Evolving Rodents: An upcoming crowdfunding campaign aims to introduce a novel approach to mouse interaction. The Rotary Mouse replaces the traditional scroll wheel with a rotating mechanism, which, according to its developers, is designed to feel as natural as possible.

The Rotary Mouse from Airra Labs will debut on Kickstarter at a significant discount and with free shipping to early backers. The company says there is real demand for the product, despite its unconventional design.

The campaign is scheduled to open on June 17. The device replaces the traditional vertical scroll wheel with a built-in rotary mechanism, which it claims delivers up to 2.5x faster scrolling compared to a standard mouse. The company also says the system is smoother and more natural than conventional scrolling wheels.

While the rotary design may appear unfamiliar, it is intended to improve scrolling performance and ergonomics. The wheel includes tactile feedback and supports both traditional vertical scrolling and a counterclockwise rotation mode. Additional features include a range-of-motion function designed to reduce finger strain.

Airra Labs argues that the mouse wheel has remained largely unchanged for the past 30 years. Users still flick small wheels to browse web pages, documents, spreadsheets, source code, and more. In recent years, some mice have adopted “infinite scroll” modes, but the company describes them as slow, repetitive, and difficult to control.

Indeed, using the wheel on my Logitech G502 Hero felt weird and cumbersome at first, but I have since adapted my scrolling habits and I don’t plan on changing anytime soon.

The Rotary Mouse is intended to reduce scrolling strain and make content navigation significantly faster. Airra claims the rotary wheel can outperform infinite scrolling by offering more precise control across different use cases, including video editing, software development, spreadsheets, document work, and long-form web browsing. The wheel can also function as a miniature steering control for driving simulators.

The project appears to be ready for production, but Airra is turning to crowdfunding to cover costs such as injection molding, certification, packaging, and warranty support. The company says mass production is expensive and requires additional operational expertise.

The Rotary Mouse is expected to be priced between $49 and $109. However, early crowdfunding backers will receive a discount of up to 65% off the final retail price. The wireless mouse supports a DPI range of 800 to 1600, suggesting it may have limited appeal among competitive gamers.

Something to look forward to: Competition in the APU market could intensify when Intel and Nvidia reveal the fruits of their announced collaboration against AMD. While no official timeline has been confirmed, reports suggest users could get a first look at Intel-Nvidia chips within the next 18 months.

Tech journalist Erdi Özüağ reports that Intel’s current roadmap targets the first quarter of 2028 for its initial x86 processors with Nvidia RTX graphics, with a possible reveal at CES. The long-anticipated chips are expected to challenge AMD’s high-end mobile APUs and bolster both companies’ broader platform ambitions.

Intel’s plans are likely still in flux this far out, but its current roadmap puts the chips in early 2028 barring delays. Intel and Nvidia officially confirmed they were developing these chips in September 2025, though specific details have been scarce since.

– Erdi Özüağ (@fx57) June 15, 2026

What is known is that the x86 SoCs will combine Intel CPU cores with Nvidia RTX GPU chiplets – likely using a high-bandwidth interconnect – across a range of devices. Reports suggest Serpent Lake will pair Intel’s Titan Lake CPU cores with a GPU based on Nvidia’s next-generation Rubin architecture.

The chip is rumored to be fabricated on TSMC’s N3P process node and is expected to support LPDDR6 memory, providing the bandwidth necessary for both high-end gaming and AI workloads. Its primary target appears to be AMD’s Strix Halo APUs in the high-end laptop segment.

Intel and Nvidia’s partnership could also shake up the mobile gaming market. AMD’s Hawk Point, Strix Point, and Strix Halo chips have powered the category in devices like the Steam Deck, Asus ROG Xbox Ally, and Lenovo Legion Go. More recently, however, MSI’s Claw 8 EX AI+ marked the debut of Intel’s Arc G3 Extreme – a Panther Lake design built on the 18A process node – giving Intel a foothold in the handheld gaming space. The chip also supports Intel’s XeSS upscaling technology.

Özüağ also reports that Apple’s negotiations with Intel over the use of its 18A foundry node are progressing, with the Cupertino company looking to reduce its dependence on TSMC and respond to political pressure to expand domestic manufacturing.

Initial shipments are not expected before the second or third quarter of 2027, and all plans remain subject to Intel’s ability to improve yield, performance, and cost competitiveness on its 18A process.

In 2017, Respond.io set out to solve a simple problem: businesses couldn’t keep up with customers who had moved to messaging apps. Today Respond, with its customer conversation management software, has become one of the tech success stories of Malaysia.

The startup, headquartered in Kuala Lumpur, has raised a $62.5 million Series B round led by Camber Partners, with participation from Endeavor Catalyst and existing investors. It last raised a $7 million Series A in 2022. The company has grown to $35 million in annual recurring revenue (ARR), growing 169% year-over-year, at a 30% profit margin, it tells TechCrunch.

Co-founder and CEO Gerardo Salandra, who worked at IBM and Google before joining Runtastic, a fitness tracking app that was sold to Adidas in 2015, founded Respond in Hong Kong in 2017 alongside Hassan Ahmed (CTO) and laroslav Kudritskiy (COO). The team relocated the business to Malaysia two years later.

The platform helps mid- to large-sized B2C businesses drive revenue from customer conversations across multiple messaging channels including WhatsApp, Instagram, TikTok, Messenger, Line, Telegram, WeChat, voice calls and web chat. It also uses AI agents to automatically handle high volumes of customer inquiries, qualify leads and close sales without human intervention.

Salandra described its core customers as “high-consideration” businesses, where customers need to talk to someone before buying, such as healthcare, automotive, retail, education and travel. “You don’t go to a website, put your credit card, and buy a car; you chat with someone, you ask a lot of questions,” he said. Its sweet spot is companies with 200 to 10,000 employees.

The rise of AI has raised an obvious question for platforms like Respond: Can tools like ChatGPT simply replace what they’ve built?

Salandra thinks his foothold is strong enough to stop such encroachment, should it come. The company is currently processing 2 billion messages per quarter.

“If I just look at the numbers, every day that AI becomes more prominent, we grow faster,” he told TechCrunch. “We are not seeing what the public SaaS markets are seeing.”

Part of that comes down to pricing, he said. Unlike enterprise software competitors that charge per seat, Respond charges based on the volume of customer conversations, meaning it doesn’t matter whether a human or an AI is answering. “When fewer humans use your product, they make less money,” he said. “But we don’t charge like that.”

The incumbent platforms, particularly those dominant in North America and Europe, were built around email and phone calls. “The platforms that exist, they bolted on messaging as a second thought. They’re very email focused, they’re very call focused, but when it comes to messaging, it’s an afterthought,” Salandra said.

That volume of message data creates a feedback loop, according to the CEO. More messages mean better AI. Better AI attracts more customers. More customers generate more messages. “This is what we call the data flywheel,” Salandra said. He added that the head start matters for any upstart AI company, as well. “Because we started so long ago and we have such a strong foundation, we can provide better AI compared to someone who just entered into the messaging space.”

With the new capital, Salandra said the company plans to pursue hiring, organic growth and acquisitions. The CEO has two types of buying targets in mind: bolt-on technology that fits into its existing ecosystem, and established teams with strong customer bases in strategic markets like Europe and North America. “Imagine how many months I can save if I find the right company that maybe already has the clients and the team,” he said. “I can save myself six months to a year through an acquisition.” He confirmed the company is already in talks with a couple of potential targets.

The geographic push makes strategic sense. Respond currently generates roughly 30% of its revenue from APAC, 30% from Latin America, and 20% from the Middle East and Africa, leaving North America and Western Europe at just 20%. But Salandra says those regions are now its fastest-growing. “They took longer to make the change, but now they’re moving very rapidly into messaging channels,” he said, adding that he expects both regions to become the company’s largest segment within two to three years.

Despite the fresh injection of capital, Salandra is cautious about what comes next. “We don’t want to be a growth at all costs company,” he said. “Even with this money, we’re going to be very disciplined.” But Salandra has bigger plans in mind. “My favorite outcome?” he said. “Ringing the bell at Nasdaq.”

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

Need some help with today’s Mini Crossword? Read on. And if you could use some hints and guidance for daily solving, check out our Mini Crossword tips.

If you’re looking for today’s Wordle, Connections, Connections: Sports Edition and Strands answers, you can visit CNET’s NYT puzzle hints page.

Read more: Tips and Tricks for Solving The New York Times Mini Crossword

Let’s get to those Mini Crossword clues and answers.

Mini across clues and answers

1A clue: Worker who sticks a thick mix on bricks
Answer: MASON

6A clue: Volvos or VWs
Answer: AUTOS

7A clue: “Wow, that’s wild!”
Answer: CRAZY

8A clue: Peaceful place
Answer: HAVEN

9A clue: Key above ~ on a Mac
Answer: ESC

Mini down clues and answers

1D clue: ___ 1 (speed of sound)
Answer: MACH

2D clue: Emanating glow
Answer: AURA

3D clue: Fend (off)
Answer: STAVE

4D clue: Seeps slowly
Answer: OOZES

5D clue: Boy band with the 2000 hit “It’s Gonna Be Me”
Answer: NSYNC

offbeat

Worried that an unexpected strike could take out critical orbital systems, Pentagon researchers want to know how fast the industry thinks it could launch replacements

War may never change, but its domains evolve, and DARPA is looking for ideas to ensure space infrastructure destroyed in future orbital skirmishes can be rapidly replaced. 

DARPA, on Friday, put out a request for information for an initiative to develop what it’s calling Rapid Reconstitution of Space Capabilities. 

“Other nations seek to position themselves as leading space powers while undermining the stability and tranquility that allows space to benefit all nations,” DARPA said, suggesting that the US would never dare deploy space weapons that could destabilize the tranquility of Earth orbit.

“Space is an increasingly contested environment, presenting a multitude of threats to U.S. space assets,” DARPA added. “Therefore, there is a strategic need to be able to quickly respond to disrupted assets and reconstitute degraded space capabilities.”

While we don’t know if the US has any weapons in space – we asked but didn’t get a response – other countries certainly are striking an aggressive posture.

Both Russia and China have reportedly blown up their own defunct satellites in recent years to demonstrate their space warfare capability, and the US Space Force has noticed what appears to be China experimenting with orbital satellite dogfighting maneuvers. The US has also accused Russia of developing anti-satellite weaponry that may or may not involve orbital nukes, leading the US to update its fleet of satellites designed to keep an eye out for potential nuclear launches. 

“U.S. competitors are implementing a sustained effort to develop a broad range of offensive counterspace capabilities through a variety of anti-satellite (ASAT) weapons, including direct attacks on satellites, jamming and spoofing of signals, and continued cyberattacks on satellite and ground infrastructure,” DARPA noted in Friday’s announcement. 

Pointing to the 2023 Space Force tactically responsive space exercise Victus Nox, which saw the USSF launch a space vehicle into orbit just 27 hours after getting the word, DARPA said it wants more of the same, but hopefully faster.

“DARPA Strategic Technology Office seeks information supporting technical solutions and operational concepts and strategies to enable rapid, responsive, cost-effective reconstitution of any lost or degraded space capabilities resulting from attacks,” DARPA explained, adding that it’s not looking for anything more than ideas at this point, but is willing to entertain anyone in the US with a good idea, be they laboratory or private outfit. 

According to the announcement, DARPA wants ideas that would get degraded operations restored in “hours to weeks,” and offer the same turnaround time for cases of surging demand as well as asset loss. 

“Possible solutions could be realized with reconfigurable, software-defined, multifunctional, and multi-mission payloads, as well as proliferated/mesh architectures and rapid on-orbit deployment concepts,” the Pentagon research arm said. 

“Rapid space capability reconstitution is a complex task,” DARPA added, so don’t expect this research to move anywhere near the speed of DARPA’s eventual rapid reconstitution rockets. 

Then again, America just minted the world’s first trillionaire, and he’s a space guy – maybe ask him how to launch rockets quickly? Surely his ideas would be grounded in good sense, right? 

Rivian CEO RJ Scaringe said the company will ship supervised point-to-point self-driving on all of its second-generation vehicles and the R2 later this year, describing the capability as “very similar to Tesla’s FSD.” Speaking at the Masters of Scale event in Anaheim on Thursday, Scaringe laid out a three-stage autonomy roadmap: supervised point-to-point driving in 2026, eyes-off unsupervised driving in 2027, and a commercial robotaxi service with Uber beginning in 2028.

The announcement represents a significant jump from Rivian’s current driver-assistance system. Universal Hands-Free, which rolled out in late 2025, handles steering and speed on roughly 3.5 million miles of marked roads in the US and Canada. It does not navigate turns, traffic lights, roundabouts, or parking lots.

Point-to-point driving would extend the system’s capabilities to handle complete journeys from origin to destination, similar to what Tesla’s Full Self-Driving Supervised already attempts. The leap from highway lane-keeping to full urban navigation is the hardest problem in autonomous driving, and no company has solved it without significant constraints.

“Later this year, we’ll have full supervised point-to-point, which will be very similar to Tesla’s FSD,” Scaringe said. “And that’ll roll out to all of our Gen 2 vehicles and, of course, R2.” He did not specify a month or quarter for the rollout, and Rivian has not publicly demonstrated the point-to-point system in an uncontrolled environment.

The comparison to Tesla is deliberate but architecturally inexact. Tesla’s FSD relies exclusively on cameras, while Rivian’s platform integrates 10 external cameras, five radar units, 12 ultrasonic sensors, and a high-precision GPS receiver. Rivian began delivering R2 SUVs earlier this month, and future R2 models will add a roof-mounted LiDAR sensor and the company’s custom RAP1 processor, a 5nm chip delivering up to 1,600 trillion operations per second.

The pricing undercut is sharper than the technology comparison. Rivian’s Autonomy+ package costs $2,500 as a one-time purchase or $49.99 per month, compared with Tesla’s FSD at $8,000 or $99 per month. Whether the lower price reflects a competitive strategy or a difference in capability remains to be seen, given that Rivian’s point-to-point system does not yet exist as a shipping product.

Rivian’s autonomy software is built around what the company calls a Large Driving Model, a foundational AI system trained end-to-end through reinforcement learning. The LDM maps raw sensor input directly to vehicle trajectory, analysing multiple driving paths and selecting the optimal one using a technique called Group-Relative Policy Optimization. The approach mirrors the end-to-end neural network philosophy Tesla adopted with FSD v12, though Rivian’s multi-sensor hardware gives the model a wider range of input data to work with.

The 2027 eyes-off milestone is where the roadmap becomes commercially consequential. Supervised driving, regardless of how capable, still requires a human to watch the road. Tesla has been promising unsupervised FSD for years and has pushed the timeline repeatedly, most recently to Q4 2026 at the earliest. Scaringe has said Rivian targets Level 3 autonomy by 2028 and Level 4 by 2030, timelines that no autonomous driving company has consistently met.

The commercial centrepiece of the roadmap is the $1.25 billion deal with Uber announced in March. Uber committed an initial $300 million investment, with the remainder contingent on Rivian hitting autonomous performance milestones through 2031. The deal calls for Uber or its fleet partners to purchase 10,000 fully autonomous R2 robotaxis, with an option for up to 40,000 more beginning in 2030. Commercial deployment is planned for San Francisco and Miami in 2028, expanding to 25 cities by 2031.

Those targets depend on Rivian achieving something it has not yet demonstrated: a vehicle that can drive itself without human supervision. The company’s Gen 3 autonomy platform, which will power the robotaxi programme, is still undergoing validation. The initial R2 production run launched without the Gen 3 hardware, meaning the robotaxi-grade vehicles are at least one hardware generation away from production.

Scaringe framed the self-driving push as essential to Rivian’s long-term economics. The company posted a net loss of $3.63 billion in 2025 despite achieving its first full-year positive gross profit at $144 million. Autonomy, if it works, transforms the revenue model from selling cars to operating a transportation platform. But the gap between announcing a roadmap at a conference and shipping a reliable autonomous system is where most self-driving timelines have historically broken down.

Retroid has brought back the 12GB RAM configuration of the internet-famous Pocket 6, returning it to the lineup at the same $279 price point as the discontinued version.

The original 12GB RAM variant was pulled from sale earlier this year after Retroid cited the rising cost of RAM and storage, leaving the Pocket 6 available only in its 8GB RAM configuration.

However, the updated model comes with the catch of reduced internal storage. That storage reduction is the most consequential change in the new configuration, with the returning 12GB model shipping with 128GB of UFS 3.1 storage rather than the 256GB that the discontinued version carried, effectively halving the built-in capacity at an unchanged retail price.