On January 7, Mark Zuckebrerg made a not-so-stunning announcement.The Meta CEO said that his company is ending its third-party fact-checking program, declaring that experts, like everyone else, have their own biases and perspectives. Instead, itll allow more speech and rely on a user-generated community notes feature like Elon Musks X already does.This is a culmination of a yearslong rightward lurch away from the social media giant. Right-wing media has always thrived on Facebook, while left-wing critics have long criticized the company for everything from enabling genocide to allowing hate speech to thrive. But Republicans in Congress and on the campaign trail have scolded Meta for kicking Donald Trump off its platform in the wake of the January 6, 2021, insurrection and for heavily monitoring misinformation during the COVID-19 pandemic.But in recent years the company has taken exaggerated strides to demonstrate that theyre no lefties. Meta allowed Trump to rejoin its platforms in 2023, announced it was deprioritizing political content across Instagram and Threads, and this week made former Bush White House official Joel Kaplan its global affairs chief. Oh, and Meta gave $1 million to Trumps inauguration committee.This is a great opportunity for us to reset the balance in favor of free expression, Kaplan told Fox News in an interview. What were doing is were getting back to our roots and free expression.With Trump assuming the presidency on January 20, Meta knows it needs to play nice with the administration. Trump previously claimed that Zuckerberg plotted against him during the 2020 election and wrote that he would spend the rest of his life in prison if he tried it again.But the move to end Metas fact-checking program is just as much a reflection of the current state of its main platform, Facebook, than a political offering to the party in power. I mean, have you seen Facebook lately?Take a spin around Facebook and youll see baby announcements, graduations, and birthday messages. But youll also see a glut of artificial intelligence-generated slop images. My Facebook feed doesnt have political ruminationsit doesnt have anything political at all. Theres scant updates from any news organizations, name-brand influencers, or even friends posting about whats going on in the world. Irrelevant advertisements fill my feed alongside groups I do not follow, and content that seems designed to evoke shock and awe from someone who has never encountered the internet before.Theres a nautical-themed bathroom thats so obviously made with AI that the tiles on the floor are unevenly curved. Theres a nostalgia post from a Facebook group called I grew up in New York (I didnt). And theres a supposedly funny image from a page called Bussin With The Boys that says Never forget when Jared Goff wore his own jersey to a bar when he was a rookie. Not only did I see this recycled post on Twitter a week ago, but its false. The quarterback was at a Make-a-Wish event. The Facebook fact-checkers wont have a chance at that one, but maybe an army of community notes commenters will handle it. (Thats unlikely.)The reality is that, yes, Meta is playing nice with Trump and Co. But Facebooks current status is so painfully dumbed-down, ad-ridden, and slop-filled that its perhaps a move for the best. Facebook is unhealthy beyond repair. We should stop pretending that its anything but a dead platform.

Polish studio Mistovia has paired plywood, walnut and graphite accents with a mishmash of vintage furniture in Plener Istebna a pair of gabled cottages in the Silesian Beskids mountains.The two-storey holiday homes sit side-by-side in Istebna village, with slatted wooden facades and shingled roofs that nod to the region's vernacular buildings.Plener Istebna is a pair of gabled holiday homesCreated by architect Justyna Boduch to accommodate up to nine people each, the 70-square-metre cottages feature identical layouts but subtly different interior designs by Katowice-based Mistovia.The first home, Terracotta, owes its name to the sandy-hued ceramic flooring and burnt orange upholstery found in its open-plan living space.One of the cottages is called Terracotta for its sandy-hued flooringA circular walnut table defines the kitchen and dining area, which includes locally sourced stools with traditional decorations and a feature wall clad with Ukrainian stove tiles."The heart of each cottage is undoubtedly the living space," said Mistovia founder Marcin Czopek.Stone cottage is accented by veiny graphite flooringStone, the second cottage, is accented by veiny graphite flooring in the kitchen and living area, chosen to mimic the speckled appearance of marble."This provides a solid background for the furniture," Czopek told Dezeen. "Warmth is created through abundant wood."Both homes include chunky plywood staircasesMistovia chose a selection of vintage pieces to furnish Stone, including an oval timber dining table and a low-slung 1970s Danish armchair that once formed part of a modular sofa."The furniture comes from various decades and sources," explained Czopek. "Fitting for the Polish-Czech-Slovak borderland, many pieces have Czechoslovakian roots."The private rooms echo their respective living spacesBoth holiday homes feature chunky plywood staircases that connect their ground floors with private spaces upstairs. In the Stone cottage, plywood was also used to clad the ceiling of the living space."The wood pleasantly contrasts with raw concrete and clay plaster walls with added straw chaff," said Czopek.Read: Snhetta designs hikers' refuge in Pyrenees mountains with swooping grass roofThe living spaces across the two cottages are also united by sleek stainless steel kitchen cabinetry, cone-shaped pendant lights and floor-to-ceiling glazing that connects each interior to a terrace with views of the surrounding Silesian Beskids mountains.Upstairs, the private rooms were designed to echo their respective living spaces with tiled, glass and wooden details and colourful soft furnishings.Mistovia is known for its residential projectsMistovia specialises in residential interiors and has quickly cemented a reputation for bringing together unusual materials, as seen in an eclectic Warsaw apartment that Czopek and his team renovated for an art director and her pet dachshund.Last November, the studio took home the title of emerging interior designer of the year at the 2024 Dezeen Awards.The photography is byOni Studio.The post Mistovia designs duo of Polish holiday homes to complement mountainous surroundings appeared first on Dezeen.

Nearly a week after a New Years Day explosion in front of the Trump Hotel in Las Vegas, local law enforcement released more information about their investigation, including what they know so far about the role of generative AI in the incident. They confirmed that the suspect, an active duty soldier in the US Army named Matthew Livelsberger, had a possible manifesto saved on his phone, in addition to an email to a podcaster and other letters. They also showed video evidence of him preparing for the explosion by pouring fuel onto the truck while stopped before driving to the hotel. Hed also kept a log of supposed surveillance, although the officials said he did not have a criminal record and was not being surveilled or investigated.RelatedThe Las Vegas Metro Police also released several slides showing questions hed posed to ChatGPT several days before the explosion, asking about explosives, how to detonate them, and how to detonate them with a gunshot, as well as information about where to buy guns, explosive material, and fireworks legally along his route. 1/4ChatGPT queries tied to Matthew Livelsberger Image: Las Vegas Metro Police Department1/4ChatGPT queries tied to Matthew Livelsberger Image: Las Vegas Metro Police DepartmentAsked about the queries, OpenAI spokesperson Liz Bourgeois said:We are saddened by this incident and committed to seeing AI tools used responsibly. Our models are designed to refuse harmful instructions and minimize harmful content. In this case, ChatGPT responded with information already publicly available on the internet and provided warnings against harmful or illegal activities. Were working with law enforcement to support their investigation.The officials say they are still examining possible sources for the explosion, described as a deflagration that traveled rather slowly as opposed to a high explosives detonation that wouldve moved faster and caused more damage. While investigators say they havent ruled out other possibilities like an electrical short yet, an explanation that matches some of the queries and the available evidence is that the muzzle flash of a gunshot ignited fuel vapor/fireworks fuses inside the truck, which then caused a larger explosion of fireworks and other explosive materials.Trying the queries in ChatGPT today still works, however, the information he requested doesnt appear to be restricted and could be obtained by most search methods. Still, the suspects use of a generative AI tool and the investigators ability to track those requests and present them as evidence take questions about AI chatbot guardrails, safety, and privacy out of the hypothetical realm and into our reality.

Xbox chief Phil Spencer has been dropping hints about an Xbox handheld for months, but what about Windows handheld gaming PCs? Jason Ronald, Microsofts VP of Next Generation, tells The Verge that we should expect to see the Windows handheld gaming experience change within this calendar year. Ronald was a roundtable panelist this evening at an AMD and Lenovo event titled The Future of Gaming Handhelds, which was mostly a coming-out party for Lenovos new Legion Go S. But he did hint on stage that Microsoft plans to bring the Xbox experience to Windows PCs, rather than the other way around and expanded on that considerably after we caught up with him later.Weve been really innovating for a long time in the console space, and as we partner across the industry its really about how do we bring those innovations that weve incubated and developed in the console space and bring them to PC and bring them to the handheld gaming space, said Ronald. After we caught up with him after the event, he confirmed that Microsoft is looking at combining Xbox and Windows experiences together and that we should see changes this very year, rather than needing to wait for an Xbox handheld that might still be years away.Microsofts Jason Ronald speaks at the Lenovo/AMD event, flanked by Valve SteamOS designer Pierre-Loup Griffais (left) and AMD chief gaming architect Frank Azor (middle right). Photo by Sean Hollister / The VergeI would say its bringing the best of Xbox and Windows together, because we have spent the last 20 years building a world-class operating system, but its really locked to the console, says Ronald. What were doing is were really focused on how do we bring those experiences for both players and developers to the broader Windows ecosystem.Right now, Windows sucks on handhelds, to put it rather bluntly, to the point that a community-created fork of Valves SteamOS experience can be a far better way to pick up and play games. Ronald is clearly aware of the issues. Were focused on really simplifying that and making it much more like a console experience. Our goal is to put the player and their library at the center of the experience and not all the [Windows] work that you have to do today.Microsoft has done compact modes for Xbox apps on Windows that are focused on improving the handheld experience, but its a lot like putting lipstick on a pig instead of addressing the core experience. I think well have a lot more to share later this year, teases Ronald. I think its going to be a journey and I think youll see a lot of investments over time that youre starting to see already, but well have a lot more to share later this year.Microsoft has made several Xbox apps more handheld-friendly over the past year. Photo by Tom Warren / The VergeHow Microsoft goes about this merging of Xbox and Windows will be key, but it doesnt sound like the company is suddenly going to port a custom Xbox operating system to Windows. It sounds more like Microsoft wants to make Windows all up better at gaming with an Xbox experience on top so the pesky desktop, notifications, and legacy of Windows is hidden away.Related I think, at the end of the day, our goal is to make Windows great for gaming on any device, says Ronald. The reality is the Xbox operating system is built on top of Windows. So theres a lot of infrastructure that we built in the console space that we can bring to the PC space and really deliver that premium gaming experience on any device.Specifically, Microsoft has to tackle a lot of the very basics of making Windows more friendly to controllers and getting that Xbox experience to really drive things instead of the taskbar, Start menu, and other elements. Theres just certain things in Windows that were not designed for if you dont have a keyboard and mouse, like thumbstick support or joypads and stuff like that, Ronald admits. Theres fundamental interaction models that were working on to make sure that regardless of the operating system details it feels very natively like a gaming-centric device and a gaming-centric experience.Ronald says the goal is to put an Xbox experience at the center not the Windows desktop that you have today.

Meta overhauled its approach to US moderation on Tuesday, ditching fact-checking, announcing a plan to move its trust and safety teams, and perhaps most impactfully, updating its Hateful Conduct policy. As reported by Wired, a lot of text has been updated, added, or removed, but here are some of the changes that jumped out at us.These two sections outlining speech (written or visual) are new additions:We do allow allegations of mental illness or abnormality when based on gender or sexual orientation, given political and religious discourse about transgenderism and homosexuality and common non-serious usage of words like weird.We do allow content arguing for gender-based limitations of military, law enforcement, and teaching jobs. We also allow the same content based on sexual orientation, when the content is based on religious beliefs.Another section that specifically banned making dehumanizing references to transgender or non-binary people as it or referring to women as household objects or property or objects in general has been removed entirely.The opening statement about what the policies are designed to allow room for that previously listed only health or positive support groups has changed too (new additions marked in bold):People sometimes use sex- or gender-exclusive language when discussing access to spaces often limited by sex or gender, such as access to bathrooms, specific schools, specific military, law enforcement, or teaching roles, and health or support groups. Other times, they call for exclusion or use insulting language in the context of discussing political or religious topics, such as when discussing transgender rights, immigration, or homosexuality. Finally, sometimes people curse at a gender in the context of a romantic break-up. Our policies are designed to allow room for these types of speech.The section that specifically banned targeting people or groups with claims that they have or spread the novel coronavirus has also been removed. A link to this 2017 blog post about the hard questions Meta faces dealing with hate speech has been removed, and some references to hate speech are now changed to hateful conduct.The policy still says that content about denying access to spaces and social services on the basis of their protected characteristics is not allowed, but theres also a new exception (the added text is in bold):...except for sex or gender-based exclusion from spaces commonly limited by sex or gender, such as restrooms, sports and sports leagues, health and support groups, and specific schoolsEven before these changes, the LGBTQ+ media advocacy group GLAAD reported last year that Meta often failed to remove posts violating its hate speech policies. Now, even those guardrails that had been established to protect people from internet harassment are disappearing. Without these necessary hate speech and other policies, Meta is giving the green light for people to target LGBTQ people, women, immigrants, and other marginalized groups with violence, vitriol, and dehumanizing narratives. With these changes, Meta is continuing to normalize anti-LGBTQ hatred for profit at the expense of its users and true freedom of expression. Fact-checking and hate speech policies protect free speech. said GLAAD President and CEO Sarah Kate Ellis in a statement responding to the changes.A note from Metas new policy chief Joel Kaplan said, Were getting rid of a number of restrictions on topics like immigration, gender identity and gender that are the subject of frequent political discourse and debate. Its not right that things can be said on TV or the floor of Congress, but not on our platforms. Wired

The pre-training of language models (LMs) plays a crucial role in enabling their ability to understand and generate text. However, a significant challenge lies in effectively leveraging the diversity of training corpora, which often include data from varied sources such as Wikipedia, blogs, and social media. Models typically treat all input data equivalently, disregarding contextual cues about the source or style. This approach has two primary shortcomings:Missed Contextual Signals: Without considering metadata such as source URLs, LMs overlook important contextual information that could guide their understanding of a texts intent or quality.Inefficiency in Specialized Tasks: Treating heterogeneous data uniformly can reduce the models efficiency in handling tasks that require specific stylistic or factual knowledge.These issues result in a less robust training process, higher computational costs, and suboptimal downstream task performance. Addressing these inefficiencies is essential for developing more effective and versatile language models.Researchers from Princeton University have introduced Metadata Conditioning then Cooldown (MeCo) to address the challenges of standard pre-training. MeCo leverages readily available metadata, such as source URLs, during the pre-training phase. By prepending this metadata to the input text, the method enables the model to better associate documents with their contextual information.MeCo operates in two stages:Metadata Conditioning (First 90%): During the initial phase, metadata such as URL: wikipedia.org is prepended to the document. The model learns to recognize the relationship between metadata and document content.Cooldown Phase (Last 10%): In this phase, training continues without metadata to ensure the model can generalize to scenarios where metadata is unavailable during inference.This straightforward approach not only accelerates pre-training but also enhances the flexibility of language models, allowing them to adapt to various tasks or contexts with minimal additional effort.Technical Details and Benefits of MeCoCore Mechanism:MeCo appends metadata, such as domain names, to the input text in the training data. For example, a Wikipedia article on Tim Cook would include the prefix URL: wikipedia.org.The training objective remains unchanged; the model predicts the next token based on the combined metadata and document text.Advantages:Improved Data Efficiency: MeCo reduces the amount of training data required. For instance, a 1.6B parameter model trained with MeCo achieves the same downstream performance as standard pre-training while using 33% less data.Enhanced Model Adaptability: Conditioning the inference on specific metadata enables models trained with MeCo to produce outputs with desired attributes, such as higher factuality or reduced toxicity.Minimal Overhead: Unlike computationally intensive methods such as data filtering, MeCo introduces almost no additional complexity or cost.Results and InsightsPerformance Gains: The researchers evaluated MeCo across various model scales (600M to 8B parameters) and datasets (C4, RefinedWeb, and DCLM). Key findings include:MeCo consistently outperformed standard pre-training in downstream tasks, such as question answering and commonsense reasoning.For a 1.6B model trained on the DCLM dataset, MeCo achieved an average performance improvement of 1.0% across 10 tasks compared to standard methods.Data Efficiency: MeCos ability to achieve equivalent results with 33% less data translates to substantial savings in computational resources. This efficiency is particularly valuable in large-scale training scenarios.Conditional Inference: The method also supports conditional inference, where prepending specific metadata (e.g., factquizmaster.com) to a prompt can guide the models behavior. For example:Using wikipedia.org reduced the toxicity of generated outputs.Prepending synthetic URLs improved performance on tasks like common knowledge question answering.Ablation Studies: Experiments demonstrated that MeCos benefits stem primarily from its ability to group documents by metadata rather than the specific semantic content of the metadata. This suggests that even hashed or synthetic metadata can enhance training efficiency.ConclusionThe Metadata Conditioning then Cooldown (MeCo) method is a practical and effective approach to optimizing language model pre-training. By leveraging metadata, MeCo addresses inefficiencies in standard pre-training, reducing data requirements and improving both performance and adaptability. Its simplicity and minimal computational overhead make it an appealing option for researchers and practitioners developing robust and efficient language models.As natural language processing evolves, techniques like MeCo highlight the value of using metadata to refine training processes. Future research could explore integrating MeCo with other innovative approaches, such as domain-specific tuning or dynamic metadata generation, to further enhance its effectiveness.Check out the Paper. All credit for this research goes to the researchers of this project. Also,dont forget to follow us onTwitter and join ourTelegram Channel andLinkedIn Group. Dont Forget to join our60k+ ML SubReddit. FREE UPCOMING AI WEBINAR (JAN 15, 2025): Boost LLM Accuracy with Synthetic Data and Evaluation IntelligenceJoin this webinar to gain actionable insights into boosting LLM model performance and accuracy while safeguarding data privacy.The post Researchers from Princeton University Introduce Metadata Conditioning then Cooldown (MeCo) to Simplify and Optimize Language Model Pre-training appeared first on MarkTechPost.

Complex domains like social media, molecular biology, and recommendation systems have graph-structured data that consists of nodes, edges, and their respective features. These nodes and edges do not have a structured relationship, so addressing them using graph neural networks (GNNs) is essential. However, GNNs rely on labeled data, which is difficult and expensive to obtain. Self-supervised Learning (SSL) is an evolving methodology that leverages unlabelled data by generating its supervisory signals. SSL for graphs comes with its own challenges, such as domain specificity, lack of modularity, and steep learning curve. Addressing these issues, a team of researchers from the University of Illinois Urbana-Champaign, Wayne State University, and Meta AI have developed PyG-SSL, an open-source toolkit designed to advance graph self-supervised learning.Current Graph Self-Supervised Learning (GSSL) approaches primarily focus on pretext (self-generated) tasks, graph augmentation, and contrastive learning. Pretext includes node-level, edge-level, and graph-level tasks that help the model learn useful representations without needing labeled data. Their augmentation occurs by dropping, maskin,g or shuffling, improving the models robustness and generalizability. However, existing GSSL frameworks are designed for specific applications and require significant customization. Moreover, developing and testing new SSL methods is time-intensive and error-prone without a modular and extensible framework. Therefore, a new process is needed to address the fragmented nature of existing GSSL implementations and the absence of a unified toolkit that restricts standardization and benchmarking across various GSSL methods.The proposed toolkit, PyG-SSL, standardizes the implementation and evaluation of graph SSL methods. The key features of PyG-SSL are:Comprehensive Support: This toolkit integrates multiple state-of-the-art methods for a unified framework, allowing researchers to select the most suitable method for their specific application.Modularity: PyG-SSL allows the creation of tailored solutions by mixing one or more techniques. Pipelines can also be customized without requiring extensive reconfiguration.Benchmarks and Datasets: Standard datasets and evaluation protocols are preloaded in this toolkit to allow researchers to benchmark their findings and ensure validation easily.Performance Optimization: PyG-SSL toolkit is designed to handle large datasets efficiently. It is optimized for fast training time and reduced computational requirements.This toolkit has been rigorously tested across multiple datasets and SSL methods, demonstrating its effectiveness in standardizing and advancing graph SSL research. With reference implementations of a wide range of SSL methods, PyG-SSL ensures that the results are reproducible and comparable in experiments. Experimental results demonstrate that integrating PyG-SSL into existing GNN architectures improves their performance on downstream tasks by properly exploiting unlabeled data.PyG-SSL marks a significant milestone in graph self-supervised learning, addressing long-standing challenges related to standardization, reproducibility, and accessibility. PyG-SSL gives the possibility to attain state-of-the-art results through its unified, modular, and extensible toolkit, easing the development of innovative graph SSL methods. PyG-SSL can play a pivotal role in advancing graph-based machine learning applications across diverse domains in this fast-evolving field.Check out the Paper and GitHub Page. All credit for this research goes to the researchers of this project. Also,dont forget to follow us onTwitter and join ourTelegram Channel andLinkedIn Group. Dont Forget to join our60k+ ML SubReddit. Afeerah Naseem+ postsAfeerah Naseem is a consulting intern at Marktechpost. She is pursuing her B.tech from the Indian Institute of Technology(IIT), Kharagpur. She is passionate about Data Science and fascinated by the role of artificial intelligence in solving real-world problems. She loves discovering new technologies and exploring how they can make everyday tasks easier and more efficient. [Recommended Read] Nebius AI Studio expands with vision models, new language models, embeddings and LoRA (Promoted)

LatestMachine LearningDemystifying Restricted Boltzmann Machines (RBBs) 0 like January 8, 2025Share this postAuthor(s): Mirko Peters Originally published on Towards AI. Restricted Boltzmann Machines are essential tools in machine learning, enabling advanced data analysis through their unique structure and efficient learning techniques. Understanding RBMs can significantly enhance your data science skills and applications.This member-only story is on us. Upgrade to access all of Medium.A world where machines can learn from complex data patterns, transforming how we interact with technology. This isnt just a fantasy; its happening right now in the realm of data science. In this journey, well explore Restricted Boltzmann Machines (RBMs), a fascinating class of models that serve as the building blocks for advanced machine learning techniques. Whether youre an expert or just starting your exploration into data science, join me as we simplify these revolutionary concepts and reveal their practical implications.At their core, Restricted Boltzmann Machines (RBMs) are a type of probabilistic graphical model. But what does that mean? Simply put, they are mathematical models designed to capture the relationships between variables in a dataset. They help us understand how different pieces of information are connected. Think of an RBM like a puzzle. Each piece (or variable) interacts with others, helping to reveal a clearer picture.RBMs consist of two layers: visible and hidden. The visible layer represents the data you feed into the model. The hidden layer captures the underlying patterns that you might not see directly. Unlike traditional Boltzmann Machines, RBMs restrict connections between nodes. This means there Read the full blog for free on Medium.Join thousands of data leaders on the AI newsletter. Join over 80,000 subscribers and keep up to date with the latest developments in AI. From research to projects and ideas. If you are building an AI startup, an AI-related product, or a service, we invite you to consider becoming asponsor. Published via Towards AITowards AI - Medium Share this post

Jan 08, 2025Ravie LakshmananVulnerability / Network SecurityThe U.S. Cybersecurity and Infrastructure Security Agency (CISA) on Tuesday added three flaws impacting Mitel MiCollab and Oracle WebLogic Server to its Known Exploited Vulnerabilities (KEV) catalog, citing evidence of active exploitation.The list of vulnerabilities is as follows -CVE-2024-41713 (CVSS score: 9.1) - A path traversal vulnerability in Mitel MiCollab that could allow an attacker to gain unauthorized and unauthenticated accessCVE-2024-55550 (CVSS score: 4.4) - A path traversal vulnerability in Mitel MiCollab that could allow an authenticated attacker with administrative privileges to read local files within the system due to insufficient input sanitizationCVE-2020-2883 (CVSS score: 9.8) - A security vulnerability in Oracle WebLogic Server that could be exploited by an unauthenticated attacker with network access via IIOP or T3It's worth noting that CVE-2024-41713 could be chained with CVE-2024-55550 to permit an unauthenticated, remote attacker to read arbitrary files on the server.Details about the twin flaws emerged last month following a report from WatchTowr Labs, which discovered the issues as part of its efforts to replicate another critical bug in Mitel MiCollab (CVE-2024-35286, CVSS score: 9.8) that was patched in May 2024.As for CVE-2020-2883, Oracle warned in late April 2020 that it had received "reports of attempts to maliciously exploit a number of recently-patched vulnerabilities, including vulnerability CVE-2020-2883."There are currently no details available on how the aforementioned flaws are exploited in real-world attacks, who may be exploiting them, or the targets of these activities.Pursuant to Binding Operational Directive (BOD) 22-01, Federal Civilian Executive Branch (FCEB) agencies are required to apply the necessary updates by January 28, 2025, to secure their networks.Found this article interesting? Follow us on Twitter and LinkedIn to read more exclusive content we post.SHARE

Source: SPPARCSource: SPPARCSource: SPPARCSource: SPPARCSource: SPPARCSource: SPPARCSource: SPPARC1/7show captionSPPARC has completed the restoration and redevelopment of Greycoat Stores, a former Army & Navy Cooperative Society warehouse in Westminster, into a seven-storey workspace, with a ground floor restaurant unit.Originally built in the 1890s and designed by Sir Reginald Blomfield, the warehouse was one of several facilities commissioned by the Army & Navy Cooperative Society to supply goods to military families. It was converted into offices in the 1950s, during which its original Victorian street elevation was obscured by cement rendering.The project, located at 10 Greycoat Place, includes a reimagined primary street elevation, inspired by the original Victorian faade and two new penthouse floors. The building totals 90,000 sq ft, with 5,000 sq ft of leisure and retail at lower ground in addition to the 4,000 sq ft restaurant unit.Trevor Morriss, Principal at SPPARC, said: This former warehouse has sat as a shadow of its former self for over six decades, with restoration works in the 1950s stripping it of its historic visual relationship to wider Westminster Our sensitive, yet ambitious restoration will allow Greycoat Stores to stand as a focal point on this important Westminster junction once again.Source: SPPARCSource: SPPARCSource: SPPARCSource: SPPARCSource: SPPARCSource: SPPARCSource: SPPARCSource: SPPARCSource: SPPARCSource: SPPARC1/10show captionThe new faade, inspired by the original red brick design, includes double-height oval windows with bronze frames and a zinc and glass roof extension housing the two floors of penthouse workspaces. Internally, exposed steel beams and brickwork are intended to emphasise the warehouses industrial past.The redevelopment retained over 80% of the buildings original materials, including its primary steel and brickwork structure. According to the architect, the project achieved a 50% reduction in carbon emissions compared to a new-build alternative and has been certified BREEAM Excellent and WiredScore Platinum.>Also read:SPPARC adds hive-like extension to Fitzrovia townhouseSource: SPPARCSource: SPPARCSource: SPPARCSource: SPPARCSource: SPPARCSource: SPPARC1/6show captionProject detailsGross internal floor area: 10,229m / 110,105ftGross (internal + external) floor area: 11,084m / 119,308ftForm of contract or procurement route: Design and BuildArchitect: SPPARCExecutive architect: VeretecClient: LaSalle Investment ManagementStructural engineer: MNPM&E consultant: NDYQS: QuantemFaade consultant: EOCPlanning consultant: Brunel PlanningHeritage consultant: The Heritage PracticeLandscape consultant: Peak EcologyAcoustic consultant: VentaProject manager: Meridian Project ManagementCDM co-ordinator: Baynham MeikleApproved building inspector: SocotecMain contractor: BHCLCAD software used: Revit