Three of the eight are dedicated directly to Lionel Messi — hero of the last World Cup that saw Argentina wipe France off the board. Not all of the sets have turned up locally yet, but we shouldn’t be waiting too long. South Africa’s Lego distributor doesn’t typically stray too far from the dollar value, so that’s a win.

Messi’s feeling bricked up

Three of those attributed to the king of football are the Lego Editions Lionel Messi – Soccer Highlights, Soccer Legend, and Celebration, available for R600, R1,600, and $200 (∼R3,300), respectively. The first of those is a simple collection of football-related things with a minifig, set atop a shrine dedicated to the man himself.

If you’re looking to actually build Messi, you’ll need to open that wallet a bit further. The 958-piece Soccer Legend set sees buyers build a Bionicle-looking Messi, with his name and number right behind him. The R3,000+ Celebration set looks like a picture in brick form, meant to hang on your wall just like a poster would. It’s got fewer pieces than you might expect — 1,427 — seeing as it isn’t a usual 3D set.

And then there’s Cristiano Ronaldo… the second-best to ever do it. He’s only got two sets to choose from, though, offering the same Highlights and Soccer Legend sets as Messi. We admit these look a smidge cooler than the Argentinian counterpart, simply because Ronaldo strikes a more photogenic pose. Even Kylian Mbappe and Vinicius Jr. are being celebrated, though only via the cheaper ‘Highlights’ sets made for, uh, superfans.

Of course, those Soccer legend sets aren’t locked into the pose on display. If you aren’t impressed with the default look, they can be readjusted to better suit whatever vibe your shelf is going for. Again, Ronaldo’s set comes out on top, with the footballer being thrust into an epic-looking bicycle kick. You be the judge:

And now for the rest

While not technically new, there are three more sets paying tribute to the FIFA World Cup. One of those is a 1,498-piece replica of the football in play, setting your wallet back by R2,400. Next up, there’s an even larger model of the World Cup trophy, consisting of 2,842 pieces at a pretty hefty R3,600. Lastly, a World Cup 2026 emblem is up for grabs at R470, its pieces lying rather flat. Just like its design.

All in all, not a good time to be the bank account of a football nerd. A couple of the sets, like the trophy or the football, are already available in South Africa. The rest that we could find locally are all up for pre-order now and should be ready on 1 May. Unless you’re the minimum age for a Lego set, there should be more than enough time to get ’em assembled. That’ll kick off on 11 June 2026.

The post Lego’s new FIFA World Cup sets could get a little Messi appeared first on Stuff South Africa.

There’s a chance this one will also pass us by. The two phones have turned up in Malaysia so far, with no official communication on where they’ll appear next. Based on the phones’ available specs, it’ll be a pity if they don’t make it to SA.

Knight to X5d

That’ll be for affordability reasons, if nothing else. The base model phone retails for around R1,550, with the Plus upping the cost to just R1,900. Even accounting for markups, neither phone would break the R2,500 barrier. When you look at the specs, it’s easy to see why. Still, they’re Honor phones, and that suggests a little more quality than you’ll otherwise get at this price point.

Both the X5d and the Plus look to use the same chassis. The image above shows both models, and the rear panels (and the dimensions and 186g weight) are identical. Both use the same 6.74in 720 x 1,600 LCD screen, MediaTek Helio G81 chipset, and 4GB of system memory. 15W wired charging and a 5,000mAh(ish) battery are also pretty common.

Differences come in from available storage — the X5d Plus has 128GB to the base model’s 64GB. The Plus also seems to have 4GB of virtual RAM. It’s not clear whether both phones include that feature. The camera setups are a little different, too. Both phones have the same 0.08MP secondary, but the Plus opts for a 50MP main sensor over the X5d’s 13MP effort.

Honor’s X9d has launched in South Africa previously, but its lower-spec versions are a little harder to find. At the projected prices for this lot, however, having these on the market will make many local customers happy.

Source

The post Honor launches X5d, X5d Plus smartphones but you can consider them a re-release appeared first on Stuff South Africa.

With usernames now tied to WhatsApp accounts, users can do away with dishing out their phone number to just anyone. WABetaInfo was the first to report on the addition, which was spotted yesterday. It reckons the rollout has taken so long — upwards of three years — because WhatsApp wants to ensure that the rollout is stable from the word go. That means adjusting the code bit by bit to make sure every feature is compatible with usernames.

Patiently waiting

Even being part of WhatsApp’s exclusive beta system isn’t enough to guarantee access to usernames, it seems. Stuff’s own beta-enabled Android device wasn’t privy to the feature as of this morning. To find out if you’re one of the lucky few, visit Settings and locate the dedicated Username option. Tap it to get started.

There are a couple of rules users need to follow before selecting a username. First, usernames cannot start with “www.” nor end with any sort of domain name (i.e. ‘.com’ or ‘.net’). Seems fair. Next, a chosen username (3-35 characters) must also contain at least one letter, to avoid bots filling out the list with only symbols and numbers. Chosen characters can only include lowercase letters, numbers, periods, and underscores.

Read More: WhatsApp’s real-time noise cancellation aims to make calls more bearable

More important is the availability of a user’s chosen username. Usernames currently in use across Facebook or Instagram cannot be chosen here, unless the user can prove they own those accounts as well. Completing this verification will require the user to link their WhatsApp account to Meta’s larger Accounts Center.

As an added measure of security, the feature is launching alongside ‘ username keys’. These are entirely optional, and make it impossible for any rando to search up your username and begin chatting at any moment. Whilst setting up a username, users can set a four-digit code that others will need when adding contacts. The user must share this key with anyone they hope to begin chatting with.

Source

The post WhatsApp usernames finally surface, ditching phone numbers for good appeared first on Stuff South Africa.

It’s a lot more conventional-looking when compared with the company’s Zone headphones or its robotic ambitions. It’s also got more in common with its portable air purifier range than Dyson’s actual fans. So what if it looks like it fires UV light at your face?

HushJet takeoff

The Cool Mini might just live up to its title, though. The pricey little handheld features five wind speeds, with the quickest promising airflow of up to 25m/s. Converted to more understandable units, it pumps out air at 90km/h. That’s thanks to the 65,000rpm brushless motor the British company has stuffed inside the unit.

Dyson’s newest product uses a rechargeable battery that sticks to USB-C for charging. An included dock means you can keep it in operation at your desk while the six-hour battery is refilled. That’s less important for South Africans right now, what with the season change, but in the summer months? It should make life very comfortable.

Details of Dyson’s HushJet Mini Cool can be found almost everywhere online, except for the company’s website. The company does have a local presence, but its airflow range doesn’t include any of the HushJet range. It could be a while before we see the Mini Cool appear in South African hands.

The post Dyson’s HushJet Mini Cool wants R1,700 to blow air on your face appeared first on Stuff South Africa.

Science and technology editor Eric Smalley interviewed Jeffrey Skolnick, eminent scholar in computational systems biology at Georgia Institute of Technology, and Benjamin P. Brown, assistant professor of pharmacology at Vanderbilt University.

Skolnick has developed AI-based approaches to predict protein structure and function that may help with drug discovery and finding off-label uses of existing drugs. Brown’s lab works on creating new computer models that make drug discovery faster and more reliable. Below is a condensed and edited version of the interview.

Let’s start with the big picture. How is AI changing biomedical research and drug discovery, and what is the potential we are talking about?

Skolnick: The upside, potentially, is very large. One of the frustrating things about drug discovery is that, in spite of the fact that the people doing it are extraordinarily intelligent and have done an extraordinarily good job, the success rate is very low. About 1 in 5 drugs will have negative health effects that outweigh its benefits. Of the ones that pass, roughly half don’t work.

In drug development, there are several key issues: Can you predict which target is driving a particular disease? Once this target is identified, how can you guarantee the drug is going to work and isn’t simultaneously going to kill you?

These are outstanding problems in drug discovery in which AI can play an important, though not 100% guaranteed, role. Unlike us, AI can look at basically all available knowledge. On a good day it makes strong and true connections called “insights,” and on a bad day it does what is called “hallucinating” and sees things that are weak and probably false.

At the end of the day, many diseases do not have a cure. Most diseases are maintained, such as high cholesterol or autoimmune conditions. A treatment for cancer might buy you five years, and now you’re in Stage 4 and you’ve exhausted all the standard care drugs. AI can play a role in suggesting alternatives where there are none.

Let’s give some basic definitions here. When we use the word drug, we’re talking about a wide range of therapies. Can you explain the range – we’ve got small molecule drugs, biologics, gene therapies, and cell therapies.

Brown: We have fairly large molecules in our bodies called proteins. They are like machines that carry out specific functions and interact with one another. Oftentimes, when we’re trying to treat a disease, we’re trying to alter functions of specific proteins. Many drugs, like aspirin and Tylenol, are small molecules that can fit into a protein and change its function. Fundamentally, drugs don’t have to just interact with proteins, but this is a major way in which our current repertoire of medications works.

There are also proteins that act like drugs, such as antibodies. When you receive a vaccine for a virus, your body is basically given instructions on how to develop antibodies. These antibodies will target some part of that virus. Your body is creating these big molecules, much bigger than aspirin, to go and interact with foreign proteins in a different way. Gene therapy is a larger step beyond that.

So these modalities – molecule, protein, antibody or gene – are very different types of molecules. They have different scales and rules, so the way you approach designing and discovering them varies widely.

Can you briefly explain artificial neural networks, and what the “deep” in deep learning means?

Skolnick: AlphaFold, developed by DeepMind, involved understanding how neural networks worked. They built a network with a lot of inputs, which are stimuli, and outputs with different weights, similar to how your brain actually works. These simple connections, or neurons, have reinforcement learning.

They also created sophisticated neural networks, such as transformers, which do specific things like a special-purpose tool that can learn, and they added a mechanism called “attention,” which amplifies critical details. Super neural networks with transformers is what we call deep learning. These now have literally billions, if not trillions, of parameters.

Essentially, these machines can learn higher-order correlations between events, meaning the patterns of conditional interactions that depend on the properties of multiple things simultaneously. In these higher-order correlations, AI has the potential to see previously unknown things that are embedded in petabytes (a unit of data equivalent to half of the contents of all U.S. academic research libraries of biological data.

AlphaFold, which predicts three-dimensional, bioactive forms of a protein, has millions of sequences and a couple of hundred thousand structures. It can tell you, based on a particular pattern, what small molecule to design that sticks to a protein to induce some kind of structural shift.

How is this technology being used in biomedical research to understand molecular dynamics or, essentially, the biological processes involved in health and disease?

Brown: In 2013, there was a Nobel Prize for molecular dynamics simulations, computational tools that help you understand the motions of molecules as they move according to physics. There’s a huge body of scientific research built around those ideas.

AI and deep learning are large right now, but it’s worth mentioning that for the last decade and a half, people have been using much smaller machine learning algorithms to help design drugs. A lot of the ideas, such as [using machine learning for virtual screening], are not new and have been in practice for a while.

With AlphaFold’s technologies to help people design proteins and predict their structure, we’ve changed how we think about a lot of these problems. We have this new repertoire of approaches to build ideas around and to start thinking about drug discovery.

From 20 years ago to now, what has today’s AI technology done in terms of scale of change in this process?

Skolnick: A lot of diseases, like cancers, are caused by a collection of malfunctioning proteins. AI now allows us to start to think conceptually about how these diseases are organised and related to each other.

Diseases tend to co-occur. For example, if you have hyperthyroidism, you’re very likely to develop Alzheimer’s. Kind of weird, right? We can look at pieces, but AI can look at all the information, integrate the collective behavior and then identify common drivers. This allows you to construct disease interrelationships which offer the possibility of broad-spectrum treatments that could treat whole collections of diseases rather than narrow-spectrum treatments.

Relatedly, AI can also help us understand disease trajectories. Diseases that tend to co-occur often present themselves consecutively. You have disease 1, which gives you disease 2, which then gives you disease 3. This suggests that if you go back to the root with disease 1, you may be able to stop a whole bunch of stuff. You can’t analyse millions of trajectories and millions of data points without a tool, so you couldn’t do this before.

This holds a lot of promise, but one must also be careful not to overpromise. It will help, it will accelerate, but it is not a substitute yet for real experiments, real clinical validation and trials.

• Jeffrey Skolnick is a Regents’ Professor; Mary and Maisie Gibson Chair & GRA Eminent Scholar in Computational Systems Biology, Georgia Institute of Technology
• Benjamin P. Brown is an Assistant Professor, Department of Pharmacology, Vanderbilt University
• This article first appeared in The Conversation

The post AI is reengineering drug discovery by speeding up testing and scanning petabytes of data for connections between diseases appeared first on Stuff South Africa.

You only care about your self(ies)

But not an awfully extravagant one. Sure, those carrying around a flagship Samsung or Apple device likely need not worry about waning selfie quality, but hardcore social media girlies walking around with a 2020 phone in their pocket? The Snap’s S$115 price might make it worth picking up. Better than buying a new phone, anyway.

The screen, which hooks up via your device’s USB-C port, isn’t just a portal into your phone’s rear camera. The 3.5in display is a touchscreen, allowing you to zoom in and frame up your shot before snapping it. There’s another version, co-developed with Amiro, going for S$130 that’ll add a light with 5 brightness levels into the mix.

Read More: Insta360’s new 5.4-gram Quick Reader 512GB SSD means fewer memory card swaps

Insta360 had to get the proportions just right if it hopes to truly enter this selfie-screen market. It’s a smidge too thick for our tastes, measuring 6.8mm thin — just a little bit thinner than an iPhone 17. It’s clearly been designed with Apple in mind, connecting to the phone magnetically if your iPhone is MagSafe compatible. If you’re rocking an Android or something other device, you’ll need to use the supplied magnetic ring.

It’s worth mentioning that the Snap Selfie Screen supports Android phones that offer DisplayPort Alt Mode — something that’s worth checking out if you’re seriously invested. Whether it’s worth the roughly R1,470 it’ll take to stick this in your pocket is entirely on you and your selfie habits. It’s still unclear if it’ll see a South African launch, but with the brand’s history on 180by2, it isn’t out of the realm of possibility.

The post Insta360 releases the Snap, because what your phone really needs is another screen appeared first on Stuff South Africa.

“Starting May 20, 2026 – 14 to 18 years after their initial launches – we are discontinuing support for Kindle devices released in 2012 or earlier,” reads the email sent out to some Kindle owners.

Amazon isn’t feeling Kindle

Let’s get something straight. After 20 May, your Kindle (whichever pre-2012 model it may be) will continue to function, though not what we’d call ‘normally’. But it will function. Owners will be unable to “purchase, borrow, or download additional books on them.” You know, the reason we buy e-readers from Amazon in the first place.

Fear not, for setting sail on the high seas may still be the way to go here. Sideloading appears to have been left untouched, allowing users to hook up their e-reader (via micro-USB, the horror) and add books manually. How you attained those books shouldn’t be of any concern to Amazon since it doesn’t want your money.

So you can still read on your Kindle. Fine. But the email didn’t stop there. It also brought forth a warning not to factory reset your Kindle beyond that date. Doing so will brick the device for good. Amazon’s exact words: “If you deregister or factory reset these devices, you will not be able to re-register or use these devices in any way.”

That’s really great, Amazon. Brick the device that folks have spent their hard-earned money on. Even all these years later, the family’s first-gen Kindle Keyboard (it’s essentially an heirloom at this stage) is still going strong. Amazon’s latest update will change that, not only for our model, but for six others just like it as well.

Kindle models released before 2013:

• Kindle (1st Generation) – 2007

• Kindle (2nd Generation) – 2009

• Kindle DX (2nd Generation) – 2009

• Kindle Keyboard (3rd Generation) – 2010

• Kindle Touch (4th Generation) – 2011

• Kindle 4th Generation – 2011

• Kindle 5th Generation – 2012

Fortunately, picking up a Kindle isn’t your only option when it comes to choosing a new e-reader. Kobo, in particular, is making strides to become Kindle’s largest software competitor, even if the hardware occasionally falls short. Hell, even Kindle’s by-the-book 2024 model will only set you back a couple of grand in 2026.

The post Amazon will pull support for some older Kindles in May, putting your e-reader at risk  appeared first on Stuff South Africa.

On an I’m Back Roll roll

The concept is sound, since the company making this version has already done something similar. But the solution was previously hosted outside of analogue cameras. This version looks like a roll of film, which stretches across the camera’s rear and plops a new digital sensor right where it’s supposed to be.

The sensor, in this case, is a Sony IMX571 APS-C affair. The add-in, according to the project’s creators, will support multiple 35mm camera systems. Some internal modification might be required, mostly the removal of the camera’s pressure plate, but everyone from Canon to Pentax to Leica should have functional models. The Kickstarter page showcases the Nikon F3 and the Leica M as examples.

It’s almost identical in design to a roll of film in size, but the I’m Back Roll integrates most of a camera’s internals. At least 64GB of storage, the battery, sensor, and other parts needed to use vintage bodies and lens assemblies pop into a camera’s rear panel. It’s all a bit thicker than using film, of course. The setup needs 4mm of space, which is why that pressure plate has to go.

Read More: Rewindpix is a modern compact camera with vintage ambitions

There’s really only one tricky bit. Since the I’m Back Roll can’t communicate with old cameras, there’s an external remote. Before a shot is taken, users must activate the remote, which starts up the sensor. After that, there are roughly two seconds before the sensor must be turned on again. Users have that long to trigger the camera’s physical shutter. So… not for high-speed, critical shooting then. The Bluetooth remote is similar to other remote camera shutters.

It would still be a cool thing to experience. The I’m Back Roll was initially looking for just over $44,000 to get production rolling. To date, with 37 days of funding to go, it’s made ten times that amount. But previous Kickstarter success doesn’t always mean that every project will go off flawlessly.

Still, if you’d like to chance it, upgrading your old hardware will cost at least R7,500 ($455). It’s a bit much for an old Kodak, but if you’ve got a vintage Leica lying around, it might be time to get it out of storage.

The post The I’m Back Roll hopes to convert old film cameras into fully-digital shooters appeared first on Stuff South Africa.

What Apple will name the line-up when the ‘Neo’ name inevitably goes stale, we don’t know. But for now, it’ll probably skate by as the MacBook Neo once again. For now, the Fruit Company is deciding whether it should ramp up production for the current model, to keep up with rising demands, or simply let the parts run out.

Neo MacBook Neo

That makes sense. The Neo, unlike its MacBook brethren, uses an iPhone 16 Pro chip — the A18 Pro — rather than a typical M-something. That stock is reportedly leftover from its iPhone run, according to Ben Thompson at Stratechery. That’s essentially free money, seeing as it chucked an old chip into an older chassis and called it a day. Not that we mind. Our review model, packing a mere 8GB of RAM, handled well enough.

Before the Neo became a surprise hit, Apple had hoped to recycle the game plan next year, using leftover A19 Pro chips to power the sequel. Now, however, the company is deciding whether to stick with the initial idea of building 5-6 million A19 Pro chips or upping that figure to accommodate the Neo 2 from the get-go.

Read More: Apple’s iPhone Fold reportedly targets a release alongside the iPhone 18

The news does at least tell us a sequel is on the way, and will almost certainly sport the A19 Pro that’s currently housed in the iPhone 17 Pro. Culpan mentions that this would still be a “binned” 5-core version of the chip, which typically has six cores. He also mentions that Apple could keep the 6-core GPU and sell it at a premium.

More importantly, though, the chip has 12GB of RAM inside, meaning a potential Neo 2 would reap the benefits. The current MacBook Neo ships with 8GB of physical memory — a bit of a letdown for power-budget buyers. No potential launch dates are mentioned, but we’re betting Apple will want to capitalise on March in the future.

Source

The post MacBook Neo 2 is already in the works, with the usual upgrades expected appeared first on Stuff South Africa.

The Xiaomi Mijia Capsule Coffee Machine isn’t confirmed for South Africa yet, but its appearance on the company’s global website makes it a distinct possibility. It targets Nescafé and Keurig as competitors, rather than the snooty folks at De’Longhi, but that’s fine. Not everything has to be properly bespoke. Besides, Xiaomi already made one of those.

Xiaomi how to make coffee

The company’s Mijia machine uses coffee capsules (or pods, depending on the brand you purchase). They look to be the generic sort, so coffee fanatics could stick their preferred brand into the Mijia’s magnetic holder. Assuming the Chinese company isn’t also getting into the pod-making business, anyway. The machine itself doesn’t do much that others of its type don’t already do. The main question is: How fast can it do it?

Xiaomi reckons that the Mijia can heat up in 25 seconds, spitting out one of two coffee sizes (40ml or 110ml) shortly afterwards. That’s thanks to a 1,360W heating system, with pressure provided by a 19-bar ULKA pump. The attached tank holds 650ml of liquid, or around six full cups, before you need to top up.

The machine controls for temperature (it bloody better) via an algorithm and an NTC sensor. Once you’ve chosen your size (selected via a long-press on the Mijia’s single button), it’ll remember it until further notice. After that, the process becomes: load capsule, poke button, sip magic bean juice.

There’s no sign of the Mijia machine on sale anywhere just yet, with launch details and pricing still a mystery. But if you’ve already stocked up on the company’s smarter home tech, chucking a coffee machine into the mix probably won’t hurt.

The post Xiaomi is coming for your morning cup of coffee appeared first on Stuff South Africa.