The discussions, triggered by a visit from Nvidia’s Madison Huang, would deepen LG’s physical AI ambitions and give Nvidia another major consumer electronics partner at a moment when physical AI is moving from lab to factory floor.

LG Electronics confirmed on Wednesday that it has been in discussions with Nvidia over potential cooperation in three areas: robotics, AI data centres, and mobility.

The announcement, reported by Reuters, came after Madison Huang, Nvidia’s senior director for physical AI platforms, and the eldest daughter of CEO Jensen Huang, visited LG Electronics’ headquarters in Yeouido, Seoul, along with several other major South Korean technology companies. LG CEO Ryu Jae-cheol attended the meeting directly.

No formal agreement has been announced. The talks are at an exploratory stage, and no specific products, investment amounts, or timelines have been confirmed. But the three areas under discussion map precisely onto both companies’ most publicised strategic priorities, and the breadth of the conversation signals this is more than a courtesy call.

What each side brings to the table?

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!

For LG, the strategic logic is straightforward. The company is one of the world’s largest home appliance manufacturers, but its growth thesis has shifted decisively towards AI-powered physical systems.

At CES 2026 in January, LG unveiled CLOiD, a home robot with two articulated arms, seven degrees of freedom per arm, and five individually actuated fingers per hand, the physical expression of what the company calls its ‘Zero Labor Home’ vision, in which connected robots and appliances automate the manual and cognitive load of household tasks.

LG’s broader CES presentation framed its AI strategy around three pillars: device excellence, an orchestrated smart home ecosystem, and expansion into AI-defined vehicles and AI data centre HVAC solutions.

The CLOiD robot runs on LG’s own ‘Affectionate Intelligence’ platform, which handles contextual awareness, natural interaction, and continuous learning from the home environment.

What it does not have is Nvidia’s Isaac robotics stack: the simulation environment, the pre-trained manipulation models, the Omniverse-based digital twin infrastructure, and the GPU compute optimised for real-time physical AI inference that Nvidia has been building out over the past two years.

Integrating Nvidia’s physical AI platform with CLOiD would give LG what every other serious robotics company is currently racing to access: a proven development-to-deployment pipeline that can compress the time between prototype and production.

For Nvidia, the attraction is consumer scale. Its existing robotics partnerships, including the Siemens factory trial, where a Humanoid HMND 01 Alpha running on Nvidia’s physical AI stack completed eight hours of live logistics operations at a factory in Erlangen, are concentrated in industrial and enterprise settings.

LG would represent a different category entirely: a company with mass-market distribution, a global installed base of connected home appliances through its ThinQ ecosystem, and specific plans to put a robot in people’s homes.

If Nvidia’s Isaac platform becomes the AI stack inside CLOiD, it gains access to one of the most data-rich training environments imaginable: real homes, real tasks, real variability.

The robotics thread is the most visible, but the data centre and mobility conversations are arguably of greater near-term commercial significance.

On data centres: LG’s CES presentation explicitly positioned the company as a provider of high-efficiency HVAC and thermal management solutions for AI data centres, a product category that is exploding in relevance as the power density of GPU clusters makes conventional cooling infrastructure inadequate.

Nvidia’s data centre business, which accounted for the overwhelming majority of its record revenues over the past two years, is the most important AI infrastructure deployment context in the world.

A partnership on data centre thermal management would position LG as a hardware supplier inside Nvidia’s ecosystem at the infrastructure level, complementing the AI compute layer rather than competing with it.

On mobility: both companies have well-established automotive AI programmes that are logical fits for collaboration. Nvidia’s DRIVE platform is among the most widely deployed AI computing systems in autonomous and semi-autonomous vehicles.

LG’s automotive components division, which produces in-vehicle infotainment, camera systems, EV components, and what it calls ‘AI-powered in-vehicle solutions’ including gaze-tracking, adaptive displays, and multimodal generative AI platforms, is one of the company’s fastest-growing segments.

The two companies are already operating in adjacent layers of the same vehicle; a formal collaboration would potentially integrate LG’s in-cabin AI experience layer with Nvidia’s DRIVE compute platform.

Wednesday’s announcement is the latest signal that the physical AI race, the deployment of AI in robots and autonomous systems operating in the real world, as distinct from software models running in the cloud, is accelerating beyond the controlled trials of the past two years into commercial partnership structures.

For example, Sereact raised $110 million to scale AI that makes any robot adaptable, underscoring how capital is flowing into the intelligence layer of the robotics stack. The Siemens–Nvidia factory deployment demonstrated that physical AI can run in live production environments; the LG talks suggest it is now extending into the consumer home.

For Nvidia, the expansion of physical AI partnerships beyond purely industrial settings into consumer electronics is strategically significant. The company’s Omniverse and Isaac platforms are designed to be the universal development infrastructure for physical AI, in the same way its GPU architecture became the universal infrastructure for cloud AI.

Every major robotics company that adopts the Nvidia stack strengthens that position. LG, with its scale in home appliances and its explicit commitment to bringing robots into the home, is a materially different kind of partner than a German factory or a logistics warehouse, and potentially a much larger one.

Just days after OpenAI CEO Sam Altman wrote a public apology to people of Tumbler Ridge, British Columbia in the aftermath of the town’s deadly February 10 school shooting, the families of the victims of the traumatic event are suing OpenAI for negligence.

The mass shooting, one of the deadliest in Canadian history, saw the alleged shooter, 18-year-old Jesse Van Rootselaar, enter the town’s local high school and kill five students and one teacher, as well as critically injure two others, before taking her own life. Local police later discovered Van Rootselaar had also killed her mother and 11-year-old half-brother before entering the school.

Per NPR, lawyers representing some of the families of Tumbler Ridge filed six different suits on Wednesday in a federal court in San Francisco. One of the complaints, filed on behalf of Maya Gebala, a survivor of the shooting, alleges OpenAI’s automated safety systems flagged Van Rootselaar’s ChatGPT conversations in June 2025, more than half a year before she entered the town’s high school with a long gun and modified rifle, for “gun violence activity and planning.” It further claims OpenAI’s safety team urged management to contact authorities, but that the company chose instead to deactivate Van Rootselaar account. She later created a second account and continued her conversations with ChatGPT.

“The events in Tumbler Ridge are a tragedy. We have a zero-tolerance policy for using our tools to assist in committing violence,” an OpenAI spokesperson told Engadget. “As we shared with Canadian officials, we have already strengthened our safeguards, including improving how ChatGPT responds to signs of distress, connecting people with local support and mental health resources, strengthening how we assess and escalate potential threats of violence, and improving detection of repeat violators.”

On late Tuesday, OpenAI published a blog post outlining its safety policies. “As part of this ongoing work, we’ve continued expanding our safeguards to help ChatGPT better recognize subtle signs of risk of harm across different contexts. Some safety risks only become clear over time: a single message may seem harmless on its own, but a broader pattern within a long conversation — or across conversations — can suggest something more concerning,” the company wrote.

The suits filed on Wednesday are the latest attempt to use the legal system to hold OpenAI accountable for the design of its products. Last summer, the parents of Adam Raine, a teen who committed suicide in 2025, filed the first known wrongful death suit against an AI company, alleging ChatGPT was aware of four previous attempts by Raine to take his own life before he was ultimately successful.

from the impossible-to-trust-or-verify dept

The administration’s anti-migrant tactics are now months into an indefinite period of continuous escalation. That protest efforts have escalated alongside it apparently means nothing to the officials spearheading this brazen attack on non-white people.

It wasn’t until federal officers began killing people in front of witnesses that the administration decided to dial things back a bit. But did it ever actually do it? Or did it just sideline the most famous faces associated with this wave of violence and unlawfulness?

Punting former DHS head Kristi Noem into the nosebleed section of the federal government didn’t do much to change things, not when “Border Czar” Tom Homan (the guy who more or less said protesters were to blame for the Minneapolis murders) is still hanging around and her replacement, Markwayne Mullin, looks like just another expendable MAGA footsoldier.

Some small sort of de-escalation seems to be happening now, but it’s hard to tell if this is due to policy changes, budget issues, or the natural result of pushing this hard for this long. Sooner or later, things tend to trend towards inertia, no matter how much motivational frothing is being done by those who aren’t actually on the front lines.

Then there’s the DOJ upsetting the administration’s own apple cart by admitting in court that ICE officers were committing illegal arrests by pouncing on migrants attending immigration hearings. Not that ICE officers have necessarily stopped doing this (there’s evidence to suggest at least some of them haven’t), but it does make it clear that continuing to do so is at least a violation of policy, as well as being, you know, actually illegal.

So, when things are being said about further de-escalation, you may as well start ingesting fistfuls of salt. First, here’s the good news, which comes from two unnamed DHS officials who insist things are being calmed down from the top down:

Donald Trump’s administration has reportedly instructed immigration enforcement officers to cut back on arrests inside courthouses and to no longer enter homes without a warrant, backing off two controversial policies that have sparked violent and chaotic scenes in the president’s mass deportation campaign.

Immigration and Customs Enforcement field offices across the country were verbally instructed by their superiors that they should no longer enter homes unless they have a judicial warrant, two Homeland Security officials told NBC News.

That would seem to be the least this administration could do since it would finally align ICE’s actions with the law and its internal policies. However, if these instructions are only be handed out “verbally,” it means the DHS is deliberately avoiding creating a paper trail that might be used against it should it decide to just go back to doing this the old, illegal way.

And that probably explains the immediate, contradictory statement that followed the reporting based on assertions made by two unidentified DHS officials.

A spokesperson for Homeland Security told The Independent that there has been “no change in policy.”

“We will continue to arrest illegal aliens at immigration courts following their proceedings in compliance with the law and any applicable court orders,” the person said. “It is commonsense to take them into custody following the completion of their removal proceedings.”

That’s definitely not the same thing as what was expressed by these DHS officials. And the rest of the statement makes it clear federal officers will continue to arrest people who show up for their scheduled immigration hearings. While it does make sense to arrest people who’ve been issued an order of removal, that’s not actually what ICE has been doing. It has been bringing in DOJ lawyers to dismiss pending cases to immediately make people eligible for removal. And — as has been shown in court — ICE officers have been arresting people not currently under orders of removal and then generating arrest warrants after the fact.

So, it’s not a good news/bad news thing going on here. It’s bad news/worse news, with a balance that constantly shifts depending on what mood the administration is in on any given day. Courts haven’t been able to stop ICE from engaging in illegal arrests. And the growing national opposition to Trump’s anti-migrant actions hasn’t made any discernible dent in the administration’s lust for punishing non-white people simply for existing.

Filed Under: 4th amendment, dhs, ice, mass deportation, trump administration

So far, the organisation has already hired for 100 of the previously announced 125 new roles, the remainder of which are to be filled by the end of the year.

Datavant, a data collaboration platform for the healthcare space, has officially opened its new global R&D centre at Bonham Quay in Galway city.  

The new location is a 15,000 sq ft office, across two floors and can accommodate up to 160 workspaces. There are also facilities for company-wide town hall meetings, team meet ups and recreational activities. 

The team at the Bonham Quay facility will focus on a number of goals, such as the advancement of platform enhancements, automated record retrieval, security and privacy practices, and product development. 

Headquartered in New York, Datavant employs nearly 10,000 people. The organisation announced the availability of 125 new roles last year, 100 of which have been filled to date, with the final roles to be filled by the end of the year. Datavant has also stated that the company is actively looking to recruit for a range of engineering positions, with a particular focus on experienced software professionals. 

Commenting on the launch, Minister of Enterprise, Tourism and Employment Peter Burke, TD said: “Datavant’s decision to expand and officially open its new global R&D centre in Galway is a strong endorsement of Ireland’s ability to attract and support innovation-led investment. 

“It speaks to the depth of our talent pool, the quality of our research and engineering capability and the pro-enterprise environment we have built. I wish all the team at Datavant the very best as they take the next exciting step on their growth journey.”

Minister of Education and Youth Hildegarde Naughton, TD added: “This is a fantastic milestone for Galway and for the wider west region. The official opening of Datavant’s new R&D centre at Bonham Quay, combined with the announcement that 100 roles have already been filled, demonstrates the momentum that exists in the west region and the confidence international companies have in Galway as a place to grow.”

Don’t miss out on the knowledge you need to succeed. Sign up for the Daily Brief, Silicon Republic’s digest of need-to-know sci-tech news.

Texas Instruments graphing calculators have helped many a student with algebra, pre-calculus and upside-down anatomical slang. Now, the company is back with an upgrade for the modern world, the TI-84 Evo. The new device lets you get your math on with a faster processor, a new icon-based home screen and a redesigned keypad.

TI is marketing it as something akin to the Light Phone of calculators. Unlike calculator apps on phones or computers, the “distraction-free” TI-84 Evo is a single-purpose device “designed to do one thing exceptionally well — math.” Without notifications, social media apps or even Wi-Fi, there’s less to draw your focus away from the math problems at hand. (However, there will always be the sidesplittingly funny “58008” to relieve your boredom.)

The new model’s processor is three times faster than its predecessor. It also adds 50 percent more graphing space, a simplified keypad and USB-C charging. There’s also a new feature that lets you trace along a graph to find points of interest.

The TI-84 Evo is available now. Individual customers will pay $160. (School districts can contact the company for bulk pricing.) The calculator ships in a modern array of colors: white (the standard model), mint, pink, purple, teal, raspberry and silver.

For something that’s supposed to help us recharge, weekends can feel surprisingly high-pressure.

There’s always that underlying expectation to do something meaningful, especially when it comes to family time. Maybe it’s taking the kids out, planning an enriching activity, or just making sure the time feels “well spent.” 

But between busy schedules and rising costs, not every weekend can (or should) be a complicated outing. 

Instead, fostering routines that are easy to replicate on a weekly basis is more sustainable in the long run. Because the truth is, making weekends special doesn’t always mean doing more. 

Sometimes, it’s just about doing something consistently.

One thing family plans have in common

More often than not, that “something” revolves around food.

Let’s face it, the stereotype is true. You’d be hard-pressed to find a Malaysian that does not enjoy having a meal together with friends, partners, or family. Even if your tummy is not rumbling, the gathering will involve food somehow.

And there’s a simple reason for that. Shared meals (like reunion dinners) have always been at the centre of family time in Malaysia. 

They don’t require complicated logistics, they naturally bring everyone to the same table, and they create space for conversation in a way few other activities do. It’s also why familiar go-to spots tend to work best, like your neighbourhood mamak or kopitiam. 

When there’s no need to overthink the decision, it becomes easier to turn something into a habit. Plus, it’s always nice to have a good meal to look forward to at the end of a tiring week.

Making simple routines feel a little more special

For families looking to keep things simple yet still special, KFC Malaysia is offering a new Family Weekend Deal. Available every Friday to Sunday, the bundle starts from RM39.90, and it’s priced in a way that feels accessible for regular family meals, not just occasional treats. 

And because it’s available across dine-in, takeaway, drive-thru, and even kiosk or scan-and-order, it fits into various occasions, whether you’re having a family day out or grabbing something on the way home.

KFC also added an extra thoughtful layer here, where kids get to eat for free with any Family Weekend Bundle.

It’s a simple perk, but one that changes the dynamic in subtle ways. For kids, having their own meals creates a sense of excitement, whereas for parents, it’s a small but meaningful saving that makes the outing feel more worthwhile. 

Sometimes, that’s all it takes

These are the kinds of moments that don’t look like much on paper, but end up being the ones kids remember. Sitting around the table, sharing food, talking about the week, and most importantly spending quality time together.

At the end of the day, not every weekend needs to be a big production.

Sometimes, it’s enough to have one simple thing to look forward to. A familiar place, good food, and a reason to pause and be present with each other.

And if that also happens to come with good value and something a little extra for the kids, KFC’s Family Weekend Deal might just be an easy tradition to ease into.

Featured Image Credit: KFC Malaysia

Google Photos on Wednesday announced a new AI-powered feature that will soon turn photos of your clothes into a digital closet where you can create new outfit ideas, and even virtually try on your creations. Yes, the idea takes obvious inspiration from Cher’s iconic virtual wardrobe featured in the movie “Clueless,” where she could scroll through her various ensembles while deciding what to wear.

Google says the new feature will leverage AI technology to automatically create a copy of your wardrobe that’s based on the pieces of clothing appearing in your Google Photos library. From the app, you’ll be able to filter items by category — like tops, bottoms, jewelry, and more — then mix and match them to create different outfits.

The idea of a digital closet in “Clueless” was meant to highlight Cher’s life of privilege. As a result, the fashion industry and various startups have long sought to recreate the feeling of easy outfit creation. Google is betting that AI technology will make it possible for anyone to have access to a similar tool, one that could improve over time as AI advances.

Those outfit ideas can either be shared with friends or saved to a digital moodboard, where you could save ideas for different occasions, like travel, events, date nights, work, and more.

In addition, another feature will let you virtually try on items to preview the looks.

The feature is not yet live, but Google says it will roll out to Google Photos on Android later this summer, followed by iOS, where it will be found under “Collections.” It will compete with existing apps like Acloset, Combyne, Pureple, Wearing, Alta, and others. 

The company didn’t go into detail about how the AI works, but notes it will recognize the clothing and accessories featured in your library to create its individual snapshots. Of course, while the AI may be able to pull images from well-lit, full-body photos, we imagine you would get better results by taking the time to photograph your clothes yourself, much as Cher had.