Feed aggregator

Amazon is reportedly testing package-delivery bots in one of its facilities. Will they talk to your Ring doorbell?

Artificial intelligence has been growing by leaps and bounds, fueled by excitement around AI’s current capabilities and potential to drive further efficiencies, improvements and innovation.

McKinsey last year reported that 65% of organizations regularly use GenAI, nearly double from 10 months prior, and a new 2025 McKinsey report indicates that three times more employees are using GenAI for a third or more of their work than their leaders imagine. Meanwhile, a new KPMG survey suggests that 68% of business executives expect to spend $50 million to $250 million on GenAI over the next 12 months – that’s up 45% from the first quarter of 2024.

But despite that growth, if AI were a person, it would still just be a child approaching puberty.

In AI’s next stage of life, it will learn more about what to do and not do, making it more understandable and predictable. AI is not yet completely reliable and trustworthy; according to a recent report, only one-third of U.S. businesses said the majority of the outputs of their AI models are accurate. As a result, businesses are still unsure if they should use AI to make key decisions and act independently, which means they must provide the proper care and feeding to ensure their AI is enterprise-ready.

Companies are also getting more pragmatic about AI. After investing big on AI experimentation, they now expect to drive real business results with artificial intelligence, so ROI is becoming very important. The recent excitement about China’s DeepSeek, which reportedly has capabilities on par with U.S. models but works at a fraction of the cost and requires far less energy, illustrates how important cost and sustainability considerations around AI have recently become.

At the same time, enterprises are keenly aware they must continue to innovate to remain competitive, and 2025 will usher in exciting new technologies to help enable that. All of that means now is the time to force AI to grow up faster so that it is production-ready for the enterprise while balancing that “enterpriseness” with innovation and business value.

Here’s a crib sheet on how to grow your AI into a trusted and enterprising young adult.

Go back to basics to make AI enterprise-ready

Governance, reporting and security are critical in enterprise environments. But these important considerations, and all they entail, are often overlooked or undervalued when it comes to AI.

As businesses make the leap from experimental to more production-level AI deployments, it is crucial for enterprises to address governance, reporting and security to meet compliance requirements, protect their own and their customer data, and built trust.

Given the level of complexity involved, that can be daunting. But it is essential to accelerate the maturity and adoption of enterprise-ready versions of AI. Understand that you don’t have to go it alone. Collaborate with a partner with deep expertise in technology and your sector, turnkey solutions, products with baked-in scalability and sustainability, and a methodical approach. Together, you can advance the “enterpriseness” of your AI efforts and drive real business value.

Create a solid data foundation for innovation

Data is critical to AI success. The more context AI has, the better results it can deliver. To get quality AI outputs, you need high-quality data. Otherwise, it’s garbage-in, garbage-out.

That’s pretty well understood at this point. But data quality is just part of the AI challenge. The fact that data exists in silos across your far-flung enterprise, and that the bulk of that data is now unstructured, can also interfere with data quality and your company’s ability to use data effectively. Inconsistencies in data collection and stewardship create further complications.

Embrace real-time data processing capabilities and enforce data governance frameworks to ensure your systems meet quality expectations. Employ AI-powered data cleaning tools that sift through massive amounts of data because doing that manually is simply impossible.

Use data catalogs and lineage tracking systems to make it faster and easier to access and understand your data. High-quality data will help ensure the explainability of AI outcomes, which is critical to meeting internal and external regulatory compliance requirements and instilling user trust.

“Trust starts with exposure and evolves with use,” as LinkedIn co-founder and venture capitalist Reid Hoffman writes in “Superagency: What Could Possibly Go Right with Our AI Future.” “Once you learn what something is and how it functions, you begin to trust it. Trust equals consistency over time. In the context of AI, we first must develop trust in the technologies themselves – no easy feat when the technologies are somewhat unpredictable and capable of error.”

You don’t necessarily have to resolve all of your data challenges immediately. However, having at least a basic understanding of your data estate and adopting these approaches and capabilities where and when you need goes a long way in building trust and enabling success.

Be ready for what’s next: Agentic AI

To date, businesses have relied on AI primarily to analyze data to uncover trends and make predictions as well as to automate routine tasks (chatbots in customer service, for example). Much of this work has been highly reactive and typically involves some human supervision.

But now we’re starting to hear more and more about this exciting – and potentially disruptive – evolution of artificial intelligence called agentic AI. As you’re probably already aware, agentic AI systems will be able to make decisions and act autonomously with minimal human intervention.

Agentic AI is a big leap forward and represents most people’s vision for AI. It works independently, takes initiative and self-optimizes. Agentic AI will drive increasing adoption of small language models (SLMs) and often involve the collaboration of smaller AI experts that focus on particular tasks based on their specialized training.

Yet, while agentic AI creates great opportunity, it comes with undeniable risks. That makes it even more critical to prioritize accountability, explainability and responsibility by building robust frameworks to govern these systems and reduce the potential for unintended consequences.

Guiding a human from childhood through their teen years to become a responsible adult who works hard to contribute to society requires ample time and attention. The same is true with AI.

With great power comes great responsibility – and the future is bright.

We've featured the best Larger Language Models (LLMs).

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

Imagine this: a whisper-quiet Rivian EV Amazon delivery truck pulls up to your front door, and before the driver even has a chance to step out, the back of the van briskly rolls up and out jumps a robot carrying a package to your doorstep.

It sounds like the stuff of science fiction, but that reality may be closer than we think, according to a new report from The Information, which says Amazon is already running tests.

Instead of humanoid robots taking the Amazon delivery crew's jobs, the humanoid robots would work in tandem with the human drivers. The robots would sit silently in the back, and when it's time for package delivery, they'd find the box, open the door, and carefully make their way to your home.

Amazon EV Delivery Truck: what if there was a robot in the back? (Image credit: Amazon)

Amazon is reportedly running test drives and deliveries in a specially created 'Humanoid Park', which we assume would help recreate some of the obstacles the humanoid robot might encounter, like cars, people, dogs, small wild animals, lawn sprinklers, porch lamps, porch swings... you get the idea.

The Information also reports that Amazon might be using large language models (LLMs) to help the humanoid robots deal with what is likely to be a hard-to-predict environment. Houses are not all cookie-cutter, after all. The doors, steps, lawns, and even distances to the front doors are all different. It's unlikely Amazon can program these robots to anticipate every unexpected obstacle, but LLMs can train them to handle a much wider variety without knowing in advance the specifics of each one.

There's no mention of the robots talking to Amazon customers, but with Amazon's investment in Anthropic's Claude AI, a chatty Amazon humanoid delivery bot seems possible.

Choose your bot

As for which robots Amazon might be using, it's already been testing Agility Robotics' Digit in a factory setting. I think it's unlikely, though, that Amazon uses that robot to complete home deliveries. Its odd appearance, which includes bird-like or backwards legs, might be somewhat off-putting for the average Amazon customer.

There's a larger issue with this kind of robot. I saw Digit in action at CES 2025 and, while impressive, it was painfully slow at most tasks. If you've ever seen your Amazon delivery person hop out of the truck, jump into the back, dash out with your package, and run to your front steps before capturing a photo for proof of delivery and then racing off to the next house, you know that an Amazon humanoid delivery robot better be fast.

Boston Dynamics all electric Altas can do a cartwheel, but could it deliver a package? (Image credit: Boston Dynamics)

There are other humanoid robot options out there from companies like Boston Dynamics and its all-electric Atlas. It's got speed and cartwheeling skills (see above), but there's no timeline for commercial adoption. Unitree G1 can do karate, but it's likely too small to carry larger Amazon packages.

According to the Information Report, there's a sense that the hardware may be the easy part, as, once they have the training in place, they can easily transfer it from one version of the Amazon delivery robot to another, ostensibly better one, all while improving that delivery algorithm.

I've asked Amazon for comment and will update this story if and when the company responds.

Amazon does have some robot delivery experience. It's been using drones (essentially flying robots) to deliver packages since 2022. Those mostly fly smaller packages and then carefully lower them to your backyard before flying off. It's not a service that Amazon offers everywhere.

None of this means Amazon's humanoid robots will be delivering packages anytime soon. Testing and development could take years, but if you add in the possibility of future driverless Rivian EVs, the combination of a self-driving car in the front and a ready-to-deliver robot in the back is a potentially cost-efficient combination for Amazon. How delivery personnel and customers might feel about it is another matter.

You might also like

• The best steak I've ever had was made by an AI robot grill
• Sorry, Figure AI, but your robot won't be cool until it can ...
• Robot abuse is still the path to the future – even if I hate it

It's the first known call between the two leaders since Inauguration Day — and the first time they've spoken since tariffs began ratcheting up.

(Image credit: Andy Wong)

As the various deadlines for net-zero approach, sustainability initiatives are climbing the boardroom agenda. In fact, two-thirds of businesses are expected to boost sustainability spending in 2025.

Cutting emissions will require a concerted effort from every department, and IT is an increasingly important voice in this conversation. Data storage, in particular, is emerging as a central element of an organization's efforts to decrease their carbon footprint. This is because a huge amount of data currently stored by UK businesses counts as “single-use data”. These are files, email conversations and other data that is generated but never used again – despite being stored where it continues to use energy and budget.

Recent NetApp research found that the overwhelming majority of IT leaders (92%) are aware of the environmental and financial impact of single-use data. The challenge is that AI – famously data and energy hungry – is a competitive advantage that businesses are unwilling to leave on the table.

As a result, organizations up and down the nation are facing a conundrum between the commercial and productivity benefits of AI, and staying on track with their sustainability progress to net-zero. So, how can businesses solve this challenge?

The invisible problem of data

When it comes to pollution, the environmental impact is often tangible and plain to see. Think of overflowing landfills and dirty rivers. But when the “landfill” resides in cyberspace, it’s mostly invisible and therefore easy to miss.

Single-use data might seem harmless, but the carbon footprint of unused data is a challenge that must be addressed; especially in our burgeoning AI age when 38% of data stored in the UK never gets accessed again after creation. This remains a substantial challenge for several reasons.

Firstly, around a third of IT leaders struggle to identify which data is single-use and which may be useful for future projects or training AI. On top of this, many are also feeling overwhelmed by the amount of accumulated data already, as sifting through stored archives requires more resource and budget than IT departments have available.

But challenging as this exercise may be, it is an important one. Not only because it will help cut emissions and save costs by reducing how much data businesses store, but an audit like this can also be massively beneficial in helping organizations get their house in order for AI.

Managing AI’s data appetite

It was, and still is, inevitable that AI will prompt an exponential rise in the data generated and stored. It is estimated that businesses will see their data estates grow by an average of 41% percent due to AI projects. For businesses that are already storing decades of data, this growth will be huge.

It’s easy to imagine a scenario of businesses hoarding data and plugging it all into AI tools as part of their implementation projects. Single-use data could become rife, costing businesses thousands as well as resulting in needless carbon emissions. Data management will be instrumental in avoiding this scenario. After all, the quality of the data used for AI is perhaps the most important factor in determining the success of AI initiatives.

Businesses must therefore ensure that their data is high quality, relevant, and useful before they begin their AI journey. Proper data infrastructure not only sets up AI projects for success, but it can help businesses stay on top of their emissions by enabling a strategic, deliberate and carefully considered approach to data growth.

Making storage more sustainable for AI

There are several approaches businesses can take to make their data storage more sustainable, even as AI implementation accelerates. Data lifecycle management policies are a fantastic place to start, which can include automated tiering and archiving of less frequently accessed datasets to more energy-efficient storage solutions.

This significantly reduces energy consumption. Elsewhere, data compression algorithms and deduplication efforts can help too. By reducing the physical storage footprint, businesses can reduce emissions by lowering the associated energy demands for both storage and retrieval.

Companies can reduce emissions by reforming their physical infrastructure too. For example, they can opt for energy-efficient data centers, choosing those with the latest cooling technologies that minimize power consumption and use renewable energy sources. Optimizing data center locations to take advantage of cooler climates or readily available green energy can also significantly reduce the carbon footprint.

Finally, businesses should not overlook the importance of responsible hardware lifecycle management. The reuse, refurbishment, and proper recycling of storage components is crucial to minimize electronic waste, further contributing to sustainability objectives.

Finding balance in a complex landscape

Ultimately, the convergence of net-zero targets and the promise of AI puts businesses at a critical juncture. The unchecked accumulation of data for AI’s sake risks undermining hard-won sustainability progress.

To navigate this challenge successfully, UK businesses need to find their own balance between innovation and responsibility. This begins with a commitment to data hygiene – regularly reviewing what is stored and why it is being kept – all while exploring various infrastructure options that align with sustainable practices.

Different organizations will require different approaches depending on their sector, size, and AI goals. What is clear is that the foundation upon which AI initiatives are built must incorporate energy efficiency and sustainability from the beginning. By taking a thoughtful, measured approach, businesses can harness the transformative power of AI while contributing towards a sustainable digital future.

We've featured the best green web hosting.

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

Looking for a different day?

A new Quordle puzzle appears at midnight each day for your time zone – which means that some people are always playing 'today's game' while others are playing 'yesterday's'. If you're looking for Thursday's puzzle instead then click here: Quordle hints and answers for Thursday, June 5 (game #1228).

Quordle was one of the original Wordle alternatives and is still going strong now more than 1,100 games later. It offers a genuine challenge, though, so read on if you need some Quordle hints today – or scroll down further for the answers.

Enjoy playing word games? You can also check out my NYT Connections today and NYT Strands today pages for hints and answers for those puzzles, while Marc's Wordle today column covers the original viral word game.

SPOILER WARNING: Information about Quordle today is below, so don't read on if you don't want to know the answers.

Quordle today (game #1229) - hint #1 - VowelsHow many different vowels are in Quordle today?

• The number of different vowels in Quordle today is 5*.

* Note that by vowel we mean the five standard vowels (A, E, I, O, U), not Y (which is sometimes counted as a vowel too).

Quordle today (game #1229) - hint #2 - repeated lettersDo any of today's Quordle answers contain repeated letters?

• The number of Quordle answers containing a repeated letter today is 2.

Quordle today (game #1229) - hint #3 - uncommon lettersDo the letters Q, Z, X or J appear in Quordle today?

• No. None of Q, Z, X or J appears among today's Quordle answers.

Quordle today (game #1229) - hint #4 - starting letters (1)Do any of today's Quordle puzzles start with the same letter?

• The number of today's Quordle answers starting with the same letter is 0.

If you just want to know the answers at this stage, simply scroll down. If you're not ready yet then here's one more clue to make things a lot easier:

Quordle today (game #1229) - hint #5 - starting letters (2)What letters do today's Quordle answers start with?

• E

• K

• V

• S

Right, the answers are below, so DO NOT SCROLL ANY FURTHER IF YOU DON'T WANT TO SEE THEM.

Quordle today (game #1229) - the answers

(Image credit: Merriam-Webster)

The answers to today's Quordle, game #1229, are…

• ELUDE
• KHAKI
• VISTA
• SMOKY

When I have four letters in a word and I’m searching for the fifth I try to put together common letter combinations with what I have. It doesn’t always pay off, but today i had I-S-T-A and L-U-D-E locked in, which made finding VISTA and ELUDE easier.

Meanwhile, for KHAKI it was just trying to think of words that included a K and a H – a very small gang with the letters I had available.

How did you do today? Let me know in the comments below.

Daily Sequence today (game #1229) - the answers

(Image credit: Merriam-Webster)

The answers to today's Quordle Daily Sequence, game #1229, are…

• SCION
• MISSY
• ARISE
• CUMIN

Quordle answers: The past 20

• Quordle #1228, Thursday, 5 June: CHIDE, RABBI, GUSTY, LANCE
• Quordle #1227, Wednesday, 4 June: BANAL, STOUT, SEDAN, HIPPO
• Quordle #1226, Tuesday, 3 June: FUGUE, SYRUP, FLACK, WORST
• Quordle #1225, Monday, 2 June: THINK, BELLE, CRONE, BOULE
• Quordle #1224, Sunday, 1 June: POINT, MERIT, WHOOP, APHID
• Quordle #1223, Saturday, 31 May: CRUMB, ELFIN, DRIER, QUITE
• Quordle #1222, Friday, 30 May: RAJAH, CAUSE, BLACK, ETUDE
• Quordle #1221, Thursday, 29 May: CRIER, DRAPE, STRUT, NEIGH
• Quordle #1220, Wednesday, 28 May: HELLO, BEADY, VIGIL, PURER
• Quordle #1219, Tuesday, 27 May: TWEET, RANGE, POPPY, RADAR
• Quordle #1218, Monday, 26 May: BLEAT, HOWDY, ASIDE, SCOOP
• Quordle #1217, Sunday, 25 May: OCEAN, AMBER, PIPER, GLEAN
• Quordle #1216, Saturday, 24 May: HUSKY, HEIST, FOGGY, POLAR
• Quordle #1215, Friday, 23 May: SHIRE, GIANT, AWAIT, CAPER
• Quordle #1214, Thursday, 22 May: LOSE, GLOVE, STINT, EXCEL
• Quordle #1213, Wednesday, 21 May: NOVEL, CHOSE, DIRTY, DONUT
• Quordle #1212, Tuesday, 20 May: DECOY, SHAKE, MAPLE, PURER
• Quordle #1211, Monday, 19 May: LINK, HANDY, DITCH, WAIVE
• Quordle #1210, Sunday, 18 May: QUACK, ROACH, PURGE, DOWNY
• Quordle #1209, Saturday, 17 May: STRIP, RANGE, UNITE, GEESE

Looking for a different day?

A new NYT Strands puzzle appears at midnight each day for your time zone – which means that some people are always playing 'today's game' while others are playing 'yesterday's'. If you're looking for Thursday's puzzle instead then click here: NYT Strands hints and answers for Thursday, June 5 (game #459).

Strands is the NYT's latest word game after the likes of Wordle, Spelling Bee and Connections – and it's great fun. It can be difficult, though, so read on for my Strands hints.

Want more word-based fun? Then check out my NYT Connections today and Quordle today pages for hints and answers for those games, and Marc's Wordle today page for the original viral word game.

SPOILER WARNING: Information about NYT Strands today is below, so don't read on if you don't want to know the answers.

NYT Strands today (game #460) - hint #1 - today's themeWhat is the theme of today's NYT Strands?

• Today's NYT Strands theme is… String's attached

NYT Strands today (game #460) - hint #2 - clue words

Play any of these words to unlock the in-game hints system.

• SOYA
• TRAVEL
• BACK
• WART
• LOVE
• TOYS

NYT Strands today (game #460) - hint #3 - spangram lettersHow many letters are in today's spangram?

• Spangram has 9 letters

NYT Strands today (game #460) - hint #4 - spangram positionWhat are two sides of the board that today's spangram touches?

First side: bottom, 3rd column

Last side: top, 3rd column

Right, the answers are below, so DO NOT SCROLL ANY FURTHER IF YOU DON'T WANT TO SEE THEM.

NYT Strands today (game #460) - the answers

(Image credit: New York Times)

The answers to today's Strands, game #460, are…

• SEASICK
• SLEEPER
• PINWHEEL
• ELEVATOR
• BREAKAWAY
• SPANGRAM: YOYO TRICK

• My rating: Hard
• My score: 2 hints

I really struggled today and had no idea what we were searching for, even after using two hints – SEASICK and SLEEPER seemed to suggest travel.

My third correct word, PINWHEEL, confused me even more; was it referring to the small wheels of a wheelie suitcase?

ELEVATOR and BREAKAWAY didn’t help much either, both still hinting at some vaguely travel-related group of words.

It wasn’t until the very end when I got the spangram that I finally understood what “strings attached meant”. Bravo Strands, bravo!

How did you do today? Let me know in the comments below.

Yesterday's NYT Strands answers (Thursday, June 5, game #459)

• COCKLESHELLS
• PRETTY
• MAIDS
• SILVER
• BELLS
• CONTRARY
• SPANGRAM: GARDEN

What is NYT Strands?

Strands is the NYT's not-so-new-any-more word game, following Wordle and Connections. It's now a fully fledged member of the NYT's games stable that has been running for a year and which can be played on the NYT Games site on desktop or mobile.

I've got a full guide to how to play NYT Strands, complete with tips for solving it, so check that out if you're struggling to beat it each day.

Looking for a different day?

A new NYT Connections puzzle appears at midnight each day for your time zone – which means that some people are always playing 'today's game' while others are playing 'yesterday's'. If you're looking for Thursday's puzzle instead then click here: NYT Connections hints and answers for Thursday, June 5 (game #725).

Good morning! Let's play Connections, the NYT's clever word game that challenges you to group answers in various categories. It can be tough, so read on if you need Connections hints.

What should you do once you've finished? Why, play some more word games of course. I've also got daily Strands hints and answers and Quordle hints and answers articles if you need help for those too, while Marc's Wordle today page covers the original viral word game.

SPOILER WARNING: Information about NYT Connections today is below, so don't read on if you don't want to know the answers.

NYT Connections today (game #726) - today's words

(Image credit: New York Times)

Today's NYT Connections words are…

• CONVENTION
• REPORTER
• SHADOWBOXER
• CRIMINAL
• STANDARD
• NEWSHOUND
• PHOTOGRAPHER
• ROUTINE
• EDITOR
• REGULAR
• TRENDSETTER
• DRAWBACK
• SWINDLE
• SNICKERDOODLE
• COLUMNIST
• COMMON

NYT Connections today (game #726) - hint #1 - group hints

What are some clues for today's NYT Connections groups?

• YELLOW: New York Times employees
• GREEN: Nothing unusual
• BLUE: There’s pros and there’s…?
• PURPLE: Canines

Need more clues?

We're firmly in spoiler territory now, but read on if you want to know what the four theme answers are for today's NYT Connections puzzles…

NYT Connections today (game #726) - hint #2 - group answers

What are the answers for today's NYT Connections groups?

• YELLOW: NEWSPAPER JOBS 
• GREEN: EVERYDAY 
• BLUE: WHAT "CON" MIGHT MEAN 
• PURPLE: ENDING WITH KINDS OF DOGS

Right, the answers are below, so DO NOT SCROLL ANY FURTHER IF YOU DON'T WANT TO SEE THEM.

NYT Connections today (game #726) - the answers

(Image credit: New York Times)

The answers to today's Connections, game #726, are…

• YELLOW: NEWSPAPER JOBS COLUMNIST, EDITOR, PHOTOGRAPHER, REPORTER
• GREEN: EVERYDAY COMMON, REGULAR, ROUTINE, STANDARD
• BLUE: WHAT "CON" MIGHT MEAN CONVENTION, CRIMINAL, DRAWBACK, SWINDLE
• PURPLE: ENDING WITH KINDS OF DOGS NEWSHOUND, SHADOWBOXER, SNICKERDOODLE, TRENDSETTER

• My rating: Easy
• My score: 1 mistake

My only mistake today was including NEWSHOUND rather than PHOTOGRAPHER among the list of NEWSPAPER JOBS. I can see now that it was a trap, but my thinking was that the other possibilities were all about words rather than images. 

Making this mistake did, however, throw open the ENDING WITH KINDS OF DOGS group as I spotted setter (TRENDSETTER) and boxer (SHADDOWBOXER) along with hound.

From here the green EVERYDAY group was easy to pull together – you could almost say it was ROUTINE, which was fortunate as the blue group was today’s most tricky, although obvious in the aftermath of a completed game.

How did you do today? Let me know in the comments below.

Yesterday's NYT Connections answers (Thursday, June 5, game #725)

• YELLOW: PRECIPICE BRINK, CUSP, EVE, VERGE
• GREEN: PLUMBING EQUIPMENT PIPE, PLUNGER, SNAKE, WRENCH
• BLUE: WAYS TO MODIFY A CAR'S EXTERIOR BUMPER, GRILLE, RIM, SPOILER
• PURPLE: BEST FEMALE ROCK PERFORMANCE GRAMMY WINNERS APPLE, CROW, SUMMER, TURNER

What is NYT Connections?

NYT Connections is one of several increasingly popular word games made by the New York Times. It challenges you to find groups of four items that share something in common, and each group has a different difficulty level: green is easy, yellow a little harder, blue often quite tough and purple usually very difficult.

On the plus side, you don't technically need to solve the final one, as you'll be able to answer that one by a process of elimination. What's more, you can make up to four mistakes, which gives you a little bit of breathing room.

It's a little more involved than something like Wordle, however, and there are plenty of opportunities for the game to trip you up with tricks. For instance, watch out for homophones and other word games that could disguise the answers.

It's playable for free via the NYT Games site on desktop or mobile.

• Silent Hill f officially launches on September 25 for PS5, Xbox Series X, Xbox Series S, and PC
• Konami released a new PS5 gameplay trailer showcasing combat and a variety of monsters
• Pre-orders are now open

Silent Hill f, the next mainline game in Konami's horror series, officially launches on September 25 for PS5, Xbox Series X, Xbox Series S, and PC.

The announcement comes from PlayStation's June State of Play showcase, where Konami shared a chilling story trailer featuring PS5 gameplay and a first look at third-person combat.

Silent Hill f follows Shimizu Hinako, a high school student struggling under society's pressure of expectations, as she navigates her hometown of Ebisugaoka, which has become shrouded in fog and is undergoing a disturbing transformation.

"Players will have to unravel mysteries, battle otherworldly threats, and come face-to-face with choices that will determine Hinako’s fate," Konami said.

The trailer, which you can watch below, shows the beginning of the transformation as Shimizu runs through her town, encountering various disturbing monsters, while also using a series of melee weapons to fend them off.

(Image credit: Konami)

Konami has previously stated that Silent Hill f is a standalone title in the Silent Hill timeline, suggesting it won't have any connection to any previous games in the series, and that newcomers will need no prior knowledge of the franchise to experience the horror game.

Silent Hill f will receive various editions at launch, including a physical Day One Edition, Digital Standard, and Digital Deluxe Editions.

The Digital Standard and Deluxe Editions of Silent Hill f will feature digital bonuses including outfits, soundtrack samples, artbook, and 48 hours of early access.

Both the Digital Standard Edition and physical Day One edition are now available to pre-order and include the early purchase bonus items, while the Digital Deluxe Edition comes as a digital download.

Physical pre-orders for the Day One Edition for PS5, Xbox Series X, and Xbox Series S at select retailers will also offer a steelbook.

You might also like...

• Best PS5 games 2025 - top PlayStation 5 titles to play right now
• The first trailer for 007 First Light reveals a young James Bond and it's coming to PC and console in 2026, including Nintendo Switch 2
• Nintendo Switch 2 launch day live: all the latest news and updates as the console starts arriving with fans and at retailers

• Passion.io, a major no-code app-building app, operated a non-password-protected database
• The archive contained millions of records, with a total size of around 12TB
• It was since then locked down, but users should still take care

Millions of records containing sensitive, personally identifiable information, were sitting online in yet another unencrypted, non-password-protected database, experts have warned.

Found by security researcher Jeremiah Fowler, who discovered and reported his findings to vpnMentor, the database contained 3,637,107 records, and was 12.2TB in total size.

It belongs to a company called Passion.io, a Delaware-based no-code app-building platform that allows creators, influencers, entrepreneurs, and coaches, to create websites without having any prior coding knowledge. They can also create, and sell, interactive courses.

Save up to 68% for TechRadar readers

TechRadar editors praise Aura's upfront pricing and simplicity. Aura also includes a password manager, VPN, and antivirus to make its security solution an even more compelling deal.View Deal

Locking the archive down

Fowler said that he analyzed a “limited sampling of the exposed documents” and saw internal files, images, and spreadsheet documents marked as “users” and “invoices”.

These files contained people’s names, email addresses, postal addresses, and details about payments or payouts for users and app creators.

This type of information is a treasure trove for cybercriminals. They can use it to create convincing phishing emails, tricking Passion’s users into making rash, dangerous decisions. Besides phishing, the data can be used in identity theft, wire fraud, and other types of scams.

The researcher notified Passion.io about his findings, and got a response on the same day. The database was locked down, and the company confirmed it was working on putting guardrails in place so that mishaps like this one don’t repeat.

“We’re treating this very seriously and moving fast,” the company told Fowler.

So far, there is no evidence the information is circulating on the dark web - and it's also not known if Passion.io is the one managing the database, or if the job was outsourced to a third party.

Without a thorough investigation, there is no way of knowing for how long the database remained open, or if any threat actors found it already.

You might also like

• Entire Brazilian population potentially put at risk by major data leak
• Take a look at our guide to the best authenticator app
• We've rounded up the best password managers

The best mattresses in a box eliminate the hassle of mattress shopping because they’re delivered straight to your home. These are the top options for you.

• Keanu Reeves didn't know Norman Reedus had been cast in From the World of John Wick: Ballerina
• The pair are good friends due to their shared obsession with motorsports
• Reedus was surprised that the franchise's lead star and executive producer had no idea about his involvement

Norman Reedus has revealed – amusingly – that Keanu Reeves didn't know Reedus had been cast in From the World of John Wick: Ballerina.

The Walking Dead star told TechRadar of his own surprise that Reeves had no idea about Reedus' involvement in the John Wick universe's latest entry. In fact, Reedus joked it was "weird" that the man synonymous with the gun-fu franchise – Reeves plays the franchise's titular hitman, and is an executive producer on its various projects – was seemingly kept in the dark about her hiring.

Keanu Reeves had no idea that his good friend and fellow motorbike enthusiast was in Ballerina (Image credit: Lionsgate)

Reedus and Reeves have become close friends due to their mutual love of motorsports and, in particular, motorcycles. Indeed, Reeves is as well-known for being a motorbike obsessive – so much so he co-founded his own manufacturing company in 2011. Meanwhile, Reedus regularly stars in his own travel docu-series – Ride with Norman Reedus – on US network AMC, which began in 2016 and has run for six seasons so far.

You would think, then, that Reeves would've been aware of Reedus' casting in Ballerina. Not so, the latter revealed to me when I asked if he sought advice from Reeves upon joining the Wick-Verse.

"I didn't speak to Keanu [beforehand], but I texted him after I got it," Reedus said. "He was like, 'Wait, you're in it?' I don't think he knew, which was weird to me because I thought he would!"

When I followed up to ask if that put additional pressure on Reedus to deliver a top-tier performance, he replied: "Yeah, but only because it's a John Wick movie. You don't want to be the guy who messes up the John Wick franchise. I'm a big fan of all of those films, so you want to do as good of a job as you can."

Who is Daniel, the enigmatic character played by Norman Reedus in Ballerina?

We hear you loud and clear, Mister Reedus, but I imagine fans will want to know more about Daniel, i.e. the character you're playing in the fourth John Wick film (well, chronologically, anyway – for more information on the movie series' timeline, read my guide on how to watch the John Wick movies in order).

So, without spoiling anything significant, what could Reedus tell me about the mystery individual he'll portray in Ballerina?

"He comes from one these... he has a history with a militant faction," Reedus said as he carefully considered his response. "He's someone with some sort of familial privilege, so I imagined how he grew up, the training he had to do, and then all the bad things he's probably done in his life.

"But, then there's this new element that changes Daniel's perspective on who he is and what he's doing," he added, "And so, by the time you meet him in Ballerina, he's basically willing to fight everybody to protect something other than himself."

You can find out what I thought of Reedus' character, as well as the movie as a whole, in my Ballerina review. Before one of 2025's new movies arrives in theaters on June 6, be sure to catch up on more of my exclusive coverage about it and the wider action thriller franchise below.

You might also like

• 'It's really misleading': Ballerina director Len Wiseman explains why the John Wick series' new movie shouldn't be called a spin-off film
• From the World of John Wick: Ballerina was delayed a whole year for a highly unusual but 'awesome' reason, its director says
• John Wick Chapter 5 is still in very early development, but 'there are a lot of options' on the table about what its story could be, Ian McShane teases

• ChatGPT now has three million paying customers, an increase of a million in just four months
• New "connectors" look to further integrate popular business tools with ChatGPT
• New Record Mode can listen in to your calls and offer suggestions and actions

OpenAI has revealed a range of new expansions for ChatGPT as it looks to increasingly target business users.

The AI chatbot will now be able to integrate with a number of widely-used business software tools, including the likes of cloud storage and collaboration tools such as Dropbox, Box, Sharepoint, OneDrive, and Google Drive.

These new "connectors" will allow users on ChatGPT business tiers to quickly find the data and information they need without leaving the platform, boosting productivity and saving time at work.

ChatGPT connectors

The company says connectors will help workers get quicker access to their company's key data, for example, a researcher could use the Box connector to request quarterly sales metrics from PDFs or spreadsheets stored in Box, which ChatGPT will then structure and clearly present the data from those documents, with citations, all while keeping to your organizations' existing permissions.

Along with connectors, a new Record mode in ChatGPT allows users to record and transcribe meetings or brainstorming sessions, with the tool also able to automatically generate notes, provide timestamped citations, and offer AI-powered suggestions.

The feature will integrate with your existing documents and saved files, making it easier to find out who suggested past decisions, or create follow-up actions, with sessions able to be transformed into actionable items such as a Canvas document.

And alongside basic connectors, ChatGPT is also seeing the beta launch of deep research connectors.

Available now with Github, HubSpot, Linear, as well as many popular Microsoft and Google tools, these tools are able to carry out in-depth research by trawling the web and third-party tools, so users can combine external data with company insights to produce a more comprehensive set of findings similar to a research analyst’s report.

The launch comes as ChatGPT announces it now has over three million paying customers - an increase from two million just four months ago.

OpenAI now hopes these new connectors will help boost adoption even further, with demand for AI tools growing in industries across the world.

"Every organization holds vast knowledge, but it’s often trapped in silos: sales teams have customer insights, product teams hold market data, and engineering holds technical expertise," Open AI said in its release.

"Our goal is to evolve ChatGPT into a platform that unlocks your organization's entire knowledge base—enabling each employee to continuously leverage this knowledge."

The new connector tools will be available across the ChatGPT Enterprise, ChatGPT Team and ChatGPT Edu tiers now at no extra cost - however Record Mode will only be available for Team users for the time being.

You might also like

• ChatGPT spends 'tens of millions of dollars' on people saying 'please' and 'thank you'
• ChatGPT remains the most popular AI tool in offices worldwide
• OpenAI launches a version of ChatGPT just for governments

• Nothing Headphone (1) is coming
• Nothing's big event is on July 1st
• The Nothing Phone (3) will appear at the same event

Nothing has just officially revealed when its first pair of headphones, rather than earbuds, are coming – and they'll arrive on July 1st, alongside the new flagship Nothing Phone (3).

They'll be called the Nothing Headphone (1) – yes, a singular headphone, apparently – and I'm really looking forward to these after the superb Nothing Ear (a).

These affordable earbuds really turned Nothing around when it comes to audio – before the brand's stuff had been distinctly fine, but the Nothing Ear (a) are simply fantastic for the price. We rank them among the the best earbuds around, and they pushed Sony out as our top pick in the sub-$100/£100 category.

Nothing has also announced that it's working with British hi-fi stalwart KEF on future audio products, but declined to say whether KEF is involved in the Headphone (1) in any way.

The Nothing Ear (a) buds – funky and very functional (Image credit: Future)

I suspect that KEF is not involved in the Headphone (1), and anything coming from that project is further down the road – but maybe I'll be surprised.

And either way, I'm excited for these. If Nothing can hit the same kind of mid-range price and great quality for a pair of noise-cancelling headphones as it has for earbuds, they could shake up the affordable end of the market.

And not only that, but they'll probably come with Nothing's cool see-through aesthetic, and as someone who grew up during the translucent tech craze of the early 2000s, I'm always happy to see more of it.

We have no details to speak of about these headphones, but at least we know what date to mark in the calendar to find out more about them: see you on July 1st.

You might also like

• Skullcandy launches its first premium ANC headphones to truly challenge Sony and Bose, and their cool cross-body travel bag is a great start
• The first Bluetooth 6.0 headphones are here, they're super-cheap, and they weigh less than 100g
• A Japanese audio company just launched 'world-first' triple-hybrid driver headphones, with an amazing 140 hours of battery life – plus a special edition version with J-pop star Hyde

The RedMagic 10S Pro is turning in some of the highest benchmark scores I've ever seen on an Android phone.

The actor just wrapped season 1 of his talk show on the streamer.

Looking to learn a new language or refresh your skills before traveling this summer? Here are the best language learning apps of 2025.

Commentary: It doesn't matter if Apple drops more AI at WWDC next week. I'm never turning my camera control button back on, and you should join me.

Anthropic has started a new blog called Claude Explains, discussing the capabilities of its AI models and written by that self-same AI model. The educational posts are “written” by Claude to explain how to use Claude. It's like an AI's personal diary, but with debugging tips instead of romantic exploits.

The blog is pitched as a “corner of the Anthropic universe where Claude is writing on every topic under the sun,” but that's not quite accurate. Claude may draft the pieces, but a team of human experts and editors sand and polish the rough outline to make sure they are readable and accurate, or as Anthropic calls it, a “collaborative approach."

Now, this idea isn't terrible on its face. This kind of AI-human tag team makes a lot of sense, at least when the AI is writing about itself. An article about how Claude can design a website or organize a financial report is well within Claude's wheelhouse. It's just explaining its own abilities. But a technically reasonable explanation and a few useful examples aren't a full blog post. Claude’s best work still won't always result in a coherent article, or one that a real person would want to read.

Anthropic is honest about how humans are part of the process throughout blog post production. Claude may start the car, but humans are at the wheel and navigating, lest it drive the article right into a ditch full of hallucinations and mixed metaphors. Anyone who’s used AI without guardrails knows this scenario isn’t far-fetched. AI is excellent at saying things that sound right until you try to actually apply them.

(Image credit: Anthropic)AI ghostwriting

Collaboration is certainly an efficient approach. Claude can crank out thousands of words without breaking a sweat, and if you’re using it to explain the same concepts it was trained on, it’s got a decent shot at getting things mostly right. Problems arise much more quickly when AI writers are left unsupervised, especially on subjects outside of the AI model's abilities.

The blog doesn’t proclaim the human element, so a casual reader might assume Claude is doing all the writing. That’s a branding choice, and not a neutral one. It creates a kind of halo effect, subtly bragging about how the AI breaks down data analysis and sounds like a real writer. Except it isn’t human. It’s a word blender that gets better results when someone else chooses the ingredients and adjusts the settings. And that distinction matters, especially as more people begin to trust AI-generated information in contexts far beyond technical blogs.

There's a steady stream of stories about media outlets embarrassing themselves by believing AI can replace entire content teams. The Chicago Sun-Times published AI-generated book recommendations for titles that didn’t exist, and multiple outlets have published AI-written features full of errors. And that's not even counting Apple's attempts at news summary headlines.

Claude Explains feels downright reasonable by comparison. If you’re going to use AI to produce content for public consumption, maybe keep to what it knows best. And don't leave out the humans.

You might also like

• Anthropic's new Claude 4 models promise the biggest AI brains ever
• How Claude’s 3.7's new ‘extended' thinking compares to ChatGPT o1's reasoning
• I tried Claude's new Research feature, and it's just as good as ChatGPT and Google Gemini's Deep Research features

• Intel's Nova Lake S desktop processors are rumored to utilize two separate iGPU architectures
• Xe3 will reportedly be used for graphic workloads, while Xe4 will be used for media-related tasks
• It could be another sign of processors with powerful iGPUs replacing budget discrete GPUs

AMD and Intel are on the frontlines battling it out in the CPU market for mobile and desktop platforms, and Team Blue's upcoming Nova Lake processors may offer a unique and rare architecture setup.

As reported by Wccftech, Intel's Nova Lake S desktop processors will reportedly feature a hybrid iGPU architecture, using Xe3 for graphics and Xe4 for display and media-related tasks, according to the reliable leaker @jaykihn0 on X. The rumor suggests that Intel intends to utilize Xe3's full potential in graphical workloads, while Xe4 could be essential for power efficiency.

It will be interesting to see whether the Nova Lake S processors will have the same effect as AMD's Ryzen AI Max processors (which are actively making their way to desktop systems), in terms of their iGPU capabilities. Notably, the Ryzen AI Max+ 395 provides great performance at both 1080p and 1440p in games like Cyberpunk 2077.

Regardless, if this rumor is legitimate, it looks like Team Blue is taking a big stride with GPU power across the board, especially with rumors of a new Arc B770 16GB GPU. It also raises the question once again; if Nova Lake S and Panther Lake both provide great iGPU performance for desktop and mobile platforms, respectively, could we be seeing the beginning of the end for discrete GPUs?

Analysis: Maybe it's almost the end for budget discrete GPUs...

(Image credit: Intel)

Now, let's pump the brakes a little here. There's almost no way any SoC or iGPU will be able to match high-end GPUs like the RTX 4090 or RTX 5080, at least not any time soon.

However, we've already got AMD's strongest APU that has surpassed the likes of the RTX 4060 and even the RTX 4070 in some gaming scenarios – so at the very least, I think budget low-end discrete GPUs are in danger.

Perhaps the only factor saving them is that they are indeed budget GPUs, but the proposition of a PC without a discrete GPU taking up space may be a tempting one for gamers.

It's worth noting that Strix Halo APUs were reserved for laptops and other mobile platforms, but have gradually made their way to desktops, specifically present in mini-PC builds.

It's also no secret that some of the budget GPUs like the RTX 5060 (based on reviews), aren't good enough for consumers with their 8GB configurations – so if desktop processors, SoCs, or AMD's APUs can continue to match or outperform them, I wouldn't be surprised to see a decent portion of low-end discrete GPUs fade away...

You may also like...

• Intel's Computex 2025 Panther Lake demos showcase power efficiency focus – and also hint at strong gaming performance
• Intel is seemingly preparing a huge Battlemage GPU reveal at Computex 2025 – and it's expected to be a higher-end offering
• Intel's Arc B770 GPU could go head-to-head with Nvidia's RTX 5060 – and it's expected to be revealed at Computex