Dyson has announced the PencilWash, a new ultra-slim wet floor cleaner designed to make everyday mopping easier and more hygienic.

Weighing just 2.2kg and built around a 38mm pencil-thin handle, it’s the company’s most compact wet cleaner to date.

The PencilWash is engineered to lie almost flat — up to 170 degrees, reaching as low as 15cm. This allows it to clean under sofas, cabinets and low furniture without sacrificing suction or hydration performance. Dyson says the reduced diameter handle improves in-hand comfort and natural steering. It makes it feel closer to using a broom, rather than a bulky floor washer.

The Logitech G435, priced at $39.99 (was $79.99), the type of quiet disruption budget gaming headsets need, delivers incredible lightness combined with respectable wireless performance, all without breaking the wallet. Logitech prioritized comfort when creating the G435. The thing disappears on your head during marathon gaming sessions, weighing just 165 grams (or around 5.8 ounces).

The earcups are constructed of soft, breathable fabric that keeps the heat at bay, and the headband has a thin layer of the same material stretched over some really basic padding. Users may wear the item for hours without breaking a sweat, and the only time they’ll feel tired is when it’s time to take it off. A wonderful addition is the braille indicators on the sides, which help you determine left from right as quickly as possible, demonstrating that Logitech thought about regular use.

Sale

Logitech G435 Lightspeed & Bluetooth Wireless Gaming Headset – Lightweight Over-Ear Headphones, Built-in…

• Versatile: Logitech G435 is the first headset with LIGHTSPEED wireless and low latency Bluetooth connectivity, providing more freedom of play on PC…
• Lightweight: With a lightweight construction, this wireless gaming headset weighs only 5.8 oz (165 g), making it comfortable to wear all day long
• Superior voice quality: Be heard loud and clear thanks to the built-in dual beamforming microphones that eliminate the need for a mic arm and reduce…

Connectivity is a major highlight here, with a USB dongle that supports LIGHTSPEED wifi for low-latency gaming on PC, Mac, PS consoles, and even Switch. Switch to Bluetooth for your phone or tablet, and it will handle music or calls without losing signal. Most configurations have a range of roughly 10 meters. The battery lasts about 18 hours per charge via USB-C, which is enough for a couple of nights of gaming without needing to recharge.

The sound comes from 40mm speakers that have been adjusted for a balanced sound profile. The bass has a soft, relaxing touch that isn’t overpowering; the mids are crystal clear for voices and effects; and the highs gently roll off past 9kHz, so there’s no harshness. Many people find that the audio produces huge, full-bodied sound for both games and music, especially with a little EQ tweaking using Logitech’s PC software. Volume is limited to roughly 85 decibels in some versions for safety reasons, which is unfortunate if you enjoy cranking it up loud, but better safe than sorry.

The mics are hidden into the left earcup as dual beamformers, picking up speech as crisp as a bell while reducing background noise without the use of a cumbersome boom arm. Clarity is quite impressive for the price, yet it lacks the isolation and richness found in higher-end models.

As humanity looks to the moon for science and economic opportunity in the coming years, understanding potential dangers lurking on the lunar surface could become increasingly important.

Ridges on the moon that signify moonquakes are the subject of a recent research paper, which delves into tectonic activity across the lunar maria, a vast network of dark plains that arose from ancient volcanic activity.

A team of researchers analyzed lunar formations called small mare ridges to create a global moon map, which is the first of its kind. The paper was originally published Dec. 24 in the Planetary Science Journal.

Cole Nypaver, a postdoctoral fellow at the National Air and Space Museum’s Center for Earth and Planetary Studies and one of the paper’s authors, told CNET that the ridges that were identified were formed by faults in the lunar subsurface, which are associated with moonquakes. 

“While those moonquakes are potentially hazardous for long-term lunar exploration missions or permanent outposts, they also present fantastic opportunities to learn more about the interior of the moon and how the moon formed,” Nypaver said.

The moon is shrinking 

Another of the paper’s authors is a scientist named Tom Watters. Back in 2010, Watters discovered that the moon is slowly shrinking because its core is cooling.

The moon’s contraction causes disturbances on its surface. The crust gets compressed and forces material up along faults, which creates ridges, similar to how mountains form on Earth. 

The most common of these ridges are called lobate scarps. They form on the lunar highlands, which are the bright spots we see when we look at the moon. But the small mare ridges only form in the lunar maria, which are the dark areas of the moon that contrast with the highlands.

This research is the first time scientists have documented the ridges throughout the lunar maria. In doing so, we now have a more complete understanding of the moon’s thermal and seismic history, which could give us a better idea of any potential moonquakes in the future. 

“Our results represent the most globally complete understanding of recent lunar tectonism to date,” Nypaver said. “The presence of these additional tectonic features in the lunar maria suggests that the moon may have experienced more global contraction in the recent past than previously thought.”

Moon missions

Humans setting up permanent footholds on the lunar surface have moved from science fiction to real plans for the near future. NASA’s Artemis II mission is set to launch in March at the earliest. And while this mission will only send astronauts to orbit the moon, future Artemis missions plan to land people on the lunar surface and build permanent infrastructure there.

University of Maryland professor Nicholas Schmerr helped NASA develop the Lunar Environment Monitoring Station for Artemis 3, which the crew of the third Artemis mission, currently scheduled for 2028, will deliver to the moon’s surface.

Schmerr said to CNET that this instrument will detect seismic activity in the lunar south polar region. 

“We’ll get a whole new picture of lunar seismic activity both on the South Pole and lunar farside,” Schmerr said. 

LEMS-A3 is a station designed to be self-sustaining, and Schmerr will act as the instrument’s deputy principal investigator for the mission. The LEMS-A3 will assess “tectonics-related seismicity of the region and any hazard the moonquakes (or, for that matter, impacts) could pose to future longer-lived infrastructure,” Schmerr said. 

Setting up shop

NASA isn’t the only one that’s looking to sustain long-term lunar operations. A company called Interlune also wants to set up mining operations on the moon to excavate helium-3, a valuable isotope that could be used for clean energy and quantum computers.

Elon Musk has been talking about building a moon base to launch AI satellites into orbit.

Getting up to speed on the areas of the moon that are more likely to experience moonquakes could influence where space agencies and private companies decide to build outposts in the future.

“There are several upcoming missions to the moon that will carry dedicated seismometers in hopes of detecting a moonquake from a small mare ridge or an asteroid impact on the moon,” Nypaver said. “By identifying a new population of tectonic features in the lunar maria, our work provides additional targets for those missions that seek to use moonquakes to better understand our closest celestial neighbor.”  

For the past three months, Google’s Gemini 3 Pro has held its ground as one of the most capable frontier models available. But in the fast-moving world of AI, three months is a lifetime — and competitors have not been standing still.

Earlier today, Google released Gemini 3.1 Pro, an update that brings a key innovation to the company’s workhorse power model: three levels of adjustable thinking that effectively turn it into a lightweight version of Google’s specialized Deep Think reasoning system.

The release marks the first time Google has issued a “point one” update to a Gemini model, signaling a shift in the company’s release strategy from periodic full-version launches to more frequent incremental upgrades. More importantly for enterprise AI teams evaluating their model stack, 3.1 Pro’s new three-tier thinking system — low, medium, and high — gives developers and IT leaders a single model that can scale its reasoning effort dynamically, from quick responses for routine queries up to multi-minute deep reasoning sessions for complex problems.

The model is rolling out now in preview across the Gemini API via Google AI Studio, Gemini CLI, Google’s agentic development platform Antigravity, Vertex AI, Gemini Enterprise, Android Studio, the consumer Gemini app, and NotebookLM.

The ‘Deep Think Mini’ effect: adjustable reasoning on demand

The most consequential feature in Gemini 3.1 Pro is not a single benchmark number — it is the introduction of a three-tier thinking level system that gives users fine-grained control over how much computational effort the model invests in each response.

Gemini 3 Pro offered only two thinking modes: low and high. The new 3.1 Pro adds a medium setting (similar to the previous high) and, critically, overhauls what “high” means. When set to high, 3.1 Pro behaves as a “mini version of Gemini Deep Think” — the company’s specialized reasoning model that was updated just last week.

The implication for enterprise deployment could be significant. Rather than routing requests to different specialized models based on task complexity — a common but operationally burdensome pattern — organizations can now use a single model endpoint and adjust reasoning depth based on the task at hand. Routine document summarization can run on low thinking with fast response times, while complex analytical tasks can be elevated to high thinking for Deep Think–caliber reasoning.

Benchmark Performance: More Than Doubling Reasoning Over 3 Pro

Google’s published benchmarks tell a story of dramatic improvement, particularly in areas associated with reasoning and agentic capability.

On ARC-AGI-2, a benchmark that evaluates a model’s ability to solve novel abstract reasoning patterns, 3.1 Pro scored 77.1% — more than double the 31.1% achieved by Gemini 3 Pro and substantially ahead of Anthropic’s Sonnet 4.6 (58.3%) and Opus 4.6 (68.8%). This result also eclipses OpenAI’s GPT-5.2 (52.9%).

The gains extend across the board. On Humanity’s Last Exam, a rigorous academic reasoning benchmark, 3.1 Pro achieved 44.4% without tools, up from 37.5% for 3 Pro and ahead of both Claude Sonnet 4.6 (33.2%) and Opus 4.6 (40.0%). On GPQA Diamond, a scientific knowledge evaluation, 3.1 Pro reached 94.3%, outperforming all listed competitors.

Where the results become particularly relevant for enterprise AI teams is in the agentic benchmarks — the evaluations that measure how well models perform when given tools and multi-step tasks, the kind of work that increasingly defines production AI deployments.

On Terminal-Bench 2.0, which evaluates agentic terminal coding, 3.1 Pro scored 68.5% compared to 56.9% for its predecessor. On MCP Atlas, a benchmark measuring multi-step workflows using the Model Context Protocol, 3.1 Pro reached 69.2% — a 15-point improvement over 3 Pro’s 54.1% and nearly 10 points ahead of both Claude and GPT-5.2. And on BrowseComp, which tests agentic web search capability, 3.1 Pro achieved 85.9%, surging past 3 Pro’s 59.2%.

Why Google chose a ‘0.1’ release — and what it signals

The versioning decision is itself noteworthy. Previous Gemini releases followed a pattern of dated previews — multiple 2.5 previews, for instance, before reaching general availability. The choice to designate this update as 3.1 rather than another 3 Pro preview suggests Google views the improvements as substantial enough to warrant a version increment, while the “point one” framing sets expectations that this is an evolution, not a revolution.

Google’s blog post states that 3.1 Pro builds directly on lessons from the Gemini Deep Think series, incorporating techniques from both earlier and more recent versions. The benchmarks strongly suggest that reinforcement learning has played a central role in the gains, particularly on tasks like ARC-AGI-2, coding benchmarks, and agentic evaluations — exactly the domains where RL-based training environments can provide clear reward signals.

The model is being released in preview rather than as a general availability launch, with Google stating it will continue making advancements in areas such as agentic workflows before moving to full GA.

Competitive implications for your enterprise AI stack

For IT decision makers evaluating frontier model providers, Gemini 3.1 Pro’s release has to not only make them rethink which models to choose but also how to adapt to such a fast pace of change for their own products and services.

The question now is whether this release triggers a response from competitors. Gemini 3 Pro’s original launch last November set off a wave of model releases across both proprietary and open-weight ecosystems.

With 3.1 Pro reclaiming benchmark leadership in several critical categories, the pressure is on Anthropic, OpenAI, and the open-weight community to respond — and in the current AI landscape, that response is likely measured in weeks, not months.

Availability

Gemini 3.1 Pro is available now in preview through the Gemini API in Google AI Studio, Gemini CLI, Google Antigravity, and Android Studio for developers. Enterprise customers can access it through Vertex AI and Gemini Enterprise. Consumers on Google AI Pro and Ultra plans can access it through the Gemini app and NotebookLM.

• CarGurus reportedly hit by ShinyHunters vishing attacks
• Hackers claim to have stolen 1.7 million records
• CarGurus is staying queit for now

Online car marketplace CarGurus is allegedly the latest company to fall prey to ShinyHunters’ vishing attacks.

The notorious hacking collective posted a new note on its data leak site warning CarGurus to act quickly or have their sensitive data posted on the dark web.

Everyone who loves mysteries secretly hopes that one day life will drop an intriguing puzzle into their lap for them to solve. Maybe not an Agatha Christie-type crime, but something that will send them on a real-world chase to connect the dots and land at a satisfying conclusion.

That’s exactly what happened to Katie Elkin, a retired teacher with a penchant for mysteries. “I’m 84 and I have lived a full, wonderful life,” she tells me over a video call from her home in Prescott, Arizona.

Until now, Elkin’s mysteries have largely been genealogy-based. She recounts an extraordinary story about making friends with a woman from California and discovering that their grandfathers had trained together in the Army and then shipped out to France in World War I on the same day. “That’s my whole life,” she says. “It’s coincidences.”

On this Friday in February, we’re talking about another coincidence in Elkin’s life — one of finding a phone, lost for a decade in the desert, and Elkin’s attempt to reunite it with its owner. 

Our phones are immensely personal items, serving both as memory banks that store our most precious data and as portals that connect us with every important person in our lives. These days, if we lose them, tracking technology means there’s every chance we could be quickly reunited with them, but that hasn’t always been the case. 

Those disappearances can be high-stress moments for anyone — just ask Apple about the unreleased iPhones it lost back in 2010 and 2011, which, coincidentally, were around the same time it introduced the Find My iPhone feature. But even today, recovering a lost phone means relying to an extent on the goodwill and honesty of the person who found it. Many people will choose to do the right thing in this scenario, and some — like Elkin — will go above and beyond to help out a stranger.

On a sunny day just before Thanksgiving, Elkin and her husband drove about 10 minutes west of the city to spend some time outdoors. Prescott is surrounded by national parks and ponderosa pine forest, but on this day, Elkin was headed to the desert — not for a hike, she says, but an “amble.”

Rather than taking the well-marked trail popular with hikers and ATVs, Elkin instead split off onto a lesser-known path “obliterated by the grasses and the weeds.”

It was Elkin’s dad who taught her that if she wanted to spot something, she should look for it — sage advice that’s served her well over the years. “He was always finding change,” she says. “And I can do that too. I always find animals. If we’re driving, I can see them in the woods … I’m always looking for something.”

Looking for a vague something can turn up the oddest of things, and on that particular day, the something Elkin found was a dusty, beaten 2012 Samsung Gusto 2 lying on its side, clamshell open in the scrub.

Elkin picked up the phone, thinking she would give it to a neighbor boy who liked to take electronics apart. But when she got it home, she was struck by another idea — what if she could get the phone to turn on? 

Like many of us with a drawer full of mystery cables, Elkin has kept all the cords and wires that have come with the electronics she’s purchased over the years. She dug through her stash and found a charger that fit the Gusto (she still has no idea what it was used for previously).

When CNET reviewed the Gusto 2 — a simple flip phone that came out the same year as the iPhone 5 and the Samsung Galaxy S3 — we said: “the construction seems strong enough to withstand multiple drops and endless opening and closing.” Our instincts about its potential resilience were, it turns out, correct.

“I couldn’t believe it when it came up charging,” Elkin says. It took a little while, but when the phone turned on, she was ecstatic. “I thought, ‘Oh my gosh, I wonder who this phone belongs to?’ And so that was when the mystery began.”

The quest for answers

Elkin went into the text messages and started to piece together the Gusto owner’s life, clue by clue. The owner worked in a cafe, she seemed to have family connections in Chicago, she was a renter and a keen hiker. Her name was Maddie.

The other thing Elkin noticed was that the last message was marked Saturday, May 16. It was the only evidence she had to indicate when exactly the phone might have been lost. She went to the internet and looked up which years May 16 had fallen on a Saturday. Two possible answers cropped up — 2020 and 2015.

Elkin’s internet research didn’t stop there. She took one of the commonly texted numbers in the phone and did a reverse lookup. “And bingo! I found a woman’s name that had that phone number,” she says. But when she called the number, it was disconnected.

“I said to myself, who would know where she is?” says Elkin. “Her dad would know.” She found a number listed under “daddio,” performed another reverse lookup and found the name of a man living in Chicago. “I was so excited because I was getting close,” she says.

On Dec. 30, Elkin’s birthday, she called the number, but no one picked up. She had to leave a message. “I was really disappointed, because I wanted to talk to somebody,” she says.

Ten minutes later, her phone rang, but when she picked up, it wasn’t a man on the other end of the line. “It was Maddie, the owner of the phone,” she says. “She had come to Chicago to visit her dad for the holidays.”

Elkin and Maddie talked for around 10 minutes. “She was amazed,” says Elkin. “We were both amazed.” Maddie didn’t want her phone back, but it turns out she had lost it in 2015 after hiking in the exact spot that Elkin had found it. 

The little phone that could

For a decade, the little Gusto had been lying out in the desert. Unlike some parts of Arizona, Prescott has four seasons, with all the minus temperatures, scorching heat, snowfall and summer storms that come with them. The Gusto weathered every storm, and battered and bruised as it was, it still came back to life.

We have little expectation these days that our phones will last us a long time, and we rarely get all the life out of our devices that they’re capable of offering us. Rather than seeking to get them repaired, once they fail us in one respect, we tend to seek out replacements. Most Americans hang onto their phones for an average of 2.5 years, according to a Reviews.org survey.

It turns out, though, that some phones are built to last, and the Gusto was one of them. After Elkin had spoken with Maddie, she reached out to Samsung to let them know her story. “I said to myself, ‘Does Samsung require some kudos for having a product that lasted that long?’”

Any tech company would. My own first phone, a 2002 Sagem MW 3020, gave up the ghost simply by being exposed to the concept of water while wrapped up inside a backpack on a rainy day. In spite of the best efforts of phone-makers to increase display resiliency, many people are still walking around out there with cracked screens.

For as long as we’ve had mobile phones, they’ve been vulnerable pieces of kit. But whatever secret sauce Samsung put inside the 2012 Gusto 2 shows that it was more robust than most — even though it was lying open with its main screen exposed when Elkin found it.

At the time we reviewed the Gusto 2, we gave it a score of 7 out of 10, with points knocked off for its subpar screen resolution and a smaller-than-usual headphone jack. It’s too late for us to go back and revise that score in light of what we know about how robust the phone is 14 years later, but it’s entirely possible that the “problems” we highlighted actually played into the Gusto’s long-term survival.

Elkin still doesn’t know what she’s going to do with Maddie’s Gusto, although a friend has suggested that Samsung clad it in gold and put it on a pole at headquarters. Samsung is clearly proud of the phone’s durability, having put me in touch with Elkin, but is also undecided about how to celebrate the life the Gusto 2 has lived. In spite of Elkin’s love for mysteries and my suggestion that the FBI recruit her, she isn’t about to start a detective agency to reunite other people with their lost possessions. “It’s just a hobby,” she laughs.

That’s a shame. As someone who’s lost more than one phone over the years, I would dearly love to be reunited with my missing technology, and I’m sure there’s a market for Elkin’s skills. Not every phone is as resilient as the Gusto. Most devices that have taken such a battering would likely refuse to even turn on. 

Perhaps there’s a longevity challenge for all phone-makers. I can’t promise CNET would be able to replicate this scenario in our reviews testing process, but in an age of disposable tech, it would be lovely to give extra points for truly hard-earned durability.