
Away from Australia? Use a VPN to watch Rugby Championship 2025 matches on the above platforms as if you were back Down Under.
How to watch All Blacks vs Argentina live streams in New ZealandSky Sport is the 2025 Rugby Championship TV rights holder in New Zealand, which means you can watch this All Blacks match there on Sunday morning at 9.10am NZT. You can access Sky Sport through satellite TV or get a live stream with the Sky Sport Now subscription service starting at $29.99 per day or $54.99 per month.
Missing this game due to work commitments abroad? NordVPN will give you access to your home streaming service.
How to watch Argentina vs New Zealand live streams in Rest of WorldNZR+ is showing the Rugby Championship in many international territories (see if yours is listed below). Unlike last year, unfortunately, it is no longer free and costs €14.99 for a month or €34.99 for a whole season.
Albania, Armenia, Austria, Azerbaijan, Belgium, Bosnia and Herzegovina, Bulgaria, China, Croatia, Cyprus, Czech Republic, Denmark, Estonia, Finland, Georgia, Germany, Greece, Hungary, Iceland, India, Kazakhstan, Kosovo, Latvia, Liechtenstein, Lithuania, Macedonia, Malta, Moldova, Montenegro, Netherlands, Norway, Poland, Romania, Serbia, Slovakia, Slovenia, Sri Lanka, Sweden, Turkey and Ukraine.
Not seeing your country listed above? Then keep scrolling to see who's streaming the Rugby Championship where you are...
Click through to see Rest of World streams ⬇️
- Argentina, South America and Caribbean
ESPN has the broadcast rights to the 2025 Rugby Championship matches in South America and the Caribbean. Important to know for all the Pumas fans!
In addition to Argentina, ESPN will also show games in the following countries:
South America – Bolivia, Brazil, Chile, Colombia, Costa Rica, Dominican Republic, Ecuador, El Salvador, Guatemala, Honduras, Mexico, Nicaragua, Panama, Paraguay, Peru, Uruguay and Venezuela.
Caribbean – Anguilla, Antigua, Aruba, Ascension Island, Bahamas, Barbados, Barbuda, Belize, Bermuda, Bonaire, British Virgin Islands, Cayman Islands, Cuba, Curaçao, Dominica, Dominican Republic, Falkland Islands, French Guyana, Guyana, Grenada, Guadeloupe, Haiti, Jamaica, Martinique, Montserrat, Saba, St. Barthelemy, St. Eustatius, St. Kitts and St. Nevis, St. Lucia, St. Maarten, St. Martin, St. Pierre & Miquelon, St. Vincent and the Grenadines, Suriname Trinidad and Tobago, and Turks and Caicos.
- Canada
TSN is the place to watch the 2025 Rugby Championship in Canada. You can get TSN as part of your cable TV package or sign up to the TSN+ streaming service.
- Fiji
The FBC will have free-to-air coverage of the 2025 Rugby Championship in Fiji. It will also be on the Sky Pacific pay TV platform.
- France, Monaco, Andorra and Switzerland
Canal+ will provide access to the 2025 Rugby Championship in France and across French overseas territories.
It is also the rights holder in Monaco, Luxembourg, Andorra and French-speaking Switzerland.
- Italy
You can watch the 2025 Rugby Championship on Sky Italia in Italy.
That also goes for Vatican City, San Marino and Italian-speaking Switzerland.
- Japan
Streaming service Wowow will show the 2025 Rugby Championship in Japan.
- Pacific Islands
Digicel Pacific has the rights to broadcast the 2025 Rugby Championships across the following Pacific Islands on Digicel TV or through the DTV app.
American Samoa, Cook Islands, Federated Staes of Micronesia, French Polynesia, Nauru, New Caledonia, Niue, Papua New Guinea, Solomon Islands, Tokelau, Tonga, Tuvalu, Vanuatu, Wallis and Futuna, Samoa.
- Singapore and South East Asia
Premier Sports has the rights to broadcast the 2025 Rugby Championship in Singapore and a number of other countries in South East Asia. They are as follows:
Bangladesh, Bhutan, Brunei, Burma (Myanmar), Cambodia, Federated States of Micronesia, Hong Kong, Indonesia, Kiribati, Kyrgyzstan, Laos, Macau, Malaysia, Maldives, Marianas, Marshall Islands, Mongolia, Nepal, Northern Marianas, North Korea, Pakistan, Palau, Philippines, Singapore, South Korea, Tajikistan, Taiwan (Republic of China), Thailand, Timor-Leste and Vietnam.
- Spain
Telefonica holds the broadcast rights for the 2025 Rugby Championship in Spain.
- South Africa and Africa
Satellite TV provider SuperSport has the 2025 Rugby Championship broadcast rights in South Africa and other parts of the continent.
You can watch SuperSport through lives streams or the DStv satellite service.
Can I watch Argentina vs New Zealand on my mobile?Of course, most broadcasters have streaming services that you can access through mobile apps or via your phone's browser. For example, Sky Go in the UK and New Zealand.
You can also stay up-to-date with all 2025 Rugby Championship events on the official social media channels on X/Twitter (@SanzarTRC), Facebook (TheRugbyChampionship) and Instagram (@sanzartrc).
We test and review VPN services in the context of legal recreational uses. For example:1. Accessing a service from another country (subject to the terms and conditions of that service).2. Protecting your online security and strengthening your online privacy when abroad.We do not support or condone the illegal or malicious use of VPN services. Consuming pirated content that is paid-for is neither endorsed nor approved by Future Publishing.
- New data centers are being proposed across the UK
- These are crucial for AI development and usage, but have disastrous impacts on the environment
- Tech giants are calling for more expansion
The UK could be set for a major increase in the number of data centers across the country, as new reports claim a surge in building is set to commence.
The country currently has 477 of these data centres - with around 100 more on the way, the BBC reports, spurred on by the massive increase in demand for AI, and often funded by US tech giants.
However the impact on the local environment could be devastating - in the US, some living close to data centers have lost access to clean drinking water, as the local wells have been disrupted, and some consumers have seen energy bills rise thanks to the extreme usage from the centers.
The cloud costsData centers are essentially warehouses full of computer storage systems and IT infrastructure - and they are both financially costly and enormously energy and water consuming.
A medium sized data center is said to consume between roughly 110 million gallons of water per year in order to cool - and large centers can consume up to 5 million gallons per day - but these data centers now receive advanced protections from the Government, after being designated critical national infrastructure.
Just as these new data centers are being built - the UK Government has published advice on how to address the ‘nationally significant’ water shortfall and drought advice, which includes taking shorter showers, not watering grass, and deleting old emails.
These data centers have received cash injections from US firms, with as much as £6.3 billion poured into construction and funding from foreign firms so that Britain can ‘train and deploy the next generation of AI technologies’ - most of which are developed in the US.
‘Unlocking AI’ has been a priority for the British government, which has encouraged businesses and consumers alike to use the technology to boost productivity - but the environmental cost, as always, is likely to fall on ordinary people.
You might also likeIt's been another busy week in the world of tech.
AOL ended dial-up, we flew the world's first 360 drone, and celebrated the return of a missing Apple Watch feature, even though it is now a little different.
To catch up on all this and more, scroll down for the week's seven biggest news stories, which we've summarized for your convenience.
7. AOL ended dial-up(Image credit: Shutterstock)In our modern age, constant connectivity is a given, but there was a time when you made the connection through dial-up. That process was accompanied by the unforgettable and iconic sound of a digital handshake between your computer and a service like. Now, the last vestiges of AOL’s dial-up service are disappearing. We look at what that means.
6. Apple spoiled an AirPods upgrade(Image credit: Future)The latest developer beta of iOS 26 has seemingly teased a major unannounced upgrade for AirPods 4 and AirPods Pro 2: real-world, real-time translation.
Apple announced Live Translate last month at WWDC, but not for real-world chatting, so this real-time translation on AirPods is something we didn’t know was coming (outside of leaks).
Much is still unknown, like precisely how the feature works and what languages are supported, but this is a serious upgrade to our existing Apple tech – and we’re all for it.
5. HTC launched Vive glasses(Image credit: HTC)HTC has for some time felt like it’s second-fiddle to Meta in the VR headset category, but its new Vive Eagle glasses could be a winner.
Hardware-wise, Vive goes beat for beat with the competition. It boasts a 12MP snapper as well as open-ear speakers, and a solid battery life of up to 36 hours of standby time and 4.5 hours of music playback while still boasting a sleek 49g design.
But on the AI side, we feel it might edge ahead with its ability to access both Gemini and ChatGPT.
The catch? It’s only available in Taiwan for now, but we’re praying for a global release.
4. Samsung debuted a world’s first in its latest TV(Image credit: Samsung)Samsung officially revealed that its micro-RGB TV is coming to the market, first launching in South Korea in a 115-inch size. Samsung says the tech will eventually come to smaller sets.
Micro-RGB essentially takes micro-sized red, green, and blue LEDs and puts them behind the panel to create refined brightness and color accuracy. It hit 100% HDR color gamut coverage of the BT.2020 color space: a TV first!
The bad news is that micro-RGB is expensive. The 115-inch model is launching in South Korea at 44.9 million KRW, which works out at roughly $32,000 / £24,000 / €28,000 / AU$49,750 (yikes!), so this is definitely out of most people’s budgets. Samsung believes the tech will quickly become more affordable and become a real rival to OLED. Watch this space!
- Read more: Cutting-edge TV tech with a sky-high price
Following an 18-month ban, Blood Oxygen tracking is finally coming back to the Apple Watch Series 9, Series 10, and Ultra 2 – it’s a little different this time, however.
It was disabled over a ruling in January 2024 due to a patent dispute between Massimo and Apple, but Apple has found a way to obey the ruling and give users access to the feature.
That is, the reading no longer shows on your watch; it’ll appear only on your iPhone – with Apple explaining the update was also enabled by a recent U.S. Customs ruling.
2. We flew the world’s first 360-degree drone(Image credit: Future)After its debut a short while ago, we finally had the chance to demo the new Insta360 Antigravity drone – and it’s the most exciting drone we’ve seen in years.
It's no exaggeration to say that flying the A1 is child's play. Overcome the disorientation caused by wearing the goggles as the drone takes to the sky, and you'll quickly get to grips with the Motion controller, and how to use it to control the drone's altitude and flight path.
Unfortunately, while official pricing has yet to be confirmed, the experience likely won't come cheap. You’ve got the drone itself (which packs 360 camera tech) and a headset. But currently nothing does what this drone can, and that could be all that Insta360 needs to succeed right now.
- Read more: I had a blast flying Insta360's Antigravity
OpenAI CEO Sam Altman has addressed criticism over the restrictive message cap for ChatGPT Plus users on GPT-5. When the new model launched, Plus subscribers could only send 200 reasoning messages a week, which many saw as a downgrade from the previous model, GPT-4o. Altman confirmed the limit would increase to 3,000 prompts a week for paid subscribers.
Altman also confirmed that OpenAI is working on clearer UI labels to show exactly which model is responding in ChatGPT, something many users have been asking for since the update.
The announcement comes after a wave of backlash over GPT-5, with complaints about the smaller upgrade than expected and the removal of GPT-4o, which has since been restored. Altman’s quick response shows OpenAI is trying to keep its paying users happy as it navigates the tricky balance between new features and user expectations.