Support for Microsoft 365 ends along with Windows 10PixieMe/ShutterstockAlthough Microsoft announced some time ago that Windows 10 will only be supported (free of charge) until October 14, 2025, the switch to its successor Windows 11 is only taking place slowly. According to calculations by security provider Eset, 32 million PCs in Germany are still running Windows 10. The situation is similar in other countries. One reason for this is that although the switch to Windows 11 is free, there are stricter hardware requirements which , according to studies by Lansweeper, around 50 percent of computers in Germany do not meet.Functional, but not supported To urge more users to upgrade to Windows 11, Microsoft recently announced in a blog post that Microsoft-365 apps will no longer be supported on Windows 10 devices after October 14, 2025. To use Microsoft 365 apps on your device, you will need to upgrade to Windows 11, it continued.The blog post raised numerous questions, and has since been deleted by Microsoft. However, there was no correction or explanation.What the software giant was actually getting at is shown by a support page on the subject that was updated in December. Here, too, Microsoft points out that Microsoft 365 apps will no longer be supported under Windows 10 after the end of support in mid-October. At the same time, however, the company explains that the applications will continue to work as before. However, to avoid performance and reliability problems over time, an upgrade to Windows 11 is strongly recommended.The reasoning: Microsoft 365 is subject to the Modern Lifecycle Policy, which requires that customers keep the product or service up to date according to maintenance and system requirements and use Microsoft 365 on a Windows operating system for which support is currently provided. License versions not affected The situation is somewhat clearer for Office versions with a one-time license: Based on the Fixed Lifecycle Policy, Office Home & Student, Office Home & Business or Office Professional Plus will continue to be fully supported under Windows 10 as long as they do not reach the end of support themselves. Support for Office 2016 and 2019 will also end at the same time as Windows 10.

When you think of AIs contributions to science, you probably think of AlphaFold, the Google DeepMind protein-folding program that earned its creator a Nobel Prize last year.Now OpenAI says its getting into the science game toowith a model for engineering proteins.The company says it has developed a language model that dreams up proteins capable of turning regular cells into stem cellsand that it has handily beat humans at the task.The work represents OpenAIs first model focused on biological data and its first public claim that its models can deliver unexpected scientific results. As such, it is a step toward determining whether or not AI can make true discoveries, which some argue is a major test on the pathway to artificial general intelligence.Last week, OpenAI CEO Sam Altman said he was confident his company knows how to build an AGI, adding that superintelligent tools could massively accelerate scientific discovery and innovation well beyond what we are capable of doing on our own.The protein engineering project started a year ago when Retro Biosciences, a longevity research company based in San Francisco, approached OpenAI about working together.That link-up did not happen by chance. Sam Altman, the CEO of OpenAI, personally funded Retro with $180 million, as MIT Technology Review first reported in 2023.Retro has the goal of extending the normal human lifespan by 10 years. For that, it studies what are called Yamanaka factors. Those are a set of proteins that, when added to a human skin cell, will cause it to morph into a young-seeming stem cell, a type that can produce any other tissue in the body.Its a phenomenon that researchers at Retro, and at richly funded companies like Altos Labs, see as the possible starting point for rejuvenating animals, building human organs, or providing supplies of replacement cells.But such cell reprogramming is not very efficient. It takes several weeks, and less than 1% of cells treated in a lab dish will complete the rejuvenation journey.OpenAIs new model, called GPT-4b micro, was trained to suggest ways to re-engineer the protein factors to increase their function. According to OpenAI, researchers used the models suggestions to change two of the Yamanaka factors to to be more than 50 times as effectiveat least according to some preliminary measures.Just across the board, the proteins seem better than what the scientists were able to produce by themselves, says John Hallman, an OpenAI researcher.Hallman and OpenAIs Aaron Jaech, as well as Rico Meinl from Retro, were the models lead developers.Outside scientists wont be able to tell if the results are real until theyre published, something the companies say they are planning. Nor is the model available for wider useits still a bespoke demonstration, not an official product launch.This project is meant to show that were serious about contributing to science, says Jaech. But whether those capabilities will come out to the world as a separate model or whether theyll be rolled into our mainline reasoning modelsthats still to be determined.The model does not work the same way as Googles AlphaFold, which predicts what shape proteins will take. Since the Yamanaka factors are unusually floppy and unstructured proteins, OpenAI said, they called for a different approach, which its large language models were suited to.The model was trained on examples of protein sequences from many species, as well as information on which proteins tend to interact with one another. While thats a lot of data, its just a fraction of what OpenAIs flagship chatbots were trained on, making GPT-4b an example of a small language model that works with a focused data set.Once Retro scientists were given the model, they tried to steer it to suggest possible redesigns of the Yamanaka proteins. The prompting tactic used is similar to the few-shot method, in which a user queries a chatbot by providing a series of examples with answers, followed by an example for the bot to respond to.Although genetic engineers have ways to direct evolution of molecules in the lab, they can usually test only so many possibilities. And even a protein of typical length can be changed in nearly infinite ways (since theyre built from hundreds of amino acids, and each acid comes in 20 possible varieties).OpenAIs model, however, often spits out suggestions in which a third of the amino acids in the proteins were changed.OPENAIWe threw this model into the lab immediately and we got real-world results, says Retros CEO, Joe Betts-Lacroix. He says the models ideas were unusually good, leading to improvements over the original Yamanaka factors in a substantial fraction of cases.Vadim Gladyshev, a Harvard University aging researcher who consults with Retro, says better ways of making stem cells are needed. For us, it would be extremely useful. [Skin cells] are easy to reprogram, but other cells are not, he says. And to do it in a new speciesits often extremely different, and you dont get anything.How exactly the GPT-4b arrives at its guesses is still not clearas is often the case with AI models. Its like when AlphaGo crushed the best human at Go, but it took a long time to find out why, says Betts-Lacroix. We are still figuring out what it does, and we think the way we apply this is only scratching the surface.OpenAI says no money changed hands in the collaboration. But because the work could benefit Retrowhose biggest investor is Altmanthe announcement may add to questions swirling around the OpenAI CEOs side projects.Last year, the Wall Street Journal said Altmans wide-ranging investments in private tech startups amount to an opaque investment empire that is creating a mounting list of potential conflicts, since some of these companies also do business with OpenAI.In Retros case, simply being associated with Altman, OpenAI, and the race toward AGI could boost its profile and increase its ability to hire staff and raise funds. Betts-Lacroix did not answer questions about whether the early-stage company is currently in fundraising mode.OpenAI says Altman was not directly involved in the work and that it never makes decisions based on Altmans other investments.

Amazon's AirPods Max sale delivers a $100 discount on the latest USB-C model, reflecting the lowest price on record as the retailer engages in a weekend price war with Best Buy.Get Apple's latest AirPods Max with USB-C for $449.99.Save $100 on Apple's latest AirPods Max over-ear headphones with USB-C charging at Amazon this weekend, bringing the price down to $449.99. Inventory is limited so it's recommended to order early for the best shipping date. Continue Reading on AppleInsider | Discuss on our Forums

People will realize the direct contribution of building facades to public health and human prosperity and start to spread the word.Very soon, I believe, property developers may have to treat neuroscientific findings as key information to be weighed up alongside structural-load calculations, energy efficiency, lighting, and acoustics. And the person in the street will welcome this change.Heatherwick mentions the involvement in a new research study between the University of Waterloo and his two-year-old Humanize Campaign, now the second such academic partner after the UKs Loughborough University announced a new neuroscience-informed masters program last spring. Artificial intelligence can play a role in aiding the development of his theories too, he says. Human Architecture and Planning Institute (aka theHapi) is another recent example of the growth of these ideas.

Shortly after the release of a 1-minute-long first-look trailer for the upcoming movie Until Dawn movie, a longer trailer has now been released. The trailer, which you can check out below, clocks in at around 2 minutes and 25 seconds.As the first-look trailer indicated, the Until Dawn movie wont be a direct adaptation of the game of the same name. Rather, it will take place in the same universe and feature an entirely new cast of characters with its own original story. The movie also seems to be taking a unique twist on the original games anyone can die feature by having its characters come back to life in the course of its story.As for the new trailer itself, it looks like the movie will make use of similar tropes to other horror movies that revolve around a group of teenagers trapped in a hostile place. How the movie will subvert these tropes, however, will likely be a key feature of the movie.While it gives us a quick look at the movies setup and some of its characters, all of the protagonists are seemingly killed off within seconds. Time seems to reverse, however, with the characters coming back to life but still somehow remembering the events that led to their deaths.The story of the Until Dawn movie revolves around the mysterious disappearance of Melanie. This prompts her sister, Clover, to head into a remote valley where Melanie disappeared with her friends to search for answers. In their time at the valley, the group discovers that it is being methodically stalked by a masked killer, and ultimately, the group is murdered one-by-one. Curiously, however, the group awakens at a time before their deaths, while still retaining the memories. Upon learning that they have a limited number of deaths left, the group starts looking for a way to escape these circumstances.Directed by David F. Sandberg and written by Blair Butler and Gary Dauberman, the Until Dawn movie stars Ella Rubin, Michael Cimino, Odessa Azion, Ji-young Yoo, Belmont Cameli, Maia Mitchell and Peter Stormare.The Until Dawn movie is slated for a theatrical release on April 25. In the mean time, fans of the game can also look forward to a rumoured sequel to the original game that is seemingly in the works.

The dye, chemically known as erythrosine, has been used since 1907 to give candies, drinks and other foods their vibrant red color. Andia/Universal Images Group via Getty ImagesAfter decades of debate, the Food and Drug Administration (FDA) has announced a ban on Red Dye No. 3, a synthetic food coloring thats been linked to cancer in male lab rats.The decision comes after a petition filed in 2022 by advocacy groups, including the Center for Science in the Public Interest (CSPI), and follows mounting state-level legislation, most notably a ban in California. The federal ban will take effect in 2027, giving manufacturers time to reformulate their products.This is wonderful news and long overdue, Melanie Benesh, vice president for government affairs at the Environmental Working Group, tells Andrew Jacobs and Teddy Rosenbluth of the New York Times. Red Dye 3 is the lowest of the low-hanging fruit when it comes to toxic food dyes that the FDA should be addressing.The petroleum-based dye, chemically known as erythrosine, has been approved for use since 1907 to give candies, drinks and other foods their vibrant red color. While its cosmetic use was banned in 1990 under the Delaney Clausea federal rule prohibiting additives that have been found to cause cancer in humans or animalsit has remained in food and ingested drugs.Now, the FDA is using the same rule to eliminate the dye from beverages, food and drugs. However, the agency noted that the findings tying Red Dye No. 3 to cancer are specific to male rats and not directly translatable to humans.Relevant exposure levels for humans are typically much lower than those that cause the effects shown in male rats, the agency writes. Studies in other animals and in humans did not show these effects.But consumer safety advocates have long criticized the use of Red Dye No. 3 in food while it has been banned in cosmetics and topical drugs. The dye is prohibited or severely restricted in many countries, including the European Union, Japan and Australia, yet it has remained a staple in several U.S. products.Red 3, like every food dye, only serves as a marketing tool for the food industry, Thomas Galligan, principal scientist for food additives and supplements at CSPI, tells NPRs Allison Aubrey. Its just there to make food look visually appealing so that consumers want to spend their money on it.Red Dye No. 3 is found in more than 9,200 U.S. food products, including popular candies, cereals and sodas, according to data from the U.S. Department of Agriculture cited by CSPI. Companies like Ferrara, which produces Brachs candies, and Just Born, the maker of Peeps, have already begun removing the dye from their products, while other companies have turned to alternatives like Red Dye No. 40 or natural colorants, from beets to turmeric.The FDAs decision follows a wave of state legislation targeting synthetic dyes. California banned Red Dye No. 3 in October 2023, alongside three other additives linked to health risks, and it has barred artificial dyes including Red Dye No. 40 from public school meals.California assemblymember Jesse Gabriel, who introduced the states 2023 bill, credited state-level efforts with spurring federal action.To me, this is a clear indication that our strategy of putting pressure on Washington and putting pressure on the FDA to look at these issues more closely, to step up to the plate and take their regulatory responsibility seriously, is working, Gabriel tells Berkeley Lovelace Jr. and Elizabeth Chuck of NBC News.But concerns about synthetic dyes extend beyond cancer risks. In 2012, a study concluded that, while such dyes are not a primary cause of hyperactivity and attention-related behaviors, they could exacerbate these symptoms in children. Then, in 2021, a study by the California Office of Environmental Health Hazard Assessment found that legal exposure limits for artificial dyes might not adequately protect childrens behavioral health.Despite these concerns, the FDA maintains that most children are not affected.The totality of scientific evidence indicates that most children have no adverse effects when consuming foods containing color additives, but some evidence suggests that certain children may be sensitive to them, the agency wrote in a 2023 consumer update.For now, consumers worried about synthetic dyes can check product labels for ingredients like FD&C Red #3, red 3 or Red Dye No. 3. Highly processed foods, which are more likely to contain artificial colorings, may also warrant a second look for those aiming to reduce their exposure.Revoking the authorized use of Red No. 3 is an example of the FDA using its risk and science-based authority to review the safety of products in the marketplace, Sarah Gallo,senior vice president of product policy and federal affairs at the trade groupConsumer Brands Association, says in a statement to NBC News. Food and beverage companies will continue to follow the latest science and comply with all food safety regulations to ensure safe and available choices for consumers.Advocates hope the FDA will adopt a more proactive approach to regulating additives. Speaking to CNNs Kristen Rogers, Galligan suggests the case of Red Dye No. 3 reflects broader challenges in food safety oversight. The FDA has a long way to go to reform the broken system that allowed Red 3 to remain in foods decades after it was shown to cause cancer when eaten by animals, he tells the publication.But the decision, Galligan adds, ends the regulatory paradox of Red 3.Get the latest stories in your inbox every weekday.Filed Under: California, Cancer, Food, Government, Health, Law, Medicine, US Government

Anuj Tandon has joined Bitkraft Ventures, a seed-stage investor for interactive media and games, as partner focused on the fast-growing Indian game market.Read More

Sony confirms cancellation of live service projects from Bend Studio and Bluepoint GamesNeither studio will be shuttered, but it's unclear if decision will result in layoffs News by Sophie McEvoy Staff Writer Published on Jan. 17, 2025 Sony has confirmed the cancellation of two unannounced live service titles from Bend Studio and Bluepoint Games.As reported by Bloomberg, a spokesperson said the projects were shut down "following a recent review." It reiterated that PlayStation is dedicated to making online and single-player games."Bend and Bluepoint are highly accomplished teams who are valued members of the PlayStation Studios family, and we are working closely with each studio to determine what are the next projects," the spokesperson said.Neither studio will be shut down, but it's unclear how this decision will impact employees and whether layoffs will occur.In a memo seen by Bloomberg, Sony said it is "working closely with Bend and Bluepoint" and "plan[s] to do everything we can to ensure there is minimal business impact."Bend Studio has been a first-party developer for Sony since 2000 and is known for its work on the Days Gone franchise.Bluepoint Games was acquired by Sony in 2021 as a first-party studio focused on developing remasters and remakes including Uncharted, God of War, and Demon Souls.Last September, Sony pulled its live service hero shooter Concord from sale two weeks after its August release. Its servers were then turned off on September 6.A month later, Sony shuttered Concord's developer Firewalk Studios alongside Neon Koi after evaluating its "games portfolio and status of [its] projects."

Godot 4.4 Beta Released / News / January 17, 2025 / Godot, ReleaseFollowing 7 different Dev releases [ our coverage of Godot 4.4 Dev1 and Dev2, Dev3, Dev4 and Dev5 as well as Dev6 and the addition of Jolt Physics in Dev7 ] Godot 4.4 Beta 1 is here. This release adds a few more new features on top of the previously feature packed Dev releases and represents the last major additions to the 4.4 branch, everything else from this point to release candidate should only be small fixes and improvements.Highlights of the Godot 4.4 Beta release include:Typed Dictionary support in GDScriptMetal rendering support and Metal upscaling on Apple devicesNew embeddable Game window for runtime interaction with your gameJolt Physics engine integrationREPL evaluation in debuggerPhysics based snapping for 3D object placement in the worldInteractive camera previewFavoriting of properties in the inspectorThere are of course several hundred other fixes and improvements in the Godot 4.4 Beta release.Key LinksGodot 4.4 Beta Release AnnouncementGodot 4.4 Beta Interactive ChangelogGDQuest Demo GitHubYou can learn more about the Godot 4.4 Beta release and see several of my favorite new features in action in the video below.

LLMs are essential in industries such as education, healthcare, and customer service, where natural language understanding plays a crucial role. Though highly versatile, LLMs challenge is adapting to new tasks. Most fine-tuning methods are resource and time-consuming. Moreover, the fine-tuning approach often results in overfitting or sacrificing general adaptability for task-specific performance. This is a barrier for LLMs to address dynamic new and unforeseen tasks and creates a bottleneck in the overall application.One of the most prominent methods to address these challenges is Low-Rank Adaptation (LoRA), which updates small, task-specific matrices while freezing the rest of the models parameters. Although this reduces the computational cost of fine-tuning, it has limitations, such as increased sensitivity to overfitting and the inability to scale efficiently across tasks. Moreover, LoRAs design lacks inherent compositionality, limiting its ability to integrate multiple domain-specific skills.The researchers at Sakana AI and Institute of Science Tokyo introduced Transformer, a novel self-adaptive machine learning framework for large language models. Transformer employs a groundbreaking method called Singular Value Fine-tuning (SVF), which adapts LLMs in real time to new tasks without extensive retraining. By focusing on selectively modifying the singular components of the models weight matrices, Transformer enables dynamic task-specific adjustments. This innovation reduces the computational burden associated with fine-tuning, offering a scalable and efficient solution for self-adaptation.At the heart of Transformer is the SVF method, which fine-tunes the singular values of weight matrices. This approach drastically minimizes the number of trainable parameters compared to traditional methods. Instead of altering the entire model, SVF leverages reinforcement learning to create compact expert vectors specialized for specific tasks. For the inference process, Transformer works on a two-pass mechanism: the first is to analyze what the task might be and requires, and in the second, it dynamically integrates various relevant expert vectors to produce suitable behavior. Modularly, the approach ensures efficiency in addressing such a wide array of tasks through Transformer.Transformer performed outstanding performance in extensive benchmark evaluations. For instance, the framework shows improvements of over 39% compared to baselines in visual question-answering domains. In mathematics-related problem-solving, when testing was done on the GSM8K datasets, this model showed its strength by winning more than any fine-tuning method, reaching about a 4% improvement in its performance. On programming tasks under the MBPP-pro benchmark, Transformer displayed considerable accuracy improvements for domain-specific tasks and its general performance on various types of domains. As a result, Transformer adapted efficiently to unseen tasks like ARC-Challenge and Humaneval by either maintaining or exceeding the baseline performance metrics.An important overall outcome was the SVF methods efficiency. This improved training times and reduced the need for significant computational requirements as this method used fewer than 10% of the parameters required by LoRA. For example, for the GSM8K dataset, only 0.39 million parameters were needed for SVF training versus 6.82 million using LoRA to achieve higher performance. In addition, the model demonstrated good compositionality; vectors trained as an expert for one task could be reused and added together with others for a different, unrelated task, indicating the ability to scale up this Transformer framework.The researchers achieved this leap forward by addressing core limitations in existing methods, such as overfitting and inefficiency. By leveraging reinforcement learning, the SVF method provided principled regularization, preventing performance collapse on small datasets or narrow task domains. This allowed Transformer to excel despite limited training data while maintaining task adaptability.Conclusion: A research team from Sakana AI provided a scalable and efficient solution to task-specific adaptation in LLMs. Transformer, with its SVF method, is a highly significant advancement within the field that will pave the way for computationally efficient self-adaptive AI systems that are highly versatile. This approach will answer present challenges and lay a foundation for future developments of adaptive AI technologies.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 our65k+ ML SubReddit.(Promoted) Nikhil+ postsNikhil is an intern consultant at Marktechpost. He is pursuing an integrated dual degree in Materials at the Indian Institute of Technology, Kharagpur. Nikhil is an AI/ML enthusiast who is always researching applications in fields like biomaterials and biomedical science. With a strong background in Material Science, he is exploring new advancements and creating opportunities to contribute. Meet 'Height':The only autonomous project management tool (Sponsored)