TechRadar News

OpenAI's custom AI chip isn't just another attempt to loosen Nvidia's grip on AI hardware. It's the clearest sign yet that OpenAI is adopting the same vertically integrated strategy that transformed Apple over the past decade.

When OpenAI and Broadcom recently shared new details about Jalapeño, their custom inference processor, most of the discussion focused on Nvidia. Nvidia currently sits at the center of the AI industry, supplying the graphics processors that power everything from ChatGPT to image generators and coding assistants. Any attempt to reduce that dependence is naturally headline news.

For years, Apple has enjoyed a competitive advantage from making the most important parts of its products in-house. Instead of relying on someone else's processors or designing software around third-party hardware, it designed and built its own hardware and software. Competitors spent years trying to match that integration.

With its new custom inference processor, OpenAI appears to be building more than just an alternative chip. It's developing the same kind of vertically integrated ecosystem that helped transform Apple into one of the world's most valuable companies.

The chip is only part of the plan

When Apple introduced its M-series processors, the company aimed to build Macs that woke instantly and ran cool and quiet. Customers cared that everything simply felt smoother. OpenAI appears to be chasing a similar goal, even if the product is completely different.

Instead of laptops, it wants conversations that arrive faster. Building its own processor gives it another lever to pull that competitors relying entirely on third party hardware simply do not have.

Jalapeño is simply another piece of a much larger puzzle. The processor has been designed for inference rather than training. Training is the expensive process of creating an AI model as opposed to the inference done afterward. Every time someone asks ChatGPT a question, that's inference. Those billions of everyday interactions eventually become just as important as building the model itself because they determine both performance and operating costs.

Designing a processor specifically for those workloads gives OpenAI something that off-the-shelf hardware never fully can. It can begin tailoring the hardware around exactly how its own models think and respond, a more efficient method. And every improvement, whether in power consumption, speed, or networking, saves money and improves the AI experience.

OpenAI has been careful not to oversell the timeline, with broad deployment of the new chip still some way off. This is the beginning of a strategy rather than the final result.

Nvidia's challenge

Nvidia isn't going to panic right now, nor should it. Its processors still power much of today's AI boom. Demand continues to outstrip supply in many areas, and OpenAI itself remains one of its major customers. None of that changes because one new custom processor has appeared on the roadmap. What should catch Nvidia's attention is the pattern beyond OpenAI.

Google has spent years developing Tensor Processing Units. Amazon created Trainium and Inferentia. Microsoft has invested heavily in its own AI chips, as has Meta in custom accelerators for its expanding AI ambitions. OpenAI is now following the same path. Different companies have different technical goals, but they all seem to arrive at the same conclusion: as AI becomes a bigger part of their business, they don't want to depend entirely on someone else's hardware.

Of course, Apple designing its own processors certainly did not destroy Intel overnight. But there was a shift as Apple gained more control over pricing and product direction each time it replaced an external component with one of its own. The same could happen with AI.

Plus, OpenAI said its own AI models helped accelerate parts of the engineering process during chip development. AI is actually helping to make the hardware that will power its future iterations. That feedback loop may become increasingly important as chip design grows more complex. The future of AI may belong to the companies that own as much of the underlying machine as possible, regardless of where the models themselves rank.

If Apple's history is anything to go by, OpenAI is ready to be that company.

• OpenClaw has launched iOS and Android companion apps
• Users can remotely control their self-hosted AI agents
• Initial reviews say the apps could do with a bit of refinement

OpenClaw has released its first official mobile apps, giving users an alternative to earlier reliance on messaging platforms like Telegram and WhatsApp.

Importantly, the iOS and Android apps act as companions to an existing OpenClaw installation rather than standalone assistants, much like the Codex controller OpenAI has embedded into the ChatGPT app.

The apps connect to a self-hosted OpenClaw Gateway running on their own hardware, rather than in OpenClaw's cloud, which has been key to the startup's success.

OpenClaw now available for mobile

OpenClaw described three core functionalities for the mobile apps: "native mobile apps, finally; agents in your pocket; and channels, tasks, replies on the go."

Users will be able to receive and approve actions requested by the AI remotely, and get status updates for ongoing workflows. Remote action triggering and managing are also available.

The AI agent gained popularity due to its local-first stance, allowing users to control their own Gateway and ultimately reduce their reliance on Big Tech and insecure cloud.

But while the tool itself proved a hit, selling out Mac minis across multiple regions after users scrambled to buy cost-effective hardware to run the agent, its mobile app has been a bit more of a flop, at least to begin with.

Its Android version in particular has attracted criticism over its poor interface and usability – cosmetic niggles that don't detract from its core functionality.

9to5Google reports many initial reviews are mostly negative, with the app currently on a 2.2 star rating, amid multiple reports of the app being buggy, with users calling it “unusable,” unable to pair, and “the worst app I’ve ever used in my entire life.”

The apps were notably published by the OpenClaw Foundation – an open-source project that OpenAI now supports following its hiring of OpenClaw founder Peter Steinberger, who stressed the project would remain "open and independent."

• A leaked video shows an iPhone 18 Pro being drop-tested
• If it's real, then this is a massive leak to occur so far from launch
• But the internet is mostly expressing mild disappointment at the design of the phone

There have been many iPhone leaks over the years, but undoubtedly the biggest concerned the iPhone 4, when a prototype unit of that device was left in a bar and subsequently purchased by Gizmodo. Now, it seems there might have been another big iPhone leak, and while it’s not quite as significant as that 2010 equivalent, it’s potentially not far off.

A video is circulating online that purportedly shows an iPhone 18 Pro undergoing a drop test, though the posts showing it seem to be getting rapidly deleted — it's unclear whether this is for legal reasons or because it's fake.

Still, at the time of writing, this video is still live in a post from @TechKard, among others. You can see a device that looks a lot like an iPhone 17 Pro in silver being dropped onto a surface, and coming out of that scenario more or less unscathed from what we can tell — though it doesn’t appear to be getting dropped from a particularly high height.

iPhone 18 Pro drop test from recent data breach at Apple supplier Tata Electronics.Via:@evleaks #Apple #iPhone18Pro https://t.co/9ZSojlOLSr pic.twitter.com/wdZs6dKuQPJune 30, 2026

A muted response

If this is legit, then it’s footage of a real unreleased iPhone months before launch, so that’s a big deal. But online reactions have been muted, with most posters simply expressing mild disappointment that the design isn’t being updated.

On Reddit, for example, posts include “looks just like the 17 pro lol,” “is this not just a 17?,” and “that's a nice looking iPhone 17.01 you got there.” Some posts also note that this device looks a bit thicker than the iPhone 17 Pro, but it appears as though the video might be slightly stretched, so the actual handset may not be any thicker.

Regardless, the overall reaction seems to be a collective shrug, which is rather different from the excitement generated by the aforementioned iPhone 4 leak.

That said, there’s some reason to doubt this leak. Reuters reports that various iPhone 18 Pro details and images have been shared on the dark web, so that much is probably true, and it’s assumed that this video is part of that leak. But the earliest posting we found of it came from an X account masquerading as famed leaker Evan Blass, but which in fact has nothing to do with him, casting doubts on its credibility.

Still, whether this is an authentic leak or not, we’re not expecting there to be too many visual differences with this year’s iPhones, as Apple is rumored to be overhauling the phones next year instead for their 20th anniversary.

The ultra-popular DJI Osmo Pocket 3 is the camera of choice for content creators for a reason. In testing out this on-the-go cam, we found it “delivers amazing video quality and beautiful slow-motion scenes” — and awarded it a Recommended badge.

If you’ve been waiting for the right moment to pick one up, the DJI OSmo Pocket 3 Creator Combo is $512 (was $639) at Best Buy as part of the 4th of July sale.

That’s a $117 saving on the bundle that includes the mini tripod, wide-angle lens add-on, carrying case, and ND filter set alongside the camera itself — everything you’d need to start shooting without buying accessories separately.

Today's top DJI Osmo Pocket 3 deal

The complete kit: the Osmo Pocket 3 camera with its 1-inch sensor, 3-axis gimbal, and 2-inch rotatable touchscreen, plus the mini tripod, wide-angle lens, carrying case, and ND filter set. Shoots 4K up to 120fps, portrait mode up to 3K, and captures slow-motion at 1080p/240fps. ActiveTrack 6.0 automatically keeps subjects locked in frame. Charges via USB-C. Battery rated for about 166 minutes at 1080p/30 fps.View Deal

For anyone who creates video content — whether for social media, travel, events, or professional work — the DJI Osmo Pocket 3 Creator Combo is one of the best compact cameras available right now.

The Pocket 3 was the upgrade the Pocket line needed. DJI fitted it with a new 1-inch CMOS sensor — the biggest change from the previous generation — and in our review, the improvement in low-light performance was immediately obvious.

That 1-inch sensor is what sets the Pocket 3 apart in the broader action camera market. GoPro and Insta360 alternatives at this price use smaller sensors that struggle in anything other than ideal lighting. The Pocket 3’s 1-inch CMOS captures noticeably more detail and light in indoor environments, evening shoots, and overcast outdoor conditions — the situations where most content creators actually end up filming, rather than in bright daylight, where every camera looks good.

The 4K 120 fps capability is another spec that earns its place in real-world use. At 120 fps, you can slow footage down to 25% speed in a 30fps timeline without it looking choppy — the kind of cinematic slow motion that previously required much more expensive cinema cameras.

For travel content, sports coverage, or anything with movement, it opens up editing options that just aren’t available on competitors that run 4K at 60 fps maximum.

Portrait mode at up to 3K resolution is genuinely useful for social media creators. Instead of cropping a landscape video into a vertical frame and losing half the image, the Pocket 3 captures natively in portrait orientation — the output is optimized for Instagram Reels, TikTok, and YouTube Shorts without any post-processing workaround.

In our review, we flagged two honest caveats to keep in mind before you buy. The touchscreen, while improved over the previous generation at 2 inches, is still “not the most enjoyable when navigating around settings and menus” given the limited real estate. And, like most compact cameras with high-powered sensors in a small chassis, it can get warm during extended filming sessions. Neither is a deal-breaker, but both are worth factoring in if you’re planning long continuous shoots.

The Creator Combo’s value is straightforward: the ND filter set alone typically costs $40–50 when purchased separately, the wide-angle lens add-on adds another $40–60, and the carrying case and mini tripod add further value. Getting all of it bundled at $512 — $117 off the combined retail value — is the right way to buy the Pocket 3 if you want a complete shooting kit rather than just the camera body.

Also consider

It might not be the latest of DJI's Action range, but this 4K camera shoots crisp footage, features waterproofing, and comes with a range of accessories to get you started creating pro-grade content. In our review, we called it "a polished GoPro alternative." View Deal

This price includes not just the Hero 13 Black, but also a handle, two Enduro batteries, two curved adhesive mounts, a 64GB micro SD card and a carrying case. You can get the camera standalone for $379.99, but honestly I think $50 for all those accessories is really good.View Deal

• Shop all action cam deals at Best Buy.

The DJI Osmo Pocket 4P was fully revealed on the China page of DJI's website earlier this month, and I've already taken my test sample traveling, but tech journalists like me across the globe are in an awkward position, still being under embargo with our review content.

What I can tell you now, is that the Pocket 4P is available in black or white, has a new 60mm medium-telephoto lens with a large aperture, offers DLog2 support and up to 17-stops dynamic range. The rest is under wraps, or publicly available on the DJI site by selecting the China region (中国大陆).

I can, however, share some of the footage that I've shot with DJI's first dual-lens model in its hugely popular series of Pocket vlogging cameras. I'll say it now — I love its versatility and am really impressed with the videos I have been able to create with it.

Duel of the dual lens cameras

DJI must have been hoping for a dramatic reveal of its first dual-lens vlogging camera, only for Insta360 to somewhat steal its thunder by launching the Luna Ultra first. I've already tested the Luna Ultra and said it's a super impressive first generation product.

That's the competition the Pocket 4P is going up against, and I look forward to sharing my comparisons between the two models soon. Until then, you can enjoy my Pocket 4P sample reel, above.

Like I say, I can't reveal any specs related to the video, though at the least it should be pretty obvious which clips are shot with the ultra-wide lens, and which clips are shot with the new 60mm medium-telephoto lens.

We can get a rough ballpark on pricing, too. According the DJI site, the Pocket 4P Standard kit costs CNY 3,799 (about $562 / £420), and the Vlog kit is listed at CNY 4,299 (about $636 / £475).

The single-lens Pocket 4 starts at £429 / AU$749 for the Standard kit, or £549 / AU$959 for the Creator Combo, however, so I expect the Pocket 4P to cost more than those price conversions above.

Insta360's co-founder already told me it 'won't compete on price' — it's Luna Ultra starts at $769.99 / £649 / AU$1,229.99. My prediction, therefore, is that the Pocket 4P will sit roughly in the middle between the Pocket 4 and Luna Ultra, but don't expect US availability any time soon.

I'll be bringing you my full review when I'm permitted to.

Over the past two years, most of the noise around AI has focused on the model race – whose model is bigger, faster or scoring better on benchmarks.

But as AI moves from pilots into the core of products and workflows, a familiar pattern from the early days of cloud is re‑emerging: systems are more programmable than ever, but they are also much harder to run.

And that means we now know where the most important competition in AI is shifting: from who has the “best” model to who can operate AI reliably, efficiently, and safely at scale.

AI is now hitting operational limits, not model limits

When looking at real‑world telemetry from thousands of production systems, a clear picture starts to form. Nearly 1 in 20 AI requests fails once applications reach scale, and a majority of those failures now stem from capacity limits such as rate limits, quotas and concurrency caps, rather than from model bugs or poor accuracy. That is a very different story from the benchmark charts most teams used to obsess over.

The amount of data sent per request is also climbing. Across many production estates, median users have more than doubled their token usage, while heavy users have seen volumes grow several‑fold. That growth is both a symptom of more ambitious AI use cases and a direct driver of cost and IT infrastructure stress.

You can see the impact most clearly in what many teams now describe as GPU sprawl: fragmented fleets spread across clouds and on‑prem clusters. Some GPUs sit idle while others are consistently saturated, and there is very little correlation between where GPU hours are spent and where they create business value.

The result is familiar to anyone who lived through the early adoption of cloud computing – runaway spend, unpredictable performance and capacity crises that appear out of nowhere.

How this is playing out in APAC

Across Asia‑Pacific, and especially in ASEAN, we’re currently seeing structural pressures: AI adoption is accelerating, but operational maturity is uneven.

Singapore is further along on governance and observability, driven in part by regulatory expectations and a more mature cloud landscape. Meanwhile, markets such as Indonesia, Malaysia and Thailand are moving very fast on deployment, often pushing AI into customer‑facing services while operational practices catch up.

As organizations across these markets roll out multi‑model and agent‑based architectures, they are running into reliability issues, limited visibility and inconsistent model performance. Token usage is increasing quickly, but optimization practices, such as prompt caching and context engineering, are underutilized.

That gap between readiness and deployment is already creating operational and cost debt that will be harder to unwind later.

The four operational disciplines AI teams need

With the evolution of AI resembling the early days of cloud, the good news is that we can predict, at least a little, where things are headed.

Now, the question AI leaders should be asking is this: which disciplines distinguish the teams that will cope best with this complexity?

In my view, there are four that teams working with AI need to adopt to see sustainable success:

1. Establish visibility and attribution

You cannot operate what you cannot see, and AI is no exception.

Teams need to see how GPU hours and tokens map to specific applications, teams and use cases, so they can connect that usage to latency, error rates and user impact.

That makes it possible to separate business‑critical workloads from background noise, and provide clarity into which services are driving cost or consuming capacity.

When usage is visible and attributable on a single view, decisions about where to optimize, protect capacity or dial back become much less emotional and much more data‑driven.

2. Enforce control and guardrails

Without guardrails, AI systems will consume as much capacity as you give them.

Practical controls include rate limits and budget caps, along with safeguards on agent behavior to stop unbounded retries, loops and poorly bounded workflows from exhausting shared resources.

These controls are about making consumption predictable and ensuring that one runaway experiment cannot impact core production services.

Without this discipline, AI programs tend to hit economic limits long before they hit technical ones. You end up with impressive prototypes, but unsustainable unit economics.

3. Optimize GPU utilization before scaling supply

Most teams reach for more GPUs when what they really have is a utilization problem.

GPU instances already account for a significant share of compute costs, and that proportion only grows as organizations push deeper into training and inference at scale.

But idle or underutilized GPUs create the sense of a shortage even when there is headroom in the estate. In turn, many teams can see their overall GPU bill climbing, but cannot see which workloads are driving consumption, or pinpoint the steps needed to improve efficiency.

What we learned during the early days of cloud is that in these instances, overprovisioning becomes the safest default – but then spend balloons even when there is stranded capacity in the fleet.

Treating GPU infrastructure as a first‑class system means tracking utilization so that teams can distinguish genuine capacity shortages from misallocation or fragmentation. Then, they can decide whether to free up capacity or truly add more supply.

4. Design for efficiency at the application layer

High AI costs and rates of failure come from how applications are put together, not from the models themselves.

Inefficient patterns, poor routing across providers and unoptimized prompts all drive up token usage and increase the risk of timeouts, errors and inconsistent behavior.

But with proper visibility into prompts, agents and tools in production, teams can see how requests actually flow through the system and tune for quality, latency and cost in a controlled way.

That turns the application layer from a black box into a place where efficient engineering choices are deliberate, measurable and aligned with business outcomes.

What leaders should do in the new AI race

The early days of cloud taught us that programmability without operational discipline can be as much a liability as an advantage. AI is now at a similar inflection point: the winners will not just be those with access to the most powerful models, but those who treat AI as a long‑term engineering and operations capability.

A useful test for any organization is whether it can explain where AI spend goes, how agents behave in production and which workloads it would protect first if capacity were suddenly cut.

If the honest answer is “I don’t know yet”, then the next phase of the AI journey is clear: stop chasing the next model release, and focus on building the operational foundations that will help you scale AI safely and sustainably.

We've reviewed and ranked the best business cloud storage services.

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

• Fable will feature a repopulation system if players murder too many NPCs
• Playground Games originally thought about letting players live with the consequences of their actions
• It decided the absence of NPCs would ruin some of the game's other systems

Playground Games has confirmed that if Fable players, for some reason, decide to kill every non-player character (NPC) in a settlement, they will repopulate over time.

During a press Q&A at an Xbox event adjacent to this year's Summer Game Fest (SGF) attended by TechRadar Gaming, associate game director Will Kennedy discussed Fable's reactive world, including the permanence of player actions. As a role-playing game (RPG), decisions will impact the NPCs around you, and players will have a reputation that will be remembered through their actions.

Player actions will also have consequences and be remembered, but if players decide to go on a killing spree, NPCs will later respawn through the game's repopulation system.

In fact, Playground originally thought about letting players live with this decision; however, it ultimately decided that the absence of so many NPCs would spoil other systems.

"In terms of killing NPCs, you can kill every NPC in a settlement, and the settlement will stay empty for some time," Kennedy explained. "But we do have a system that begins to repopulate in a settlement as well, but the reason that we do that is, we thought about it, there was a school of thought where you just live with that consequence literally forever, but we also thought in doing that it would cut off a lot of systems to players.

"It would be cool [at] first, but then it might actually get a bit annoying, so we may make the choice to prevent that."

We followed up with Playground Games after the event and received the following from game director and general manager of Playground Games, Ralph Fulton, on how repopulation will work: "The new NPCs who, over time, repopulate a settlement to replace NPCs who have sadly died (i.e., were killed by you…) will be full NPCs like the ones they’re replacing. They’ll have all the elements that all our Living Population NPCs have – name, job, traits, home, clothing, appearance and yes, voice – and will be fully functional in all the same ways.

In the same Xbox event interview, Fulton revealed that Fable will feature over 150,000 lines of dialogue, and over 1000 hours of voice-over were recorded this year alone.

"We've had multiple studios set up running in parallel for over 1000 hours just this year alone, just recording VO," he said, "so it's a huge machine, because the vision that the guys have for this part of the game just requires that way of writing."

Fable will officially launch on February 23, 2027, for PlayStation 5, Xbox Series X, Series S, and PC.

I'm still yet to invest in one of the best espresso machines, but if there was ever a deal to tempt me to do so, then this might just be it. It's for the Breville Barista Express Espresso Machine, which is on sale at Amazon for $499.95 (was $692.99).

This is a machine that allows you to hone your coffee skills to make drinks exactly how you like them. Whether that be a flat white, cappuccino, or a simple espresso, the Barista Express can do it all. The whopping discount brings the highly-rated coffee machine down to the lowest price it's ever been.

Today's best coffee machine deal

We've seen this coffee machine discounted several times in the past few years, but never by this amount. Now, coming in at under $500, there really has never been a better time to invest. If you like your home-crafted coffee, then you won't regret it with one of the best coffee makers we've ever tested for its lowest-ever price.View Deal

If you've been looking for a TechRadar review of this coffee machine, then I wouldn't be surprised if you couldn't find it. We actually reviewed it under its UK name, Sage the Barista Express. Whatever US or UK readers call it, it's the same machine, and we rated it an impressive 4.5 stars.

In particular, we loved how simple it was to use, the many adjustable settings to get your drink exactly right, and the "tasty espresso" that the machine is a pro at making. One of the downsides that our reviewer highlighted was the price, something that is made significantly more palatable with this deal.

If you're open to other machines, then take a browse of the best coffee makers for lattes and the best coffee makers. We've collated all our favorites based on features and budgets.

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 Monday's puzzle instead then click here: NYT Connections hints and answers for Monday, June 29 (game #1114).

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 #1115) - today's words

(Image credit: New York Times)

Today's NYT Connections words are…

• CURL
• BREEZE
• BOX
• FENCE
• SKETCH
• SKATE
• BOTTLE
• ON TAP
• CAN
• WALL
• RECRUIT
• HEDGE
• LUGE
• NEWSPAPER
• GATE
• SKI

NYT Connections today (game #1115) - hint #1 - group hints

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

• YELLOW: Things that split up places
• GREEN: Skills for snow and ice sports
• BLUE: Things that can be reused
• PURPLE: United by a word that sound like “craft”

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 #1115) - hint #2 - group answers

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

• YELLOW: DIVIDING STRUCTURES
• GREEN: PARTICIPATE IN SOME WINTER OLYMPICS
• BLUE: COMMON RECYCLABLES
• PURPLE: WHAT "DRAFT" MIGHT REFER TO

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

NYT Connections today (game #1115) - the answers

(Image credit: New York Times)

The answers to today's Connections, game #1115, are…

• YELLOW: DIVIDING STRUCTURES FENCE, GATE, HEDGE, WALL
• GREEN: PARTICIPATE IN SOME WINTER OLYMPICS CURL, LUGE, SKATE, SKI
• BLUE: COMMON RECYCLABLES BOTTLE, BOX, CAN, NEWSPAPER
• PURPLE: WHAT "DRAFT" MIGHT REFER TO BREEZE, ON TAP, RECRUIT, SKETCH

• My rating: Hard
• My score: 2 mistakes

CAN, BOX, BOTTLE and ON TAP all sounded like wine delivery methods to me, while HEDGE, BOX, BOTTLE and FENCE were surely describing someone who wouldn’t commit to an opinion.

Alas, in both cases I was very wrong, but fortunately before the situation go too dire I was able to spot four ways in which one could PARTICIPATE IN SOME WINTER OLYMPICS and the quartet of DIVIDING STRUCTURES.

With just eight tiles left I finally managed to see the COMMON RECYCLABLES while WHAT "DRAFT" MIGHT REFER TO passed me by like a a BREEZE. Fine work if you saw the purple first.

Yesterday's NYT Connections answers (Monday, June 29, game #1114)

• YELLOW: OLD TIMEY TROUBLEMAKERS MISCREANT, ROGUE, RUFFIAN, SCOUNDREL
• GREEN: CONSUME WITH GUSTO CRUSH, GUZZLE, INHALE, SNARF
• BLUE: PARTS OF A SPEAKER CABINET, CONE, MAGNET, WOOFER
• PURPLE: ENDING IN PARTS OF A TREE EMBARK, GROOT, NUDIBRANCH, STRUNK

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.

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 Monday's puzzle instead then click here: NYT Strands hints and answers for Monday, June 29 (game #848).

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

• Today's NYT Strands theme is… And … action!

NYT Strands today (game #849) - hint #2 - clue words

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

• SHOWER
• CORD
• DROP
• WASTE
• PASTOR
• CRIED

NYT Strands today (game #849) - hint #3 - spangram lettersHow many letters are in today's spangram?

• Spangram has 12 letters

NYT Strands today (game #849) - hint #4 - spangram positionWhat are two sides of the board that today's spangram touches?

First side: bottom, 4th column

Last side: top, 4th column

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

NYT Strands today (game #849) - the answers

(Image credit: New York Times)

The answers to today's Strands, game #849, are…

• WRITER
• DIRECTOR
• CAST
• PRODUCER
• CREW
• EDITOR
• SPANGRAM: THATSSHOWBIZ

• My rating: Easy
• My score: Perfect

There was no mystery about today’s search, although I was expecting to find Lights and Camera in the game somewhere.

Instead, the game ended up being a broader take, with THATSSHOWBIZ covering both stage and screen (both small and large) and the various roles that make up a production.

We could have added “audience” I suppose, as without that there is little point — although I know from a relation who works in a movie theater that even if no one turns up they still play the film the same as if it was a full house, as they are all set by a program at the start of each day.

Yesterday's NYT Strands answers (Monday, June 29, game #848)

• CLEFT
• REST
• NOTE
• BRACKET
• MEASURE
• ACCIDENTAL
• SPANGRAM: MUSICALSTAFF

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 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 Monday's puzzle instead then click here: Quordle hints and answers for Monday, June 29 (game #1617).

Quordle was one of the original Wordle alternatives and is still going strong now more than 1,400 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 #1618) - hint #1 - VowelsHow many different vowels are in Quordle today?

• The number of different vowels in Quordle today is 3*.

* 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 #1618) - 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 #1618) - hint #3 - uncommon lettersDo the letters Q, Z, X or J appear in Quordle today?

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

Quordle today (game #1618) - 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 #1618) - hint #5 - starting letters (2)What letters do today's Quordle answers start with?

• H

• D

• T

• M

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

Quordle today (game #1618) - the answers

(Image credit: Merriam-Webster)

The answers to today's Quordle, game #1618, are…

• HALVE
• DRYER
• THERE
• MINTY

I played this game without my glasses, which added a new dimension of difficulty.

The first error this caused was my using the word “ducky” instead of my starter word “duchy” and the second was missing the letter H and wasting a guess before getting HALVE.

Somehow, I managed to keep my winning streak going.

Daily Sequence today (game #1618) - the answers

(Image credit: Merriam-Webster)

The answers to today's Quordle Daily Sequence, game #1618, are…

• VIVID
• CYCLE
• CHUNK
• PLATE

Quordle answers: The past 20

• Quordle #1617, Monday, 29 June: SLURP, CRACK, CRANK, PHONY
• Quordle #1616, Sunday, 28 June: RUPEE, TOPAZ, FULLY, BEING
• Quordle #1615, Saturday, 27 June: PRINT, MARRY, SADLY, BICEP
• Quordle #1614, Friday, 26 June: JUICE, ARRAY, BONEY, SKIFF
• Quordle #1613, Thursday, 25 June: SHELF, TAWNY, HYPER, SOLVE
• Quordle #1612, Wednesday, 24 June: SOBER, ECLAT, GOOSE, NINNY
• Quordle #1611, Tuesday, 23 June: ARDOR, DADDY, SERVE, SHEAR
• Quordle #1610, Monday, 22 June: WAXEN, APNEA, CHIME, WAVER
• Quordle #1609, Sunday, 21 June: ABBOT, NOTCH, DREAD, LURID
• Quordle #1608, Saturday, 20 June: SLAIN, TAMER, VIPER, FALSE
• Quordle #1607, Friday, 19 June: ALOUD, POINT, GLOBE, GROIN
• Quordle #1606, Thursday, 18 June: LATCH, BRAWL, STEEL, CRUSH
• Quordle #1605, Wednesday, 17 June: HOIST, PLUSH, GROUP, LEMUR
• Quordle #1604, Tuesday, 16 June: SLAIN, PLUCK, PINTO, SLICE
• Quordle #1603, Monday, 15 June: GAUNT, SNEAK, ROUTE, POKER
• Quordle #1602, Sunday, 14 June: WIMPY, WISPY, VIRAL, NYLON
• Quordle #1601, Saturday, 13 June: DEALT, STEED, BELIE, GULLY
• Quordle #1600, Friday, 12 June: TENTH, SHOAL, JELLY, UNIFY
• Quordle #1599, Thursday, 11 June: GAMMA, SPILL, SALVE, RURAL

• New report claims only 2% of English schools currently have formal AI strategies
• No clear policies means the sector isn’t working together to define use cases safely
• Clear strategies, but also room to experiment, are the next steps

New data from Accenture and Teach First has revealed that, even though schools across England are already using artificial intelligence to support learning, only 2% actually have formal strategies.

Full deployment strategies aside, only 12% of the 200 secondary schools surveyed have any type of AI policy, leaving an overwhelming majority investing and deploying blindly.

As a result, most schools are ultimately conducting informal experiments with AI, but on a national level, a lack of consistency or shared learning across the education system could be putting the sector several marks behind enterprise and business counterparts.

Moving beyond experimentation needs a clear strategy

“Many school leaders are navigating this shift without clear guidance or the confidence to implement the technology effectively,” UK&I Head of Accenture Matt Prebble said.

Still, schools are far behind businesses, with 27% of C-suite and senior leaders noting they have a comprehensive AI strategy, per Gartner data (via The Times of India).

Besides surveying around 200 secondary schools, Accenture also conducted 30 in-depth interviews with school leaders, who generally believe that AI has significant potential to improve education once early niggles are ironed out.

At the moment, leaders are worried about plagiarism, safeguarding and bias. Teachers are also being considered in AI rollouts, even if formal strategies lack – nearly two-thirds (63%) cite a lack of staff confidence.

But early adopters are already starting to see results, with common use cases across education including lesson planning, generating quizzes and drafting mock exam questions.

Separate Government reporting also implies that AI can help provide tailored feedback and support personalized learning, as well as administrative tasks that would otherwise take away from staff’s teaching time.

The report also uncovers how top-down leadership can impact AI’s role within any given school, with sceptical leaders causing slower and patchier uptake.

“Building the leadership capability and providing practical support to adopt AI responsibly will be critical to ensuring its benefits are delivered consistently and equitably across the education system,” Prebble added.

Regionally, London leads the way with 29% of leaders using AI everyday, compared with just 12% in the rest of England.

Support for how to use and deploy AI is starting to be written

“Safety should be the top priority when deciding whether to use generative AI in your education setting,” the Department for Education stressed. Industry regulator Ofsted also supports the responsible use of AI.

But with artificial intelligence no longer an experimental novelty and pressures from Ofsted monitoring, schools are being urged to consider more detailed strategies as they deploy AI.

“The biggest risk is doing nothing and assuming that you can just continue as is,” an unnamed headteacher told Ofsted in a separate report.

Teach First CEO James Toop also emphasized the knock-on impact poor strategies could have on schoolchildren: “Ensuring every young person, regardless of background or where they live, can safely benefit from the opportunities AI presents must be a priority for the education system.

Accenture and Teach First set out five key priorities for schools looking to advance their AI strategies: headteachers and leaders should engage with AI more directly and visibly; policies should clearly define purpose and boundaries; early pilots should start where value is clearest before progressing; teaching staff should be given permission to experiment with use cases; and shared learning must complement formal training for a more rounded approach.

• AMD Zen 5-based EPYC 9965 is on sale on eBay for just under $6,000
• With 192 cores and 384 threads in tow, AMD's offering can handle as much as 6TB of RAM per processor for AI-centric workloads
• The CPU not only offers the highest core count in all x86 CPUs to date, but also holds its own against the competition

When AMD unveiled its new EPYC CPUs in June 2024, one particular chip caught everyone's eye, given how outrageously powerful it was compared to the competition.

The AMD EPYC 9965 CPU was also the company's most ambitious server-grade offering to date, and one that has held its own against the competition with relative ease thanks to its yet-unmatched 192-core, 384-thread x86 configuration.

At a suggested MSRP of $14,813 (now $11,988) at launch, it is also one of AMD's most expensive products across the board, with only a few of its server-grade GPUs pushing noticeably higher.

Why the eBay listings are somewhat hard to digest

Despite being unveiled over two years ago, the EPYC 9965 CPU remains largely unchallenged by the competition, thrashing most benchmarks with relative ease. It remains the flagship server chip for AMD-based systems worldwide, with Team Red shipping it globally, including China.

The chip itself is a genuine monster. The EPYC 9965 packs 192 Zen 5c cores and 384 threads onto a single SP5 package, with a 2.25 GHz base, a 3.35 GHz all-core, and a 3.7 GHz maximum boost, 384 MB of L3 cache, twelve channels of DDR5, 128 lanes of PCIe 5.0, and a 500 W default TDP, AMD's highest yet. It is the densest part in AMD's 9005 stack, built for cloud providers aiming to cram the maximum number of virtual machines into a rack.

All in all, it is designed primarily to serve hyperscalers and cloud providers, with a focus on low-power, efficient cores that dominate such setups.

Even at its reduced reference price of just under $12,000, eBay listings selling the same CPU for just under $6,000 seem inexplicable, offering a sub-50% discount over current advertised prices and over 60% off its original launch price.

This is even lower than the last time we tracked this CPU in 2025, but there might be an explanation of sorts for what would otherwise be seen as an anomaly. Most server-grade CPUs trade at prices below MSRP, with OEM and negotiated discounts that quickly add up to significant savings for large-scale deployments.

The other explanation that often applies to older server-grade CPUs that often get cycled out or decommissioned by data centers does not apply here, however; the EPYC 9965 does not have a successor yet, and it is hard to imagine a currently available CPU replacing AMD's most capable chip yet in a data center just yet

Other reasons prices are low could include excessive inventory, canceled orders, or simply someone liquidating hardware in an industry that is increasingly GPU- and memory-focused to meet its AI needs.

Most industry players prefer a direct OEM solution, especially given AMD's support for PSB-locking (secure boot) on their CPUs for certain vendors, which makes them unwilling to purchase from 3rd-party sellers. This also raises the risk profile for buyers, who could unwittingly end up purchasing a vendor-locked CPU, making for a rather annoying testing and return process.

With a shift in spending towards GPUs and memory becoming prohibitively expensive, hardly helping matters, while enterprise consumers are still buying up the EPYC 9965, prices might reflect a market reality: they are unlikely to be scouring eBay for their next 192-core monster CPU purchase, even assuming they have a compatible motherboard and memory modules to match.

EOFY is a good time to check gaming laptop deals, even if the laptop isn’t going to be used for work. Gaming laptops can be quite expensive, so a legitimate discount can make a bigger difference than it does on a basic everyday laptop.

That said, a gaming laptop can also be a powerful workstation if your work does need extra grunt. A fast CPU, dedicated GPU and plenty of RAM can help with video editing, 3D work, coding, running AI tools, creative apps or just keeping a lot of browser tabs and programs open at once.

It’s also worth checking our main EOFY laptop deals hub even after June 30, because in previous years we’ve seen some great laptop deals hang around or pop up in early July as retailers clear leftover stock.

For gaming laptops, we have deals from lower-cost RTX 5060 machines with generous RAM, through to much more expensive RTX 5090 desktop replacements.

This is a tempting price if you want a gaming laptop that can also handle study and everyday work. The Core 7 240H, RTX 5060, 32GB RAM and 1TB SSD make the spec very capable for the money, while the 16-inch 1920 x 1200 180Hz display means you can push older titles to high frame rates.View Deal

For those who want serious gaming power, the Legion Pro 7i with an Intel Ultra 9 275HX processor and RTX 5090 GPU is one of the better options going, even though it's 15% more than the price we saw on Black Friday in 2025. There are very few comparable machines at this price, making it an excellent deal if you are willing to spend big. Read our review to learn more.View Deal

The Acer Nitro V 16S is a good option for a buyer who wants current-gen gaming performance without paying big money. You still get an RTX 5060, 32GB RAM, a 1TB SSD and a 180Hz display for AU$1,777, which is a lot of hardware for the price.

The RTX 5060 is a lower-end current-gen GPU, but that’s still a good fit for 1080p or 1200p gaming, especially if you’re happy to use DLSS or turn down a few settings in the most demanding games. With RAM prices rising, 32GB of RAM is harder to take for granted at this price, so it’s good to see Acer include it rather than cutting back to 16GB.

The Lenovo Legion Pro 7i is at the other end of the scale. Our Lenovo Legion Pro 7i review described it as “perfect for a desktop replacement, just okay as a portable”, and that’s exactly how I’d frame this deal.

You’re not buying it because it’s cheap or easy to carry everywhere. You’re buying it because you want a very powerful 16-inch gaming laptop with an RTX 5090, 32GB RAM, a 1TB SSD and a 240Hz OLED screen.

Our review gave the Legion Pro 7i 4 stars and called out its excellent gaming performance, gorgeous OLED screen and useful port selection. That makes it a good fit if you want one machine for demanding games, an external monitor setup and heavier work like editing, rendering or development.

The big caveat is battery life, which our review said “is not great”, but that’s the trade-off with this kind of desktop-replacement gaming laptop.

If neither of these gaming laptop deals is quite right, our main EOFY laptop deals hub has more options across cheaper Windows laptops, MacBooks, 2-in-1s and higher-end gaming machines.

• Alogic Folio Duo offers dual-QHD portable monitor setup that works with both MacBooks and Windows-based laptops
• It is one of the first portable displays to offer full gesture control and 10-point multitouch on both Mac and Windows
• The Folio Duo also offers users the ability to charge their laptops on the go with 45W of USB-C passthrough

Portable monitors are hardly a new concept, with most products offering a sizeable upgrade in screen size, resolution, or both compared to the desktops, laptops, or MacBooks they connect to.

However some, like the Alogic Folio Duo, kick things up a notch while doubling the screen real estate, and its foldable design, coupled with an integrated cover stand, makes for a much easier-to-deploy and carry solution than many of its peers.

With a dual-QHD configuration coupled with a 400-nit display, the Folio Duo outdoes many of the devices it will inevitably be attached to, and considerably so, while weighing approximately 1.2kg.

A premium portable dual-monitor offering

The Alogic Folio Duo first appeared at InfoComm 2026, pitched as a premium dual-monitor offering for professionals who simply need more screen space without adding as much of a volumetric footprint as some of its peers.

Alogic's offering features dual IPS panels rated for 100% sRGB color accuracy, with a native resolution of 2560x2880 (or 2560x1440 per display). Prospective gamers, however, might be disappointed with the relatively tame 60Hz refresh rate, but one could argue that most would not be looking for a portable dual-monitor setup to begin with.

Like most of its peers, the Folio Duo offers USB-C connectivity but kicks things up a notch by adding USB-PD input for an optional power source and a USB-C port for the device it connects to, allowing it to charge a MacBook or Windows laptop up to 45W via USB-C passthrough.

Alogic offers a generous 2-year warranty for the unit and lists device compatibility that includes Apple, Chromebook, and Microsoft devices, with the package including a monitor cover in the box, while skipping the pressure-sensitive Active Style the company sells, even as it integrates wireless charging into the package.

The fact that it is one of the first portable dual-display offerings to support full gesture control and 10-point multitouch on both Mac and Windows might underscore the ambitious price the portable monitor currently commands: a whopping $1,299, which makes it rival the recently repriced 13.6-inch Apple MacBook Air (M5), a full-fledged computer.

For those who can afford it, however, the Folio Duo seems to be a premium portable dual-monitor offering with significantly higher resolution (and PPI), color, and build quality than the UPerfect Delta Max Touch we recently reviewed, albeit at more than twice the asking price.

• Dbrand made a Companion Cube case for the Steam Machine
• It was a tribute to the Companion Cube from Portal, and a popular idea
• Unfortunately, Dbrand went ahead and made this product without any permission from Valve, and so had to withdraw it from sale today

Dbrand has announced that its fancy case for the Steam Machine that made Valve's gaming PC look like a Companion Cube from Portal has been pulled from sale.

Why? Because, rather unfathomably, as Dbrand explained on Reddit, it didn't ask permission from Valve to make and sell this creation.

Dbrand said: "We launched around 3am on Monday, June 22nd. Overnight, it [the Companion Cube case] became the second-fastest selling product in our 15-year history, behind only the Switch 2 Killswitch."

"Shortly after, Valve's legal team reached out. They stated that the Companion Cube is Valve intellectual property, for which Dbrand does not have a license. They requested we take down the product and launch film immediately. This was entirely within their rights, and they were direct, fair, and respectful throughout."

Dbrand complied, while launching an appeal to Valve to see if there was "any way to keep the project alive: properly licensed, with their blessing, on their terms".

The short response was no, although Dbrand concedes: "Given our backwards approach of building first and asking permission later, it was a fair answer."

And that is the story of a $130 hard-shell Steam Machine case ($99 for the 'poverty' version) that was never meant to be. Refunds are being issued today, June 29, 2026, for anyone who ordered a Companion Cube case.

Analysis: permission denied

(Image credit: Shutterstock / aslysun)

This whole episode is suitably baffling considering that Dbrand apparently put over a thousand hours of engineering into making the Companion Cube, ensuring the Steam Machine fit into it perfectly, and developing 44 sets of injection-molding tools in the process. Not to mention hiring out a university campus as the set for the launch video.

Dbrand says: "By the end, we were losing money on every $99 Poverty Cube sold, but it didn't matter. This had turned into a passion project for the entire organization."

As you can imagine, there are a lot of bemused (or indeed amused) gamers on the Reddit thread calling Dbrand stupid, which the company freely admits is true. One Redditor says: "You had two jobs! Ask Valve, and make the thing."

Another asks: "How on Earth did you think you would be able to sell this thing without getting permission from Valve...???"

Well, yes. It's a good point. Some argue that Dbrand might have wanted to build up some publicity and show off all the pre-orders to persuade Valve to be more likely to say yes, but that theory falls apart because the company forgot to ask Valve – at all.

Amidst the (rightfully) withering criticisms are more than a few disappointed punters who are actually canceling their Steam Machine (they say) because they won't be able to put the gaming PC in Dbrand's enclosure.

Obviously that's a subjective matter, but I'm not seeing the appeal at all personally. The Companion Cube made the Steam Machine a good deal chunkier (when the idea is that it's a low-profile, unobtrusive living room PC), and while very nicely made by all accounts, it pushed the price up even further. And, despite ventilation obviously being considered in the design, I do wonder about the effect on the Steam Machine's cooling system on a hot day.

Still, those are all moot points now. Dbrand said this one is a "hard lesson to learn publicly," and, indeed, hopefully the company will take it to heart. It's not the first time that a Dbrand product has come off the rails like this, though. You might recall Sony sending a cease-and-desist letter, which canceled Dbrand's 'Darkplates' (faceplates for the PS5) because of the way they replicated a "protected product design". A Darkplates version 2.0 was launched with changes to address this, and it remains on sale today.

When the Darkplates were first revealed by Dbrand, it was with the prescient slogan 'Go ahead, sue us', so you've got to wonder whether this particular lesson taught by Valve is going to stick.

• Fitbit devices have a new fitness coach powered by Gemini AI
• But the AI seems to be giving people extremely questionable advice
• Users have voiced their dissatisfaction with the feature

Fitbit recently introduced a new fitness coach powered by Google’s Gemini artificial intelligence (AI), and it’s safe to say that it’s received a lot of flak from Fitbit users. Complaints have been flooding in, but surely few have been as bizarre as one recently posted to Reddit that involved some truly “unhinged” advice from Gemini.

Posting on Reddit, user bitteroldladybird started off by claiming that “The coach suggested I ditch my dog.” If that didn’t raise your eyebrows, what comes next surely will.

They continued by explaining that, “I’ve been walking my dog twice a day her whole life. Including the last year and a bit when I’ve had my Fitbit.”

But after that preamble, things start to get weird: “Recently the AI coach has been giving me feedback on my walks and it asked why my pace was so slow. I answered that I walk with my dog. This slows me down because she stops and sniffs and pees etc. Coach said it understood. Today it asked if I could ditch the dog to speed up my walks.” The user then opened the floor and asked fellow Redditors: “Has the coach given you weird or unhinged advice?”

Funnily enough, bitteroldladybird was far from the only person to relate a story like this. User KateJ95 recounted how “I got told to ditch my toddler … Turned coach off after that.” Individual_Sun2060, on the other hand, said “My coach incessantly tells me to rest, and has probably suggested I take the day off EVERY SINGLE DAY.”

User vemailangah, meanwhile, had a helpful suggestion for Fitbit’s next update: “coach sends AI robot to get rid of the dog to help you improve your walks.”

TechRadar’s own Matt Evans has had a similarly bizarre experience with Fitbit’s AI coach, explaining that it developed an obsession with a minor cold and wouldn’t let the issue go. After Evans didn’t wear his Fitbit for one day — and therefore logged zero steps or workouts — the AI chimed in with: “yesterday was a full recovery day with minimal movement.”

As Matt explained, it seemed that Gemini “really thought I spend 12 hours lying perfectly still, like a mummy in a sarcophagus.”

Latching onto any context

(Image credit: Fitbit)

Judging by the feedback from people on Reddit and here at TechRadar, it seems that Fitbit’s AI coach is tuned a little too strongly towards fitness efficiency and improvement. If it detects any sort of “hindrance” that it feels is slowing you down, it suggests jettisoning it at the earliest opportunity — even if that means ditching your beloved pup.

TechRadar’s Evans points out that this behavior could be because Gemini “just kind of latches on to any context you give it, and is designed to improve your health — occasionally to its detriment when it comes to subtlety and context.” Because you know far more about yourself than Gemini does, the chatbot has to take any cue it can in order to build a picture of your wellbeing. And if you mention something tangentially relevant in your life, Gemini has a few other resources for context.

Aside from creating the kinds of bizarre situations that we’ve seen here, this issue limits the fitness coach’s utility. Another thread on Reddit asked “Does anyone actually use the AI Coach?” and was filled with replies from people who have lost patience with the feature. “When the trial ends, I’m out. Coach is garbage,” said flanga, while realManTing shared that “I find myself yelling at it over text and I cannot wait for my trial to end.”

As the original poster in that thread put it, the coach “constantly gives me long walls of text that are either obvious, outdated or just not useful. I don’t want to read an essay every time I open the app — I just want short, actionable insights.”

It therefore seems clear that Gemini’s AI coach is not particularly popular among Fitbit users and has a worrying tendency to offer questionable advice and to irritate them. Hopefully Google can make some rapid improvements before it suggests anyone else dump their dog to record a slightly faster walk.

• AMD has 'relaunched' three processors that use legacy cores to address demand at the lowest end of its mobile and desktop CPU offerings
• The CPUs range from 2 to 8 cores and leverage its older Zen+ and Zen 2 core technology as the chip designer grapples with memory and storage shortages to serve an increasingly cash-strapped budget segment
• All three processors support Microsoft Windows 11 and leverage cheaper DDR4 memory as manufacturers aims to 'hold the fort' in the face of rising AI demand creating challenges for entry-level PC buyers

AMD's latest Ryzen CPU offering might not be one that most of its performance-centric consumers would appreciate in 2026.

The company is looking to reintroduce three budget-centric mobile CPUs that were last seen over half a decade ago, using legacy Zen+ and Zen 2 cores.

The move comes at a time when assemblers and consumers alike are grappling with increasing chip, memory and storage costs, even as lower-end options continue to grow increasingly limited for price-conscious users.

A Dual-core CPU in 2026?

AMD is reintroducing the Ryzen 3 3100U, Ryzen 5 3501U, and Ryzen 7 4700LE to consumer markets, according to reports. The CPUs offer dual, quad, and octa-core options.

While the Zen 2-based desktop-class Ryzen 7 4700LE has a listed launch date of 25th March 2026, the other two SKUs were launched more recently, in June 2026, according to AMD's own listings.

The Ryzen 3100U and the 3501U both use AMD's legacy Zen+ architecture on a 12nm process node, making them less power-efficient than the 7nm-based Ryzen 7 4700LE.

In an emailed statement to PC World, AMD stated, "The Ryzen 3100U and Ryzen 3501U are additional SKUs based on AMD’s existing Picasso architecture that were developed to support specific OEM requirements in the value segment. These processors are intended to address targeted customer demand for lower-cost solutions and will be available in limited volumes through select OEMs".

As OEM components are made directly available to laptop assemblers, AMD does not disclose prices for these models, even as rival chipmaker Intel seems to be taking a similar route.

What might rankle some users in particular is not the fact that the 3100U and the 3501U are rebadged CPUs that will still lag behind pretty much even the most basic laptop CPUs available in the last few years, but the fact that AMD might be thinking it is acceptable to offer a 2-core, 4-thread CPU in 2026 as a baseline for new, entry-level laptops.

The Ryzen 3 3100U is getting considerable ire, as Redditors are already drawing unfavorable comparisons to current-generation Athlon CPUs. While it offers the same iGPU (Vega 8), TDP (15W), and memory configuration (DDR4-2400), along with the same socket (FP5) as its quad-core sibling, it is noticeably weaker than most, if not all, recently launched mobile CPUs.

Unsurprisingly, it does make the cut for Windows 11 by offering support for fTPM, allowing for secure boot, and just pulling ahead of Microsoft's stated requirements for its OS (1 gigahertz (GHz) or faster with 2 or more cores), even as most users will struggle to do anything but the bare minimum on a laptop sporting a CPU essentially rebadged from 2019.

AMD's move, however, is predicated on the fact that rising RAM, SSD, and cutting-edge silicon prices all create a void that must be addressed for the foreseeable future, even as its target audience seems to be getting an increasingly short end of the stick as part of the solution.

Another month; another bumper crop of great tech. Over the past four weeks, we've tested dozens of new gadgets and appliances, and between our editors, testing team, and trusted freelancers, we've awarded 21 products 4.5- or 5-star ratings. I've rounded them all up below, for easy browsing — hit the 'View details' button for a summary of each one's main pros and cons, plus a link to read the full review.

In the full-marks camp, we have the Dyson HushJet Mini Cool, a tiny-but-mighty handheld fan that provides up to 6.5 hours of cooling per charge, and might have just saved our Home editor's sanity during our recent UK heatwave. There's also the JBL Xtreme 5, a beast of a Bluetooth speaker that delivers, according to our tester, "mesmerizing bass, impressive clarity, and thunderous power". Finally, we were blown away by the Philips Café Aromis espresso machine, which "doesn’t just brew excellent coffee — it makes the process fun as well". Plus, it doesn't require barista-level experience to deliver professional-quality brews.

Scroll down for more winners from this month, across all our physical product categories, including everything from flawless cameras to standout earbuds to top TVs. Note: only the products that are available in your region will display as cards.

5-star products Philips Cafe Aromis 8000 Series Bean to Cup Coffee Machine

Read our full review

Pros

• Gorgeous design
• All components feel solidly made
• Consistent brewing results
• Prepares hot and cold drinks
• Detailed customization options
• Excellent mobile app experience

Cons

• Ground coffee hatch has a tendency to steam up
• Needs to cool down completely before preparing cold drinks

Dyson HushJet Mini Cool Fan

Read our full review

Pros

• Powerful concentrated airflow
• Runs for up to 6.5 hours on a single charge
• Stylish, unusual design
• Significantly cheaper than closest competitor

Cons

• Air intake holes are easy to cover with your hand when holding

JBL Xtreme 5 Bluetooth speaker

Read our full review

Pros

• Amazing audio with exceptional bass output
• Impressive clarity and detail, especially using USB-C
• Phenomenally sturdy and waterproof
• Great value despite fairly high price

Cons

• Bulkier than its predecessor
• No mic for hands-free calls
• Battery life isn’t class-leading

4.5-star products

Insta360 Luna Ultra vlogging camera

Read our full review

Pros

• Clever detachable touchscreen simplifies and expands solo vlogging possibilities
• Two lenses with two distinct perspectives boost versatility
• Crisp video quality and attractive color profiles
• Excellent audio features and supporting accessories

Cons

• Larger than DJI rivals
• Vertical video capped at 3K
• Subject tracking performance could be refined
• Much pricier than single-lens alternatives

Dyson Find+Follow Purifier Cool fan

Read our full review

Pros

• Premium performance and feature-packed
• Impressive 350-degree oscillation
• Quiet operation and smooth airflow
• Control it with the remote, app, or your voice

Cons

• There are cheaper alternatives
• No vertical oscillation adjustment
• Only one colorway: white & gold

Ninja AutoBarista Pro Fully Automatic Espresso Machine

Read our full review

Pros

• Makes coffee and espresso beverages to your specifications
• Very easy to use
• Grind iQ takes the guesswork out of your settings
• Create up to two user profiles
• Two separate bean hoppers
• Chute for pre-ground coffee

Cons

• Fairly expensive
• Big and heavy
• Coffee isn't super hot
• You have to do regular maintenance

Meraki Espresso Machine Gen 2

Read our full review

Pros

• Stunning, pro-looking design
• High-end, sturdy accessories
• Smart features augment but don’t replace manual coffee making
• Extracts delicious, consistent coffee

Cons

• Undeniably expensive for a manual machine
• CoffeeSense only useful for Meraki beans

Ninja French Door Air Fryer and Countertop Oven

Read our full review

Pros

• 10 cooking modes cover just about any needs
• Fast and easy to use
• Glass doors so you can keep an eye on your food
• Very fast preheating

Cons

• You do need to watch your food
• Fairly pricey

Panasonic Lumix L10 compact camera

Read our full review

Pros

• No shortage of features
• Comfortable handling
• Good overall image quality

Cons

• Slightly aggressive noise reduction
• A bit on the 'plasticky' side
• Price a little too high

Leica SL3-P mirrorless camera

Read our full review

Pros

• Outstanding image quality
• Exceptional build quality and styling
• Handy top-plate display
• Excellent battery life

Cons

• Very expensive
• Tilt-only rear screen
• Heavy with a serious lens mounted

LG B6 OLED TV

Read our full review

Pros

• Brighter, bolder picture than the B5
• Excellent gaming features
• webOS 26 is a great smart TV platform

Cons

• Green tint more of an issue than on B5
• Average sound quality
• LG C6 is a bigger upgrade and isn't much more expensive

Khadas Tea Pro Magnetic Headphone Amplifier

Read our full review

Pros

• Premium look and feel
• Wide compatibility
• Compelling audio

Cons

• A little fiddly
• Display text is tiny

PowerA Protection Case for PS Portal

Read our full review

Pros

• Highly durable exterior
• Padded screen-protector provides peace of mind
• Cable passthrough feature for easy charging

Cons

• Handle could be higher quality
• May be a touch bulky for some

UGreen Qi2 25W wireless power bank

Read our full review

Pros

• Reliable performance
• Great build quality
• Integrated cable and USB-C port add flexibility

Cons

• Can get a little hot
• Price may be a turn-off

Motorola Edge 70 Fusion

Read our full review

Pros

• Fantastic-looking screen
• Premium design
• Giant, long-lasting battery

Cons

• Bluetooth woes
• Some software issues
• Only three years of updates

Anker Soundcore Liberty 5 Pro earbuds

Read our full review

Pros

• Strong noise cancellation
• Rock-solid fit
• Fun, bassy sound

Cons

• Charging case on the bulky side
• Could do with a discount
• Touchscreen can be unresponsive

Shokz OpenDots 2 open-ear buds

Read our full review

Pros

• Powerful, bassy sound
• Good battery life
• Stable enough for runners
• Customizable EQ modes

Cons

• Prone to dislodging on contact
• Expensive
• Not suitable for helix piercings

Denon Home 400 speaker

Read our full review

Pros

• Great app support for control and multi-room audio
• + Customizable sound modes that reveal layers in spatial audio mixes
• Premium build quality

Cons

• Expensive for a single speaker
• Full spatial experience depends on well-encoded Atmos source material
• No Alexa or Google Assistant

Apple AirPods Max 2 headphones

Read our full review

Pros

• Even better active noise cancellation
• Live Translation that's easier to use than on Apple's earbuds
• Adaptive Audio means meatier, punchier sound than the originals
• Even head-tracked Spatial Audio feels more accurate and crisp

Cons

• Battery life hasn't been improved
• That flimsy 'headphone bra' case remains
• Still very iOS-only (Android owners need not apply)
• No wireless hi-res codec support still

Sonos Era 100 SL wireless speaker

Read our full review

Pros

• Precise, detailed sound
• Bass seems less muddy
• Same strong design as Era 100
• More affordable

Cons

• Room correction not Android-compatible
• Unimpressive stereo
• No voice control

Klipsch The Nines II speakers

Read our full review

Pros

• Immersive and full sound
• Lots of connectivity
• Wi-Fi streaming

Cons

• Issues with Dirac Live
• Weight can make setup cumbersome

• Connected Car ruling will come into effect in 2027
• Anything with a link to China or Russia is in the spotlight
• Volvo has managed to avoid the same fate

North American fans of beautifully minimal Scandinavian design will no longer be able to purchase one of Polestar’s innovative and stylish EVs as of next year, as a new Connected Vehicle Rule can prohibit the sale of items with a “sufficient nexus” to China or Russia in the US.

The final decision falls to the Bureau of Industry and Security, part of the US Department of Commerce, which declined to grant Polestar authorization to sell vehicles in the US from model year 2027, according to Electrek.

The rule, which was instituted during the final days of the Biden administration but kept under the Trump administration, centers around national security concerns, particularly with those brands controlled by, or subject to, the jurisdiction or direction of China or Russia.

Polestar, which has its headquarters in Gothenburg, Sweden, is majority owned by one of the largest automakers in China, Geely Auto, which also happens to own Volvo.

However, its Swedish sibling was granted a waiver in May to continue its sales in the US.

“Companies from these countries may be compelled to share data or allow remote access to connected vehicles in the United States,” the notice of the rule said, according to CNN.

But to add insult to injury, none of the Polestar models that are on sale in North America are assembled in China, with the Polestar 3 built in a Volvo plant in Charleston, South Carolina, and the Polestar 4 built in South Korea.

The company has said that it will keep its 32 US-based dealerships open to continue offering aftercare and support to existing customers, as well as to clear any remaining inventory.

A statement released by the brand said existing Polestar owners and lease customers will “continue to receive the same level of support and access to service as they do today”.

Analysis: Causing more confusion for the EV market

(Image credit: Polestar)

Polestar likely won’t take the ruling to heart, seeing that the majority of its sales come outside of the US anyway, but the Department of Commerce’s decision to ban Polestar and not Volvo remains baffling.

Both brands share much of their basic platform and electrical architecture, while Polestar and Volvo’s infotainment systems are only really discernible by the fonts and graphics they use on-screen.

What’s more, both these brands run Android Automotive Operating Systems and have Google Maps installed natively for navigation, both of which were originally developed by US tech giants.

While the rule appears to have national security at its heart, it also just adds further confusion to a global automotive market that is struggling to keep up with the constantly-moving goalposts, both in terms of shifting emissions rulings and global trade legislation.