OpenAI is finally addressing one of the most frustrating things about working with files in ChatGPT. The company is rolling out two new features to help users quickly access previously uploaded files, including a Recent files menu and a dedicated Library tab.

How do ChatGPT’s new file management features work?

Until now, files in ChatGPT were largely tied to individual conversations, which meant finding them often involved going back to the original chat and scrolling through long threads. The new Recent files option in the attachment menu now lists some of the files you’ve used most recently, making it easier to jump back into ongoing work without digging through older chats.

On the web, there’s also a new Library tab in the sidebar. This acts as a central hub for all your uploaded and generated files, giving you a more organized view instead of tying everything to separate conversations. You can browse, search, and quickly attach files to new chats from this tab.

OpenAI also says ChatGPT can answer questions about files you’ve already uploaded, so you don’t need to reupload them every time you want more insights. Together, these changes make file reuse faster and far less tedious, especially if you regularly juggle files across multiple sessions.

Who’s getting access, and when?

The update is rolling out globally to ChatGPT Plus, Pro, and Business subscribers. Those in the EU, Switzerland, and the UK will have to wait a bit longer, with availability in these regions expected soon. There’s no word yet on whether these features will make their way to the free tier.

With these changes, OpenAI is continuing to position ChatGPT as more than just a chatbot, gradually turning it into a tool for managing ongoing work across conversations.

X is updating its revenue-sharing incentives to give more weight to engagement from a user’s home region, Nikita Bier, the company’s Head of Product has announced. Bier said the change in policy was to “encourage content that resonates with people in [the user’s] country, in neighboring countries and people who speak [their] language.”

Bier continued that while X appreciates everyone’s opinion on US politics, the company is hoping the new policy can “disincentivize gaming the attention of US or Japanese accounts.” The US and Japan have the largest number of users on X. Bier didn’t mention it outright, but dozens of popular accounts tweeting pro-Trump sentiments and commentaries focusing on US politics in general were revealed to be based outside the US late last year, when X rolled out a transparency feature that exposed users’ locations. Those accounts, which pretended to be from the US and garnered millions of likes, views and reposts, turned out to be based in countries like India, Kenya and Nigeria.

“X will be a much richer community when there’s relevant posts for people in all parts of the world,” Bier said. When one user responded to his post that some countries barely have any users, making it hard to earn money from the website, Bier just suggested that they should write about their day-to-day experiences. “Of course, you’re welcome to continue chiming in on America politics. We just won’t send money overseas for that content,” he said. X’s new policy will start taking effect on Thursday, March 26.

Since Donald Trump’s war on Iran started more than three weeks ago, United States military forces have allegedly attacked more than 9,000 sites, creating a climate of fear and constant uncertainty for Iranians in Tehran and across the country. Without an advanced warning system from the government, and amid the longest internet shutdown in Iran’s history, Iranians are left in an information void.

Even before Israel and the United States began dropping bombs, Iran’s lack of a public emergency alert tool and severe state-controlled digital oppression has impacted tens of millions of citizens. Since the 12-day Israel-Iran war last year, though, a group of Iranian digital rights activists and volunteers has been working to fill the gap with a dynamic, regularly updated mapping platform called Mahsa Alert. The project can’t replace real-time early alerts that could come from a coordinated government service, but the tool sends push notifications when Israeli forces warn about attacks, details some confirmed strike locations, and offers offline mapping capabilities.

“There is no emergency alert in Iran,” says Ahmad Ahmadian, the president and CEO of US-based digital rights group Holistic Resilience, which is behind Mahsa Alert and has been developing the platform since last summer. “This was where we saw the traction, we saw the need, and we continued working on it with the volunteers, with some [open source intelligence] experts, and used this to map the repression machinery ecosystem of Iran and surveillance.”

Mahsa Alert is a website but also has Android and iOS apps, which were intentionally designed to be lightweight and easy to use on any device. Given the heavy government connectivity control inside Iran and erratic access to the internet, volunteers also prioritized engineering the platform for offline use. And it can be easily updated if a user does get connectivity for a brief period by downloading APK files that contain new data. The team works to keep these updates extremely small; a recent release was 60 kilobytes, and Ahmadian says they are typically no more than 100 kilobytes.

One overlay on Mahsa Alerts plots the locations of “confirmed attacks” that Ahmadian says his team or other OSINT investigators have verified, using video footage or images that are submitted to a Telegram bot or shared on social media. There are also warnings about areas where Israeli forces have issued evacuation alerts, along with the crucial component of people submitting reports on what is happening around them.

“We have to go through a due diligence and verification process and tag them before putting them on the map,” Ahmadian says of the reported attacks and incidents, adding that the team has a backlog of more than 3,000 reports that it is working through or is unable to verify. Along with attempting to map strikes, the team behind Mahsa Alert have also plotted “danger zones” that could be at risk of attack—such as sites linked to Iran’s nuclear program or military—so ordinary citizens can stay away from them. Ahmadian claims 90 percent of attacks it has confirmed were at sites that were already present on the map. “Some of them that we can confirm, we do it because [a user] has shared a photo or they have shared some details that makes them verifiable,” he says.

The map also includes locations of thousands of CCTV cameras, suspected government checkpoints, and other domestic infrastructure. Medical facilities, such as hospitals and pharmacies, are included on the map along with other resources like the locations of religious sites and past protests.

Mahsa Alert has become more visible on global social media feeds as Iranians around the world share details from the map, encouraging people to look into the service and flagging it for friends and family who could use it as a resource. “The app went from near zero to over 100,000 daily active users in a matter of days,” Ahmadian says, adding that in total there have been around 335,000 users this year, with people first turning to the app during the Iranian regime’s brutal crackdown on anti-government protesters in January. Through the limited user information the app collects, Ahmadian claims there are signs that 28 percent of users are accessing the platform from inside Iran.

Last week’s Nvidia GTC conference highlighted new chip architectures to power AI. But as the chips become faster and more powerful, the remainder of data center infrastructure is playing catchup. The power delivery community is responding: Announcements from Delta, Vertiv, and Eaton showcased new designs for the AI era. Complex and inefficient AC to DC power conversions are gradually being replaced by DC configurations, at least in hyperscale data centers.

“While AC distribution remains deeply entrenched, advances in power electronics and the rising demands of AI infrastructure are accelerating interest in DC architectures,” says Chris Thompson, vice president of advanced technology and global microgrids at Vertiv.

AC to DC Conversion Challenges

Today, nearly all data centers are designed around AC utility power. The electrical path includes multiple conversions before power reaches the compute load. Power typically enters the data center as medium-voltage AC (1kV to 35kV), is stepped down to low-voltage AC (480V or 415V) using a transformer, converted to DC inside an uninterruptible power supply (UPS) for battery storage, converted back to AC, and converted again to low-voltage DC (typically 54 V DC) at the server, supplying the DC power computing chips actually require.

“The double conversion process ensures the output AC is clean, stable and suitable for data center servers,” says Luiz Fernando Huet de Bacellar, vice president of engineering and technology at Eaton.

That setup worked well enough for the amounts of power required by traditional data centers. Traditional data center computational racks draw on the order of 10 kW each. For AI, that is starting to approach 1 MW. At that scale, the energy losses, current levels, and copper requirements of AC to DC conversions become increasingly difficult to justify. Every conversion incurs some power loss. On top of that, as the amount of power that needs to be delivered grows, the sheer size of the convertors, as well as the connector requirements of copper busbars, becomes untenable. According to an Nvidia blog, a 1 MW rack could require as much as 200 kg of copper busbar. For a 1 GW data center, it could amount to 200,000 kg of copper.

Benefits of High-Voltage DC Power

By converting 13.8 kV AC grid power directly to 800 VDC at the data center perimeter, most intermediate conversion steps are eliminated. This reduces the number of fans and power supply units, and leads to higher system reliability, lower heat dissipation, improved energy efficiency, and a smaller equipment footprint.

“Each power conversion between the electric grid or power source and the silicon chips inside the servers causes some energy loss,” says Fernando.

Switching from 415 V AC to 800 V DC in electrical distribution enables 85 percent more power to be transmitted through the same conductor size. This happens because higher voltage reduces current demand, lowering resistive losses and making power transfer more efficient. Thinner conductors can handle the same load, reducing copper requirements by 45 percent, a 5 percent improvement in efficiency, and 30 percent lower total cost of ownership for GW-scale facilities.

“In a high-voltage DC architecture, power from the grid is converted from medium-voltage AC to roughly 800 V DC and then distributed throughout the facility on a DC bus,” said Vertiv’s Thompson. “At the rack, compact DC-DC converters step that voltage down for GPUs and CPUs.”

A report from technology advisory group Omdia claims that higher voltage DC data centers have already appeared in China. In the Americas, the Mt. Diablo Initiative (a collaboration among Meta, Microsoft, and the Open Compute Project) is a 400 V DC rack power distribution experiment.

A handful of vendors are trying to get ahead of the game. Vertiv’s 800 V DC ecosystem that integrate with NVIDIA Vera Rubin Ultra Kyber platforms will be commercially available in the second half of 2026. Eaton, too, is well advanced in its 800 V DC systems innovation courtesy of a medium-voltage solid-state transformer (SST) that will sit at the heart of DC power distribution system. Meanwhile Delta, has released 800 V DC in-row 660kW power racks with a total of 480 kW of embedded battery backup units. And, SolarEdge is hard at work on a 99%-efficient SST that will be paired with a native DC UPS and a DC power distribution layer.

But much of the industry is far behind. Patrick Hughes, senior vice president of strategy, technical, and industry affairs for the National Electrical Manufacturers Association, says most innovation is happening at the 400 V DC level, though some are preparing 800 V DC. He believes the industry needs a complete, coordinated ecosystem, including power electronics, protection, connectors, sensing, and service‑safe components that scale together rather than in isolation. That, in turn, requires retooling manufacturing capacity for DC‑specific equipment, expanding semiconductor and materials supply, and clear, long‑term demand commitments that justify major capital investment across the value chain.

“Many are taking a cautious approach, offering limited or adapted solutions while waiting for clearer standards, safety frameworks, and customer commitments,” said Hughes. “Building the supply chain will hinge on stabilizing standards and safety frameworks so suppliers can design, certify, manufacture, and install equipment with confidence.”

Gamers have long been searching for a computer that can be slipped into a coat pocket and used to complete tasks, a dream that now appears to be within reach due to a creative designer who wrapped a Razer Edge tablet within a custom 3D printed shell.

Flip the lid open and a familiar tablet screen greets you, cleanly framed in black plastic with just enough orange trim to make its intentions clear. A compact Bluetooth keyboard sits snugly in the base, and when everything is folded shut the whole thing is no bigger than a large phone, slim enough to disappear into a pocket or bag without a second glance. Those orange accents on the hinges and keycaps are a quiet reminder that this is anything but an ordinary device.

Sale

ASUS ROG Xbox Ally – 7” 1080p 120Hz Touchscreen Gaming Handheld, 3-month Xbox Game Pass Premium…

• XBOX EXPERIENCE BROUGHT TO LIFE BY ROG The Xbox gaming legacy meets ROG’s decades of premium hardware design in the ROG Xbox Ally. Boot straight into…
• XBOX GAME BAR INTEGRATION Launch Game Bar with a tap of the Xbox button or play your favorite titles natively from platforms like Xbox Game Pass…
• ALL YOUR GAMES, ALL YOUR PROGRESS Powered by Windows 11, the ROG Xbox Ally gives you access to your full library of PC games from Xbox and other game…

The project started with a Razer Edge picked up for around $80, a tablet that had largely faded from the spotlight since its release but still packed a capable Qualcomm Snapdragon G3x Gen 1 processor, plenty of RAM, and Android 12 under the hood. It came without the original controllers, but at that price it was too good a candidate to pass up.

There were already some design files floating around online for modular clamshells that could hold a phone, so it was just a matter of modifying them a little to fit the Razer Edge. Then it was simply a matter of using free editing software to make the necessary changes and printing them on a regular consumer printer. This all came together with simple screws and pins for the hinges, and a few lock sliders on the front keep the whole thing shut securely when you’re traveling.

The major challenge was getting the tablet inside without damaging it. A metal ring affixed to the rear of the Razer Edge, combined with a MagSafe-style adaptor on the case, locked everything together with powerful magnets that can be peeled away with some moderate coaxing. The tablet can leave the shell in seconds and return to becoming a tablet whenever it feels like it. The keyboard simply fits into a little tray in the bottom and automatically pairs over Bluetooth. Given its size, the layout is quite decent, and all of the shortcuts for doing daily tasks are available. When the lid closes, everything tucks neatly against the screen.

Power it up and things get interesting fast. Android 12 handles all the everyday essentials without breaking a sweat, and cloud streaming over Wi-Fi or mobile broadband opens up a much bigger games library on top of that. Emulation is where it really shines though, running GameCube titles at 720p with a solid frame rate and pushing PlayStation 2 games to 1.75 times their native resolution on many titles. Lighter PC games load up through dedicated apps and run without issue, and for anyone feeling nostalgic there is even a Windows 98 simulator tucked in there for good measure.
[Source]

The brand expects the store to break even in two years

Singaporean furniture retailer Castlery will open a showroom in New York on May 15, making it one of the very few homegrown companies to establish a permanent retail presence there. This marks the next phase of growth for the company in the United States, following six years of operating online-only in the market.

Co-founder Declan Ee called the brick-and-mortar flagship outlet, a first in the US, as a “natural progression” from its digital retail model.

“The goal was always to create a best-in-class experience for our customers… and the final piece of this experience is completed when we have an offline store,” he said.

The 3,000-square-foot showroom in Manhattan’s Chelsea neighbourhood represents a seven-figure investment on a 10-year lease. Ee’s team scouted over 200 sites over two years before choosing this one.

The showroom features 17 fully furnished room settings and a complimentary interior styling service that will advise customers on space planning, furniture selection and interior layout.

Ee told The Business Times that he expects the store to break even within 1.5 years to 2 years, or even within a year if sales are strong.

The opening of the store in the Big Apple marks Castlery’s fourth showroom worldwide, following the opening of its third in Brisbane last Aug. Its Sydney store was set up in 2024 and expanded in 2025, while its 24,000 sq ft flagship store in Liat Towers was established in 2022.

Castlery is in 5 markets, with most sales coming from the US

Castlery was founded in 2013 by Ee and his co-founders, Fred Ji, Zhou Zhiwei and Travers Tan, as a digital retail furniture brand. It currently employs more than 500 staff worldwide, with 200 in its Singapore headquarters.

To date, the brand has sold more than 1 million pieces of furniture and introduced more than 7,000 products.

The label prides itself on affordable, consistent pricing worldwide, ranging from S$399 for a swivel chair to S$2,499 for a leather recliner.

Its first overseas foray came in 2017, when it entered Australia, a market 10 times bigger than Singapore at the time, according to the founders.

The brand entered the US in 2019 during the COVID-19 pandemic as an online brand, starting with two warehouses in New Jersey and Los Angeles, California. Today, Castlery reaches all 50 states from six US warehouses, with the addition of sites in Seattle and Georgia in 2023, and then Texas and Chicago in 2024.

Ee noted that this has reduced delivery times to its US customers, many of whom rent their homes and need furniture delivered with short lead times.

“We were very aggressive in the first two to three years, when we were scaling the business online in the US,” he said.

The US currently makes up Castlery’s largest market by contributing to 65% of the company’s overall sales. Australia comes in second at 17%, followed by Singapore at 15%. The UK and Canada, where Castlery expanded online in 2025, make up the remaining 3%.

The New York store will serve as a testing ground amid evolving market conditions

With this offline expansion, Ee said Castlery will take a “measured” approach given evolving global developments and geopolitical tensions.

The New York showroom will be a testing ground for Castlery before it decides to commit to more showrooms in the country.

Well aware of New York’s competitive retail scene with players such as West Elm and Crate & Barrel that have multiple outlets, Ee acknowledged that this will give consumers plenty of options.

“There’s a lot of room for us to grow in the US, but we’re taking things step by step because one’s perspective changes after opening the first store. You get data, you see how customers react and their basket size—all these things,” he explained.

As with other foreign companies in the US, Castlery was hit hard by US President Donald Trump’s “Liberation Day” baseline tariffs in 2025. This was on top of duties on certain furniture imports, such as upholstered furniture and kitchen cabinets.

As more than half the brand’s products were being manufactured in China and then shipped directly to US customers, Castlery saw its Chinese imports slapped with the highest tariff rates of close to 30%.

Castlery has since diversified its supply chain to reduce its exposure to tariffs. It has moved some of its manufacturing from China to places such as Vietnam, Thailand, and India, leaving only about 20% of its production in China today.

After diversifying its supply chains, Ee said production costs have risen, given higher minimum-order quantities.

This has caused profits to fall by 1% to 3%, which Ee noted is not a negligible amount for a growing furniture brand that typically enjoys margins of 4% to 8%. The tariffs also created consumer uncertainty, leading to a six-month dip in sales, though they have since recovered.

Besides the tariffs, geopolitical tensions have put additional pressure on Castlery’s bottom line. Rising fuel prices amid the ongoing Middle East conflict have squeezed its profit margins.

Taking all these factors into account, Ee expects Castlery’s revenue growth for the current FY2026 ending in Mar to be “flat or in the single-digit” range, down from FY2025’s 10% to 15% year-on-year growth.

A step closer to Castlery’s global ambitions

That said, Ee is still “cautiously optimistic” about Castlery’s growth prospects.

“We control what we can. You don’t know where the wind will blow, so you build the sail to catch it,” he said.

“In our case, it’s about being close to the customer and creating products that they would want to buy, even in difficult economic times.”

The opening of the New York store brings the brand a step closer to its global ambitions.

By 2029, Ee aims to have eight to 12 showrooms in key cities worldwide, including Washington, D.C, Los Angeles, San Francisco and Seattle, as well as in Melbourne and Perth in Australia.

Ee is actively scouting for retail locations in London as well, seeing Castlery’s UK online sales double month-on-month until Nov 2025 following a pop-up it held at the London Design Festival in Sep that year.

Ee explained: “Unlike the US, there are not so many big furniture brands in the UK. So we think there’s space for us to enter the market, not to mention that the sales pick-up from customers has been very encouraging.”

Achieving its expansion plans would place Castlery “on track” to evolve from a digital-first furniture retailer into a “proper global retail brand.”

“If we’re nationwide (in a single market), it gives customers a sense of assurance that we’re not just an online challenger brand, but a serious operator.”

• Learn more about Castlery here.
• Read other articles we’ve written on Singaporean businesses here.

Featured Image Credit: Castlery

from the fair-trade-ai-training-data dept

As many of the AI stories on Walled Culture attest, one of the most contentious areas in the latest stage of AI development concerns the sourcing of training data. To create high-quality large language models (LLMs) massive quantities of training data are required. In the current genAI stampede, many companies are simply scraping everything they can off the Internet. Quite how that will work out in legal terms is not yet clear. Although a few court cases involving the use of copyright material for training have been decided, many have not, and the detailed contours of the legal landscape remain uncertain.

However, there is an alternative to this “grab it all” approach. It involves using materials that are either in the public domain or released under a “permissive” license that allows LLMs to be trained on them without any problems. There’s plenty of such material online, but its scattered nature puts it at a serious disadvantage compared to downloading everything without worrying about licensing issues. To address that, the Common Corpus was created and released just over a year ago by the French startup Pleias. A press release from the AI Alliance explains the key characteristics of the Common Corpus:

Truly Open: contains only data that is permissively licensed and provenance is documented

Multilingual: mostly representing English and French data, but contains at least 1[billion] tokens for over 30 languages

Diverse: consisting of scientific articles, government and legal documents, code, and cultural heritage data, including books and newspapers

Advertisement

Extensively Curated: spelling and formatting has been corrected from digitized texts, harmful and toxic content has been removed, and content with low educational content has also been removed.

There are five main categories of material: OpenGovernment, OpenCulture, OpenScience, OpenWeb, and OpenSource:

OpenGovernment contains Finance Commons, a dataset of financial documents from a range of governmental and regulatory bodies. Finance Commons is a multimodal dataset, including both text and PDF corpora. OpenGovernment also contains Legal Commons, a dataset of legal and administrative texts. OpenCulture contains cultural heritage data like books and newspapers. Many of these texts come from the 18th and 19th centuries, or even earlier.

OpenScience data primarily comes from publicly available academic and scientific publications, which are most often released as PDFs. OpenWeb contains datasets from YouTube Commons, a dataset of transcripts from public domain YouTube videos, and websites like Stack Exchange. Finally, OpenSource comprises code collected from GitHub repositories which were permissibly licensed.

The initial release contained over 2 trillion tokens – the usual way of measuring the volume of training material, where tokens can be whole words and parts of words. A significant recent update of the corpus has taken that to over 2.267 trillion tokens. Just as important as the greater size, is the wider reach: there are major additions of material from China, Japan, Korea, Brazil, India, Africa and South-East Asia. Specifically, the latest release contains data for eight languages with more than 10 billion tokens (English, French, German, Spanish, Italian, Polish, Greek, Latin) and 33 languages with more than 1 billion tokens. Because of the way the dataset has been selected and curated, it is possible to train LLMs on fully open data, which leads to auditable models. Moreover, as the original press release explains:

By providing clear provenance and using permissibly licensed data, Common Corpus exceeds the requirements of even the strictest regulations on AI training data, such as the EU AI Act. Pleias has also taken extensive steps to ensure GDPR compliance, by developing custom procedures to enable personally identifiable information (PII) removal for multilingual data. This makes Common Corpus an ideal foundation for secure, enterprise-grade models. Models trained on Common Corpus will be resilient to an increasingly regulated industry.

Another advantage for many users is that material with high “toxicity scores” has already been removed, thus ensuring that any LLMs trained on the Common Corpus will have fewer problems in this regard.

The Common Corpus is a great demonstration of the power of openness and permissive copyright licensing, and how they bring benefits that other approaches can’t match. For example: “Common Corpus makes it possible to train models compatible with the Open Source Initiative’s definition of open-source AI, which includes openness of use, meaning use is permitted for ‘any purpose and without having to ask for permission’. ” That fact, along with the multilingual nature of the Common Corpus, would make the latest version a great fit for any EU move to create “public AI” systems, something advocated on this blog a few months back. The French government is already backing the project, as are other organizations supporting openness:

The Corpus was built up with the support and concerted efforts of the AI Alliance, the French Ministry of Culture as part of the prefiguration of the service offering of the Alliance for Language technologies EDIC (ALT-EDIC).

This dataset was also made in partnership with Wikimedia Enterprise and Wikidata/Wikimedia Germany. We’re also thankful to our partner Libraries Without Borders for continuous assistance on extending low resource language support.

The corpus was stored and processed with the generous support of the AI Alliance, Jean Zay (Eviden, Idris), Tracto AI, Mozilla.

Advertisement

The unique advantages of the Common Corpus mean that more governments should be supporting it as an alternative to proprietary systems, which generally remain black boxes in terms of where their training data comes from. Publishers too would also be wise to fund it, since it offers a powerful resource explicitly designed to avoid some of the thorniest copyright issues plaguing the generative AI field today.

Follow me @glynmoody on Mastodon and on Bluesky. Originally published to Walled Culture.

Filed Under: ai, ai training, common corpus, copyright, open licenseing, open licensing, public domain, training data

Companies: common corpus, pleias

from the here-comes-the-comments dept

The polarization over any and all uses of artificial intelligence and machine learning continues. And, to be clear, I very much understand why this is all so controversial. Any new technology that has the chance to be transformative will also necessarily be disruptive and that causes fear. Fear that is not entirely unfounded, no matter your other opinions on the matter. If that’s you, cool, I get it.

I’ll start this off by pointing to the latest edition of the Techdirt podcast in which both Mike and Karl engaged in a fantastic discussion about the use of AI. I’ve listened to it twice now; it’s that good. And, while I found myself arguing out loud with the both of them at certain points during the podcast, despite the fact that neither of them could hear my retorts, it presents a grounded, often nuanced conversation, which we need much more of in this space.

And now, in what might be a subconscious attempt by this writer to commit suicide by comments section, let’s talk about that controversial demo of NVIDIA’s forthcoming DLSS 5 technology. What DLSS 5 does compared with previous versions of the technology is indeed new, but what is not new is the introduction of AI and machine learning into the equation. DLSS 2 and 3 had that already, in the form of pixel reconstruction and frame generation. DLSS 5, however, introduced what is being labeled as “neural rendering”, which uses machine learning to alter the lighting and detailed appearances in environments and, most importantly, character rendering over the engine on top of the 2D image output. Here’s the video demo that got everyone talking.

The backlash to the video was wide, immediate, and furious. There was a great deal of talk about the alteration of artistic intent, about whether this changed what the original developers were attempting to portray when they created the games, and, of course, industry jobs. I want to talk about the major complaint pillars seen across many outlets below, but this backlash also supposedly came with death threats foisted upon NVIDIA employees. I would very much hope we could all at least agree that any threats of that nature are completely inappropriate and absurd.

With that, here is what I’ve seen in the backlash and what I’d want to say about it.

Get your damned AI out of my games!

Perhaps not the most common pushback I saw in all of this, but a very common one. And a silly one, too. As I mentioned above, DLSS versions already used some version of AI and machine learning. That isn’t new. How it’s applied is certainly new, but that isn’t the same as the demand to keep AI entirely out of the video game industry.

And if that’s where you are, go ahead and shake your fist at the clouds in the sky. AI is a tool and, as I’ve now said repeatedly, the conversation we should be having is how it’s used in gaming, not if it’s used. That’s because its use is largely a foregone conclusion and it is an open question as to whether its use will be a net benefit or negative overall to the industry. Dogmatic purists on AI have a stance that is understandable, but also untenable. We’re too far down this road to turn around and go home. And if the tech were able to lower the barriers of entry to the gaming industry, acting as the fertilizer that allows a thousand indie studios to sprout roots, would that really be so bad for the gaming ecosystem?

I can appreciate the purists’ point of view. I really can. I just don’t see where they have a place in the conversation when it comes to gaming.

It overrides artistic intent!

Does it? If it did, then hell yes that’s bad. But if it doesn’t, then this concern goes away entirely.

DLSS 5 is built with options and customizable sliders for game developers. That’s really, really important here. At the macro level, a developer that has decided to use DLSS 5, or decided and customized how it’s used in their games, is exercising consent over their products. That should be obvious.

But then we get into really interesting questions of art, the actual artist, and the ownership of that art, because those last two are very different things. As Digital Foundry outlines:

It may even raise consent and other questions surrounding artistic integrity. On site and witnessing the demos in motion, concerns about this seemed less of a problem when the games we saw had been signed off by the studios that made them – the contentious assets we’ve seen, likewise. Nothing from the DLSS 5 reveal released by Nvidia has not been approved by the studios that own those games. But perhaps the issue isn’t just about specific approvals by specific developers on agreed DLSS 5 integrations, but rather the whole concept of a GPU reinterpreting game visuals according to a neural model that has its own ideas about what photo-realism should look like.

While we’ve seen endorsements from Bethesda’s Todd Howard and Capcom’s Jun Takeuchi, to what extent does that consent apply to the entire development team and other artists associated with the production? And by extension, there is also the question of whether now is the right time to launch DLSS 5 at a time when the games industry is under enormous pressure, jobs are on the line and cost-cutting is a major focus in the triple-A space. The technology itself cannot function without the work of game creators – it needs final game imagery to work at all – but the extent to which it could be viewed as a worrying sign of “things to come” cannot be overstated bearing in mind the reactions elsewhere to generative AI.

That strikes me as a valid and interesting ethical question when it comes to the use of this technology, but one that is probably overwrought. Individual artists who work on video games already have their artistic output live at the pleasure of the game developers they contract with. Those developers already can use this game art in all kinds of ways that the individual artist may not have had in mind when creating it, or indeed have even considered such possibilities. DLSS 5 is just one more version of that, with the main difference being that it involves AI making changes to game images. That’s an important thing to consider, sure, but there are cousins to this ethical question that we’ve all come to accept already. This strikes me more as part of the “all AI is bad all the time” crowd finding a foothold in something other than dogma to grab onto.

Developers and publishers own their games. If they want to use DLSS 5 in those games, there is little other than specific work for hire or other contractual stipulations with individual artists that would keep them from implementing it. If artists don’t like that, I completely understand that point of view, but that’s what contract negotiations and language are for.

Bottom line: I have been as vocal as anyone arguing that video games are a form of art for well over a decade now and I struggle to agree that an optional technology that has approved buy in from game developers and publishers equates to “overriding artistic intent”, writ large.

The faces in these examples look like shit, are “yassified”, or suffer from the uncanny valley effect!

Look, here we’re going to get into matters of opinion. I have to say that when I viewed the demo video myself, I had the opposite reaction. And, yes, this opens me up to claims that I am somehow a massive fan of AI-created pornography (this is where the yassified comments come in), or that I just want all the characters to look “hot” (I’m too old for that shit), or that my older age of 44 means I’ve lost touch with what video games should look like. Despite my genuine respect for the dissenting opinions here, allow me to say this: bullshit.

The caveat to all of this is that the demo revealed very little in the way of this technology working within these games in motion. It’s also certainly true that NVIDIA chose the best potential images to show off its new technology. If the DLSS 5 rendering sucks out loud in a larger in-motion game, or if the images it creates end up being inconsistent throughout gameplay, or if it does just end up looking shitty, then I’ll be right there with you with a torch and pitchfork in hand.

And here’s the other thing to consider with this particular complaint, combined with the previous one about artistic intent: do any of you use visual mods in your games? I do. A ton of them. For a variety of reasons. I have used them to alter the faces and models for games like Starfield and Skyrim, among many others. Do I need to feel bad for altering the artist’s intent? Do I need to apologize for incorporating mods to make characters and environments appear in a way that helps me better connect with the game I’m playing?

Because I’m not going to do either. And I don’t expect you to. Nor do I expect game developers that choose to use this optional technology to beg for forgiveness for their own output.

The hardware demands to run all of this are insane!

Fine, then you’ll get what you want and nobody will be able to use this technology anyway. But I don’t think that will be the case. NVIDIA knows what it will take to run this tech once it leaves the demo stage and goes into production. The idea that they would hype up technology that nobody can use strikes me as unlikely in the extreme.

Conclusion: everyone take a breath

This still strikes me as more of a “all AI is bad” crowd grasping at lots of other things to buttress their pushback than anything else. AI has plenty, plenty of potential pitfalls. Worried about jobs in the gaming industry and elsewhere? Me too! But if you’re not also looking at the potential upsides for the industry, then you’re engaging in dogma, not conversation.

Will DLSS 5 be good? I have no idea and neither do you. Will DLSS 5 alter previously released games in a way that fundamentally alters how we play these games? I have no idea and neither do you. Will it negatively impact the gaming industry when it comes to the number of jobs within it? I have no idea and neither do you.

This was a tech demo. Details on how it works are still trickling out. Most recently, there has been some clarification as to the 2D rendering nature of the technology and what that means for the output on the screen. As an early demo of the technology, feedback is going to be important, so long as it’s informed and reasonable feedback.

The technology may end up being trash and hated for reasons other than “all AI is bad all the time.” If that ends up being the case, I trust the gaming market to work that out for itself. But a lot of the hand-wringing here looks to me to be speculative at best.

Filed Under: ai, developers, dlss 5, rendering, video games

Companies: nvidia

The US Federal Communications Commission (FCC) is tasked with regulating both wired and wireless communications, which also includes a national security component. This is how previously the FCC tossed networking gear made by Huawei and foreign-manufactured drones onto its Covered List, effectively banning it from sale in the US. Now foreign-made consumer routers have been added to this list, barring explicit conditional approval on said list that would exempt them during a ‘transition phase’.

As per the FCC fact sheet, this follows after determination by an interagency body that such routers “pose unacceptable risks to the national security of the United States [..]”. This document points us to the National Security Determination PDF, which attempts to lay out the reasoning. In it is noted that routers are an integral part of every day life, and compromised routers are a major risk factor, ergo it follows that only US-manufactured routers are to be trusted.

These – so far fictional – US-manufactured consumer routers would have to feature ‘trusted supply chains’, which would seem to imply onshoring a large industrial base, though without specifying how deep this would have to go it’s hard to say what would be involved. The ‘supporting evidence’ section also only talks about firmware-related vulnerabilities, which would imply that US firmware developers do not produce CVEs.

Currently there do not appear to be any specific details on what router manufacturers are supposed to do about this whole issue, though they can continue to sell previously FCC-approved routers in the US.

Although hardware backdoors are definitely a possibility, this requires a fair bit of effort within the supply chain that should generally also fairly easily to detect. Yet after for example Bloomberg claimed in 2018 that Supermicro gear had been infested with hardware backdoors, this started a years-long controversy.

Meanwhile actually verified issues with Supermicro hardware are boringly due to software CVEs. In that particular issue from 2024 two CVEs were discovered involving a lack of validation of a newly uploaded firmware image.

All of which is reminiscent of an early 2024 White House ‘memory safety appeal’ that smelled very strongly of red herring. Although it’s easy to point at compromised hardware with scary backdoors and sneaky software backdoors hidden deep inside firmware of servers and networking devices, the truth of the matter is that sloppy input validation is still by far the #1 cause of fresh CVEs each year, especially if you look at the CVEs that are actually being actively exploited.

As for this de-facto ban on new routers being sold in the US, this will correspondingly not change much here. The best defense against issues with networking equipment is still to practice network hygiene by keeping tabs on what is being sent on the LAN and WAN sides, while a government could e.g. force consumer routers to pass a strict independent hardware and software audit paid for by the manufacturer.

Speaking as someone who used to run DIY routers for the longest time built around FreeSCO and Smoothwall Linux, there’s also always the option of turning any old PC into a router by putting a bunch of NICs and WNICs into it and run SmoothWall, OpenWRT, etc.. A router is after all just a specialized computer, regardless of what the government feels that it identifies as.