Technologies

Think Robots Are Impressive Now? Just Wait Until They Have 6G

This next-generation network technology won’t just make our phones faster; it’ll unlock new capabilities in robots, turning them into all-sensing, always-learning fleets.

Why are there so many robots at a show focused on phones? This is the question I asked myself as I roamed the halls of Mobile World Congress, on the lookout for the most exciting technology that will define the next few years.

The first and most obvious answer is that robots draw crowds. A dancing humanoid is an easy way to attract people to your booth. But to see the robots at this year’s MWC purely as a publicity stunt would be to ignore the bigger conversation happening around robots and connectivity.

Already in 2026, we’ve seen major leaps forward in robotics, with companies including Boston Dynamics and phone-maker Honor showing off humanoid robots designed for industry and home environments. But there is yet another level to unlock, and it relies on 6G — the next-generation network technology set to succeed 5G in 2030 and beyond.

On the surface, 6G and robotics might seem distinctly unrelated — beyond being technologies of a future that we’re not living in quite yet. But in this future, 6G will open new doors for humanoid robots that’ll transform them from clunky, standalone mechanical figurines into efficient fleets, where individuals will form part of an all-sensing, always-learning ecosystem.

This will happen first in industry, then in hospitality and care environments, before potentially landing in our homes. It’s an exciting prospect, but as the experts I spoke to at MWC last month cautioned, there’ll be some big leaps in technology required before they, and we, are ready for that.

The power of 6G

To understand how 6G will unlock new possibilities for robots, let’s start with the special capabilities the network technology will have.

The first is that 6G will act as a sensor network, with sensors embedded into both the robots and their environments, Qualcomm’s executive vice president of Robotics Nakul Duggal told me.

This allows the 6G radio to act like radar — constantly scanning and mapping its surroundings in real time to detect obstacles. Imagine a robot attempting to navigate a crowded environment: The 6G network should quickly and cheaply help create a kind of virtual map for it to do so safely.

Second, there’s the pure speed at which 6G will communicate vast reams of data. The 5G networks we currently use aren’t necessarily built to handle AI requests, but the 6G networks will be, providing a consistent, low-latency, relatively low-power way to process intelligence and deliver that intelligence to robots, according to Frank Long, associate director of intelligent services at deep tech research firm Cambridge Consultants.

Private 5G networks combined with edge AI (relying on devices for computing, not just the cloud) can fill the gap for now, but public networks, not so much. By contrast, Long said, «with 6G you can pretty much have that quality of service guarantee.»

Cambridge Consultants brought a demo of an autonomous humanoid robot to MWC that can pick up and place down a box based on where it sees you pointing. The gesture recognition, plus the ability to react in real time, while varying its grip to pick up something that might be on an angle, requires an enormous amount of compute power. (The demo was powered by a private 5G network.)

Whether robots are connected to the cloud, or to each other in a peer-to-peer fleet, the network will need to handle their intelligence demands at speed. For robots to be constantly talking to the infrastructure around them — and to each other — a strong, reliable uplink will be required, explained Anshuman Saxena, general manager of robotics at chipmaker Qualcomm.

He gave the example of two robots working in a retail environment where one is unloading soda cans from a truck, and another is restocking shelves. They’ll need to align on how to read the space around them to complete each task, including understanding how many cans will need placing, and when they’ll be ready.

«The only way is this robot, while shelving, goes to the back door entry of the truck that is getting unloaded and sees what is available,» said Saxena. «Or the robot that’s unloading is communicating the bigger picture to every other robot, so that we have a view of where the things are placed, so that they can plan.»

This is what’s known as long-horizon planning, where a robot isn’t just focusing on the immediate task but thinking about how that task fits into a broader context over a longer timeframe within a dynamic and unstructured environment. In other words, it’s performing the kind of ongoing mental multitasking that humans do on a daily basis, reacting at speed to what’s going on around us, while also considering what’s next. In the Cambridge Consultant demo, the robot was capable of thinking 16 steps ahead.

Meanwhile, lightning-fast 6G will help robots make split-second decisions, based on feedback not just from their own sensor-packed bodies, but from other robots and tech in the environment. «The retail stores have cameras,» said Saxena. «It’s not a robot, but it can be the eyes of the robot.»

For robots, every day will be a school day

In your own home, you might have only a single humanoid robot. But that won’t be as different from the retail scenario as you may think.

That’s because many of the devices you own, including your phone and security cameras, can already communicate with each other, and the robot will be just another one in the mix. Or maybe you’ll have one humanoid and a bunch of smaller robots designed for specific tasks.

«There is a fleet aspect in the products that we use,» Duggal said. «You don’t feel that, but that is exactly how the product is working.»

Keep in mind that your phone is both a physical object itself and all the software and data that are managed elsewhere. The phone also provides feedback to refine that software, as will the 6G-equipped robots.

«So a robot is going to be performing a certain physical task, and while it may perform it in your home, if it’s also performing the same task in many other homes, there is this aspect of learning and deployment,» Duggal said.

This continuous learning is perhaps one of the biggest challenges that 6G is expected to help solve in robotics. Robots and AI will need massive amounts of real-world data that today’s networks can’t keep up with, even for mundane tasks.

For example: picking up and serving you a cup of coffee, which involves dexterity and balance, with the added element of heat. A robotic arm might not care about the temperature. «But if it is hot, how would we react?» said Saxena. «We would just quickly leave it, which is a very fast reaction time.»

The speed of 6G networks will be essential. By the time a robot arrives in our homes, we will want to know that it shouldn’t hand us a scalding-hot drink and how to protect itself from damage.

Much of this learning might have taken place in hotels or restaurants, where overnight, robots load and unload dishwashers and reset the kitchen. The robot will bring that training into your home, where it’ll still need to further learn about your unique layout and routine. This will likely be a time-consuming process.

«It’s going to be incredibly challenging,» said Long. «Put it this way, members of my immediate family still struggle with opening the baby gate in my stairs, even after extensive training. So a robot, I think, might be a few years away from opening that baby gate.»

Readying robots for 6G… and our homes

But 6G is not expected to roll out widely until at least 2030. What are the robots that companies are already building and deploying to do until then?

They’re making the leaps and bounds they can with the networks of today. «So you’re not waiting for 6G,» Saxena said, «but when the connectivity comes along, you are talking about experiences which can be way beyond what robotics can do [today].»

While the confluence of robotics and 6G will indeed unlock some hitherto unseen next-level robotics, there is plenty that robots can learn in the meantime — particularly when it comes to improving dexterity — to prime them to take advantage of better connectivity. That’s especially true if we’re ever to consider inviting humanoids into our homes, an idea that feels, at least for now, like something worth delaying until at least the 6G-enabled 2030s — if not beyond.

Technologies

Anthropic Reins In Subscribers’ Unlimited AI Use for OpenClaw

It may be the year of the AI agent but Claude’s «all-you-can-eat buffet» is over.

Anthropic over the weekend told subscribers they’d have to pay up for heavy use of its Claude AI models to power third-party agents like OpenClaw.

Users with monthly subscriptions can still use Claude models, including Opus, Sonnet and Haiku, through these third-party agents. But you’ll have to pay via Anthropic’s API or use a «pay-as-you-go option» that will be billed separately from your Claude subscription payment.

«The $20/month all-you-can-eat buffet just closed,» wrote AI product manager Aakash Gupta on X.

At the same time, Anthropic recently announced new features that bring some of the things that made OpenClaw so popular into Claude itself. Claude can use your computer, even if you’re not at it, for example.

Why this policy matters

There has been growing tension between OpenAI and Anthropic, recently inflamed by the controversies involving contracts with the US Defense Department. But there is also tension between users who want to run autonomous AI agents constantly and the AI labs that are trying to control costs by managing the tasks their models are used for.

Claude is a chatbot that was created to be prompted by humans, not for millions of AI agents to use it for workflows. These agent tools, like Manis and OpenClaw, require much more power to run and burn through tokens faster than regular human chatting. Anthropic has already taken steps to address the demand that heavy agent users bring, like a five-hour session cap during peak periods for the models.

«We’ve been working to manage demand across the board, but these tools put an outsized strain on our systems,» Anthropic wrote in its email to customers.

OpenAI has been all-in on agentic tools. Early this year, the AI company hired Peter Steinberger, the creator of OpenClaw, with the aim of bringing AI agents to a broad audience. Steinberger was vocal about his critiques of Anthropic’s new policy, taking to X over the weekend.

«Funny how timings match up, first they copy some popular features into their closed harness, then they lock out open source,» he wrote.

(Disclosure: Ziff Davis, CNET’s parent company, in 2025 filed a lawsuit against OpenAI, alleging it infringed Ziff Davis copyrights in training and operating its AI systems.)

The future of agent compute power

Friction between heavy agent users and AI companies is likely to get worse. These AI agent tools are extremely powerful: they can run for hours, take actions across apps like Gmail, Slack and iMessage, and work autonomously much longer and faster than a human could. Because of this power, they are far more costly and require far more power to run compared to a human prompting a bot. It’s likely that AI companies will increasingly push these compute costs onto heavy users through price increases or steps like those taken by Anthropic.

Technologies

My Running Tests Left Me Feeling Like the Moto Watch Is Low-Key Catfishing

The Polar partnership and $150 price tag had me sold. Then I actually lived with it.

The Moto Watch feels like a kid trying their hardest to stand out in a sport, only to walk away with a participation trophy. Having spent years reviewing pricey fitness trackers and smartwatches, I know how rare it is for a relatively affordable $150 device to arrive with real fitness credibility, so I was genuinely rooting for this one. When Motorola announced a partnership with Polar, along with dual-band GPS and week-long battery life at this price, it sounded like a breakthrough moment. I thought this could be Motorola’s big return to relevance in wearables.

Then I actually used it for a few weeks and reality set in.

Motorola isn’t a stranger to this space. The Moto 360 helped define early Android wearables back in 2014, and made a strong impression doing so. But the years since have been relatively slow on its wearables front. This new Moto Watch is its most serious attempt at breaking through the space in a while, and the Polar partnership gives it a level of fitness-tracking street cred that’s rare at this price.

But theory and execution don’t quite align here. At $150, the Moto Watch isn’t trying to compete directly with higher-end wearables from Samsung or Google; rather, it’s trying to carve out a league of its own with this big-screen 47mm watch. And it’s no home run — yet.

The Polar partnership, tested

The Polar integration is the headline feature that had me excited to put it through the paces. The brand is synonymous with accuracy among serious endurance athletes, and its H10 chest strap is the gold standard we reach for at CNET for heart rate benchmarking on other devices.

So I took both to a college track — three miles (12 laps) — with the watch unpaired from my phone and the chest strap recording simultaneously for comparison. The watch consistently kept up, but I noticed it struggled to keep pace during my sprints.

The workout summaries showed similar numbers, which is why I prefer exporting the raw, second-by-second heart rate data to get more granular. The Polar app makes it easy to export a spreadsheet of your HR data, but the Moto Watch is running it’s own app, and there was no export option. I had to settle for comparing the snapshot of metrics that I got from the workout summary.

The graphs looked similar at first glance, with matching peaks and valleys during the laps when I picked up my pace. The average heart rate was only one beat off from the chest strap. But the watch seemed to smooth out the spikes, and the max heart rate was off by seven beats (173 bpm on the watch versus 180 bpm on the chest strap). That kind of gap is pretty standard for wrist-based tracking, which measures blood flow rather than the heart’s electrical signals. Still, you may not be getting full credit for your effort if you plan to use this as a serious training tool.

Distance tracking was another reality check. Dual-band GPS is usually reserved for higher-end sports watches, so I had high hopes that the Moto Watch would be right on track. It took a while to lock onto a satellite and dropped connection more than once during my 30-minute run. By the end, it had given me 0.15 miles of extra credit. That’s about a 5% error rate, which sounds small until you’re training for a half-marathon and your long runs keep coming back inflated. It’s fine for casual activity tracking, but this is no Garmin replacement.

Health features

Away from the track, the Polar integration holds up better. The watch monitors heart rate, blood oxygen and stress levels throughout the day, though it lacks more advanced features such as ECG or temperature tracking. Wear it to bed (if you can) and you’ll get sleep stages plus a Nightly Recharge Status, Polar’s version of a recovery or readiness score that can help guide training intensity.

But it’s just too bulky to wear comfortably while sleeping. I only wore it to bed once during my month-long testing journey because I felt like the larger size got in the way of my sleep quality. Admittedly, I’m averse to sleeping with accessories on; I don’t even wear my wedding ring to bed. Testing wearables always means making a few concessions, but the Moto Watch just didn’t make the cut for what I’m willing to put up with. It’s definitely more Garmin Fēnix 8 Pro level bulk than Pixel Watch, which I’m ok wearing to bed.

Design: It screams ‘bro’

Motorola positioned this watch as the Clark Kent of smartwatches: a fitness watch cloaked in a polished suit that can go from sweat session to the boardroom. That was the pitch. What landed on my desk, was a different picture with much less polish than I had envisioned. Strapping it on only made matters worse, because it’s 47mm watch looked (and felt) as if it had swallowed my 6.5-inch wrist.

The 1.43-inch OLED touchscreen wasn’t the problem — that was the bright spot. It’s more responsive and more vivid than you’d expect at this price, with slim bezels thanks to a cleverly positioned dial.

You also get a rotating crown for scrolling or clicks, plus a programmable side button. The aluminum case looks polished, too, but it’s easy to miss. The oversized black silicone straps run straight into the frame with no visual break, making the whole thing look like one continuous slab.

Turns out all it needed was a stylist. The desperation of having to wear this thing for weeks put me in problem-solving mode, and I realized the straps were standard width (22mm) and easily swappable with third-party bands you can buy anywhere. Once I switched them, it finally looked like the watch Motorola had sold me. It still screamed «bro,» but it was board room bro.

A battery that just won’t quit

After a three-mile outdoor run with GPS active and no phone, plus a full day of notifications popping up on its always-on display, most flagships would be down to their last breath, but not the Moto Watch. This smartwatch barely broke a sweat and finished the day at 85% battery.

With the always-on display (and no sleep tracking), I made it a full week on a full charge. Switch the screen activation from always-on to raise to wake and Motorola promises it will last 13 days, which I didn’t test, but it seems totally feasible. This is impressive even by sports watch standards.

For the right person, battery life alone could be the reason to buy this.

App, setup and smartwatch functionality

Out of the box, the watch has notifications turned off and set to raise to wake (probably to help get you to the promised 13 days of battery life). And while that might work for some people, I spent most of my first day wondering why nothing was happening on my wrist. If you like to get a heads-up on what’s going on in your phone, I suggest you dig into settings before you start wearing it.

I was skeptical because the watch runs on Motorola’s proprietary software rather than Android’s Wear OS, though it seems like a very bare-bones knockoff. Text previews come through, call notifications work and basic alert handling is fine. But there are a lot of trade-offs that left me wondering why they went rogue in the first place, especially because it still only works with Android phones. It doesn’t support message replies from the wrist, Google Assistant, NFC payments or much of a third-party app ecosystem. For replacing quick glances at your phone notifications, it works. For anyone hoping to actually interact with their phone from their wrist or use their smartwatch to pay for riding a train, it falls short.

The phone app combines health and technical features into one interface, which takes some getting used to, but it ultimately works. It’s a hybrid of Fitbit’s health widget layout and Apple’s activity ring system — almost a blatant borrow, but an effective one for visualizing daily steps, active minutes and calories.

A pricing identity crisis

The Moto Watch is priced to feel like a deal: stellar battery life, dual-band GPS, Polar-backed tracking, blood oxygen, sleep stages and a screen that outperforms its price. On a spec sheet, it punches above its weight.

But $150 is a tricky number. It’s not cheap enough to be an obvious budget pick, and it’s not capable enough to compete at Polar-level performance. The sensor limitations and lack of data export put a ceiling on what that partnership can actually deliver.

Instead, it sits at an awkward intersection, more of a first attempt at carving out something in between. The bones are good. The execution needs work.

Who is this for?

If you’re an Android phone owner who wants sportswatch-level battery life in a sleeker package, this one might be worth a second glance. It’s best suited for casual fitness trackers who want a watch that covers the basics. Serious athletes will want something more precise.

But deal-seekers could be better off with the $160 Fitbit Charge 6 for its additional features or one of the truly budget watches made by Amazfit such as the Bip 6 and Active 2. Style options are limited, and there’s no cycle tracking, so it’s also less appealing for women looking for those features.

Technologies

Today’s NYT Connections: Sports Edition Hints and Answers for April 6, #560

Here are hints and the answers for the NYT Connections: Sports Edition puzzle for April 6 No. 560.

Looking for the most recent regular Connections answers? Click here for today’s Connections hints, as well as our daily answers and hints for The New York Times Mini Crossword, Wordle and Strands puzzles.

Today’s Connections: Sports Edition is a tough one. If you’re struggling with today’s puzzle but still want to solve it, read on for hints and the answers.

Connections: Sports Edition is published by The Athletic, the subscription-based sports journalism site owned by The Times. It doesn’t appear in the NYT Games app, but it does in The Athletic’s own app. Or you can play it for free online.

Read more: NYT Connections: Sports Edition Puzzle Comes Out of Beta

Hints for today’s Connections: Sports Edition groups

Here are four hints for the groupings in today’s Connections: Sports Edition puzzle, ranked from the easiest yellow group to the tough (and sometimes bizarre) purple group.

Yellow group hint: City of Angels.

Green group hint: Winter football.

Blue group hint: Like Hemsworth, but in hoops.

Purple group hint: Cinderellas.