The Hisense A10 is only a rumour for now, but a teaser (and some leaks) point towards it having an E-Ink display in addition to its other smartphone functions. The biggest change here is an improved display, since the tech has moved on since 2022. But there’s also a general hardware upgrade. Which makes sense. Four years is a lifetime in tech.

A10 goes brrrt

According to various reports, Hisense’s new phone/reader will sport a 7in version of the Carta 1300 E-Ink display. It’s very similar to the screen found on the Boox P6 Pro. The internal specs may have seen a downgrade (kinda), thanks to various AI-manufactured pressures. The Hisense A9 had an eight-core Snapdragon chipset in 2022. The A10 will supposedly include a quad-core chip that hasn’t been detailed yet.

It may also be that Hisense doesn’t actually need that much power for this one. 6GB of RAM and 128GB of storage match up with the most recent release, while a 4,500mAh battery will keep it all powered on. Hisense is supposed to be including a camera on the phone capable of recording video in HDR 4K. It’s doubtful you’ll be able to watch it on the phone, but it’ll be there. Somehow.

Bluetooth 5.1, 5G support, and an Android operating system are also part of the A10’s package. Unfortunately, there’s no sign of when (or where) it’ll become official. As with other phones and readers like this, it’ll probably be scarce in South Africa. That’s a pity. We’d love to try it out — just to see what the fuss is about. Pricing is similarly murky. If you’re only in it for the screen, a dedicated device would seem to make more sense.

Source

The post The Hisense A10 could be the Chinese brand’s newest E-Ink smartphone appeared first on Stuff South Africa.

Now officially available in South Africa, the HUAWEI nova 15 Max lets users truly “Max Your Fun“. Combining intelligent navigation, exceptional battery life, immersive entertainment and dependable durability, it is a smartphone built to keep up with modern life. Starting at just R7 999, customers will also receive a HUAWEI Added Value Pack worth up to R7 596, including Unlimited Screen Damage Protection and HUAWEI FreeBuds SE valued at R1 499 at selected retailers.

Available in Black, Cyan and Gold, the HUAWEI nova 15 Max balances performance with style, complementing its slim design and distinctive Star Ring camera layout. At a time when smartphone prices continue to rise across the industry, the HUAWEI nova 15 Max stands out by delivering flagship-inspired features, generous memory and storage, and one of the largest batteries in its category at a highly competitive price point.

Battery life built for long days

A smartphone is only useful if it can last the distance. At the heart of the HUAWEI nova 15 Max is an impressive 8500mAh HUAWEI Super Battery, one of the largest batteries available in its category. Designed to support busy lifestyles, it helps users stay connected, entertained and productive throughout the day without constantly searching for a charger.

Real-world testing shows the device can support up to 23 hours of continuous video playback. The phone also supports reverse charging, allowing users to power compatible Huawei devices directly from their smartphone when needed. This effectively turns the HUAWEI nova 15 Max into a portable power bank, making it easy to keep compatible Huawei devices powered when you’re away from a charger.

Whether you’re commuting, travelling for business or heading away for the weekend, the HUAWEI nova 15 Max is built to keep going.

Ready for the realities of everyday life

South Africans need devices that can handle more than just office desks and coffee shop tables. The HUAWEI nova 15 Max has achieved the SGS Premium Performance Mark 5-Star Drop Resistance certification, helping reduce the risk of damage from everyday knocks and accidental drops. Precision sealing technology helps protect against dust and debris, while the IP65 rating provides resistance against dust and water exposure encountered during daily use.

Behind the scenes, on-device AI models intelligently adjust touch parameters when the screen or fingers are wet, helping maintain smooth and accurate operation in everyday situations.

From unexpected rain showers to busy days spent on the move, the device is engineered to provide added peace of mind.

Capture every moment, day or night

Life’s adventures deserve great photos, and the HUAWEI nova 15 Max is equipped to capture them. The smartphone features a 50MP Ultra Vision Camera with Huawei’s advanced RYYB colour filter array, which increases light intake by 40% compared to traditional solutions. Combined with a large 1/1.56-inch sensor and F1.9 aperture, the camera is designed to produce bright, detailed images even in challenging lighting conditions.

Whether you’re photographing a sunset over the coastline, capturing memories around a braai with friends or documenting your latest adventure after dark, the HUAWEI nova 15 Max delivers vibrant colours, impressive detail and natural-looking results. The camera is also supported by AI-powered features such as AI Best Expression, helping users select and refine the best moments from a sequence of shots, ensuring natural-looking smiles and more polished group photos.

Smarter navigation when it matters most

Whether you’re navigating Johannesburg traffic, exploring Cape Town’s winding roads or travelling somewhere entirely new, reliable navigation has become an essential part of everyday life.

The HUAWEI nova 15 Max introduces AI-Powered Navigation, a feature designed to maintain stable route guidance even when GPS signals become weak or inconsistent. Powered by Huawei’s self-developed AI fusion algorithm, the system intelligently calculates driving trajectories and helps maintain navigation continuity in challenging environments such as tunnels, dense urban areas and multi-level roads.

The result is a smoother and more dependable navigation experience that helps drivers stay confidently on course when conventional positioning systems may struggle.

Display and sound designed for entertainment anywhere

Whether you’re catching up on your favourite series, watching videos during your commute or navigating outdoors, the HUAWEI nova 15 Max is designed to deliver an immersive viewing experience.

Its large 6.84-inch OLED Eye Comfort Display combines vivid colours with smooth responsiveness and intelligent viewing adaptation. With up to 4000 nits of dynamic peak brightness, the display remains clear and easy to view even under bright sunlight, making it ideal for navigation, streaming and everyday use.

Entertainment on the HUAWEI nova 15 Max extends beyond the display. Symmetrical stereo dual speakers deliver powerful, immersive sound, while the expansive soundstage makes everything from movies and music to podcasts and video calls clearer and more engaging. Whether you’re relaxing at home or sharing content with friends, the audio experience is designed to be heard and enjoyed.

Built to keep up with your lifestyle

With its massive 8500mAh battery, SGS-certified durability, capable camera system and immersive display, the HUAWEI nova 15 Max is designed to deliver premium experiences without the premium price tag.

Available from just R7 999 through the HUAWEI Online Store, HUAWEI Experience Stores (HES) and selected retailers, customers will also receive a HUAWEI Added Value Pack worth up to R7 596, including Unlimited Screen Damage Protection and HUAWEI FreeBuds SE valued at R1 499 at selected retailers.

The post The HUAWEI nova 15 Max combines massive battery power with everyday durability appeared first on Stuff South Africa.

It’s a much bigger deal if you recall playing games on the original translucent green Xbox, back before it started picking up suffixes like PhDs. The limited-edition hardware will be presented in what Microsoft calls OG Green. Back in 2001, that was just called… green.

X25 to life

South Africa will hopefully have a decent shot at the revamped console this time around, but we’ll have to wait till closer to November to find out. The console doesn’t look like it’ll be substantially different from the standard Series X. It’s no massive loss if we are skipped. It’s still a bummer, though.

The X25 will launch with the standard Series X hardware, plus a terabyte of storage. An Xbox Wireless Controller X25 Special Edition is included in the bundle. Folks who prefer to hold their nod to gaming history can also grab the controller separately. The console’s startup will nod towards the original hardware, with a green X illuminating when it’s powered on.

The controller is also themed around the original Xbox input device, though it won’t be as large and chunky. It will hark back to the original colour scheme from the ‘Duke’, of course. Otherwise, it’ll be Microsoft’s Series X controller all the way. Sometimes it’s better to just look historical. You could also… you know, buy a knockoff of the OG Xbox input.

Microsoft also appears to have easter eggs planned. Or, as Xbox puts it, “You’ll also discover a few hidden surprises throughout, as a thank you to the community.” Pricing and pre-order details should be along shortly, though the company hasn’t said exactly when this will be.

The post Microsoft to launch X25 anniversary edition Series X and controller appeared first on Stuff South Africa.

NASA’s Artemis program is an extensive collaboration, with folks like SpaceX and Blue Origin standing alongside… well, Prada. Technically, the luxury brand isn’t working on its own, having partnered with Axiom Space to help create the Axiom Extravehicular Mobility Unit spacesuit. Both have now unveiled the underwear that astronauts will wear while out in the vast coldness of space. It’s called the Liquid Cooling and Ventilation Garment, because humans need life support while on the lunar surface. The ‘Prada’ part of the partnership is mostly so regular people will pay attention to it.

Okay, fine, the brand also has some scientific input. The underwear “draws on Prada’s expertise in engineered knitting and innovative design concepts, resulting in a next-generation garment developed through advanced 3D modelling techniques that maintain cooling and ventilation while enhancing comfort during up to eight-hour spacewalks.” In keeping with Prada’s tendencies here on Earth, they’re also bound to be ferociously expensive. No, you can’t buy a set for yourself.

A Nintendo Switch 2 with a swappable battery to launch in the EU (send it here too, please)

Nintendo’s Switch 2 will soon have a new version of the console on the market. It’s destined for the European Union, mostly because the EU loves to throw regulations at companies. The new console will include a swappable battery to comply with one of these regulations. Specifically, “from February 18th, 2027, batteries integrated into certain appliances and sold in the EU must be easily replaceable by end-users at any time during the lifetime of the product.”

Nintendo doesn’t say how these consoles will differ from the original hardware, but a swappable battery is needed to comply. The gaming giant has a little under a year to come up with a plan, and it seems to have one. There are detailed plans to designate the console specifically for the EU market. How easy it’ll be to perform the swap remains to be seen. If it’s as simple as swapping an old Nokia dumbphone’s battery, we’d quite like to see them launch here as well. Extending a gaming session by dropping a new battery in would be awesome. Nintendo would probably also make a bundle selling battery packs.

Netflix has a football game launching later this week

We’ve been expecting it for some time, but Netflix has finally announced the launch of its FIFA football game for subscribers. FIFA World Cup Launch Edition will launch on the same day the World Cup decimates global productivity on 11 June. It’ll be playable on smartphones, with basic touch controls letting players while away the hours before the next kickoff. It’ll be free for Netflix subscribers, of course. It’ll also be exclusive to the platform, so you may have to jump through some illegal hoops to play it another way.

According to Netflix, “you’ll be able to play as any of the 48 teams in the tournament, travel to the 16 real-world stadiums, and take control of any of the 1,248 players in the Cup. So as the drama of the sport plays out, you can share every goal, save and celebration right on the couch with your friends and family.” It’s an ambitious release, given how finicky football simulator players tend to be. But this one’s intended for the everyday player, not the guys and girls who live and die by their Fantasy Premier League rankings.

NASA’s X-59 supersonic plane finally goes… well, supersonic

Supersonic flight is on the way back, with NASA’s X-59 plane doing the legwork. We’ve been tracking its progress for some time, and there’s finally an update. The X-59 ‘quiet’ supersonic plane has finally broken the sound barrier. The test flight took place over the weekend, lasting 81 minutes. In that time, the plane reached a maximum speed of 1,147km/h. That’ll get you just about anywhere in South Africa in 60 minutes, no matter where your starting point is.

There are more tests to come, with the X-59 expected to eventually reach Mach 1.4 (around 1,500km/h) before the testing really begins. Once that milestone is reached, it’ll ” eventually fly over several U.S. communities, enabling NASA to gather data about how people may perceive its quiet thump.” The ‘quiet thump’ is the big deal here. Should the design allow for high-speed travel without the massive shockwave, we could reach our destinations faster than ever… assuming we can afford tickets on whatever comes next.

Source

The post Light Start: Prada in space, Switch 2’s EU case, Netflix’s soccer ace, and the X-59 sets pace appeared first on Stuff South Africa.

The digital divide refers to the gap between individuals and households who have access to the internet, and those who do not. The digital divide can restrict education attainment, economic opportunity, the ability to adapt to rapidly changing employment environments, healthcare access, social inclusion, and overall quality of life.

While digital technology will bring about many environmental, social and economic gains, the pathway to South Africa’s digital future is not without challenges. The country needs to make the benefits inclusive and equitable.

As a researcher at the Gauteng City-Region Observatory (GCRO), I analyse urban development with the aim of providing evidence for policies. I recently explored Gauteng’s digital divide to understand how it might shape inequalities in the future. I asked whether residents of Gauteng – South Africa’s most populous province and a regional economic power house – have equal access to opportunities in our digital futures.

My findings show that there’s a marked digital divide in Gauteng. It is spatially concentrated and characterised by social inequality in terms of race and household income.

These findings matter because digitalisation and digital transformation are increasingly affecting the shape of the economy and society. Not having access reduces opportunities. Maximising the benefits of digital futures depends on reliable and affordable connections to the internet for everyone.

The measuring

South Africa’s digital divide is a function of extreme social inequality. Many residents still lack the financial means to access the internet, or live in areas with poor internet connectivity. Large parts of our society are unable to participate in digitisation and digital transformation and benefit from it.

I used Quality of Life Survey 7 (2023/24) data to explore how access to home internet varies in Gauteng. This survey series is one of the largest and longest-running social surveys in South Africa. It collects information from adult respondents in Gauteng to measure quality of life and understand the successes and challenges of the province.

The 13,795 survey respondents were asked whether their household had access to a selection of things that were in good working order. The list included things like a microwave oven or air fryer; a smartphone; a television; a personal computer, laptop or tablet; a car; fibre-based home internet; or other home internet connection.

In this analysis, I focused on the last two assets. Does a household have access to a fibre-based home internet connection or access to another home internet connection (Wi-Fi, home-based 5G, LTE connection or any other internet connection that is used in the household)?

Connectivity

Among all the survey respondents, 46% lived in households with home internet connections. The remaining 54% of respondents lived in households without any home internet connection.

The research also showed that 85% of respondents lived in households with a working smartphone. This means that most households had other means to access the internet from home. This can be through mobile networks or access to municipal Wi-Fi networks. However, mobile data is expensive and smartphones are somewhat limited when it comes to remote work or online learning.

A home internet connection is also very important for survey respondents with access to a resource like a laptop. About 39% of survey respondents live in a household with a personal computer, laptop or tablet, but 25% of these respondents do not have access to home internet. This reduces the potential value of having a laptop because the laptop cannot be connected to home internet in order to do remote work, pursue online qualifications, or just get useful information.

Spatial divide

Drilling further into the survey results shows that access to home internet is uneven across wards in Gauteng. In suburban areas like Centurion, Midrand and Randburg, more than 80% of households have home internet. Suburbs in South Africa are low density residential areas where households typically have above average incomes.

In low-income communities like Hammanskraal, Soweto and Katlehong, there are many wards where only 40% of households have home internet. Similarly, in parts of Mamelodi, Sebokeng and Daveyton, less than 20% of households in a ward have access to home internet.

The spatial patterns are substantially influenced by infrastructure and service coverage (5G and LTE coverage), the infrastructure rollout plans of fibre installers, and household income.

For example, fibre infrastructure rollout is driven by the private sector and requires space on the road verge. This means that rollout is focused on areas where there is guaranteed demand and where it is practically feasible to install fibre lines on road verges.

Why it’s important

The digital divide is deeply associated with socio-economic inequality.

Only 39% of black African respondents lived in households with home internet, compared to 87% of Indian/Asian respondents and 86% of white respondents.

Access to fibre-based home internet is further skewed. Only 18% of black African respondents lived in households with fibre-based home internet, compared to 74% of Indian/Asian respondents and 70% of white respondents.

Similar differences were visible between households with lower or higher income. Only 20% of households in the lowest monthly income bracket had home internet, compared to more than 80% of households in the top monthly income brackets. Once again, access to fibre home internet was even further skewed.

Only 5% of households in the lowest monthly income bracket benefit from having fibre, compared to more than 60% of households in the top monthly income brackets.

What should be done?

The digital divide needs to be narrowed if Gauteng is to follow a path of inclusive growth. This analysis can enable policymakers and community leaders to make strategic decisions about inclusive digital futures.

The results also show that closing the digital divide will require partnerships between the private and public sectors.

• Christian Hamann is a Researcher, Gauteng City-Region Observatory
• This article first appeared in The Conversation

The post Internet access is unequal in South Africa’s economic powerhouse: survey shows race and income mark the digital divide appeared first on Stuff South Africa.

When I was growing up, teachers would assign research papers that required going to the library, or later, searching for relevant material on the internet. If the paper was going to turn out well, we students needed to patiently comb through piles of material, weaving what we found into a coherent argument that was well-supported with evidence.

Unbeknownst to us at the time, our teachers were giving us a chance to develop our patience.

That chance is rapidly disappearing with increased use of artificial intelligence tools. Now you can have an AI do everything from school assignments to legal writing, sermon preparation, vacation planning, work emails and academic research. Researchers are already documenting how using AI tools in these contexts likely erodes critical thinking skills.

But what hasn’t been appreciated is AI’s effect on patience. As a philosopher who has written extensively about virtue, including the virtue of patience, I am especially concerned about what people can do to resist this trend.

What is patience, and why is it important?

Patience involves responding calmly when it is taking longer than you want to accomplish your goals.

When I am stuck in a traffic jam, or the checkout line is barely moving, I might wish that I was meeting my goals faster, but my calm demeanor is a sign that I am being patient. If I react to delays like these with frustration or anger, that is a sign that I am being impatient.

The same applies in the case of doing research. If it is taking me awhile to find everything I need, that can test my patience. But if I react to such a delay with calmness, I avoid frustration or anger and hence impatience.

Philosophers, theologians and educators have long considered patience an important character trait to cultivate. It is a virtue that contributes to well-being. More specifically, researchers have linked it to a variety of good outcomes, including healthier lifestyles, greater emotion regulation, more fulfilling relationships, increased caring about equity and justice, increased cooperation, greater purpose in life, lower depression and higher life satisfaction.

Why AI tools erode a capacity for patience

AI tools are helping foster a culture of immediacy, thereby diminishing the capacity for patience. Admittedly, we already started down this path with the dawn of the internet and the launch of fast and easy search engines. But now, AI instantaneously delivers fully developed answers, further reducing the delays once experienced as people searched, assessed and integrated information from various sources.

The training in patience that people used to get from thorough research and investigation is being replaced by a growing sense of impatience with thinking that takes time and effort. And this impatience doesn’t just stop with research. It extends to writing as well.

Research on AI and patience is still in its infancy. But my conclusions about these impacts rest on plausible inferences from what researchers know more generally about cognitive psychology. For instance, psychologists have long understood that people’s expectations change due to repeated use and exposure to something.

This adaptation explains why the hourlong train ride to work can start out as exhausting, but become part of your daily routine. Or you might initially be impressed by how fast your new computer is, but after a while you take it for granted and get frustrated if loading a PowerPoint presentation takes even a few moments.

Hence using AI tools is likely to recalibrate what feels normal to you. In particular, it is likely to normalize getting immediate, fully formed answers to your questions. This shift, I contend, makes people increasingly impatient with the very tasks of research and investigation that helped train us to become more patient in the past.

One concrete illustration of this change is with students. If a professor gives an assignment involving interpreting an author’s text and then developing a critique of the author’s position, students today are very tempted to offload the patient work of interpretation and critique to an AI.

Or consider sermon preparation. Pastors normally take hours a week to examine the original language for their text, consult commentaries, develop illustrations and examples, and deliberate about practical applications. Now, this process can all be done in a matter of seconds using AI, and one study found that a majority of pastors are using it for sermon preparation. There is no patience training happening here.

What can be done?

There are ways to cultivate patience in the age of AI tools, but they will not be easy. Here are three:

• Deliberately choose a slower path. Select this option because it comes with intellectual struggles, not in spite of them. Don’t rely on AI summaries or shortcuts, but try to come up with the answers on your own. This choice needs to be deliberative since the default human tendency is to take the easier route. But the long-term benefit is worth the short-term cost.
• Design your environment. Remove AI tools from your surroundings and carve out dedicated time free of distractions and notifications. Reading and writing take time, and by being willing to invest that time and not get impatient with how long it is taking, you can cultivate patience.
• Encourage and reward intellectual engagement. Institutions such as schools and churches have a structural role to play. The more such institutions can resist integrating AI tools into every aspect of their operations, and instead incentivize human intellectual engagement even at the expense of efficiency, the better as far as patience is concerned.

There is one other hopeful suggestion. Patience can be developed in lots of different areas of life that have nothing to do with research and which are less susceptible to AI incursion. Working on a craft project, detailing a car, weeding a garden, practicing your basketball shot, lifting weights – all these activities can foster patience too. The more this character muscle is strengthened, the more it will be available to use in many different areas of your life.

• Christian B. Miller is a Professor of Philosophy, Wake Forest University
• This article first appeared on The Conversation

The post Eroding a virtue: AI trains people to expect instant answers – and that’s bad news for patience appeared first on Stuff South Africa.

South Africa’s 50 public technical and vocational education and training (TVET) colleges are, in the main, struggling institutions.

In many, throughput rates – how many students qualify in the expected time – are low. Some lecturers are under-qualified and under-resourced. Relationships with employers, which are crucial for the type of training that these colleges offer, are uneven.

Colleges are hard-pressed to provide training to young people with weak schooling behind them and no clear path to employment ahead. The youth unemployment rate is almost 44%.

The response to problems in the sector has been reform: rename the colleges, restructure them, give them new governance models, new qualification types, new funding arrangements. Over 30 years of democracy, South Africa has done all of these things repeatedly. It has not worked.

And now there’s another round of changes being rolled out. There is little clearly documented explanation of what the new system is and how it will work in practice. But colleges have been instructed that most current qualification offerings will be phased out and replaced by new “occupational” qualifications.

In 2024, I wrote a paper tracing the history of the technical and vocational training sector, drawing on published literature, my research on skills development and my own involvement in South Africa’s education and training policy processes. The paper sets out why the sector is not working and what it needs to succeed.

In my view, based on the history of the sector, there is a serious risk that the latest reforms will make things worse.

Thirty years of the same mistake

South Africa’s policy vision and funding model for TVET colleges has, like that of many other countries, been to base funding on student enrolment for programmes that are linked to employer demand. It assumes colleges will respond to what employers want, and channel young people into jobs.

It has a long and largely unsuccessful track record, with problems in many countries – most extensively documented in Australia and the UK, the originators of the broad policy model.

The problem is structural. Funding institutions only through enrolments in specific programmes provides no institutional stability. It creates no incentive to invest in equipment, lecturers, or long-term relationships with employers. It treats colleges as if they were competing as private training providers.

When the programmes that attract funded enrolments change – as they do, repeatedly – colleges are left with stranded staff, obsolete equipment, and no financial buffer. And when new funding is made available, for new programmes, they don’t have lecturers who can teach them.

Private institutions tend not to offer manufacturing-related programmes – those are expensive. They focus on business-related programmes, which are cheaper.

Consider the National Technical Education Diploma (Nated) qualifications, the government-funded programmes that colleges have provided for decades. First, they were to be phased out. Then, when the National Development Plan created TVET enrolment targets, colleges were told to expand them. Colleges have built up staffing around them and enrolled students in them.

Now, the Department of Higher Education and Training has instructed colleges to phase them out. What replaces them are “occupational qualifications”.

The occupational qualifications problem

The department defines an occupation as

a set of jobs whose main tasks and duties are characterised by a high degree of similarity (skill specialisation).

The theory behind occupational qualifications is sound: link qualifications to specific occupations, make workplace experience part of the qualification, and graduates will have credentials that employers recognise and value.

The framework has thousands of occupations.

The problem – and here is where our new research (not yet published online) is indicating an uncomfortable finding – is that many of the “occupations” to which these new qualifications are linked do not really exist in workplaces and labour markets. And there is little publicly available information about them.

Some “occupations” have special skills that need special training, and others are really just jobs.

For example, in our research (not yet online) across 53 food and beverage manufacturing plants, we found that there are artisan trades like millwrighting, fitting and turning, and electrical work which fit the idea of an occupation. But machine operators don’t fit that description. Yet machine operators are among the new qualifications to be offered. The employers we visited don’t need those qualifications. They would rather hire someone they can train themselves, to use the equipment in their plant.

Training in a “knowledge module” like “personal mastery and interpersonal relationships” is not specific to the “occupation” of operating a machine.

You cannot create an occupation by developing a qualification for it. It works the other way: the occupation must exist before you create a qualification for it.

This is not an abstract concern. Colleges are now being instructed to gain accreditation to offer these qualifications, to hire staff to teach them, to find workplace placements for students doing them – all on the assumption that there is a real occupational destination at the end.

For artisans, this assumption holds: there are real occupations that translate to opportunities in the workplace. But for the majority of new occupational qualifications being developed, far more analysis is needed.

What institutions actually need

Colleges cannot become strong institutions through enrolment-driven funding alone, any more than a school can become strong by being paid per pupil with no base funding for teachers or classrooms. And calling qualifications “occupational” does not mean that they will lead to work where there is no meaningful occupation in labour markets or workplaces.

Institutions need a stable core – employed lecturers, maintained equipment, administrative capacity – that allows them to function as institutions rather than as collections of projects cobbled together from different funding streams.

Some of them may be better off offering second-chance matric (secondary school leaving certificate) programmes instead of narrowly focused programmes where there are few real opportunities for employment in the surrounding areas, and no way colleges can find work placements for their learners.

Pockets of genuine excellence exist in the current system: colleges with good employer relationships and real employment outcomes for graduates. What they have in common is principled management, experienced staff, and enough stability to build relationships over time. The system should be trying to replicate those conditions.

In my view, what needs to happen is this:

• colleges should be funded with a core institutional grant, and enabled to provide a mix of training that reflects their local economic contexts
• occupational qualifications should be rolled out only where employers need them.

Otherwise the latest reforms risk repeating the errors of the past 30 years. Colleges and young people deserve better than that.

• Stephanie Allais is a Faculty member, Centre for Researching Education and Labour, University of the Witwatersrand
• This article first appeared in The Conversation

The post Reforms to South Africa’s technical colleges keep failing students and employers: why? appeared first on Stuff South Africa.

Major event venues are increasingly using the technology. Some, like Madison Square Garden, use it for surveillance purposes, and some, like Citizens Bank Park in Philadelphia, to offer visitors optional ticketless admission.

Adoption of facial recognition technology is increasing, becoming more prevalent in daily life, from public buses to public buildings. The Transportation Security Administration has deployed the latest facial recognition technology at security checkpoints at numerous airports. The agency says the new system will be used in cities across the U.S. that are hosting World Cup 2026 soccer matches.

The growing use of facial recognition has broadened concerns about accuracy and bias. But in my research studying facial recognition technology in the Vision Lab at the University of Dayton, I’ve found that advanced deep learning models have made face recognition systems more accurate and reliable. The AI models, trained on hundreds of millions of face images, are more than 99% accurate in controlled environments – settings such as cellphones, airports and border checkpoints.

Facial recognition basics

Facial recognition involves three steps: locate a face in an image or video frame, create a faceprint that catalogues salient features – including the shape of the face and landmark points such as eyes, nose and mouth – and record the texture of the skin. Then it compares the faceprint to those in a database, which may be inside a smartphone or at a bank or hospital, to verify a person’s identity or allow access.

In the physical world, these systems are faster and simpler than requiring people to show IDs. In the online world, they are easier than entering a login name and password. Facial recognition also significantly reduces the possibility of forgery or fraud when compared with ID cards or passwords.

Improvements in the technology have come from a variety of research projects. FaceNet, a deep learning model developed by Google, has upgraded recognition of faces that are partly covered or hidden in images. DeepFace, a landmark AI-powered facial recognition system developed by Facebook AI Research, achieves the same high level of verification shown by humans.

NeoFace, a highly accurate AI-powered algorithm developed by NEC, is built into Mobile Fortify, the mobile facial recognition system used by U.S. Immigration and Customs Enforcement to identify people.

Reducing false positives and negatives

Real-world conditions such as poor lighting, difficult viewing angles, extreme facial expressions, concealment by face masks or sunglasses, and poor image quality can still hamper performance, leading to faulty identification. False positives and false negatives are the two primary errors. False positives are when a person is incorrectly matched to a different person in a database. False negatives are when an individual is not found in a database, even though their image exists there.

False positives are more critical in security and safety applications. They can lead to wrongful accusations, discrimination or detention. In 2025, a 50-year-old woman in Tennessee was arrested and put in jail for six months based on an AI-powered facial recognition system that incorrectly tied her to a North Dakota bank fraud investigation. False negatives may prompt authorities to deny services to people who qualify for them.

Accuracy can suffer if models are trained on data that does not reflect real-world demographics. A 2025 study showed that systems trained on public databases in which people with darker skin tones are lacking leads to lower recognition accuracy. This kind of unintentional bias in training data may lead to misidentification of women, people of color and young and old people. One report found that facial recognition systems used by 42 U.S. government agencies falsely identified African and Asian American faces 10 to 100 times more often than white faces, in some cases leading to wrongful arrests.

Accuracy also deteriorates when people are wearing heavy makeup and for young children and old people because their landmark features tend to change more quickly than adults of other ages. Balancing datasets by collecting more representative images across age, gender and ethnicity, and frequently updating databases, can improve accuracy and produce fairer results.

Adjusting images before they are sent for matching – for example, changing brightness levels – can improve accuracy, too. People squint their eyes when they are in dark or very bright light. Advanced processing software can mimic this human trait to improve the facial recognition system’s ability to extract facial features from the image.

A full face from partial data

Humans are good at identifying a person even if part of their face is covered by sunglasses or a face mask. The brain assigns more significance to the exposed details. If facial recognition programs can learn to do the same, that would reduce false positives and false negatives, including when cameras only capture part of a face.

Facial dynamics can help, too. It may be difficult for someone to instantaneously recognize a middle school friend they haven’t seen for many years, but if the old friend smiles, that change in expression can immediately improve recall.

Researchers are developing a facial recognition method for doing this, known as volumetric directional patterning. It captures the subtle movements of facial muscles, as well as eyelid blinks, in consecutive frames of a video. It tracks how facial landmarks shift over time, as well as the context in which a face is being observed, which can improve recognition accuracy.

Researchers are also creating more accurate AI-powered three-dimensional systems that can capture the precise geometry of a face, including features such as contours of the eye socket, nose and chin. This kind of work could lead to anti-spoofing techniques that prevent facial recognition systems from falling for fake faces that are generated by computers and their human operators.

Fewer mistaken identities

Setting aside questions of privacy and cybersecurity and lingering issues of bias, one thing is clear: Facial recognition technology is improving. And that promises fewer errors – and fewer of the serious consequences that come with them.

---

• Vijayan Asari is a Professor of Electrical and Computer Engineering, University of Dayton
• This article first appeared in The Conversation

The post Powerful AI is making facial recognition better at identifying you appeared first on Stuff South Africa.

“Your child likely has an ear infection,” the father learns. “Your symptoms could indicate a cardiac condition,” the woman reads.

Those are helpful answers – and there’s a good chance they’re correct. Artificial intelligence is approaching, and in some cases exceeding, doctors’ ability to make accurate diagnoses. An April 2026 study found OpenAI’s o1 model had a 78% accuracy rate on complex diagnostic cases published in the New England Journal of Medicine and also outperformed experienced doctors when diagnosing actual emergency room patients. Similarly, ChatGPT, working on its own, outperformed physicians in diagnosing complex cases, a 2024 study found – even when the physicians were able to use ChatGPT themselves.

Making a correct diagnosis, though, is only half a doctor’s job. The other half is knowing what to do about it – in other words, deciding how to manage a patient’s health condition.

I am a doctor and medical educator studying how doctors make these decisions, a process known as management reasoning, and how doctors in training develop this ability. For clear-cut health concerns, an AI diagnosis may be enough for someone to get the care they need – a little numbing cream for a baby’s gums, say, or an appointment with a cardiologist.

But uncertainty is common in clinical practice. Often, knowing what ails a patient is necessary but not sufficient for determining how to care for them. And how to manage a patient, even after the diagnosis is settled, is a complex question.

Diagnosis categorizes, but management prioritizes

Experienced doctors do not assess each patient from scratch. Over years of practice, they build mental shortcuts called illness scripts.

Illness scripts are more than symptom checklists. They capture what a disease typically looks like, who tends to get it and how it most often progresses. When a doctor sees a new patient, they match what they observe against these mental scripts – a process of categorization and pattern recognition.

When a patient appears with a familiar pattern of signs and symptoms, a doctor calls up the matching mental script almost without thinking. This frees them to notice elements that don’t quite align: a symptom that doesn’t fit, or a detail in the patient’s history – a recent trip abroad, an unusual exposure at work – that points toward a different diagnosis.

It’s not surprising that AI is good at this pattern-matching process. Large language models like ChatGPT work in a similar way. They predict what word should come next in a sentence based on patterns learned from enormous amounts of text, including the medical literature. In that literature, the word “pneumonia” reliably follows certain symptom patterns: fever, say, combined with a cloudy patch on a chest X-ray. Pattern matching, at this level, is essentially the same thing a doctor does when fitting a patient’s symptoms to an illness script.

But deciding what to do next – what tests to run, what treatments to try, what to monitor and what to follow up on – works differently. Instead of one right answer, a doctor faces multiple reasonable options. The art of medical management is prioritizing which among these options is best for the patient in front of you.

The human advantage

So how does a doctor go from diagnosing a patient to figuring out how best to care for them? The answer is almost always, “It depends.”

Consider two men, Marcus and Tomás, both 68, both just diagnosed with early-stage prostate cancer. Their biopsies show the same thing: a slow-growing tumor confined to the prostate.

Both are offered the same two management options. Treat now, with surgery or radiation, accepting the risks of urinary incontinence and changes to sexual function. Or monitor closely with regular tests and biopsies, treating only if it grows. A study that followed more than 82,000 men with early-stage prostate cancer for 15 years found that fewer than 3 in 100 died of their prostate cancer regardless of which path they chose, though men who chose monitoring were about twice as likely to see their cancer spread.

AI can present both options alongside those statistics. What a doctor brings is knowledge of the person sitting across from them.

Marcus has no other significant health conditions. His doctor knows this, and knows Marcus well enough to know that uncertainty sits badly with him. For a patient without other pressing health concerns, a slow-growing tumor has time to progress and become something more serious. Both management paths are genuinely reasonable, but Marcus cannot live with waiting. Knowing cancer is in his body, watched but untreated, is not something he can set aside. He chooses treatment.

Tomás has advanced heart failure, something his doctor has been managing alongside him for years. She knows that his heart condition poses a more immediate threat to his health than this slow-growing tumor does. She knows, too, that he watched a friend go through radiation and come out diminished. Treating aggressively would mean bearing real costs for a benefit that may never arrive. She recommends active surveillance. For Tomás, it is the right answer and a relief.

Different management decisions are the norm in medicine. The right path for any patient depends on who that patient is and what they value, and on a doctor’s judgment about where the evidence is reliable and where genuine uncertainty remains.

Judging risk and uncertainty

To decide how to manage a patient’s condition, a doctor first considers evidence from the medical literature and then applies the available management options to the patient’s particular circumstances. This requires honest communication , shared decision-making, jointly navigating risk and acknowledging uncertainty.

Some risk can be measured. For chest pain, doctors use scoring tools that estimate a patient’s short-term likelihood of a heart attack based on their symptoms and test results. AI can likely work through those numbers faster than most doctors.

Read More: Why people are turning to AI first for relationship advice — and why they shouldn’t

But risk and uncertainty at the bedside or in the clinic are difficult to measure. Scoring systems and practice guidelines are designed for the average patient – an idealized person, who does not exist. And both doctors’ and patients’ sense of risk and uncertainty are shaped by their experience. For many patients, this includes a long and justified history of mistrust in the healthcare system.

AI does not know what you have been through or what risk trade-offs you are willing to accept. It cannot acknowledge uncertainty the way a good doctor can, returning to it with you as your circumstances change.

This is where diagnosis and management part ways. The father of the feverish toddler probably got a useful answer: AI has seen enough feverish toddlers in the medical literature to make a reasonable call. But knowing what to do next, including when to stop watching and start worrying, is a conversation best had with your doctor.

• Andrew Parsons is an Associate Professor of Medicine, University of Virginia
• This article first appeared in The Conversation

The post Dr. ChatGPT is getting remarkably good at diagnosing health problems ‑ but actual doctors are still better at weighing treatment options appeared first on Stuff South Africa.

That probably isn’t something you’ll have to worry about. Not that this will be too expensive (it will be), but that Audi is only making 499 of them, with deliveries starting next year.

“With the Audi Nuvolari, we are accelerating technological progress,” says Gernot Döllner, Chairman of the Board of Management of AUDI AG. “It shows what is possible when the focus is on technology, performance, and execution through teamwork — and when we achieve progress together.”

Will the Audi Nuvolari juice be worth the squeeze?

Audi’s new hybrid powertrain starts with a 4.0-litre V8 biturbo engine, which produces 588kW and 730Nm of power and torque, respectively. It’s joined by three “axial flux” electric motors, each producing 110 kW. There are two of those over the front axle, which together produce an astronomical 2,150Nm of torque. The electric engines draw from a lithium-ion battery with a gross capacity of 7.3 kWh. Altogether, the Nuvolari boasts a maximum system output of 736kW.

If you’re ever in a hurry, it will do 0-100km/h in 2.6. If you’re in a big hurry, 0-200km/h will happen in a claimed 6.8 seconds. It maxes out at “more than” 350km/h. So… 351km/h?

Since entering the F1 scene this year, it seems like Audi was looking for an excuse to use F1 tech in its next street car. The Nuvolari uses bits of active aerodynamics that work as a system to adjust downforce, drag, and aerodynamic balance in response to driving conditions, which Audi says maximises the vehicle’s stability and precise control.

The best bit of this active aerodynamic system is the deployable adaptive rear wing. It’s capable of operating in three conditions: closed, low downforce, and high downforce, and generating up to 400kg of downforce in certain conditions. Then there is the breaking system that sports an “energy absorption capacity of up to 2.8 megawatts,” which Audi says is “on par with a current Formula 1 car.”

The interior is suitably sharp. It features Audi’s new signature colour, Titanium, with minimal distractions and controls “directly within the driver’s field of view”. One of those is a manual DRS button (Drag Reduction System), like the F1 cars use.

While the design might not be everyone’s cup of tea, we don’t think it will receive the same backlash as Ferrari’s latest vehicle did.

Source

The post The Nuvolari is Audi’s first hybrid supercar that you won’t own appeared first on Stuff South Africa.