Rent the Runway’s Mark Walsh discusses the tech ecosystem in Galway and how he found himself in the role of director of engineering.

It was “half accidental and half strategic, with the accidental being the bigger half, as my kids would say”, joked Mark Walsh, when asked what led him to become the director of engineering at Rent the Runway’s Galway location.  

“On the accidental side, I began my tech career in a start-up in Spiddal in Galway, and spent a decade working as a software engineer,” he said. “Then came the advent of DevOps and a wave of emerging technologies beyond Java, which pulled me into different areas, cloud, solution architecture and even product management-type roles. 

“People management arrived completely by accident about six years ago, not as any great career strategy, and it has evolved from there into what I do today.”

Prior to joining the organisation he was commuting to a different role, four hours there and back, which he explained took a toll on his family life. Moreover, he had always planned on returning to Galway.

“I wanted to return to Galway. Crucially though, I was focused on a destination, not just another job. Rent the Runway was somewhere I already knew through former colleagues, it had the kind of opportunities that genuinely motivated me and it felt like somewhere I could add something meaningful rather than just fill a seat.”

What is the current engineering landscape like in Ireland?

I have to be honest and say I can’t speak adequately for the tech landscape much beyond Galway, as that would be overstepping what I actually see day to day. What I will say is that from a Galway perspective, the picture is more positive than the broader headlines might suggest.

There is a real depth of engineering talent here, and a sense that you don’t need to be in the capital city to work on interesting, complex problems at scale. The mix of homegrown companies maturing alongside international organisations that have put down genuine roots has created a stronger local ecosystem than people sometimes give it credit for. It is not without its challenges, but the opportunity feels real.

What does a typical day look like, if there is such a thing?

There genuinely isn’t one, which I think is part of what keeps it interesting. As a US company, our natural overlap with our US team members falls in the afternoon Irish time, so that tends to be where the collaborative work lands: cross-team meetings, planning sessions, working through anything that needs real-time conversation across time zones. Mornings are generally freer, which I try to protect for focused work, Galway-based one-to-ones, or anything that benefits from uninterrupted thinking time.

The nature of the work itself shifts depending on where we are in the cycle. We run quarterly planning and within that there are shorter-term operational necessities running alongside medium-term delivery work and longer horizon strategic efforts. As a platform team with an SRE mindset, that means BAU maintenance such as managing upgrades across our existing portfolio, future-looking initiatives addressing our strategic vision such as platform migrations and responding to emergent support or security issues as they arise.

What is the current focus of your work and its potential long-term impact?

The intent is always the long-term health of Rent the Runway as a platform and a business. In practical terms that means a few things running in parallel: reducing operational overhead and complexity where it has built up over time, managing risk and compliance thoughtfully and continuing to advance our existing capabilities while staying open to where new opportunities might emerge.

What I find genuinely motivating about engineering at this level is that the decisions you make about architecture, tooling and how teams are structured have a compounding effect. Good calls made now create headroom for the business to move faster later. Poor ones create drag that someone, usually a future version of your team, has to unpick.

So while the immediate focus is always on what needs to happen this quarter, the framing I try to bring to it is, what are we making easier or harder for ourselves two or three years from now?

How does the field of research impact your role at Rent the Runway?

When I hear the word research I tend to map it back to engineering practice as a whole. My teams are engaged in continuous research, whether we label it that way or not, because the technology landscape never stops moving.

The old saying that today’s problem was yesterday’s solution is as true now as it has ever been, arguably more so. My teams are always engaged in some form of discovery work: reducing technical uncertainty, evaluating emerging approaches, making sure that what we build is robust, secure and fit for purpose beyond the immediate delivery. That process of deliberate inquiry is baked into how our engineering teams operate, even when it doesn’t carry the formal title of research.

Have you any advice for professionals considering a research-oriented role in the engineering space?

My advice is, even if it sounds old-fashioned, given the pace of change right now, don’t abdicate your thinking completely to your tooling.

The engineers and knowledge workers I see navigating uncertainty well are those grounded in the fundamentals, reasoning through complex problems, communicating clearly about risk and sitting with ambiguity rather than rushing to the first solution. Those qualities are always in demand, at least with me.

Staying current matters and I am not suggesting otherwise. But understanding why you are reaching for a particular approach, rather than just something that appears to work, has always been important. Value your judgement and decision-making rigour first, the tools will follow.

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.

The Greatness phishing-as-a-service (PhaaS) platform has expanded from credential phishing to adversary-in-the-middle attacks and device-code phishing targeting Microsoft 365 accounts.

The platform has been active since at least mid-2022, targeting Microsoft 365 users in the United States, Canada, the UK, Australia, and South Africa.

It evolved over the years and now targets multiple platforms, including Microsoft 365, iCloud, Yahoo, and Google Workspace.

Currently, it is sold for $289 per month to cybercriminals over a Telegram channel with thousands of subscribers.

In a recent campaign observed by researchers at email security company ZeroBEC, Greatness operators abused the RingCentral communications platform to bypass email security filters on the recipient side.

RingCentral is a communications platform used by businesses for services such as cloud calling, messaging, and voicemail.

In the Greatness phishing activity, the attacker impersonated the platform by claiming their emails came from service@ringcentral[.]com, targeting actual users of the service.

These emails used fake voicemail and performance-review notifications as lures to entice recipients to open them.

Although the messages originated from an unknown IONOS mail server, failed SPF and DMARC checks, and had no DKIM signature, they were still accepted by the receiving systems because RingCentral was whitelisted.

Moreover, the emails included a fraudulent banner claiming that the sender had been verified by the organization’s safe-sender list, which helped reduce suspicion at the human level.

ZeroBEC explains that the tactic achieved a Spam Confidence Level (SCL) of -1 on Microsoft Exchange, allowing them to bypass the normal email filtering stages.

Clicking the button embedded in those emails took victims to the Greatness infrastructure, where they were routed either through a Microsoft adversary-in-the-middle (AiTM) phishing flow that captured an MFA-approved authentication token or through a device-code phishing flow.

Post-compromise, the attacker replayed Microsoft 365 authentication tokens from VPS and commercial VPN infrastructure to access the compromised accounts.

They then enumerated Outlook mailboxes, Teams conversations, SharePoint sites, OneDrive files, contacts, calendars, and registered applications through Microsoft Graph, with access persisting for more than two weeks in some cases.

It should be noted that RingCentral recently disclosed a data breach incident which was claimed by threat actor ShinyHunters.

“This incident has affected data for a limited portion of RingCentral customers, and we are communicating with affected customers directly,” explained the company in a security bulletin published July 28.

ZeroBEC comments that it’s likely that cybercriminals using Greatness got a list of valid targets, users of the RingCentral platform, from that incident, though a connection cannot be confidently made.

The researchers recommend auditing safe-sender lists and replacing blanket domain exclusions with rules requiring valid email authentication.

Also, hunt for Greatness infrastructure and suspicious MFA-approved Microsoft 365 sign-ins from hosting or VPN addresses.

If compromise is suspected, administrators should revoke all access and refresh tokens, review OAuth consent, Microsoft Graph activity, and access to Microsoft 365 services.

Security teams log 54% of successful attacks and alert on just 14%. The rest move through your environment unseen.

The Picus whitepaper shows how breach and attack simulation tests your SIEM and EDR rules so threats stop slipping by detection.

Get the whitepaper

Jon Prosser’s latest Front Page Tech (FPT) lays out the most complete picture yet of what he calls the iPhone Ultra. He presents it as a book-style foldable that arrives this September alongside the iPhone 18 Pro and Pro Max, and positions it as the clear premium model of the year. Base iPhone 18 and Air models wait until later, with two color options being offered at launch: black / grey, and white / silver.

Closed, this phone feels completely familiar, somewhat wider than previous iPhones but still fits easily into your pocket, measuring approximately 9 to 9.5 millimeters thick. When you open it up, it’s only 4.5 mm thick, which is significantly thinner than the existing iPhone Air. Prosser claims that the difference will be noticeable the first moment someone picks up both handsets, owing to the titanium build and painstaking engineering that has gone into making the open phone feel as strong as a rock. Many users may even be tempted to forego a case entirely, as Apple proved with the Air by performing a few flashy (but probably not very rigorous) drops.

Samsung Galaxy Z Fold8 Unlocked Cell Phone + $350 Gift Card, 256GB Graphite

• PRE-ORDER NOW: Get an Amazon gift card when you pre-order Samsung Galaxy Z Fold8. You will receive an email once your gift card is available after the…
• CREATED FOR YOUR CONTENT: Any way you turn it, the new Galaxy Z Fold8 gives you a fuller viewing experience. Designed with your content in mind, this…
• WORLD’S LIGHTEST FOLD¹: Galaxy Z Fold8 is the newest foldable phone for content on the go. Unfolded, it fits movies and books for an immersive…

In terms of this phone, the hinge receives the most attention, with Prosser revealing that it is constructed on a liquid-metal design that relies on custom 3D-printed pieces and a unique plate that disperses any strain or stress. The crease depth is roughly 0.15 millimeters, so it’s not fully gone, but it’s less obvious in regular use, with the inner display looking practically seamless. This is the same architecture that allows the phone to be so thin while still feeling so robust.

Screen sizes follow a nice pattern, with a compact outer panel and a significantly larger inner one. Expect something around 5.5 inches on the exterior and 7.8 inches on the inside, with a larger aspect ratio and more tablet-like design than a regular phone. When closed, it behaves similarly to any other iPhone, but when opened, it transforms into a tiny tablet that you can still easily carry in one hand. According to Prosser, Apple actively pushed developers at WWDC to stop designing for the same old fixed sizes and orientations, and code strings for fold state and angle degrees are already appearing in iOS 27 builds to match the updated hardware.

Cameras are kept very practical rather than striving for the ‘largest and flashiest’ approach, with a dual setup on the rear consisting of a 48-megapixel main and ultra-wide lens lying on a horizontal plateau reminiscent of the Air. There is no telephoto lens in sight since there just isn’t enough room. When the phone is closed, the front-facing selfie camera handles calls and other functions. Some prototypes have placed the camera underneath the display, allowing it to disappear while the device is open, yet the physical Camera Control button remains, albeit taking up interior space.

Power comes from an A20-series chip paired with a large 12 gigabytes of RAM, which should be sufficient for Apple’s AI tasks. There is also an in-house C2 modem, which enhances efficiency and provides satellite communication. The battery capacity appears to be the highest of any iPhone so yet, built around some high-density cells in the 5,800 milliamp-hour area. Face ID is no longer available because there is no physical space for that hardware in this tiny body, however Touch ID returns in the power button.

Software receives as much attention as hardware. Split screen mode now appears on the large inner screen, and it is a long-awaited feature that makes the larger display more functional than just something to show off; running two programs side by side is the new normal. Everyday operations, such as watching a video while answering messages, no longer require awkward workarounds and instead become second nature.

The price is rumored to be far higher than $2,000, and may even exceed $2500 for those with higher end storage levels, putting the Ultra in a league far above the Pro Max. Apple has not confirmed anything, and has even sued Prosser for some previous leaks he did, but the sheer volume of matching code references, WWDC clues, and all the physical facts he has matching up is keeping the whole story alive.

It’s also highlighting classic titles from the past two and a half decades on Game Pass.

Apple has introduced new beta firmware for AirPods, AirPods Pro, and AirPods Max, so that developers can continue testing out inbound iOS 27 features.

Apple will be introducing a number of new features for select AirPods models this fall, as part of the iOS 27 and other 27-gen operating system releases. Ahead of that release, it is testing out the features via beta firmware builds.

The new firmware, build 9A5336b, replaces the previous build, 9A5314b, which Apple seeded on July 7. It can be downloaded to the AirPods Pro 2 and AirPods Pro 3, AirPods 4, and AirPods Max 2.

It is intended to be used only by developers, not the general public. A public version is expected within days.

To install the beta, users need to be running on iOS 26, iPadOS 26, macOS 26, or later releases. There is an option under the AirPods settings interface to enable the firmware installation.

Once enabled, the update itself happens automatically, when you place the AirPods on charge near the host device.

Among the changes users can expect from the AirPods firmware is a new EQ setting in iOS 27. Other changes to make the management of AirPods easier are also anticipated to arrive.

I didn’t expect Samsung to be the one crossing the finish line first here, given how much noise Dolby Vision 2 made when it was announced (September 2025). 

The company has confirmed its next HDR format, and it’s arriving on a popular OTT platform this month, beating Dolby Vision 2 to the punch.  

So what exactly is Samsung launching?

Called HDR10+ Advanced, Samsung’s new HDR format is arriving on Prime Video this month, and the company’s 2026 TV models should support it. The genuinely surprising part for me is that this format only got announced back in November, months after Dolby Vision 2, and Samsung still managed to ship it first. 

Now, Samsung hasn’t said exactly which TV models qualify, or how much Prime Video content we’ll get at launch. However, a promotional image confirms the Amazon series Ride or Die will support it.

Advertisement

Under the hood, this update bundles six separate upgrades. It includes a brightness boost built for today’s punchier TV panels, genre-specific picture tuning, more precise local dimming for mini-LED sets, and, last but not least, improved color accuracy.

At their core, both formats are chasing the same idea. They’re after smarter, scene-by-scene motion smoothing that handles fast action without slipping into that dreaded “soap opera” look older TVs are known for. 

How does this actually compare to Dolby Vision 2?

Samsung’s version goes by the name Intelligent Motion Smoothing. It’s paired with a gaming-focused mode aimed partly at cloud gaming services. Dolby Vision 2’s version is called Authentic Motion, just so you know. 

The real issue here isn’t about the similarities or the differences, in my opinion. It’s about a chaotic format war breaking out between players in the TV industry. 

Hisense, Philips, and TCL are all lining up behind Dolby Vision 2, perhaps to close the gap with premium brands that often go with proprietary chip-level picture processing and the tiered flexibility (Dolby Vision 2 vs. Dolby Vision 2 Max) that could help them differentiate between value and premium products.

Advertisement

Peacock and Canal+ have also committed to stream that format once it’s actually ready, as the format needs a software update to enable it. LG, on the other hand, is sitting this fight out entirely. Instead, it’s building its own separate system with Prime Video. 

Samsung avoided Dolby’s licensing fee while shipping a competing format first. That suggests real financial motivation behind the rush.

from the be-careful-what-you-wish-for dept

I don’t know but when your best argument is “we got locked out of our accounts for supporting violent insurrectionists,” maybe it’s time to stop arguing.

But if there’s anything Trump likes, it’s being a plaintiff in a lawsuit. So, that’s exactly what the Trump Organization argued, in hopes of pressuring Capital One into a presumably hefty settlement.

The case centers on roughly 385 accounts tied to the Trump Organization, Eric Trump and a collection of affiliated businesses — including a winery, a bottled-water company and a golf course developer — that banked with Capital One for more than a decade before the accounts were shut down in mid-2021.

In an amended complaint filed earlier in July, the Trump-affiliated companies insisted the closures had nothing to do with financial crime or money-laundering, but everything to do with politics.

The Trump Organization alleges Capital One moved to distance itself from Donald Trump after the Capitol riot and that the bank’s cited anti-money-laundering rationale was invented after the fact to cover for that decision.

Advertisement

If you need a reason to “distance” your company from persons or organizations, that’s a pretty good reason. If you’ve chosen to align yourself with violent insurrectionists and rely on them for financial support, then it’s pretty much “you’ve made your bed, now please take your business elsewhere.”

But what Capital One is alleging in its response to the Trump Organization’s lawsuit somehow makes the Trump Organization look even worse than it did when it first filed its lawsuit alleging some weird form of political persecution.

According to Capital One’s motion to dismiss [PDF], this had nothing to do with acts that were unsavory (buddying up with violent insurrectionists) and everything to do with stuff that is indisputably illegal, rather than merely unseemly.

The SAC [second amended complaint] concedes that Capital One’s decision to terminate the accounts was expressly permitted by the governing agreement and instead rests on vague allegations of political discrimination that are not supported by any of the documents attached to the SAC. To the contrary, those documents and Plaintiffs’ own allegations make clear that Capital One closed Plaintiffs’ accounts for anti-money laundering (“AML”) reasons. The closures were the result of months of analysis and a careful review by Capital One’s AML team in accordance with bank policies and regulatory guidance. Capital One never publicized the termination decision nor its confidential internal process giving rise to the closure, and it permitted Plaintiffs several months (and granted several extensions) to find new banking services, which they did.

It wasn’t because you guys wanted to destroy democracy! It’s because you seemed like you were engaged in actual crimes!

In any event, as Trump and his Trumpians surely know, private companies can terminate accounts at a moment’s notice for any reason they choose to do so. That’s the bargain consumers agree to when utilizing corporate services. Just because it happens to you doesn’t make it immediately actionable. If you don’t like the terms of the agreement, don’t agree to it. Take your business elsewhere. Otherwise, deal with it and take your business elsewhere when you’re told this particular place of business is no longer an option.

But let’s not lose sight of the main thing here. The Trump Organization walked into court insisting it was the victim of “political discrimination.” Now, the organization is facing the considerable possibility of limping out of court looking like the grifters we’ve always assumed them to be.

And claiming Capital One had a legal obligation to tell the Trump Organization it was being investigated for suspected money laundering does nothing but allow Capital One to permanently link the organization to money-laundering allegations on the public record. No one suspected of criminal activity is due a head’s up, whether it’s from a private entity or a government agency.

As Capital One points out, no matter whose name is on the letterhead, the institution’s obligations during this money laundering investigation were to the government, not to the aggrieved party hoping to turn this into a paycheck.

Plaintiffs’ argument that Capital One should have proceeded with the contractually-permitted termination process differently and given Plaintiffs an opportunity to explain suspicious transactions fails for the additional reason that Capital One had no duty to do so. As a federally regulated financial institution, Capital One is subject to the Bank Secrecy Act (“BSA”) and its implementing regulations, under which a bank’s BSA-mandated compliance obligations are “owed to the United States and not private bank customers.”

The Trump Organization now has multiple self-inflicted gunshot wounds in its feet. It could stop the bleeding by conceding defeat and agreeing to the dismissal. But if history has proven anything, it’s that Trump never knows when to stop. And while this isn’t Trump himself suing, it’s safe to assume the Trump Organization won’t accept defeat. But it probably should. If it insists on keeping this case alive, there’s a good chance some more rounds of discovery will be necessary. And when that happens, all the details supporting Capital One’s money laundering investigation are going to come out. I, for one, hope this litigation survives this motion to dismiss.

Filed Under: corruption, debanking, donald trump, insurrection, january 6, money laundering, political violence, trump administration

Companies: capital one, trump organization

Most conversations about generative artificial intelligence in schools eventually zoom in on using AI in the classroom. Before districts redesign teaching and learning around AI, they may need to answer a more fundamental question: Can schools afford an AI-first future?

The question sounds strange because generative AI is often presented as software with free and low cost tiers to individual users. Teachers open a browser window, type a prompt, and receive a response in seconds. The experience feels almost weightless and as simple as a Google search. The infrastructure behind that interaction is much more complicated.

A useful way to think about generative AI is to remember the large desktop computers that once sat in school computer labs. Students interacted with a monitor and keyboard, but much of the important work happened elsewhere inside a massive tower packed with hardware.

Today’s AI systems operate similarly, except the tower has been replaced by massive data centers located hundreds or thousands of miles away — and increasingly in some cases, just a few miles away. 

Cost of Compute

An explanation is in order. How do chatbots and the hardware behind them work? Think of the chatbot prompt as the remote control. The hardware stored at the data center is the wiring within a television, and the chatbot’s output is what appears on screen as you watch and flick through channels. 

Every student prompt, teacher-generated lesson plan or AI-assisted feedback comment depends on specialized processors, networking infrastructure, electricity, water, and increasingly scarce computing capacity.

Most discussions about AI in education begin after those systems are already in place. However, a growing body of research suggests schools should pay closer attention to the infrastructure itself.

Researchers studying AI adoption in education have largely focused on classroom implementation, AI literacy and governance. Stanford’s review of the evidence base for AI in K-12 education found that adoption continues to outpace rigorous evidence about educational outcomes. At the same time, UNESCO and other organizations have increasingly emphasized governance, transparency and human oversight as schools experiment with AI tools.

A separate body of research examines the infrastructure that makes those tools possible. Urban planners, computer engineers and environmental researchers have begun documenting the physical footprint of artificial intelligence. Their work points to a reality that is largely invisible to educators: generative AI is both software and hardware that requires robust infrastructure to support and scale. 

Research by Xiaofan Liang, PhD on data centers describes how AI expansion increasingly shapes land use, energy systems, local planning decisions and community development. Research by Shaolei Ren, PhD on power and water demand demonstrates that large-scale AI deployment carries substantial resource requirements that extend well beyond the technology sector. Researchers and policymakers are now examining how data center growth affects electricity demand, water consumption, electrical grid capacity, and environmental sustainability. 

According to estimates cited by the Congressional Research Service, U.S. data centers consumed about 176 terawatt-hours of electricity in 2023, roughly 4.4% of all U.S. electricity consumption. Using average residential electricity consumption estimates from the U.S. Energy Information Administration, that’s enough electricity to power nearly 17 million American homes for a year. The map below shows where the United States sits in the world’s energy picture and why AI’s growing appetite for power matters.

Attribution: Hannah Ritchie, Pablo Rosado, and Max Roser (2020) – “Energy Production and Consumption” Published online at OurWorldinData.org. (archived on May 18, 2026).

Traditionally, districts purchase educational technology such as learning management systems, assessment platforms and instructional software through licensing agreements that can often be forecast years into the future. But generative AI operates differently.

Unlike traditional software, which becomes cheaper to distribute as it scales, generative AI continues generating costs each time users engage with the system. Industry observers increasingly point to what’s called “inference costs,” which are the computing resources required to generate responses. These are some of the major costs of LLMs for consumers and one of the central economic challenges facing AI companies.

For schools, how can a district plan for these costs, and what happens when the costs far exceed expectations? Put another way, it’s unclear whether generative AI is financially feasible for schools. 

Many districts are currently experimenting with AI through pilot programs, limited licenses or AI features embedded within existing products. There are few examples of what universal access would actually cost. 

What would it mean for every student and their teachers to have access to generative AI every day? Before we address this question, there is another cost variable to consider: data privacy.

Many educators and parents have expressed concerns about student information flowing into commercial AI systems. One response has been to advocate for private deployments, district-controlled systems or locally hosted models that offer greater oversight and protection.

Those approaches may provide stronger governance, but they also require additional investment. That makes student data privacy a matter of policy and infrastructure. The more control schools want over data, the more likely they are to encounter costs related to storage, cybersecurity, hardware, networking and technical expertise.

Understanding the Generative AI Market

Meanwhile, the broader market continues to evolve.

OpenAI, Anthropic and other major AI companies are still competing to define the commercial landscape. Product offerings change frequently. Pricing models continue to evolve. Infrastructure investments remain enormous.

The result is a technology ecosystem with long-term economics that remains uncertain at precisely the moment schools are being encouraged to integrate it more deeply into teaching and learning. This uncertainty arrives during a challenging financial period for many districts.

Federal ESSER funding has expired. States continue debating educational technology spending priorities. District leaders face growing pressure to justify technology investments while responding to staffing shortages, student mental health concerns, and academic recovery efforts post-COVID-19 school shutdowns.

Against that backdrop, AI presents a different kind of procurement question: Do districts understand the long-term commitments they may be making when AI becomes embedded in curriculum, assessment and daily operations?

There is still one more cost factor to consider: community impact around data centers. Data centers are expanding rapidly across the United States. Local governments and residents are increasingly debating the benefits and tradeoffs associated with new facilities. Questions about energy demand, water consumption, environmental exposure and land use have become common features of public meetings and planning discussions.

For educators, these debates may seem distant from classroom practice. But every discussion about AI in schools ultimately depends on the infrastructure being built in communities across the country.

Schools are currently debating how to integrate AI into teaching and learning while the infrastructure, economics and governance systems required to support large-scale adoption are still taking shape.

Before schools decide how deeply AI belongs in classrooms, they may need a clearer understanding of how much it costs and if it’s feasible to maintain the systems that make an AI-ready classroom possible.