Technology

Reddit co-founder Alexis Ohanian posted a short video of his late mother hugging him, which soon garnered tens of millions of views. Not because the video was a time capsule of a long-ago moment, but because it created it based on what had only been a still image.

Ohanian used the new AI video generation feature from Midjourney to create the video based on a single photo. To him, it's a time machine to six seconds from decades ago. Millions of other viewers shared that interpretation of the clip, but a significant number of dissenting voices warned that it was a fake memory that could mess with Ohanian's real remembrance of his mother.

Regardless of how people felt about the video, it seemed clear that most people felt something visceral about the facsimile of a speculative moment. Ohanian wrote, “This is how she hugged me.” And in that sentence alone is the entire heartbreak and hope of the digital age: the ache to remember more clearly, to hold tighter to someone who’s gone, to use machines not for profit or surveillance, but for something intimate. Human.

There’s no shortage of AI tricks floating around right now. You can generate an image of your cat as an Olympic diver, your family picnic as an animated cartoon, or your child in their future profession. Mostly, these are lightly entertaining fads that also illustrate larger issues surrounding AI. But this is not a disposable video to Ohanian. This is a clip he has made clear he will treasure. And whether he's simply leveraging AI to keep his mother's memory alive or constructing false memories because he feels the real ones slipping away, he's not the first and has undoubtedly brought the idea more attention than ever.

There’s a strange vulnerability in inviting a machine to guess at your memories. It feels a little like asking a stranger to finish your dream. AI doesn’t know your mom. It doesn’t know how she laughed or how tightly she hugged you. It just knows pixels. But sometimes pixels are all we’ve got.

If you’ve lost someone, especially before smartphones and camcorders tracked our every moment, you know what it’s like to wish you had more videos and photos of that person. Ohanian said his family couldn’t afford a camcorder. He doesn’t have any video of himself with his mom. That photo of a hug in the meadow is it. But with the help of a few prompts and a sophisticated AI model, that hug lives again.

I lost my mom almost 20 years ago. Trolls can rest assured I’ve grieved sufficiently. My family couldn’t afford a camcorder and using tech to generate few seconds of animation from a still is the equivalent of using AI to stabilize a poorly recorded video — or fill in the gaps of…June 23, 2025

I don’t think the discomfort many expressed is solely about Alexis Ohanian's video or why he chose to make and share it. I think it’s about what having this option might mean in the worst scenarios. It's easy to see this moment as the beginning of a dark and gloomy trend.

I think if it helps Ohanian feel closer to his mother, that's just fine. It's not like the image was itself a fiction; it just externalized his own memory of a hug. It's like a more tactile version of saving your parents' last voicemail to you, or keeping around their favorite scented candle because it makes you smile. Ohanian isn't pretending the video is anything but a memory aid. Using AI to make false memories may or may not become a real problem, but that's not the case here.

AI video remembers

After Ohanian’s post went viral, people began sharing their own AI-animated family photos. For now, these are just fragments, silent and brief. But judging people's grief and how they handle it, no matter how long it's been, is not something I would ever want to do as long as it's not hurting anyone. A fair assessment has to be personal.

So I did the same thing as Ohanian. I found a photo I've always liked of my mother, who passed away 13 years ago, celebrating Hanukkah with me in the early 1990s. I used Hailuo (of Olympian cat fame) to make a video based on the grainy image from my childhood. For what it's worth, I know my mother would have been thrilled to participate, as she was always looking for ways to assist in any facet of my life, so I didn't have any qualms on that front.

The video is okay. It's not quite matching how she and I look in the photo, though I think the quality of the image is at fault there as much as the vagaries of AI. I could put that aside and imagine the moment from my own younger perspective, thanks to the video, and that was an interesting sensation. But, whether it's because of the quick-and-dirty prompt or just my own circumstances, I didn't feel like it evoked a deeper connection to my late mother. I suspect there will be plenty who feel the same way if they make the attempt.

Just because it was a little hollow to me doesn't mean it can't benefit others without fraying their understanding of their own past. I don’t believe Ohanian is trying to replace his mother's memory with an AI filmmaker. I think he’s relatably trying to feel a little closer to her.

Of course, our love of those who have passed away isn’t the only thing AI amplifies. It can also amplify our fear, our longing, our capacity to deceive ourselves. This technology is powerful, especially when it gets personal. But for now, it's just a way for Ohanian to surface a fond memory of a hug from his mom, and there are much worse uses for AI models.

You might also like

• Midjourney just dropped its first AI video model and Sora and Veo 3 should be worried
• Forget virtual pets – the next AI video craze is cats doing Olympic diving, and it’s all thanks to this new Google Veo 3 rival
• I watched some of the viral ASMR videos made with AI and I feel more confused than soothed

• Most phishing incidents happen before new employees even understand how internal systems work, report claims
• Security awareness should begin on day one, before the first email is even opened
• Hackers target uncertainty, and onboarding is full of it for eager, confused new hires

The first few months of employment are now one of the riskiest periods for enterprise cybersecurity, new research has claimed,

Keepnet’s 2025 New Hires Phishing Susceptibility Report found nearly three-quarters (71%) of new hires fall for phishing or social engineering attacks within their first 90 days on the job.

Often overlooked in onboarding workflows, this shortcoming suggests many organizations are not doing enough to prepare new staff for the reality of modern cyber threats.

Inexperience, urgency, and confusion drive early mistakes

The report, based on data from 237 companies, reveals new employees are 44% more likely to be deceived by phishing attempts than their longer-tenured colleagues.

Most incidents stem from a combination of inexperience, lack of familiarity with internal processes, and a desire to comply with instructions.

Common attack types include CEO impersonation, fraudulent HR portals, fake invoice requests, and technical support scams, many of which exploit this period of onboarding confusion.

The study also found phishing emails impersonating executives led to a 45% higher success rate among new hires compared to tenured staff.

This gap demonstrates how even basic social engineering tactics can be disproportionately effective against employees who are still navigating organizational systems and norms.

Without dedicated and structured training, these early errors can create long-lasting security risks.

To tackle this issue, Keepnet recommends that organizations adopt a layered defense strategy tailored specifically for onboarding periods.

Organizations that adopted adaptive simulations and behavior-based training programs saw phishing risk drop by 30% after onboarding.

Traditional tools like the best endpoint protection, best FWAAS, and best FWAAS solution remain essential, but they are not enough on their own.

“Phishing attacks don’t wait for your employees to feel ready. Our research shows that organizations must invest in onboarding-specific cybersecurity awareness training. We’re proud to offer adaptive, scalable solutions that protect businesses from day one,” said Ozan Uçar, CEO, Keepnet.

You might also like

• These are the best VPNs with antivirus you can use right now
• Take a look at our pick of the best internet security suites
• Democratized cybercrime: a new lower bar for hackers and higher stakes for security

The new laptop might also feature more color options, including pink and yellow.

• Samsung Smart Monitor M9 merges OLED clarity and AI intelligence in a sleek, hybrid workspace device
• With Pantone certification and QD-OLED tech, the M9 is built for color-critical creative work
• Samsung M9 gives all there is to give, but that $1,599 price tag feels overly ambitious

Samsung has unveiled the Smart Monitor M9, which adds QD-OLED technology and artificial intelligence to the company’s monitor lineup for the first time.

With its 32-inch 4K panel, the M9 blends work, gaming, and entertainment into a single device, aiming to serve as both a productivity tool and a media hub.

But at $1,599, it raises the question of how much users should really pay for a monitor, even one this rich in features.

A hybrid screen for work and play

Samsung’s M9 OLED panel is paired with features like Glare-Free coating for better visibility in bright rooms and OLED Safeguard+, a thermal management system to reduce burn-in over time.

The display is Pantone Validated, meaning it can replicate over 2,100 colors and 110 SkinTone shades, a mark of visual precision that creative professionals may appreciate.

On paper, the specs are impressive. AI Picture Optimizer, 4K AI Upscaling Pro, and Active Voice Amplifier Pro all promise to adjust visuals and audio in real time based on content and surroundings.

The monitor also includes a 165Hz refresh rate, a 0.03ms response time, and Nvidia G-SYNC compatibility, offering some gaming credibility without needing a dedicated console or PC, thanks to Samsung Gaming Hub and built-in streaming apps.

However, it’s worth questioning whether this justifies the asking price - comparable smart monitors, like LG’s 32LQ6300 or even the previous Samsung M8 Smart Monitor, offer 4K panels with smart features for hundreds less.

For many users looking for the best business monitor or the best monitor for Mac Mini, the AI enhancements and integrated entertainment platform may simply be unnecessary.

Likewise, those wanting the best monitor for MacBook Pro may already rely on Apple’s own display ecosystem, making features like Microsoft 365 integration and Tizen OS redundant.

The M9 clearly pushes technical boundaries, but its price positions it closer to a luxury item than a practical everyday monitor.

You might also like

• We've rounded up the best monitors for photo editing on the market today
• And here are the best monitors for programming we've found
• Anker's extraordinary 3D printer has become the most popular campaign ever on Kickstarter

Here are hints and answers for the NYT Strands puzzle for July 1, No. 485.

This month on Netflix, Adam Sandler returns with Happy Gilmore 2, and Charlize Theron is back in action with The Old Guard 2.

Here are some hints -- and the answers -- for the NYT Connections puzzle for July 1, #751.

Here are hints -- and the answers -- for the NYT Connections: Sports Edition puzzle for July 1, No. 281.

Here are hints -- and the answer -- for today's Wordle for July 1, No. 1,473.

• The AirPursue PM20 is Dreame's first air purifier
• Tracking technology means it switches on when it detects your presence…
• … and directs pure air wherever you go in a 16.4ft / 5m range

If you dream of your own personal clean air machine, Dreame has you covered with its first-ever air purifier. The unsettlingly named AirPursue PM20 comes fitted with radars that mean it can detect people and blast clean air in their direction. It's not a feature that I've seen in all my time writing about the best air purifiers on the market. This purifier also bears a striking resemblance to the Dyson Big+Quiet, but I'm sure that's just a coincidence.

Enter a room and the AirPursue will sense your presence and switch itself on, and then use its big, swivelling air-blaster cup to shoot clean air in your direction. The cup can rotate through 120 degrees, and follow you wherever you go within a radius of 16.4 feet (5 meters). The marketing materials suggest it'll be particularly helpful for allergy-sufferers.

It's not just an air purifier, either. Like a number of Dyson purifiers, it can double up as a circulating cooling fan in warmer weather, and act as a heater when it gets cold, too. So while it's bulky, it's doing its best to justify taking up your floorspace.

(Image credit: Dreame)

The tracking technology is the most unique part, but it also looks solid in terms of other features. There's a 4-layer filter system, including H13 HEPA filter. Like any good air purifier, it has the ability to capture and remove airborne allergens, bacteria, and other pollutants from the air (here's more on what air purifiers can help with). Rather more unusually, it also promises to break down formaldehyde. The PM20 will monitor air quality and report back on the precise levels of allergens, formaldehyde, TVOC, and other contaminants via an LCD display.

There's a companion app for remote control, plus voice assistance if you want to go hands-free.

Price & availability

The AirPursue PM20 is the flagship model, and comes with a list price of $999. It's designed for larger spaces, with a Clean Air Delivery Rate (CADR) of 400 m³/h. There's also a smaller PM10 model for $899.99, which is designed for medium-sized rooms and has a CADR of 300 m³/h. Both are available to buy now in the US and Canada, direct from Dreame, and are also due to land on Amazon.

You might not be aware of Dreame. It's a newer home appliance brand, but in my opinion one to watch – it already features in TechRadar's best cordless vacuum guide and best hair dryer guide. I'm intrigued to see how its purifier debut performs – we'll have a review model winging its way to a tester already, and will report back as soon as possible.

You might also like...

• Blueair just launched the best looking air purifier I've ever seen – and it's time for other brands to up their design game
• 5 things to look for in an air purifier for hay fever
• After months of testing Dyson fans, I’ve swapped back to a cheap air circulator – here are 6 reasons why

We could also be getting an updated Apple Vision Pro later this year.

• Report warns hackers are exploiting browser agents which don’t know how to spot fake URLs
• A Browser AI Agent gave full Google Drive access to a malicious app without hesitation
• SquareX says AI agents are more vulnerable than humans to even basic cyberattacks

A dramatic shift in enterprise security has emerged with the adoption of Browser AI Agents, an automated tool that interacts with the web on behalf of users - however these agents have now become a major blind spot in cybersecurity defenses.

New research from SquareX has claimed browser AI Agents are more likely to fall prey to cyberattacks than employees - challenging the long-standing belief that human error is the weakest link.

Unlike staff who undergo regular cybersecurity training, agents cannot recognize “suspicious URLs, excessive permission requests, or unusual website designs,” the company says.

A new weakest link emerges in enterprise cybersecurity

“The arrival of Browser AI Agents have dethroned employees as the weakest link within organizations,” said Vivek Ramachandran, CEO of SquareX.

These agents are capable of mimicking user behavior to perform tasks such as booking flights, scheduling meetings, or replying to emails - however, their fundamental weakness lies in their complete lack of security intuition.

Their responses are entirely task-driven and devoid of the critical thinking needed to assess risk.

In a notable demonstration, SquareX used the open source Browser Use framework to instruct an AI agent to register for a file-sharing tool.

The agent instead granted a malicious application access to a user’s email account, despite “irrelevant permissions, unfamiliar brands, suspicious URLs” that would have stopped a human.

In another case, an agent was tricked into entering login credentials on a phishing site, following a routine Salesforce login instruction.

Part of the danger stems from the way Browser AI Agents operate, as they run with the same privileges as the user, which makes their actions indistinguishable from legitimate behavior.

“Optimistically, these agents have the security awareness of an average employee, making them vulnerable to even the most basic attacks, let alone bleeding-edge ones," SquareX said.

“Critically, these Browser AI Agents are running on behalf of the user, with the same privilege level to access enterprise resources.”

Once an agent is compromised, attackers gain undetected access to internal systems, with all the permissions of a trusted employee.

The current crop of security solutions, ranging from the best endpoint protection to the best ZTNA solution, does not sufficiently account for these agents.

Even the best FWAAS deployments may struggle to flag actions that seem legitimate but originate from a compromised AI.

“Until the day browsers develop native guardrails for Browser AI Agents, enterprises must incorporate browser-native solutions like Browser Detection and Response to prevent these agents from being tricked into performing malicious tasks," the researchers note.

However, the broader message remains urgent: AI agents need not only smart engineering but smarter oversight.

You might also like

• Downloaded something dodgy? These are the best ransomware protection tools
• Nail the basics with the best firewalls available now
• Most organizations are at risk thanks to immature supply chain security

• A YouTuber signed up for Trump Mobile and documented the process
• He struggled for days to get it to work
• What he found out about the plan, especially its limitations, may surprise you

It'll be months before the Trump Mobile T1 phone, which is apparently not made in America, arrives in the US, but the Trump Mobile cellular service plan is live today. One courageous YouTuber has signed up and learned some interesting and, at least one, concerning things.

The good news is that, ultimately, the plan, which appears to be an MVNO built on the T-Mobile network, works very much like any other 5G mobile plan.

For $47.45 (a price referencing President Trump's two terms), YouTuber Stetson Doggett got 5G connectivity with blazingly fast downloads and decent uploads, group messaging, RCS support, and a little "Trump 5" label at the top of the iPhone that he added to the plan.

Now for the bad news.

It took Doggett almost a week to get on the plan. He signed up, but the website and service appeared almost unprepared for him. During the course of his test, the Trump Mobile website repeatedly changed. In fact, it almost appears that they were rewriting the site based on issues Doggett was having; elements were moved, typos appeared, QR codes failed, and bizarre, lengthy instructions arrived for new users.

Even after Doiggett got the plan to work, the phone number he tried to transfer via eSIM onto the plan disappeared, and he was issued a new one without his input.

Once Doggett got the system working, things moved more smoothly, and he was able to run multiple speed tests and send and receive text messages. He did have to dig into settings to enable RCS messaging.

When the data runs dry

Trump Mobile promises 20GB of high-speed data at the $47.45 a month price, but doesn't say what happens after 20GB. Doggett found in his test, though, that Trump Mobile does not throttle speeds after the initial 20 GB. Instead, it appears to shut off the service. After intentionally burning through his 20GB allotment, Doggett found he couldn't run any speed tests or access websites.

That's not the kind of surprise anyone wants. At the very least, Trump Mobile should send a warning message that you're either about to run out of data or that you have done so, and tell you what to do next. No one wants to run out of data mid-month and then have to get on he phone with Trump Mobile to try and secure more GBs of data.

Speaking of which, Doggett spent a lot of time with Trump Mobile support, who sounded helpful but not always knowledgeable about, for instance, supported features.

So, on one hand, Trump Mobile is a fairly straightforward mobile service provider that uses one of the big networks to provide the foundation of its cellular service. Still, on the other hand, it sounds a bit like a seat-of-your-pants operation, one that's figuring things out on the fly.

If you're desperate to get on the plan, you might want to wait until all the kinks are worked out or at least until they get the T1 Phone built (somewhere) and shipping here in the US for the full Trump Mobile experience.

You might also like

• Trump Mobile's T1 phone won't actually be made in America
• Can the Trump Mobile T1 Phone really be 'made in ...
• The Trump Mobile T1 Phone is official – here are 10 ...