Elementary school is tough.

There are playground politics, multiplication tables and learning to read.

Imagine dealing with all that in a new language — or even a whole new country.

That’s the added challenge for kids who are learning English at the same time they’re learning everything else as their peers.

It’s an issue that Sarah Walters and her colleagues were determined to tackle in Troy City Schools, a public school district made up of nine campuses roughly an hour north of Cincinnati. The area is home to an automotive manufacturer that brings some employees — and their families — over from Japan.

Roughly 3 percent of the 4,000 students have primary languages like Spanish, Ukrainian and Japanese, a relatively small population compared to the most recent national average of 11 percent.

But that small group is making big gains. Looking to close the literacy gaps that have plagued schools since the pandemic, the district took a big swing to increase literacy among its English learners. It trained 116 staff members — including every elementary teacher, intervention specialist, paraprofessional and principal — in the Orton-Gillingham approach, which folds movement and touch into reading instruction.

They say it’s paying off.

Walters, a literacy instructional support specialist, says that helping multilingual students master their grasp on English is vital. Like any other student, the foundation that they lay in reading and math will affect their learning from that point on.

“We want to help the students continue to thrive, and really everything that we’re thinking about with our student services is equitable learning opportunities,” Walters says.

Moving Toward Equity

Federal data shows that English learners’ achievement scores lag far behind their peers on average, and have made little improvement over the past two decades.

Troy City Schools was eager to close widening literacy gaps that surfaced after the onset of the pandemic, Walters says, which was particularly hard on English learners like those at Concord Elementary. A big hurdle was phonics, the letter sounds that make up words.

“We were seeing a lot of student frustration and wanting to give up,” Walters recalls. “Students being very withdrawn, those social-emotional impacts.”

Back in 2020, English-language instruction was inconsistent and fragmented across classrooms.

Yet, even with the desire to boost English learner scores, the program took some time.

Following the pandemic, Troy City Schools mulled over the changes for three years before it had enough funding to deliver on it, according to Danielle Romine, director of elementary teaching and learning for the district. The effort was funded through post-COVID relief grants and budget allocations made by the district’s leaders.

As a literacy specialist, Walters became certified in the Orton-Gillingham method through the Institute for Multi-Sensory Education. She’s now responsible for supporting and training staff to successfully use the techniques.

Walters says teachers and staff were trained to utilize drills that connect literacy concepts through visuals, sound and movement. Students might use flash cards as a visual element or tap their fingers to each letter as they spell out a word. Students also learn the origin and history of words to strengthen their ability to decode them. For example, a “red word” is one that does not follow phonics rules.

“Our multilingual learners love it because no longer are they being told, ‘That’s just the way it is,’” Walters says.

After an initial summer training on the Orton-Gillingham approach, teachers spoke so highly of the method that requests for training grew among staff.

Initial Promise

“In a school district, if you want to get something out, just tell a teacher, because it [will] spread like wildfire,” Romine says.

And the data are showing promising results, Walter says. The district-wide third grade reading proficiency had plummeted to 56 percent in 2021-22 but had risen to 81 percent by 2023-24 — slightly higher than its pre-COVID achievement rate. The most recent state data shows Concord Elementary far surpassed its target goal for English proficiency among multilingual students.

Walters has heard from teachers who say that the approach has helped some English learner students make lightning-fast gains in reading. One educator told her that two students from Japan who joined the elementary school in the fall were conversing in English by December. Another student’s phonics diagnostic score shot up by 38 points in the same timeframe.

Now, the district is working to spread the method beyond its own campuses.

“Eventually, our goal is to support the entire community, or the entire county because Sarah having that training [enables her] to support teachers from other districts, as well,” Romine says.

But for English learners, ensuring they’re on grade level in reading goes beyond measuring their success in the classroom.

Walters says that the district is thinking about long-term learning for children who, for example, may be in the U.S. for a few years before returning to Japan.

Now, the district is working to spread the method beyond its own campuses.

“Eventually, our goal is to support the entire community, or the entire county because Sarah having that training [enables her] to support teachers from other districts, as well,” Romine says.

But for English learners, ensuring they’re on grade level in reading goes beyond measuring their success in the classroom.

Walters says that the district is thinking about long-term learning for children who, for example, may be in the U.S. for a few years before returning to Japan.

“We want students to have success across math, science, everything,” Walters says. “So it’s important that we get them up to speed as quickly as possible, because those long-term impacts could really be harmful for them. That early literacy is key.”

A man with ALS has been using a brain implant to speak independently for more than 3,800 hours over the past two years, producing nearly 2 million words with an average speed of 56 words per minute. The study, published Monday in Nature Medicine by researchers at the University of California, Davis, represents the longest sustained demonstration that a brain-computer interface can function as a practical daily communication tool outside a laboratory. Casey Harrell, the 47-year-old participant, has used the system to return to full-time work as an environmental advocate.

The implant consists of four microelectrode arrays placed in Harrell’s left precentral gyrus, the brain region that coordinates speech, recording activity from 256 cortical electrodes. Machine learning algorithms built into a software platform called BRAND, developed by UC Davis postdoctoral fellow Nicholas Card, translate that neural activity into English-language phonemes, then map those phonemes to words and sentences. The system reads out the decoded text in a synthesised version of Harrell’s pre-ALS voice.

In controlled testing with a 125,000-word vocabulary, the system scored over 99% word accuracy. In daily use outside the lab, Harrell rated 92% of sentences as accurate or mostly correct. During the study period, he communicated more than 183,000 sentences.

The 💜 of EU tech

The latest rumblings from the EU tech scene, a story from our wise ol’ founder Boris, and some questionable AI art. It’s free, every week, in your inbox. Sign up now!

“The key thing to me is that it’s enabling everyday communication for a guy who wants to talk but can’t,” neurosurgeon David Brandman, who implanted the device in 2023 and co-led the study, told The Register. “Despite being paralysed, he has gone back to work full time and has meaningful conversations with his daughter who’s never heard the sound of his voice.”

The study’s significance lies not just in accuracy but in independence. Previous BCI systems required researchers to be present in the patient’s home whenever the device was in use, or required the patient to travel to a lab. Harrell’s system is operated by his home care team, with no researcher support needed.

Based on the study’s timeline, he averaged more than five hours of daily use.

The UC Davis team is part of BrainGate, the consortium of universities and the US Department of Veterans Affairs developing brain-computer interfaces for speech restoration, computer control, and movement recovery. The hardware itself is not custom-built, using existing microelectrode arrays produced by Blackrock Neurotech. The breakthrough is in the software, specifically the BRAND platform’s machine learning algorithms that decode attempted speech from neural signals in real time.

Brandman compared the current state of BCI technology to early pacemakers, which in the 1950s required external wiring to large batteries or wall power. Seventy years later, pacemakers are implanted in outpatient procedures. “We’re at the early stages of this kind of technology,” Brandman said.

Harrell is still wired to external computers, but the UC Davis team’s AI advances combined with hardware miniaturisation work at companies like Neuralink, Synchron, and Paradromics point toward a future where the setup is far less cumbersome.

The competitive landscape in BCI is accelerating. Neuralink has implanted devices in at least 21 patients under research protocols but lacks commercial approval. China approved the first commercially available invasive BCI earlier this year.

Other approaches to restoring speech for people with ALS use AI voice conversion rather than brain implants, but those methods require the patient to retain some vocal ability.

What distinguishes the UC Davis work is its demonstration that a BCI can cross from laboratory experiment to sustained, practical daily tool. The 3,800 hours of brain recording also constitute the largest individual neural dataset with single-neuron resolution ever collected, according to co-principal investigator Sergey Stavisky, which will inform future improvements to the decoding algorithms.

The system remains an investigational device, limited by federal law to research use, and has been tested on a single patient. Whether the results generalise to other ALS patients, or to people with different neurological conditions, is not yet known. Scaling the technology from a clinical trial to a prescribed medical device will require regulatory approval, hardware miniaturisation, and cost reduction that could take years.

“I want desperately to not be unique or special, because that will mean I no longer have the disease or that everyone that has the disease like me can get this prescribed to them,” Harrell said through his BCI system.

SYSTEMS

Running low on memory, can’t afford more? The House of Zen’s latest acquisition puts an AI spin on flash-based memory expansion

With no end in sight to the memory crunch, AMD thinks that AI, the main cause of the shortage, could be part of the solution. This week, the House of Zen acquired predictive memory startup Mext for an undisclosed sum, setting the stage for a world where bots decide which data to put into RAM and which to store in less-expensive flash.

Founded in 2023, the Mext proactive memory platform uses machine learning algorithms and learned heuristics to proactively offload “cold” memory to flash storage, and, based on data access patterns, restore it before its needed again.

Modern flash arrays are already approaching main memory in terms of aggregate bandwidth, but swapping to disk still imposes a stiff latency penalty.

Mext claims it can expand the effective memory of a system by 2 to 4x using flash, which gig for gig is still vastly less expensive than DRAM.

This flash memory is exposed to the operating system like regular memory simply by running the Mextd daemon.

Memory tiering is nothing new and has seen various reincarnations over the years with some being software based and others, like Intel Optane persistent memory, using special 3D XPoint memory tech co-developed by Micron.

Mext stands out for its use of machine learning to migrate data from hot memory to cold storage almost like a branch predictor — something AMD has an awful lot of experience with.

Mext isn’t using one model to decide when to shuffle your data. Instead it uses a series of heuristics, long short term memory, and modern transformer architectures depending on which combination renders the best results.

“This approach has the potential to reduce infrastructure costs, improve resource utilization, and help customers more effectively scale general-purpose and AI workloads,” Dan McNamara SVP of AMD’s compute and enterprise AI biz wrote in a blog post this week.

Beyond enterprise applications, the technology could have implications for AI serving.

Modern mixture of experts (MoE) models are, as their name suggests, comprised of multiple sub-models.

For each token predicted, a different selection of experts may be used. In practice an LLM may use some experts more frequently and others rarely.

We wouldn’t be surprised to see AMD use Mext’s prediction algorithms to offload infrequently utilized experts from HBM to slower system memory, enabling enterprises to take advantage of larger more capable models with fewer resources.

That’s just speculation of course, but we’ve reached out to AMD for comment; we’ll let you know if we hear anything back. ®

After multiple delays, Shrek is back, and he is bringing the whole family time! DreamWorks has released the first trailer for Shrek 5, and it is exactly as chaotic as you would hope. The film marks the franchise’s return after 17 years since Shrek Forever After in 2010, and it hits theaters on June 30, 2027.

The wild moments in the Shrek 5 trailer

Mike Myers, Eddie Murphy, and Cameron Diaz are all back as Shrek, Donkey, and Fiona. The trailer opens with a storybook recap of the original story, before Donkey interrupts to announce it is time for a makeover.

The gang, now joined by Shrek and Fiona’s three grown-up kids, head to a new city called Further Further Away, which is sketchier than Far, Far Away ever was.

Advertisement

The trailer wastes no time in serving up a Gingerbread Man with two gumdrops strapped to his butt, proudly declaring he is “caked up like a friggin’ bakery,” followed by some very enthusiastic twerking.

There is also a melting, grimy snowman lurking in a back alley asking, “Wanna date a snowman?” taking a dig at Frozen’s Olaf.

Where is Zendaya’s character Felicia in the Shrek 5 trailer?

Zendaya voices Felicia, Shrek and Fiona’s daughter, and her involvement in Shrek 5 has been confirmed. However, she is completely absent from the trailer footage. Shrek can be seen in the jail cell with Fiona and his two sons, Fergus and Farkle, but Felicia is nowhere to be found. Whether Felicia got kidnapped or simply had better weekend plans is anyone’s guess.

Why are fans upset about the Shrek 5 trailer?

Not everyone is thrilled with how the trailer looks, though. Many fans have taken issue with the updated animation style, which makes Shrek and Donkey look noticeably more detailed and lifelike than in the previous films.

its actually insane that a movie from 15+ years ago has infinitely better lighting, skin textures than movie coming out in 2027. how did technology regress this badly?? they need to throw out the new assets and bring back the classics. #fixourshrek #shrekisnotminions #notmyshrek

— Lol Hekek (@LHekek99269) June 16, 2026

Comments on Universal’s Instagram ranged from “we want the old Shrek back” to “do what they did with Sonic,” a reference to how Paramount famously redesigned Sonic the Hedgehog after fans revolted over his original movie look. It’s too early to say if Universal would be forced to do the same.

Advertisement

If animated trailers are your thing right now, Pixar also just dropped the first trailer for Gatto, and Disney unveiled its first look at Hexed, the studio’s most exciting original film in years.

On Wednesday morning, workers poured hydrogen peroxide into the Lincoln Memorial Reflecting Pool in Washington, DC.

The treatment is the latest attempt by the Interior Department to control an algae bloom that has turned the pool bright green, despite President Donald Trump’s costly renovation to make it “American flag blue” in time for the nation’s 250th anniversary. Hot temperatures and climate change are among the risk factors that could be driving the outbreak.

The Trump administration spent more than $14 million to update the pool ahead of celebrations across the US capital. The work was done under a no-bid contract by a company that has never worked for the federal government. (It has, however, worked on President Donald Trump’s golf courses, according to The New York Times.)

Algae began forming less than a day after the updated pool was unveiled last week. An Interior Department spokesperson told CNN that the bloom was due to “residual algae from the supply lines, which have been sitting dormant for eight weeks.”

One of the issues appears to be the water source. The Reflecting Pool usually draws water from the nearby Tidal Basin, which is often filled with algae. During periods with high amounts of algae, the water supply switches to municipal drinking water. The US Interior Department didn’t immediately respond to questions from WIRED about which water source is currently feeding the pool.

Another issue is the weather. High temperatures create “a perfect storm for [algae] to bloom,” says Hans Paerl, a former professor at the University of North Carolina’s Institute of Marine Sciences. Stagnant water, Pearl says, makes the problem even worse. “Lakes and reservoirs around the world—they all have this problem during this time of year.” Hotter-than-normal weather is expected to hit Washington, DC, to end the week, which could make controlling the bloom more challenging.

Paerl also points out another, related driver that the Trump administration has shown it’s not in a rush to resolve: climate change.

“It’s just getting hotter, and these blooms are expanding globally—they’re moving up into higher latitudes,” he says. “It’s clearly a temperature effect allowing them to optimize their growth.”

Beyond the hydrogen peroxide, the Interior Department is “deploying high-tech nanobubble ozone technology” to keep algae at bay, an agency spokesperson tells WIRED in an email.

Trump has touted the work on the pool, which includes painting it blue and fixing leaks, while also saying the company the government hired would be able to do repairs “in much less time, for much less money.”

After alleging that Apple supplier Tata contaminated farm water and threatening to close its iPhone factory, Indian regulators have now dropped the matter altogether.

As a result of Apple’s supply chain diversification plans, which aim to reduce reliance on China, a quarter of all iPhones sold worldwide are now made in India. Tata, which manufactures iPhone backplates, is a relevant factor in the process, but its production efforts have come under scrutiny.

In 2025, farmers near Tata’s facility in Tamil Nadu complained to authorities that factory wastewater was contaminating their land and open wells. After conducting five inspections from December 2025 through May 2026, the Tamil Nadu Pollution Control Board (TNPCB) threatened to shut down Tata’s factory.

On Tuesday, however, the regulator seemingly backpedaled. In a statement to Reuters, Tata explained that it had addressed all contamination concerns and that its facility in Tamil Nadu was no longer under TNPCB scrutiny.

As the company “satisfactorily addressed all queries” made by the Tamil Nadu regulator, it has now “dropped any further course of action on this issue.” Neither Apple nor the TNPCB commented on the matter, though.

Initially, the Tamil Nadu Pollution Control Board argued that “groundwater in the open wells located in the adjacent agricultural lands” was contaminated. It was alleged that Tata’s rainwater harvesting pond ended up overflowing because of wastewater discharged from the factory.

Tata was reportedly made aware of the contamination on December 23, 2025, but the company did not respond.

On Saturday, however, Tata revealed it had conducted an independent sample analysis, which indicated the company was in full compliance with regulations. All parameters were reportedly within the prescribed limits.

Now, the company says the TNPCB has reached the same conclusion, as “the reports of its own analysis of recently collected water samples from Tata Electronics’ manufacturing facility in Hosur, Tamil Nadu, do not indicate any contamination”.

Though the regulator ultimately conceded, it’s still unclear if farm water in the Tamil Nadu region was actually contaminated.

It’s unclear how and why Tata’s findings differed from the results of the initial tests conducted by the TNPCB. Either the regulator somehow made the same mistake five times in a row, or it was pressured into dropping its threat against Tata.

Tata’s role in the Apple supply chain

While Apple’s most significant supply chain partner, Foxconn, assembles the iPhone in India, parts for the devices also come from Tata.

In 2024, the company entered into a $1B partnership with Pegatron, another key Apple supplier, to expand iPhone manufacturing in Tamil Nadu. The year prior, Tata purchased Wistron’s Karnataka facility.

With its Tamil Nadu plant now receiving an all-clear from regulators, Tata’s iPhone component production will continue. Its manufacturing efforts may even increase, as Apple seeks to reduce its long-standing reliance on China.

• GM enters large-scale energy storage through a sodium-ion battery partnership
• Sodium-ion batteries promise cheaper storage without complex cooling systems
• Peak Energy supplies storage systems while GM builds sodium-ion cells

General Motors (GM) has announced a partnership with energy storage firm Peak Energy in a move marking a notable shift in the automaker’s battery strategy.

Under the agreement, GM will manufacture sodium-ion (Na-ion) battery cells for stationary energy storage systems serving utilities, data centers, and other large electricity users.

After years of user demand, Instagram has rolled out a feature that allows users to further personalize their profiles’ look. The newly released Instagram Grid Reordering feature allows users to shuffle their posts on their grid without deleting or republishing them. The new feature will be quite convenient for users to highlight their best content. Alongside this innovative feature, the company has launched other creative features.

How To Reorder Your Instagram Grid?

1. Open the Instagram app and go to your profile page.
2. Long-press any post on your grid.
3. Tap Reorder Grid from the available options.
4. Drag and drop posts to their preferred positions.
5. Repeat the process for any other posts you want to move.
6. The changes will be automatically saved, and the new layout will be viewable to all who visit your profile page.

Other New Creative Features Coming to Instagram

Grid Reordering is not the only update Instagram plans to roll out. It has also unveiled other updates aimed at encouraging creativity within the app. With the new Spotify integration, users can share their favorite songs on Instagram Notes, making it easy to showcase their musical preferences.

The platform will also launch new handwritten fonts for Stories and Reels. Instagram created these fonts from handwriting samples submitted by content creators. The new fonts are scheduled to launch on June 20.

Meta Ray-Bans have been under increased public scrutiny following revelations about the facial recognition work Meta has been doing on its smart glasses. Consumers are rightly wary of products that could convert wearable tech into everyday surveillance devices. 

In early June, an investigation by Wired exposed how Meta had quietly embedded code for dormant facial recognition software under the internal designation “NameTag.” The feature, if rolled out, could have allowed Meta smart glasses to biometrically identify anyone in view — in real time, without consent — using a stored digital faceprint. The code, which was never made active for users, was removed a day later. 

The Electronic Frontier Foundation’s Threat Lab verified the initial findings and reported that Meta reversed course following public blowback. But the privacy nonprofit noted that Meta deleting the code “does not equal a permanent change of heart.”

Now, just a week after Meta removed the code, the company is facing new questions about its facial recognition software prototype. A new investigation by Wired uncovered that Meta partnered with Rank One Computing, a supplier for the US military and law enforcement agencies, for its biometric identification technology. Wired said it uncovered a software license tying the Pentagon vendor to the Meta AI app, the same one used for Meta’s smart glasses products.

The license agreement would authorize Meta to use Rank One’s military-grade facial recognition and “liveness detection,” which confirms whether someone is seeing a live person or a mask or photo. This business relationship, as Wired pointed out, “shows how thin the line has grown between the surveillance technology sold to law enforcement and the military and the consumer products sold to everyone else.”

According to Wired, Rank One Computing declined to comment on the findings. The Denver-based firm, which earns roughly 80% of its revenue from government clients, didn’t respond to a request from CNET for comment.

A Meta spokesperson told CNET that it has made no final decisions on facial recognition software for Meta Glasses, but would not confirm whether the tech giant is licensing a military-grade engine for its glasses. In an emailed statement, Meta noted: “Nothing has shipped to consumers and no final decision has been made on what to do here, if anything. If we do decide to roll something out, we will take a thoughtful approach and do so with full transparency.”

Meta’s facial recognition controversies

In our previous coverage, CNET noted a dangerous precedent if Meta’s glasses store biometric face data in an embedded database architecture that can compare new faceprints to existing ones. At the time, a Meta spokesperson responded that the company is “not building a central face database.” 

In late 2021, under public pressure, Meta announced plans to shut down its efforts to build a central facial recognition database on the Facebook platform. By that point, the company said, about 600 million users globally had already opted into the software, which could identify faces in photos and videos for tagging people on the social media site. Meta later settled a 2024 lawsuit filed by the state of Texas over the collection of facial recognition data for $1.4 billion. 

Earlier this year, the New York Times reported that Meta was developing software for its smart glasses to identify people, presumably using data from its social networks, such as Facebook and Instagram. The article cited an internal memo from Meta that said political tumult in the US would distract critics from the feature’s release.

Privacy advocacy groups such as the EFF have long spelled out the harms of facial recognition technology, as biometric-enabled public surveillance severely undermines anonymity. Facial recognition technology also has a disproportionately negative impact on marginalized groups, as it can track movement, misidentify people of color and lead to wrongful arrests. Businesses and governments can also abuse faceprints without consent, creating risks for identity theft and cybersecurity.