On Friday, Apple dropped the bombshell news it was suing OpenAI over the alleged theft of trade secrets, claiming that OpenAI stole Apple’s confidential data and engaged in efforts to learn proprietary information while recruiting former Apple employees.
In accusing OpenAI of stealing secrets about Apple’s unreleased products, Apple revealed that a former employee allegedly siphoned reams of sensitive files from the company’s shared network folders, weeks after leaving Apple for a job at OpenAI.
In its complaint, Apple says the former employee, a system electrical engineer named Chang Liu, allegedly “exploited a rare, previously unknown authentication bug” that allowed access to the company’s network. The bug is classified as a zero-day vulnerability, meaning that Apple had no time to fix it before it was allegedly exploited.
Apple has since fixed the bug and said it terminated the employee’s access once it learned of this “security breach.” In its complaint, Apple said the bug could have allowed a “few other” people to access data on its network, but alleged that only Liu exploited the bug to steal Apple’s confidential information while no longer an employee, citing a check of its server logs.
Advertisement
The disclosure, while light in detail, highlights the challenges that organizations face with protecting sensitive corporate data after employees no longer work there. Companies often move to immediately cut off departing staff from further access to protect any sensitive information from leaving, including inadvertently. Companies that fail to fully decommission their employees’ accounts can face future security lapses, data breaches, or malicious actions by disgruntled staff.
Apple spokespeople did not respond to an email from TechCrunch with questions about the security vulnerability, how it was exploited, and when the company decommissioned the employee’s credentials.
“LOL… so funny.”
In the complaint, Apple alleged that Liu took “dozens of Apple’s confidential hardware-related files” over the course of several weeks while as a new OpenAI employee.
Apple said the files contained “detailed information about unreleased products, engineering presentations, technical specifications, and proprietary project data.”
Advertisement
The company claims Liu failed to return the Apple-issued work laptop he had previously used to access Apple’s network, suggesting it was once able to send and receive files from Apple’s internal systems. The complaint said that Liu allegedly claimed to have “another computer.” While he was at OpenAI, Liu also allegedly misused the access of an acquaintance, Yu-Ting Peng, a then-Apple employee who later went to work for OpenAI. Liu allegedly used Peng’s Apple-issued work laptop “while she was still employed at Apple and he was not.”
Apple said that during February 2026, Liu “tried to access Apple’s network storage — a cloud-based file repository containing Apple’s confidential engineering files, project documentation, and other proprietary information.”
Liu had allegedly discovered that he “still could access Apple’s network repository after leaving Apple, the result of a then-unknown authentication vulnerability.”
Apple did not describe the authentication “bug” that Liu allegedly used to access Apple’s network. However, authentication bugs generally refer to flaws in the login process that allow improper access to systems or data, either because of a weakness in how the login mechanism works or due to a misconfiguration, such as overbroad permissions or not decommissioning the login credentials of a former employee.
Advertisement
Apple wrote in its complaint that when Liu learned he had unauthorized access to Apple’s systems, he did not report the bug to Apple under his employment agreement obligations, nor did he return his Apple-issued work laptop.
The complaint added that Liu also failed to “delete the program that allowed the access” to Apple’s network. The company did not say what program or app that Liu allegedly used to access Apple’s systems. It’s not uncommon for employees to have tools, such as a work-approved VPN or remote-viewing app, that allow them to access sensitive data from outside of the company’s offices using their credentials.
Given that Liu was previously granted credentials to Apple’s network as an employee, TechCrunch asked Apple when the company decommissioned Liu’s access, but we did not hear back.
Once Liu allegedly gained access to the network share, he wrote to Peng: “LOL, I found out I can access the [network storage], so funny.”
Advertisement
Apple filed its suit in the U.S. District Court for the Northern District of California in San Jose, and has demanded a jury trial. OpenAI previously said it has “no interest in other companies’ trade secrets.”
The case, if it proceeds, could begin this year.
When you purchase through links in our articles, we may earn a small commission. This doesn’t affect our editorial independence.
Prior to the tournament, Singapore’s Mediacorp broadcaster announced that it would stream 28 World Cup 2026 matches completely FREE on both TV and online via its MeWatch streaming platform. While that’s nowhere close to the likes of the UK’s BBC/ITV or Australia’s SBS On Demand, which are streaming all 104 matches for free, it’s still a significant jump from the nine matches that were streamed free in 2022.
The good news for football fans in Singapore tuning in for the business end of the tournament is that all four remaining matches, including the two semifinals, the third-place playoff, and the all-important FIFA World Cup final, are available to watch free of charge in the country.
But can you watch these free streams from anywhere in the world? And what if you’re a foreigner travelling in Singapore during these matches? Can you still access your usual free World Cup coverage? Follow our guide below to find out where to watch FIFA World Cup 2026 live streams in Singapore, including free streams.
Advertisement
How to watch World Cup 2026 in Singapore
Mediacorp has made all 104 matches available live and on demand via its MeWatch streaming platform. A FIFA World Cup Pack costs S$118, including taxes.
However, if you’re tuning in now, there’s no need to get a subscription, as the remaining four games are free-to-air on MeWatch. The platform is also showing highlight videos of all 104 matches for free.
However, you can only access World Cup 2026 live streams on Zee5 with an Indian IP address. Abroad right now? Here’s how to unblock Zee5 from anywhere in the world…
Advertisement
Can I watch FIFA World Cup 2026 for free in Singapore
As mentioned earlier, while the entire tournament might not have been free-to-air, the remaining four matches, including the third-place clash, are FREE on the MeWatch streaming platform.
And if you’re a cable TV loyalist, you can watch all the action live and free on Mediacorp’s Channel 5 TV channel.
But free streams aren’t just limited to Singapore’s borders. Lots of streaming platforms from around the world are offering free World Cup streams. So, if you’re traveling in Singapore during the tournament, you can use a VPN to connect to a server in your home country and watch your usual free streams.
Advertisement
Here are the countries showing every game of the FIFA World Cup 2026 for free:
Abroad? Can’t access your free stream? Unblock your free World Cup stream with Norton VPN – more on that below.
Watch FIFA World Cup 2026 live streams from anywhere
If you’re traveling, you might discover your usual World Cup live stream is suddenly unavailable due to geo-restrictions.
Advertisement
Don’t worry, that’s exactly where a VPN can help. A virtual private network lets you connect to servers around the world so you can securely access your usual World Cup coverage as if you were back home.
Tuesday, July 14 — Semi-finals France vs Spain: 3 p.m. ET / 12 p.m. PT / 8 p.m. BST
Advertisement
Wednesday, July 15 — Semi-finals England vs Argentina: 3 p.m. ET / 12 p.m. PT / 8 p.m. BST
Saturday, July 18 — Third-place playoff TBC vs TBC: 5 p.m. ET / 2 p.m. PT / 10 p.m. BST
Sunday, July 19 — Final TBC vs TBC: 3 p.m. ET / 12 p.m. PT / 8 p.m. BST
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.
SpaceX just released a new documentary that drops straight into the control rooms and launch site during the final stretch before its biggest rocket flew again. Titled Critical Path, the 34-minute film follows engineers at Starbase in Texas through the intense days leading up to Flight 12 on May 22. It is the second episode in their ongoing series and stays tightly focused on the real work required to get the first Version 3 Starship and Super Heavy off the ground.
Version 3 was a total revamp from the start for both stages. It was a completely different ballgame with a new upper stage, a new booster, updated Raptor engines, and a totally new launch pad designed to handle the strains of flying more regularly. The documentary demonstrates how all of those modifications resulted in a whole new set of dependencies that all had to fall into place on time. Engineers begin by explaining the critical path in simple terms. Simply said, it is the longest chain of jobs, with each step dependent on the previous one. Any delay in that chain means that the entire launch is pushed back. They describe the ongoing discipline of keeping an eye on those links without becoming bogged down in second-guessing, which would only stifle growth.
NASA rocket model kit – Launch into a creative project with the LEGO Icons NASA Artemis Space Launch System model building project for adult space…
What’s in the box? – This creative building set includes everything you need to craft a multistage rocket with 2 solid-fuel boosters, an Orion…
Features and Functions – This NASA-themed rocket model features retractable launch tower umbilicals, rocket support and crew bridge, detachable…
The story begins with a 10 engine static fire on the booster. The crew was looking for strong evidence of how the new pad and engines would perform together under stress. An early attempt was cut short when sensors detected unexpected vibration on the flame diverter. They modified things and then completed a full-duration burn on the second attempt. When the clean data began to arrive, there was a noticeable sense of relief.
We get a close-up view of the new Pad 2 infrastructure. This is a gigantic flame diverter that can route the exhaust from 33 engines. Water deluge systems are on standby, ready to convert into steam and remove heat from the concrete and steel. The design is based on the hard lessons learned from earlier flights that ruined the original pad. Crews had never attached the quick-disconnect propellant arms or flowed fuel on this pad previously. Small pressure blips and mild movement in those arms during testing were carefully examined.
One launch attempt was canceled when those peculiarities reappeared in the closing minutes. The crew labored late one night to find out a solution, and then added a simple welded hard stop to limit any undesired motion, which held on the next attempt. Another sequence depicts what occurs when something breaks on a large moving item; following the static fire, a link in the chain on one of the tower arms collapsed during retraction. The catch mechanism has to be fixed immediately away. Spare parts arrived from the other side of the nation, and crane men worked all night. After 30 hours, the arm was back up and working nicely.
For this flight, twenty-two objects were placed into the bay, including mass simulators and two modified Starlink satellites equipped with cameras. With a total mass of 37.5 metric tons (the heaviest payload the Starship has ever carried), the team had to double-check every last surface to ensure that nothing would come loose at the worst possible time, and recovery planning is not overlooked; in fact, it receives a lot of attention. One of the mission’s key objectives was to acquire some extremely high-quality imagery of the heat shield during re-entry; the splashdown site featured the most buoys, drones, and support vessels observed in the Indian Ocean to date. This meant that they needed all of that hardware to collect reliable data on how the vehicle performed under the extreme heat of re-entry to create an accurate model.
When the launch day arrived, the first countdown had to be halted due to a quick-disconnect pressure reading that did not match. Engineers felt it was best to stand down and focus on the remedy, which they did, implementing it overnight and hoped for the best. They returned the next evening, fully prepared, and liftoff went off without a hitch. All 33 engines started up cleanly and supplied the power they were designed to. One of them shut down briefly during the ascent, but the car continued to climb. Hot staging, the key step in which the stages separate, was a breeze; nothing got stuck up or caused us any problems. The upper stage placed the payload in orbit and then performed a controlled re-entry, complete with a little banking maneuver to test the rear flaps and see how they would react when things got hot. Needless to say, it splashed down exactly on target in the Indian Ocean.
Televisions are evolving yet again. This time, the new kid on the block is mini RGB, which emits red, green, and blue lights through an LCD panel instead of just white or blue lights. According to Hisense, one of the first brands to debut an RGB LED TV, the technology is an attempt to produce “pure colors directly at the source.”
The intended result is LED TVs with better colors, improved contrast, and much higher brightness—all designed as a proper alternative to OLED. Whether it’s a giant leap forward, though, depends greatly on the kind of movies and shows you watch and your willingness to tweak a few settings. I tested five new models to get a taste of what’s to come and to determine which (if any) you should buy.
The first RGB TVs came out in 2025, but it’s really in 2026 that the technology is getting wider distribution across sizes and prices. As I stated above, an RGB TV has conventional LED backlighting but uses a red, green, and blue backlight module rather than the standard white or blue LEDs. But it’s not quite that simple. Samsung and LG call it “micro RGB,” while TCL and Hisense say “mini RGB.” The tech works roughly the same, even if micro RGB uses smaller LEDs. Sony uses the term “True RGB” and claims there’s no difference between mini RGB and micro RGB.
Advertisement
My goal in testing, though, was to cut through the marketing spin and just put each model through a few benchmarks, watch the same movies, and stay glued to the World Cup no matter which model I was testing. The surprising discovery is that mini RGB (which is the term I’ll use for all of them) is noticeably vivid, has excellent contrast, and looks bright and clear even during the daytime. Mini RGB televisions are also excellent for off-angle watching.
That said, I don’t think you have to put that OLED up on Facebook Marketplace just yet. Mini RGB is an evolution in tech, but it also means manufacturers can keep using LCD panels. OLED was a sea change because individual pixels can emit color or be turned off entirely. In my side-by-side tests, mini RGB is a smart upgrade but isn’t the ultimate display tech. While the costs are comparable for now, I expect mini RGB televisions to come down in price soon.
Why Choose a Mini RGB Over Other Models?
The name mini RGB would imply that it is all about color—specifically red, green, and blue. Yet, the way the technology works by shining those colors through an LCD panel means there is a lot more control over contrast and brightness as well.
Advertisement
In my tests, I found settings for brightness, clarity, contrast, and picture mode had a much more obvious impact than on a normal LED or QLED. I equate this to a sports car. You can drive a BMW M5 on a side street at low speeds, but until you enable track mode, adjust the suspension, and perform other tweaks, you won’t really know what the car can do. Mini RGB is similar in that it’s highly customizable.
How Much Do RGB TVs Actually Cost?
Anything brand-new to the market will likely cost more than we might expect. While the Hisense UR9 RGB Mini-LED was the cheapest by far at only $1,999 for the 65-inch model, every other mini RGB and micro RGB costs closer to $4,000. That’s about $1,300 more than an OLED at the same size. The flagship OLED models from LG and Samsung tend to cost around $2,700 for a 65-inch. LED and QLED models are even cheaper, running as low as $500.
How Easy Are They to Mount?
Advertisement
As with any modern television, mounting one on a wall is fairly easy since there’s so much information online about how to do it. I’ve found YouTube videos that explain exactly what to do, even covering how to install an electrical box. Mainly, you have to use a stud finder and make sure you’re mounting the brackets into a stud and not just sheetrock, but even that process only takes a few minutes.
The one exception is if you go with a larger-size mini RGB model like the TCL RM9L RGB-Mini LED. If the television weighs over 100 pounds, it changes the ball game in terms of using multiple studs and adding extra mounting brackets.
The Best Overall RGB TV
LG makes high-end televisions that tend to be a bit pricey, but they’re often worth the extra expense. The LG Micro RGB Evo is no different. Priced at $4,500 (that’s with a $500 discount right now) for the 75-inch model I tested, this is one brilliantly colorful and impressive television.
Setup and install were simple. It’s just a matter of inserting the legs and screwing them in tight. My only complaint here was the legs were a bit pointy. For connectivity, the LG Micro RGB Evo has four HDMI 2.1 ports, an Ethernet port, one digital optical, a coaxial connection, and two USB 2.0 ports. WebOS is a capable streaming platform, if a bit confusing and bloated with too many apps and advertisements.
Advertisement
Movies and shows looked stunning and vivid on this television, once I got used to tweaking the picture modes. By default, some of the stock settings (e.g., Filmmaker mode) made movies like Awake on Netflix look too dark. Tron: Ares on Disney+ was vivid with deep blacks and reds.
While the LG Micro RGB Evo worked perfectly fine for console gaming, it had some trouble with a gaming laptop. The variable refresh rate setting, which LG calls Motion Booster, did not work correctly when I used an Alienware 16X Aurora laptop. However, at the native 165-Hz refresh rate, this model is a game changer—Crimson Desert looked absolutely stunning.
London Tech Week’s focus on AI – from a £12 million investment in AI for SMEs to AI bootcamps for graduates and more – has reflected the pressure to compete in an AI-era.
As this digital revolution progresses, the job economy is changing, but the mantra that AI is taking our jobs is simply not correct and potentially fueled by an undercurrent of classicism.
Vincent Huguet
CEO and co-founder of Malt.
When the Luddites famously started to break the first machines of the industrial revolution in 1811 in England, fearing for their job as textile artisans, the “Bourgeoisie” would describe them as “ignorant workers”, with no understanding of basic economics.
Latest Videos From
More than two hundred years later, with the rise of GenAI, it is no longer the blue-collar workers who fear for their job, but the white-collar workers. This time it is the “bourgeois” who live in the anxiety of an uncertain world.
Since ChatGPT introduced AI into the everyday lexicon, it has been clear that we would experience an unprecedented revolution. The rhetoric that immediately began to dominate social discourse has been that AI tools would render most jobs insignificant.
Advertisement
Furthermore, whilst other technological revolutions ended up being creative destruction, ‘this time it was different’.
But is that really the case? Or are we more fearful, more concerned about destroying the status quo, because this time it’s a different ‘class’ of people being impacted? This time it’s the desk workers, not the physical laborers, who risk losing jobs, and suddenly there is alarm.
Sign up to the TechRadar Pro newsletter to get all the top news, opinion, features and guidance your business needs to succeed!
Advertisement
Artificial Intelligence relies on humans – and more humans than ever
AI is a human creation and still relies on humans to evolve. First, we have those who build the infrastructure, like data centers, which accounted for almost all of the United States’ GDP growth in the first half of 2025 (according to Harvard economist Jason Furman).
Then, we have those who train the models, which still need to be constantly retrained. Even if models are able to train themselves eventually, there is no consensus that human intervention in training will become obsolete, because human behavior and the entropy of organizations are in a constant state of flux and evolution.
And even when trained, AI constantly needs to also understand the “context” in which it is prompted to perform efficiently. AI then needs to be deployed. Managing security, defining guardrails for agents, understanding how to use AI and tracking agentic AI’s actions, all comes with inherent challenges.
Advertisement
The CIOs of the largest global corporations are already investing hundreds of millions of pounds to understand this. Startups based in San Francisco – a city I recently visited where 95% of out-of-home ads were about AI agents – are focused entirely on resolving these problems for large enterprises.
The fact that both Anthropic and OpenAI have launched their own consulting companies is proof that managing AI complexities in the coming years will be the biggest source of growth for all consulting and outsourcing companies of the world.
Advertisement
Sourcing the right human talent in the AI era is the biggest challenge
Software engineering is a job category where GenAI – perfectly trained on open-source code and GitHub repositories – can now code better than even the most experienced developers.
Additionally, developers in AI labs – with privileged access to “tokens” on Claude Code or OpenAI Codex – now develop 100% of the time without writing a single line of code. Nonetheless, when asked about their biggest challenges, all AI startups would point to recruitment.
A report by the UK’s National Foundation for Education Research showed a 50% increase in tech job adverts between 2019/20 and 2024/25, with entry-level roles particularly affected. However, we’re now seeing a surge in demand driven by Gen Z, according to Employment Hero’s March Jobs Report.
This demand for AI expertise is reflected in a new Malt Tech Trends Report, which analyzed 1.2 million searches of tech freelancers in 2025. It reveals that AI is now the second most-in-demand skill, irrespective of company size, industry, or project type. More specifically, demand for freelancers with agentic AI expertise exploded by 5,800% in just twelve months.
Advertisement
Observer of the AI revolution, Andrew Ng, explains that if, for example, a team of 3 developers builds 10 times faster, then they need more designers or product managers to fuel the creative process. Doing more faster, with fewer people creates more work to fuel and execute the output.
More people are echoing the same rhetoric as Ng, calling out the phenomenon of ‘AI washing’, whereby companies have justified mass redundancies with AI disruption. In reality, in many cases, they were either adapting to geopolitical and economic uncertainties or had simply employed too many in the crazy post-COVID bull market.
The AI job apocalypse is not yet here… Still, the fear is real and needs to be understood
Software engineering is a perfect example of a job category that has constantly evolved. Since the inception of computer science, programming has become progressively more about “natural language”. Whilst there were 50,000 developers worldwide in the 1960s, today there are almost 50 million. Undoubtedly, the eradication of barriers to entry to build software increases that number tenfold.
Advertisement
History, data, and observation shows us that the AI job apocalypse is not yet here. Still, the fear is real and needs to be understood. The reason every science fiction novel paints an inhospitable world and unattractive paradigm is because the human mind always fears change. We assume the worst.
AI transformation, like all transformations, will be a cultural change first. And it’s companies, not professionals, who are most at risk if they fail to adapt. If one thing will be different in this digital revolution, compared to the last (arguably comparable is the advent of the internet), it’s the rate of change.
CEOs will have to be imaginative, change org charts and processes, admit they are not omniscient, take risks, and invest in training. Schools and universities also face the challenge of teaching soft skills: how to adapt to live and work in a more uncertain world. Because we can only harness top-tier AI talent if we understand how to truly adapt to change.
Independent professionals – those who create their own roles – from freelance developer to fractional manager and strategic consultants – have already redefined work.
Advertisement
On average, freelancers spend 4 hours a week on upskilling and keeping up with the job market and already have the habit of switching from one client project to another. They were the first to adapt to AI and realize that a job is more than just a bundle of tasks.
As Jensen Huang, CEO of Nvidia, recently said, if someone were to observe him at work, we would conclude that his day consists of tasks like making hundreds of calls and sending emails. AI will replace, augment, and improve these tasks. But it will not take Jensen’s job.
This article was produced as part of TechRadar Pro Perspectives, our channel to feature the best and brightest minds in the technology industry today.
Advertisement
The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/pro/perspectives-how-to-submit
Keep yourself open to, “OK, I’m gonna look around. Is there a Luddite poster here? I’m at my local bookstore. OK, is there some information here?” It’s this idea of social infrastructure, switching people over from the idea of, “OK, I’ll look at an Instagram page, and here’s all the events” to the idea of, “I actually have to leave my house.” Be on the lookout. We’ll be out here. Tompkins Square Park, probably, other parks, privately owned public spaces. We have events all over this city. It’s amazing. Knicks in five. I love New York.
Gowanus, we like to play a little game on each show.
I love games.
We’re very proud of it. It’s called Control, Alt, Delete. So I wanna know what piece of tech you would love to control, what piece you would alter or change, and what you would delete, or vanquish from the earth.
Advertisement
Wow.
Are you ready?
Control, Alt, Delete. OK, great.
Let’s start with control.
Advertisement
This is so funny that you’re asking me this. I would say the servers for the internet.
The internet was billed as something that was, OK, we are going to have a free exchange of knowledge across the world. It’s going to help diversity and globalization and all this stuff, and then, boom. Military technology, right?
I’m trying to think about which technologies do I feel have a real positive impact, but have a negative sort of profit incentive.
Let me give this one. I would alter the way that social media platforms have a centralized system instead of a federated system. I think that’s pretty solid, and this was almost in the creation of Twitter, they were about to do this. And then Jack Dorsey was like, “No, we should have it be more centralized.”
And finally, delete.
I can’t wait. This is my favorite one.
Advertisement
What would you delete?
AI. AI data centers. Done. Done. Immediately done. Boom. Gone. I mean, come on, guys.
Come on. Especially when they bill it as, “OK, we’re gonna make an AI to fix all of our climate catastrophe, all of our income inequality, but first, we have to absolutely drain the natural resources of the Earth.” I mean, it’s just ridiculous. It just makes me so frustrated. I think the data center is honestly one of the worst materializations of this Big Tech oligarch world that we live in.
Where like the whim of Mark Zuckerberg wanting a Meta Hyperion data center in Louisiana is legitimately going to use multiple times the amount of energy that New Orleans uses. We can’t sustain that as a planet, truly. It doesn’t matter how interesting the technology is or what it could do. These are resources that we need now for life to continue. We need fresh water. We need land to grow food. We need ecosystems where birds and bees and wildlife can coexist, and I’m not very interested in trading that for a chatbot, even if the chatbot tells me it loves me and we can kiss through the phone.
Advertisement
How to Listen
You can always listen to this week’s podcast through the audio player on this page, but if you want to subscribe for free to get every episode, here’s how:
If you’re on an iPhone or iPad, open the app called Podcasts, or just tap this link. You can also download an app like Overcast or Pocket Casts and search for “uncanny valley.” We’re on Spotify too.
Apple has finally introduced its first public betas of iOS 27, iPadOS 27, macOS 27 Golden Gate, watchOS 27, visionOS 27, and tvOS 27.
Since the introduction of the 27-generation operating systems at WWDC in early June, Apple has been testing developer beta builds. A few rounds later, and it has started to do the same with the public beta.
A public beta differs from a developer beta in that the software has been tested enough to be less of a risk to end users. While a developer beta has the potential to cause problems for testers, the public variant is from a later stage with the bigger potential issues out of the way.
While the public beta is generally safer to use than a developer version, it’s not entirely safe. There’s still the risk of data loss and other issues, just that the chance is smaller.
Advertisement
AppleInsider and Apple strongly advise against users installing test operating systems or beta software onto primary or “mission-critical” hardware. That warning stands for both developer betas and public beta builds.
Just like the developer counterparts, public beta users should really install the operating systems onto secondary, non-essential hardware. Also, they should maintain sufficient backups of their critical data at all times.
We don’t say this for fun. Every year, someone on our editorial team ends up suffering because of a problem with the developer betas. We have also heard countless stories of people going through the same issues, in varying degrees of magnitude.
Don’t be like us. If you must try out the new operating systems, do it on the public betas instead of the developer ones.
The open statement says that leaders in this space must act now to understand the economics of transformative AI and steer the tech in the right direction.
Almost 200 economists and technology leaders have signed a statement warning of the risks posed by AI if it is to be left ‘unchecked’ in the coming years. Many of the world’s experts are concerned that AI is reaching a stage where it is too powerful and needs to be guided in a more human-focused direction.
The statement, which is titled “We Must Act Now,” was organised by economists Erik Brynjolfsson, Ajay Agrawal, Anton Korinek and Tom Cunningham and was signed by a range of people close to the issue.
This includes several Nobel laureates, the chief economists of Open AI and Anthropic, Jack Clark, a co-founder of Anthropic, Eric Schmidt, the former CEO of Google and experts from Cambridge University, Stanford, Harvard and Oxford, among others.
Advertisement
The statement said, “AI may become radically more powerful over the next 10 years. This could drive an unprecedented transformation of our economy, larger than the Industrial Revolution, but unfolding over a vastly shorter time frame. It could bring risks, including large-scale job displacement, as well as opportunities such as major gains in living standards.
“Economists, policymakers and technology leaders must act now to understand the economics of transformative AI and to build the incentives, guardrails, and institutions needed to steer AI in a direction that complements humans and benefits society.”
The statement is reflective of a landscape in which more and more people are becoming concerned about AI’s potential to eliminate employment, impact the economy and affect how we live our lives.
In early July, Microsoft announced it is laying off 4,800 people, including 3,200 from its gaming division Xbox, as the company aims to cut costs and flatten its organisational structure in response to AI and a changing landscape.
Advertisement
In June, new research from Boston Consulting Group (BCG), found that for many organisations, AI is fundamentally reshaping the nature of work, leadership and how employees experience the workplace. While there were positive elements to the research, many contributors also found an increase in ‘cognitive load’, creating a paradox’ where AI is making work better and harder simultaneously.
Don’t miss out on the knowledge you need to succeed. Sign up for the Daily Brief, Silicon Republic’s digest of need-to-know sci-tech news.
Somewhere right now, a customer is repeating themselves. They are explaining their problem for the third time, to the third person, because the organization on the other side has no shared memory of the previous two conversations. It is an infrastructure problem that AI is making harder to ignore.
Ahmed Bashir
It is also becoming impossible for policymakers to ignore. Just in April, the Mayor of London launched a new AI and Jobs Taskforce to examine how AI is changing work across the capital, signaling that the conversation has moved well beyond investment announcements and into the harder question of what AI does inside organizations.
Advertisement
It is also shining a spotlight on a memory crisis inside modern business.
Latest Videos From
AI is accelerating work, not clarity
As UK organizations rush to deploy AI in the workplace, many are layering it onto fragmented systems that were never designed to preserve institutional memory in the first place.
According to research published in Harvard Business Review, knowledge workers toggle between applications and tools roughly 1,200 times per day, a pattern known as “toggling tax”. That figure alone tells the story: we aren’t short of tools, but there is no coherence among them.
Advertisement
The result is a new kind of productivity paradox. Work is moving faster, but clarity is not improving.
This is where much of the current enterprise AI conversation unravels. A surprising amount of what is marketed as AI today still relies on humans to do the synthesis work themselves. The system retrieves documents. It summarizes conversations. It surfaces links. But employees still carry the burden of reconstructing meaning, and so do the customers and end-users waiting on the other side of those decisions.
Sign up to the TechRadar Pro newsletter to get all the top news, opinion, features and guidance your business needs to succeed!
Advertisement
Notably, when these types of AI tools do the retrieval, but humans skip the synthesis, the output feels hollow. That creates a trust and credibility problem – not just for the individual, but for AI as a category. People start associating “AI-assisted” with “low-effort”.
When context is lost internally, the effects aren’t invisible. They surface as slow responses, repeated requests for information that customers already provided, support experiences that feel fragmented, and sales teams reconstructing account history manually before every renewal, escalation or executive review.
Stateless systems cannot preserve organizational memory
The AI models themselves are becoming more capable, but the organizational foundation beneath them remains fragmented.
Advertisement
Most AI systems today are fundamentally stateless. They generate outputs based on temporary context windows rather than durable organizational memory. Every interaction requires the system to repeatedly reconstruct understanding from fragments.
Consider how databases work. We do not recompute everything from scratch every time a query arrives. We cache and index, then preserve relationships between entities, because continuously recomputing context is computationally irrational.
Yet much of enterprise AI is still being deployed exactly this way and the industry has started mistaking activity for intelligence.
Advertisement
What I believe organizations should focus on is whether they have structured, durable memory that lets AI and humans reason from the same shared context. Without that foundation, AI outputs remain generic.
Most collaboration systems multiply this problem in two ways. First, they encode knowledge into naming conventions and tribal memory – the kind that lives in channel names nobody can decode and folder structures only three people understand. New employees are not learning the business, they are learning the conventions.
Second, even when information exists, it remains inaccessible. The same decision appears as “PostgreSQL migration”, “database move Q3”, and “backend infrastructure change” across three different channels. They are semantically identical but textually invisible to any system trying to surface it.
This problem becomes even more acute in distributed organizations. I don’t believe you can build modern global companies on a “you had to be there” culture. Yet many businesses still operate as though important context naturally transfers through proximity and synchronous communication.
Advertisement
Search is not the same as understanding
Search was designed to discover information, whereas modern enterprise work requires systems that understand the relationships in data.
A customer escalation is not just a support ticket. It is connected to product decisions, engineering discussions, account history, contractual obligations, and revenue impact. A sales opportunity is tied to customer sentiment, historical support patterns, product usage, and internal stakeholder alignment.
Traditional collaboration systems flatten these relationships into disconnected channels and documents, whereas AI knowledge graphs preserve them.
Advertisement
Researchers call this a transactive memory system: the collective understanding of who knows what, how decisions were made, and how work is coordinated across teams. The same logic now extends to AI. Intelligent systems can participate in that process too by encoding context, surfacing relevant history, and routing knowledge to the right people at the right time.
Britain’s productivity problem is becoming an AI problem
The Office for National Statistics has consistently flagged weak productivity growth as one of the UK’s most persistent economic challenges. Since 2010, UK productivity has grown at 6.2%, compared with roughly 10% across the euro area and nearly 15% in the United States over the same period. AI is increasingly being positioned as a mechanism to help close that gap.
But productivity does not improve because your business has added more AI agents to the workflow. If every important decision still requires humans to manually reconstruct fragmented context, organizations just accelerate confusion.
Advertisement
What UK businesses need are systems capable of preserving context, maintaining institutional memory, and grounding AI systems in trusted organizational knowledge. Better AI infrastructure starts with a simple question: Does your organization remember anything? For most, the honest answer is no.
This article was produced as part of TechRadar Pro Perspectives, our channel to feature the best and brightest minds in the technology industry today.
The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/pro/perspectives-how-to-submit
Marc Porat sat with a red notebook in 1989, drawing what no one else could see. A little rectangular piece of glass with a touch screen, phone, fax, messages, video, games, ticket purchases, and apps delivered over the air. He named it the Pocket Crystal. It would feel like a piece of jewelry you carried every day, something with the comfort of a seashell and the pull of a crystal. At Apple, where he worked, the idea landed with John Sculley. Resources stayed scarce. So in May 1990 the project left Cupertino and became its own company. Bill Atkinson and Andy Hertzfeld, two of the original Macintosh wizards, signed on. General Magic was born.
They called the location after a line from Arthur C. Clarke. The idea is that any sufficiently advanced technology appears magical. You had a bunch of veteran Mac users and some hungry new developers crowded into Mountain View offices. Joanna Hoffman was in charge of marketing because she was one of the first people on board. Susan Kare developed the icons for the new operating system, and Megan Smith joined shortly thereafter. Meanwhile, a young whippersnapper named Tony Fadell walked in from the street. There were even rabbits bouncing around on the floors, as well as a parrot or two, presumably released by its owners when they went for the day. Some folks were even sleeping off while resting their heads on their desks. You could bet that at any minute, someone would start a water battle. However, the energy was fantastic. Everyone was confident they were onto something major, specifically the next item after the Mac.
WHY IPAD — The 11-inch iPad is now more capable than ever with the superfast A16 chip, a stunning Liquid Retina display, advanced cameras, fast…
PERFORMANCE AND STORAGE — The superfast A16 chip delivers a boost in performance for your favorite activities. And with all-day battery life, iPad…
11-INCH LIQUID RETINA DISPLAY — The gorgeous Liquid Retina display is an amazing way to watch movies or draw your next masterpiece.* True Tone…
Magic Cap was the name of the operating system. When you first booted it up, you were in a virtual area that appeared to be a real office. There was your calendar in the corner, and your inbox was simply waiting for you. Walk down the virtual hallway and you’ll come across a variety of handy rooms, including a library, a game room, and even a downtown business center where you may purchase new software. Messages were decorated with stickers and animated characters, and those little faces evolved into the emojis we all know and love today. You could navigate with a stylus. Software modems handled connections without the need for additional hardware, and early versions of what we now know as USB connectors appeared. To keep things light, the hardware had to do significantly less work.
Telescript was the brainchild in charge of all the sophisticated elements. When you leave your smartphone, a digital “gentleman” will journey across networks and return with answers to your questions. Jim White and his team developed a language that enabled programs to move from one machine to another, just as humans go between cities. They discussed the “Telescript cloud” before anyone knew what it was. AT&T built PersonaLink on top of it because agents needed somewhere to go.
Advertisement
Before you knew it, some of the biggest players were lining up to participate in the action. Sony, Motorola, Matsushita, Philips, AT&T, and later NTT, Toshiba, and France Telecom also joined in. Each provided money and appointed their top executives to a special council. Sony’s John Sculley and Norio Ohga were among the best performers. The Alliance swiftly became the industry’s largest collection of business players, prompting antitrust specialists to establish new rules for the meetings. In 1993, the New York Times named General Magic Silicon Valley’s most-watched startup of the year.
Finally, after all of the excitement, the hardware was released in 1994. Sony shipped the Magic Link for a cool $800. The device looked like a grey brick with a stylus, a small monochrome screen, and a built-in modem that required a phone connector. Motorola followed up with the Envoy, which added a wireless radio to the equation. Both used Magic Cap. You could email, fax, or even page somebody if that was your preference. Keep your contacts and calendars up to date. Play some games and send some files over with IR, as the device was essentially a magic wand. However, nothing like existed previously. Of course, sales were small, with the majority of the units going to friends and relatives. Battery life was a joke, and performance was sluggish. Had no internet (yet) and no cell data worth noticing. To make matters worse, Apple had recently released the Newton the year before, which had likely stolen some of General Magic’s thunder.
An IPO in February 1995 nonetheless managed to raise 96 million dollars. But that was only the beginning; the stock had more than doubled on the first day, and it appeared like cash was flowing in. The engineers were practically unstoppable, and new gadgets popped up left and right. Later that year, Portico was introduced as a voice service that anyone could access using any old phone. An 800 number would then read out your email, calendar, and messages in a polite, calming voice, almost as if you had your own personal assistant. By the time they reached a peak of 2.5 million users, they had already created MyTalk, which has earned a permanent home in the Smithsonian. However, the initial notion of such ‘dream devices’ never really took off. AT&T chose to discontinue PersonaLink in 1996. By 1997, the hardware partners had essentially stopped producing. The stock fell precipitously, prompting layoffs. It all came to an end in September 2002, when activities ceased, and by 2004, they had been totally liquidated. Paul Allen ended up purchasing the majority of the patents.
On more than one occasion, I’ve embarrassed myself by brewing coffee outdoors and spilling a freshly made cup onto an unsteady camping table. Not to mention, light packers would scoff at the weight of my coffee gear — a necessary sacrifice to avoid instant coffee. Savoring high-quality joe in the open air feels special, though, hence why I bring a coffee-making setup every time.
Finally, I’ve found the easiest option: The MiiR Pourigami. Put together, the pyramid-shaped dripper fits atop any trusty travel mug. Taken apart, this Miir Pourigami resembles a card holder, slim enough to fit into my pants pocket. It functions like other pour-over setups, meaning I can still dial in tasting notes. If you’re like me and think about coffee no matter the circumstances, this nifty setup lets you play barista in any environment.
A look at the Miir Pourigami
Advertisement
Yup, that’s a coffee maker in my hand.
Nikita Ephanov/CNET
Right out of the box, the Pourigami looks sleek. Disassembled, the brewer consists of three thin stainless-steel trapezoids. I can’t imagine the pieces bending or chipping — crucial, as I’m prone to breaking camping equipment. Handily, the dripper stores flat, occupying a rectangular size smaller than 10 by 16 centimeters. Weighing just shy of 8 ounces, the brewer isn’t featherweight, but it offers great portability nevertheless. Contained in an unassuming synthetic case, the Pourigami seamlessly fits into any bag.
Honestly, I’m terrible at paper origami, but assembling this brewer into the pyramidal shape is a breeze. It only takes me about 20 seconds to slip the three indents into the respective slots — there’s no confusion to the construction. The completed dripper holds steady without a wobble and comes apart just as easily.
Put together, the interior forms a triangular pyramid shape that can accommodate any #2 cone-shaped filters. I find that Miir’s own filters, available for purchase online function most reliably, creating steady streams without slipping. Not to mention, the brand’s paper-based filters are compostable, a small but satisfying environmental win. A compatible filter is easy to find, making the Miir Pourigami simple to set up and get to brewing.
Advertisement
Brewing with the Pourigami
The Pouragami functions much like other pour-over coffee devices.
Miir
If you’ve used pour-over vessels like a V60, Chemex or a Kalita Wave, the Miir Pourigami is familiar territory. The dripper requires a hot water source, the aforementioned paper filter, and a cup or carafe to catch the coffee. A kitchen scale and thermometer help brew with utmost accuracy, but I’ve produced solid cups while eyeballing proportions outdoors.
It’s best to follow a brewing ratio to extract the most out of the brewer, especially when familiarizing yourself with its flow. The Miir brand suggests 21 grams for single-origin beans and 23 grams for blends, each extracted with 300 milliliters of water. Using water heated to 90 degrees Celsius, I’ve found these proportions reliable, as long as extraction is completed by the three-and-a-half-minute mark.
Advertisement
The Pouragami functions similarly to a Chemex but with better portability.
Taylor Martin/CNET
Compared to my V60, the Miir Pourigami takes longer to drain, so a coarser grind helps keep water moving. As a result, the vessel is best suited for full-bodied cups of medium- and dark-roasted coffees. The grind quality is crucial: You’ll want a coarse yet uniform consistency. I’ve used both the portable MiiR Coffee Hand Grinder and the Baratza Encoreto great success; I would avoid utilizing a blade grinder for this setup, though. Away from home, I’ve asked coffee shops to grind beans — I request a consistency one click coarser than a V60. Pre-ground coffee is the most convenient way to brew on the move.
The Pourigami’s steep interior makes saturating coffee grounds easy – no need to carry a gooseneck kettle alongside. I’ve used jet-boil-powered camping kettles and even cooking pots to make excellent cups of coffee, making sure to use hot water that’s off the boil. As long as I’m timing the process, using the Miir Pourigami is undemanding.
Advertisement
What does Pourigami filter coffee taste like?
The coffee I brew turns out light-bodied, but rarely weak or watery.
Nikita Ephanov/CNET
As with all pour-over coffee, the beans strongly influence the flavor. I’ve produced the best-tasting MiiR Pourigami cups using medium-roast blends — think grocery-store beans like Stumptown’s Holler Mountain Blend. Such bags respond well to the requisite coarser grind and are forgiving in outdoor scenarios. The Miir Pourigami translates gentle notes of acidity and sweetness, seldom leaning into burnt flavors. The coffee turns out light-bodied, similar to other filter setups, but I’ve never brewed a cup that tastes weak or watery.
If there’s one downside to this dripper, it’s that the coffee occasionally turns out too acidic, a sign of under-extraction. For this reason, I avoid brewing delicate light roasts with the Pourigami — not that I’m packing fancy beans for camping anyway. A bold, balanced medium roast cup hits the spot after a night in a tent.
Advertisement
Is the Pourigami worth it?
Coffee makers don’t get much simpler and more portable than the Pouragami.
Nikita Ephanov/CNET
At home, I’m not often assembling my Pourgami, instead settling on the trusty espresso machine or extracting delicate cups of V60. When I’m brewing outside of the house, though, the Pourigami is my top choice. In addition to camping, I’ll pack the brewer away in my suitcase for air travel, making the occasional cup on the go.
Before acquiring a model, I used to camp with a bulky plastic V60 dripper, which I inevitably fractured among camping equipment. The AeroPress certainly fares better in terms of durability, but it can be difficult to find a sturdy surface suitable for firm plunging. Compact and durable, the Miir Pourigami wins on logistical ease, making it easy to incorporate into a car-camping, backpacking, or even a bike-packing setup.
Advertisement
Sold for $35, it’s a thoughtfully designed coffee gadget that justifies the price point. The inventive design isn’t a trade-off for coffee quality. I’m happy to use the Pourigami several days in a row – the steel material is a breeze to clean. Whether at home or on the move, the brewer doesn’t occupy much space, making it a reliable favorite.
You must be logged in to post a comment Login