At a TikTok Trends event at SXSW London, the company also revealed the content that's keeping us glued to our screens.

The FDA has launched the generative AI tool, Elsa, agency-wide to help its employees with everything from clinical reviews to investigations. Sure, we're living in a time of widespread disinformation and pushbacks against science, but why not rush th

In technical fields like information technology, definitions are fundamental. They are the building blocks for constructing useful applications and systems. Yet, despite this, it’s easy to assume a term’s definition and wield it confidently before discovering its true meaning. The two closely related cases that stand out to me are “security” and “privacy.” I say this with full awareness that, in my many writings on information security, I never adequately distinguished these two concepts. It was only after observing enough conflation of these terms that I resolved to examine my own casual treatment of them. So, with the aim of solidifying my own understanding, let’s properly differentiate “information security” and “information privacy.” Security vs. Privacy: Definitions That Matter In the context of information technology, what exactly are security and privacy? Security is the property of denying unauthorized parties from accessing or altering your data. Privacy is the property of preventing the observation of your activities by any third parties to whom you do not expressly consent to observe those activities. As you can see, these principles are related, which is one reason why they’re commonly interchanged. This distinction becomes comprehensible with examples. Let’s start with an instance where security applies, but privacy does not. Spotify uses digital rights management (DRM) software to keep its media secure but not private. DRM is a whole topic of its own, but it essentially uses cryptography to enforce copyright. In Spotify’s case, it’s what constitutes streaming rather than just downloading: the song’s file is present on your device (at least temporarily) just as if you’d downloaded it, but Spotify’s DRM cryptography prevents you from opening the file without the Spotify application. The data on Spotify (audio files) are secure because only users of the application can stream audio, and streamed content can’t be retained, opened, or transmitted to non-users. However, Spotify’s data is not private because nearly anyone with an email address can be a user. Thus, in practice, the company cannot control who exactly can access its data. A more complex example of security without privacy is social media. When you sign up for a social media platform, you accept an end-user license agreement (EULA) authorizing the platform to share your data with its partners and affiliates. Your data stored with “authorized parties” on servers controlled by the platform and its affiliates would be considered secure, provided all these entities successfully defend your data against theft by unauthorized parties. In other words, if everyone who is allowed (by agreement) to have your data encrypts it in transit and at rest, insulates and segments their networks, etc., then your data is secure no matter how many affiliates receive it. In practice, the more parties that have your data, the more likely it is that any one of them is breached, but in theory, they could all defend your data. On the other hand, any data you fork over to the social network is not private because you can’t control who uses your data and how. As soon as your data lands on the platform’s servers, you can’t restrict what they do with it, including sharing your data with other entities, which you also can’t control. Both examples illustrate security without privacy. That’s because privacy entails security, but not the reverse. All squares are rectangles, but not all rectangles are squares. If you have privacy, meaning you can completely enforce how any party uses your data (or doesn’t), it is secure by definition because only authorized parties can access your data. Mobile Devices: Secure but Not Private Casually mixing security and privacy can lead people to misunderstand the information security properties that apply to their data in any given scenario. By reevaluating for ourselves whether a given technology affords us security and privacy, we can have a more accurate understanding of how accessible our data really is. One significant misconception I’ve noticed concerns mobile devices. I get the impression that the digital privacy content sphere regards mobile devices as not secure because they aren’t private. But while mobile is designed not to be private, it is specifically designed to be secure. Why is that? Because the value of data is in keeping it in your hands and out of your competitor’s. If you collect data but anyone else can grab your copy, you are not only at no advantage but also at a disadvantage since you’re the only party that spent time and money to collect it from the source. With modest scrutiny, we’ll find that every element of a mobile OS that might be marketed as a privacy feature is, in fact, strictly a security feature. Cybersecurity professionals have hailed application permissions as a major stride in privacy. But whom are they designed to help? These menus apply to applications that request access to certain hardware, from microphones and cameras to flash memory storage and wireless radios. This access restriction feature serves the OS developer by letting users lock out as much of their competition as possible from taking their data. The mobile OS developer controls the OS with un-auditable compiled code. For all you know, permission controls on all the OS’s native apps could be ignored. However, even if we assume that the OS developer doesn’t thwart your restrictions on their own apps, the first-party apps still enjoy pride of place. There are more of them; they are preinstalled on your device, facilitate core mobile device features, require more permissions, and often lose core functions when those permissions are denied. Mobile OSes also sandbox every application, forcing each to run in an isolated software environment, oblivious to other applications and the underlying operating system. This, too, benefits the OS vendor. Like the app permission settings, this functionality makes it harder for third parties to grab the same data the OS effortlessly ingests. The OS relies on its own background processes to obtain the most valuable data and walls off every other app from those processes. Mobile Security Isn’t Designed With You in Mind The most powerful mobile security control is the denial of root privileges to all applications and users (besides, again, the OS itself). While it goes a long way toward keeping the user’s data safe, it is just as effective at subjecting everything and everyone using the device to the dictates of the OS. The security advantage is undeniable: if your user account can’t use root, then any malware that compromises it can’t either. By the same token, because you don’t have complete control over the OS, you are unable to reconfigure your device for privacy from the OS vendor. I’m not disparaging any of these security controls. All of them reinforce the protection of your data. I’m saying that they are not done primarily for the user’s benefit; that is secondary. Those of you familiar with my work might see the scroll bar near the bottom of this page and wonder why I haven’t mentioned Linux yet. The answer is that desktop operating systems, my preferred kind of Linux OS, benefit from their own examination. In a follow-up to this piece, I will discuss the paradox of desktop security and privacy. Please stay tuned.

Satellite view of coral reefs in New CaledoniaShutterstock/BEST-BACKGR​OUNDS There might be an upside to the loss of coral reefs. Their decline would mean oceans can absorb up to 5 per cent more carbon dioxide by 2100, researchers estimate, slowing the build up of this greenhouse gas in Earth’s atmosphere. “It is a beneficial effect if you only care about the concentration of CO2 in the atmosphere,” says Lester Kwiatkowski at Sorbonne University in Paris, France. But the decline of corals will also reduce biodiversity, harm fisheries and leave many coasts more exposed to rising seas, he says. Advertisement How much the world will warm depends mainly on the level of CO2 in the atmosphere. So far the land and oceans have been soaking up around half of the extra CO2 we have emitted. Any factors that increase or decrease these so-called land or ocean carbon sinks could therefore have a significant impact on future warming. It is often assumed that corals remove CO2 from seawater as they grow their calcium carbonate skeletons. In fact, the process, also known as calcification, is a net source of CO2. “You’re taking inorganic carbon in the ocean, generally in the form of carbonate and bicarbonate ions, turning it into calcium carbonate and that process releases CO2 into the seawater, some of which will be lost to the atmosphere,” says Kwiatkowski. Unmissable news about our planet delivered straight to your inbox every month. Sign up to newsletter This means that if reef formation around the world slows or even reverses, less CO2 will be released by reefs and the oceans will be able to absorb more of this greenhouse gas from the atmosphere – a factor not currently included in climate models. Observations suggest coral reef calcification is already declining as rising seawater temperatures cause mass coral bleaching and die-offs. The higher level of CO2 is also making oceans more acidic, which can make it harder to build carbonate skeletons and even lead to their dissolution. Kwiatkowski and his team took published estimates of how corals will be affected by warming and ocean acidification and used a computer model to work out how this might change the ocean sink in various emission scenarios. They conclude that the oceans could take up between 1 and 5 per cent more carbon by 2100, and up to 13 per cent more by 2300. This doesn’t take account of other factors that can cause reef decline such as overfishing and the spread of coral diseases, says Kwiatkowski, so might even be an underestimate. On the other hand, the work assumes that corals aren’t able to adapt or acclimatise, says Chris Jury at the University of Hawai’i at Manoa, who wasn’t involved in the study. “If the worst-case or even medium-case scenario in this study comes to pass, it means the near-total destruction of coral reefs globally,” says Jury. “I think that with consideration of realistic levels of adaptation and acclimatisation by corals and other reef organisms, the authors might come to different conclusions under a low to moderate level of climate change.” If Kwiatkowski’s team is correct, it means that the amount of emitted CO2 that will lead to a given level of warming – the so-called carbon budget – is a little larger than currently thought. “I think we would like our budgets to be as accurate as possible, even if we’re blowing through them,” says Kwiatkowski. Journal reference:PNAS DOI: 10.1073/pnas.2501562122 Topics:

Key Takeaways Dell and Nvidia will together provide architecture for the next set of supercomputers, named Doudna, for the US Department of Energy. Dell will focus more on sustainable hardware and cooling systems, while Nvidia will provide its AI architecture, including the Vera Rubin AI chips. The US Department of Energy wants to focus more on the sustainable development of AI, hence the choice of environment-conscious companies. The US Department of Energy said that Dell’s next batch of supercomputers will be delivered with Nvidia’s ‘Vera Rubin’ AI chips, marking the beginning of a new era of AI dominance in research. The said systems are expected to be 10x faster than the current batch of supercomputers, which HP provided. The supercomputer will be named ‘Doudna,’ after Jennifer Doudna, a Nobel Prize winner who made key contributions in CRISPR gene-editing. Supercomputers have been instrumental in key scientific discoveries in the last few decades and also played a big role in the design and maintenance of the U.S. nuclear weapons arsenal. And now, with the introduction of artificial intelligence, we’re heading towards a new decade of faster and more efficient scientific research. It (supercomputers) is the foundation of scientific discovery for our country. It is also a foundation for economic and technological leadership. And with that, national security – Nvidia CEO, Jensen Huang Dell Going All in on AI This isn’t the first time Dell and Nvidia have come together to develop newer AI solutions. Back in March 2024, Dell announced the Dell AI Factory with NVIDIA, an end-to-end enterprise AI solution designed for businesses.  This joint venture used Dell’s infrastructure, such as servers, storage, and networking, combined with NVIDIA’s AI architecture and technologies such as GPUs, DPUs, and AI software. Image Credit – Dell For instance, the Dell PowerEdge server uses NVIDIA’s full AI stack to provide enterprises with solutions required for a wide range of AI applications, including speech recognition, cybersecurity, recommendation systems, and language-based services. The demand for Dell’s AI servers has also increased, reaching $12.1B in the first quarter of 2025, with a total backlog of $14.4B, which suggests a strong future demand and order book. The company has set a bold profit forecast between $28.5B and $29.5B, as against the analysts’ prediction of $ 25.05 B. With Doudna, Dell is well-positioned to lead the next generation of supercomputers in AI research, invention, and discoveries. Focus on Energy Efficiency Seagate has warned about the unprecedented increase in demand for AI data storage in the coming few years, which is a significant challenge to the sustainability of AI data centers. Global data volume is expected to increase threefold by 2028.  Image Credit – DIGITIMES Asian The data storage industry currently produces only 1-2 zettabytes (1 zettabyte equals 1 trillion gigabytes) of storage annually, which is much lower than what would be required in the next 4-5 years. At the same time, Goldman Sachs predicts that power requirements will also go up by 165% by 2030 due to increasing demand for AI data centers. This calls for a more sustainable approach for the supercomputing industry as well.  Dell will use its proprietary technologies, such as Direct Liquid Cooling, the PowerCool eRDHx, and Smart Flow design in the Doudna, ensuring energy efficiency. Direct Liquid Cooling (DLC) increased computing density by supporting more cores per rack, which reduces cooling costs by as much as 45%. Dell’s PowerCool eRDHx is a self-contained airflow design that can capture 100% of the heat generated by IT systems. This reduces the dependency on expensive chillers, as eRDHx can work in usual temperatures of 32 and 36 degrees Celsius, leading to 60% savings in cooling energy costs. Lastly, the Dell Smart Flow design improves airflow within IT components and reduces the fan power by 52%. This leads to better performance with fewer cooling requirements. Besides this, Dell plans to incorporate Leak Sense Technology. If a coolant leak occurs, the system’s leak sensor will log an alert in the iDRAC system so that swift action can be taken. As per a report titled ‘Energy and AI’ by the IEA, the data center electricity demand will increase to 945 terawatt-hours (TWh) by 2030. For comparison, this is more than the total electricity consumption of Japan today. The US alone will consume more electricity in 2030 for processing data than for manufacturing all energy-intensive goods combined, including aluminum, steel, cement, and chemicals. Therefore, the need to develop sustainable AI data centers and supercomputers cannot be highlighted enough. Dell’s technology-focused, sustainable approach can be a pivotal point in how efficiently we use AI in the next decade. The US Department of Energy’s choice of Dell also seems to be a conscious move to shift towards companies that give importance to sustainability and can vouch for the long-term viability of research-intensive AI setups. Krishi is a seasoned tech journalist with over four years of experience writing about PC hardware, consumer technology, and artificial intelligence.  Clarity and accessibility are at the core of Krishi’s writing style. He believes technology writing should empower readers—not confuse them—and he’s committed to ensuring his content is always easy to understand without sacrificing accuracy or depth. Over the years, Krishi has contributed to some of the most reputable names in the industry, including Techopedia, TechRadar, and Tom’s Guide. A man of many talents, Krishi has also proven his mettle as a crypto writer, tackling complex topics with both ease and zeal. His work spans various formats—from in-depth explainers and news coverage to feature pieces and buying guides.  Behind the scenes, Krishi operates from a dual-monitor setup (including a 29-inch LG UltraWide) that’s always buzzing with news feeds, technical documentation, and research notes, as well as the occasional gaming sessions that keep him fresh.  Krishi thrives on staying current, always ready to dive into the latest announcements, industry shifts, and their far-reaching impacts.  When he's not deep into research on the latest PC hardware news, Krishi would love to chat with you about day trading and the financial markets—oh! And cricket, as well. View all articles by Krishi Chowdhary Our editorial process The Tech Report editorial policy is centered on providing helpful, accurate content that offers real value to our readers. We only work with experienced writers who have specific knowledge in the topics they cover, including latest developments in technology, online privacy, cryptocurrencies, software, and more. Our editorial policy ensures that each topic is researched and curated by our in-house editors. We maintain rigorous journalistic standards, and every article is 100% written by real authors.

Interviews Love, Death + Robots – Volume 4: Tim Miller (Creator & Director) & Jennifer Yuh Nelson (Supervising Director) By Vincent Frei - 02/06/2025 Earlier this year, Tim Miller spoke to us about his animated anthology Secret Level. Now, he returns to discuss the latest season of Love, Death + Robots. Jennifer Yuh Nelson talked about season two of Love, Death + Robots in 2021. She later worked on The Sea Beast, before returning once again to the anthology universe. What was your overall vision for the fourth season of Love, Death and Robots and how did it evolve from previous seasons? Tim Miller (TM) // We have the same strategy as every volume – we try to pick the best stories we know of and provide a mix that is hopefully appealing to everyone. There are a lot of variables to consider. Including genre, tone, and style of animation, such as stop motion, CG, and 2D. We try not to have two stories that are too similar. For example, if there’s already a military sci-fi story, we avoid selecting another one. We like to mix humor, horror, sci-fi, fantasy, and anything else that we think might be interesting from either a story or animation perspective. How did you approach the balance between experimenting with new styles and maintaining the signature identity of the show? TM // Honestly, we just try and follow our gut. What we think is interesting as filmmakers, animators, and storytellers will also be interesting to the animation community and fans alike. So, we keep an eye out for new voices, filmmakers, and new ways of doing things to keep things interesting. I’m not sure we have an identity of the show. In fact, I think if we did have an identity, it would be that we don’t have an identity… but we try and do whatever we think is interesting. Jennifer Yuh Nelson (JYN) // The fortunate thing about LDR is that the signature itself is experimenting with new styles. The trick is finding new aggressively experimental styles that still communicate to a mass audience. The stories are key to that. If the story is engaging, even to an audience that doesn’t usually gravitate to animation, then you can make it looks as weird as you want. What are some of the key challenges you faced while overseeing this season and how do you tackle them? TM // This season, there was a lot going on in the animation community that created some challenges with getting work done, whether studios were too full or ceased to exist entirely. Everyone struggled with budgets. But I didn’t feel like it was a problem with our show but rather a problem with the entire industry. People were struggling. And then it’s just always difficult when your ambition is high, your budgets are reasonable but still challenging, and you have to wrangle hundreds of people to get on board with your vision. (JYN) // These shows take a long time to make. R&D for a look that doesn’t exist can take a lot of trial and error. For example, Emily Dean, who directed Very Pulse of the Machine last season, did For He Can Creep this season. She had a cool angle of making her episode look like lithography. That was very very hard, but somehow Polygon, the studio that made both shorts, came through with it. And I think it turned out very well. Can you talk about how you selected the different animation studios for this season? What made you decide to work with the studios involved? (JYN) // We’ve been very fortunate to have worked with amazing people and studios these last few seasons, so it made sense to float some stories by them again. But it really comes down to the stories, and how each leans towards a certain technique. For example, How Zeke Found Religion was holding a slot where we wanted something 2D. We went to Titmouse because they were great with pushing the boundaries of 2D animation, and they suggested Diego Porral as a director who could bring a modern edge. How do you ensure each studio’s unique visual style complements the story and tone of each episode? TM // I know this sounds a little mystical and I don’t mean it to be, but I think the story speaks to you about style. Some things just feels right, and you have an innate concept of what would be the best version of the story, whether it’s stop motion, CG, 2D animation, or even live action. When you start thinking about the story in a creative way, a style becomes apparent. Which is not to say there aren’t many ways to do things and tell stories, but we feel a best version becomes clear. (JYN) // We do a lot of research, not just into what the studios have done before, but also into what they wish to do but haven’t had the chance to do. Often it’s just a matter of getting to know them and seeing if they have a philosophy of pushing for experimentation and risk. Then we try to support them as much as possible in their creative R&D. You both directed episodes for this season, what was that experience like? How did it differ from your work as overseeing directors? TM // For me, it’s really just trying to create the best story and I love working with the artists and trying to be open to what everybody brings to the table because everybody wants to do the best possible episode they can. I try and be open to letting people help carry that load. The best thing about being a director is that you get to pick and choose between all the great ideas that everybody has and shape the narrative by getting the benefit of everyone’s expertise and talent. (JYN) // It’s a different mindset. As a Supervising Director, I help. As a Director, I do. On episodes I’m not directing, I am deciphering that director’s ambition and pushing for whatever is required to make that absolutely great. On an episode I’m directing, every choice and image has to go through my brain so it’s more a reflection of my personal taste. Plus I tend to storyboard a lot more on my own episodes since it’s a way for me to communicate to the crew. I storyboard a lot on other episodes, but mainly to help figure out problems here and there. It also doesn’t come out of those director’s budgets so the free storyboarding is often welcomed. How did you choose the episode you worked on yourself and what aspects of it made it resonate with you both? TM // In my case with “The Screaming of the Tyrannosaur,” it was really by default. I had written the episode for Zack Snyder but Zack was too busy, and by that time I’d already fallen in love with the story, so I figured, why not just do it myself? As for “Golgotha,” I always loved the story. It was very efficient and short, which is hard to find in a story – it felt like a full meal. It has a beginning, middle, and end and it resolves in a satisfying way. “Golgotha” had all of that, plus it was funny. (JYN) // Spider Rose was on the story wall since the beginning. It was one of the “special” ones- very hard, ambitious, uncomfortable. Over the seasons we offered it to different directors and they veered away from it for one reason or another. But it glowed with a complexity that’s rare in a short story. I think that’s because it was written as an exploration for a far larger world that Bruce Sterling was developing. For me, it was the raw emotionality that drew me in. It’s how I understand how to communicate any story. And I love the way Spider Rose draws you in with emotion then shivs you with it. Were there any episodes in this season that particularly pushed the boundaries of what you had done before? How did that push happen? TM // I think “How Zeke got Religion” pushed the boundaries of 2D animation. The amount of detail and action that the guys at Titmouse were able to pull off was truly astonishing. Once again, Robert Valley outdid himself with 400 boys. The action scene at the end was one of my favorite pieces of animation in all of Love, Death, and Robots. (JYN) // Golgotha, Tim’s episode is live action, which is a rarity for the show. There was one live action episode in season 1, but none since. It is primarily an animation series, but nowadays, the line is so fuzzy that it seemed to make sense. How do you balance creative freedom with the thematic unity required for a show like Love, Death, and Robots? TM // There isn’t really a thematic unity. We’re just trying to create the best version of each of the episodes. They don’t tie into each other, they don’t relate to each other, they aren’t supposed to be about either Love, Death or Robots – the title is a meant to be a “catchall” that could hold ANY story or visual art we thought might be cool. Hopefully, the overall assemblage feels like a balanced meal with a little bit of something for everybody. But thematically speaking, again, I think our theme is that there is no theme. (JYN) // We try to set the foundation with a good story, based off the many short stories Tim has read over the years. Then the HOW of what that story becomes is the wooly Wild West. The directors and studios are fully encouraged to push all the boundaries of how to make these as innovative, impractical, and beautiful as they want. And, since each short is under 15 minutes, the studios we choose can be as experimental and scrappy as each story demands. Looking at the overall direction of the season, was there any unexpected moments or surprises that stood out to you during production? TM // Yeah, I think the color palette for “Zeke” was a shock to me in a wonderful way because it was completely unexpected and nothing I would ever do as a director but boy did I love it. And I think that “Can’t Stop” was an interesting addition. We wanted to do a music video from volume 1 onward, and this was the moment that we took to do it. I think it’s the greatest concert video ever made. (JYN) // Why do we have so many cats and babies? I’ve no idea. But when we saw the first giant baby shots in 400 Boys, it was a rare joy. They walk like babies, real babies. And somehow that was both accurate and terrifying. Looking forward, where do you see the show heading in future seasons, are there any new themes or concepts you’d love to explore? TM // So many directors in the industry have asked if they could play in our sandbox, and I would like to expand our reach to get some established names. Not that we don’t want new talent – we will always want that – but it would be great to have some really fantastic directors who have accomplished big movies come and play with our stories. I also think there’s a version where we bring in some content that may have existed in other mediums like comic books and perhaps tell some larger stories that take more than one episode to tell. Truthfully, I’ve already got a some really interesting stories picked out for the next few seasons, — of course those will change as the show evolves, but they’re fascinating stories that explore the whole arc of history… past, present, and future and some of the big challenges that humanity is facing today. I’d be lying if I didn’t mention that many of them explore the future of what mankind will become with the advent of AI and how artificial intelligence and humanity’s future intersect. (JYN) // Often themes only show up afterwards. There is a bit of a “herding cats” energy to the show that promises surprises in the production process. But the point of a show like this is that it is surprising. It has its own energy, and sometimes we just have to listen to it rather than dictate. If you had the opportunity to create any kind of story for Love, Death, and Robots, what would your dream narrative and what type of animation style would you envision for it? TM // Well, I have to say that I love high-end 3D animation, and that’s what Blur does for a reason. And secondly, I’d like to do a kind of story that could be live action and has some vast scope to it, but we choose to do it in animation because we get more value from using the techniques that animation brings. We can tell a bigger story, with more scope, and more action than we would using any other methodology…. and it competes favorably with live action in terms of the kind of audience that comes to watch it. Not just fans of animation, but fans of good cinema. (JYN) // I’d love to see an anime episode, like a Tsutomu Nihei fight scene, or something by Katsuhiro Otomo. A big thanks for your time. WANT TO KNOW MORE?Blur Studio: Dedicated page about Love, Death + Robots: Volume 4 on Blur Studio website. © Vincent Frei – The Art of VFX – 2025

Modern software engineering faces growing challenges in accurately retrieving and understanding code across diverse programming languages and large-scale codebases. Existing embedding models often struggle to capture the deep semantics of code, resulting in poor performance in tasks such as code search, RAG, and semantic analysis. These limitations hinder developers’ ability to efficiently locate relevant code snippets, reuse components, and manage large projects effectively. As software systems grow increasingly complex, there is a pressing need for more effective, language-agnostic representations of code that can power reliable and high-quality retrieval and reasoning across a wide range of development tasks.  Mistral AI has introduced Codestral Embed, a specialized embedding model built specifically for code-related tasks. Designed to handle real-world code more effectively than existing solutions, it enables powerful retrieval capabilities across large codebases. What sets it apart is its flexibility—users can adjust embedding dimensions and precision levels to balance performance with storage efficiency. Even at lower dimensions, such as 256 with int8 precision, Codestral Embed reportedly surpasses top models from competitors like OpenAI, Cohere, and Voyage, offering high retrieval quality at a reduced storage cost. Beyond basic retrieval, Codestral Embed supports a wide range of developer-focused applications. These include code completion, explanation, editing, semantic search, and duplicate detection. The model can also help organize and analyze repositories by clustering code based on functionality or structure, eliminating the need for manual supervision. This makes it particularly useful for tasks like understanding architectural patterns, categorizing code, or supporting automated documentation, ultimately helping developers work more efficiently with large and complex codebases.  Codestral Embed is tailored for understanding and retrieving code efficiently, especially in large-scale development environments. It powers retrieval-augmented generation by quickly fetching relevant context for tasks like code completion, editing, and explanation—ideal for use in coding assistants and agent-based tools. Developers can also perform semantic code searches using natural language or code queries to find relevant snippets. Its ability to detect similar or duplicated code helps with reuse, policy enforcement, and cleaning up redundancy. Additionally, it can cluster code by functionality or structure, making it useful for repository analysis, spotting architectural patterns, and enhancing documentation workflows.  Codestral Embed is a specialized embedding model designed to enhance code retrieval and semantic analysis tasks. It surpasses existing models, such as OpenAI’s and Cohere’s, in benchmarks like SWE-Bench Lite and CodeSearchNet. The model offers customizable embedding dimensions and precision levels, allowing users to effectively balance performance and storage needs. Key applications include retrieval-augmented generation, semantic code search, duplicate detection, and code clustering. Available via API at $0.15 per million tokens, with a 50% discount for batch processing, Codestral Embed supports various output formats and dimensions, catering to diverse development workflows. In conclusion, Codestral Embed offers customizable embedding dimensions and precisions, enabling developers to strike a balance between performance and storage efficiency. Benchmark evaluations indicate that Codestral Embed surpasses existing models like OpenAI’s and Cohere’s in various code-related tasks, including retrieval-augmented generation and semantic code search. Its applications span from identifying duplicate code segments to facilitating semantic clustering for code analytics. Available through Mistral’s API, Codestral Embed provides a flexible and efficient solution for developers seeking advanced code understanding capabilities.  vides valuable insights for the community. Check out the Technical details. All credit for this research goes to the researchers of this project. Also, feel free to follow us on Twitter and don’t forget to join our 95k+ ML SubReddit and Subscribe to our Newsletter. Sana HassanSana Hassan, a consulting intern at Marktechpost and dual-degree student at IIT Madras, is passionate about applying technology and AI to address real-world challenges. With a keen interest in solving practical problems, he brings a fresh perspective to the intersection of AI and real-life solutions.Sana Hassanhttps://www.marktechpost.com/author/sana-hassan/Off-Policy Reinforcement Learning RL with KL Divergence Yields Superior Reasoning in Large Language ModelsSana Hassanhttps://www.marktechpost.com/author/sana-hassan/This AI Paper from Microsoft Introduces WINA: A Training-Free Sparse Activation Framework for Efficient Large Language Model InferenceSana Hassanhttps://www.marktechpost.com/author/sana-hassan/Apple and Duke Researchers Present a Reinforcement Learning Approach That Enables LLMs to Provide Intermediate Answers, Enhancing Speed and AccuracySana Hassanhttps://www.marktechpost.com/author/sana-hassan/National University of Singapore Researchers Introduce Dimple: A Discrete Diffusion Multimodal Language Model for Efficient and Controllable Text Generation

Every drop counts: How space tech innovations are revolutionising Earth observation  Innovation News Network

The Krusty Burger is a health inspector's nightmare, responsible for spawning the Krusty Burger, the Ribwich, The Clogger, and the Steamed Ham (per Principal Skinner). It comes with seven minifigures, including Krusty the Clown, Sideshow Bob, Homer Simpson, and Officer Lou. You can purchase it for $209.99 on June 4, although LEGO Insiders can order the set via Early Access beginning on June 1. You can sign up for free here.LEGO The Simpsons: Krusty BurgerOut June 1 for LEGO Insiders, and June 4 for everyone else.$209.99 at LEGO StoreTo learn more about the set and its creation, we interviewed its designer. Ann Healy is a 39-year-old Senior Model Designer who's worked with the LEGO Group for the past six years. Here, lightly edited for clarity, are her thoughts on creating LEGO The Simpsons: Krusty Burger, the first new LEGO Simpsons set in nearly a decade. What other LEGO sets have you had a hand in designing over the past six years? Healy: I worked for three years on the LEGO Friends line. A few years ago, I got the opportunity to work on a Disney favorite of mine: The Sanderson Sisters’ Cottage set from Hocus Pocus. That project took years in the making to come to market, and I’m quite proud of it. More recently, I have been working on the LEGO Icons team. From last year’s products, I designed the McLaren MP4/4 & Ayrton Senna set and the Poinsettia LEGO Botanicals set.What drove the decision to revive the Simpsons brand after such a long hiatus? Healy: We saw that even after a 10-year stretch since a new Simpsons set, there was still a lot of love for The Simpsons amongst LEGO fans. Our market research showed that there is huge brand loyalty for The Simpsons, appealing to people globally. Designing the Krusty Burger in LEGO brick form serves as a natural continuation of our Simpsons line and gives us the opportunity to make new, never-before-seen LEGO minifigures.LEGO designers often build in their free time, working towards builds that they hope will one day become sets.“Coincidentally, the original Simpsons House and Kwik-E-Mart LEGO sets were some of the last sets I purchased before I knew I would be coming to work at The LEGO Group. In my first week working here, I found an early prototype of the Krusty Burger set on a shelf in our office. For context, LEGO designers often build in their free time, working towards builds that they hope will one day become sets. It gave me hope that someday, I might get the chance to work on something like that. Five years later, to my surprise, our Head of Marketing pulled that same prototype off the shelf and said it was time to revisit The Simpsons! I volunteered right away as a lifelong fan of the Simpsons. I couldn’t believe my luck!Describe the process of designing and building this set as best you can. How many times did you have to build the entire model in the course of finalizing it?Healy: I started with refining and updating the old prototype, figuring out the general layout and developing a building-instruction flow. I worked digitally first, then built the model physically. From there, I would go back and forth between digital and physical, seeing if the model would work with real bricks. We have design team check-ins where I showed progress on the model and got suggestions from our Creative Lead and the other model designers. I had several check-ins with the IP partner, where I would show them the latest updates and they could give feedback on the model and mini figures. I also had internal review meetings with our Model Governance and Building Experience teams, where we build the model, review every construction step, evaluate the play experience and test the strength of the model.I did not keep track of every time I rebuilt the model, but I would estimate at least 20 times. The last time I built the Krusty Burger set was when the first production boxes arrived from the factory. I built it one last time for quality assurance!Krusty Burger isn't a fully realized location in the show in the same way that Moe's Tavern is, or the Kwik-E Mart is, or the Simpsons’ house is. Which TV episodes did you use to map out the Krusty Burger and determine how it's set up? Healy: Disney was a great partner in sending us reference images and layouts when available. Also, I watched as many episodes of The Simpsons as possible, so I could translate the TV animation into a real, physical, toy playset. Season 20, Episode 21, “Coming to Homerica,” is the episode we used the most in reference, because it features the Mother Nature Burger that we reference several times in this LEGO set. The “Ribwich” from Season 14, Episode 12, “I'm Spelling as Fast as I Can,” is also featured. Other referenced episodes of note include: Season 6, Episode 15, “Homie the Clown” – Homer performs at a Krusty Burger Season 12, Episode 13, “Day of the Jackanapes” – Bart and Sideshow Bob go to the Krusty Burger. Season 10, Episode 1, Lard of the Dance” – Homer tries to sell grease from the restaurant.Season 7, Episode 15, “Bart the Fink” – Features the IRS Burger takeover.Season 19, Episode 1, “He Loves to Fly and He D'ohs” – Homer hides in the Krusty Zone ball pit. Season 15, Episode 10, “Diatribe of a Mad Housewife” – Shows the Krusty Burger drive thru sign. Season 21, Episode 4, "Treehouse of Horror XX" – Features the Krusty Burger in “Don't Have a Cow, Mankind.”What's your favorite deep cut Easter Egg in the build? Healy: Above the Krusty Burger drive-thru window, there is a kitchen display screen, showing that someone has placed an order for 700 burgers. This is a reference to the Season 5 episode “Boy-Scoutz 'n the Hood,” in which Krusty has unwisely opened a Krusty Burger oil rig in the middle of the sea.What was the most challenging aspect of designing this build? How were you able to overcome it? Healy: The most rewarding challenge of designing this set was building the Krusty Burger Sign with LEGO bricks. The IP partner paid a lot of attention to the proportions of Krusty’s face within the sign, giving us reference material and tips on how to best capture his likeness. We even got hand-drawn doodles of the Krusty the Clown from them! One of my model-designer colleagues used to work as an illustrator and used his skills to capture Krusty’s features in brick-form even more accurately. In addition, it was challenging to get the heavy Krusty Burger sign to stand up and remain stable on a single axle. Our Element Lead on the design project suggested we use a new element she was developing, an axle sleeve, to increase the weight the rod could hold upright. This new element did the trick.The most rewarding challenge of designing this set was building the Krusty Burger Sign with LEGO bricks.“Is this a one-off revival of the Simpsons brand? Or is this the beginning of a longer partnership, where we can expect to see more LEGO Simpsons themed sets in the near future?Healy: I cannot speculate on future products, per the LEGO Group’s and our partners' policies. Nonetheless, as a huge Simpsons fan myself, I am hopeful that fans have a positive response to this set!LEGO The Simpsons: Krusty Burger, Set #10352, retails for $209.99, and it is composed of 1635 pieces. You can purchase it on June 4 for the general public, or starting on June 1 for LEGO Insiders. You can sign up for LEGO Insiders for free here. And stay tuned! We will be building, photographing, and reviewing LEGO Krusty Burger later this month,Kevin Wong is a contributing freelancer for IGN, specializing in LEGO. He's also been published in Complex, Engadget, Gamespot, Kotaku, and more. Follow him on Twitter at @kevinjameswong.

Facepalm: In the latest example of why we shouldn't blindly put our faith in online map services, German autobahns and highways were plunged into chaos when Google Maps incorrectly showed that vast sections of the roads were closed. The incident's timing was especially bad as it coincided with the start of a holiday break. German motorists likely felt disheartened at the sight of all the stop signs on Google Maps on Thursday. The Guardian reports that major roads in western, northern, south-western and central parts of the country were shown as closed. Even parts of Belgium and the Netherlands appeared to have ground to a halt. The situation was exacerbated by the incident taking place at the start of a four-day break for the Ascension holiday, when many Germans were travelling. It led to a huge number of Google Maps users heading for alternative routes to avoid the non-existent closures. Somewhat ironically, this caused huge jams and delays on these smaller roads. Drivers not relying on Google Maps – and any Google users who decided to check another service or the news – didn't have to deal with these problems. Apple Maps, Waze, and the traffic reports all showed that everything was moving freely. The major highways were likely quieter than usual as so many Google Maps users were avoiding them. The apparent mass closure of so many roads caused panic among those who believed Google Maps' warning. Some thought there had been a terrorist attack or state-sponsored hack, while others speculated about a natural disaster. // Related Stories When asked about the glitch, which lasted around two hours, Google said the company wouldn't comment on the specific case. It added that Google Maps draws information from three key sources: individual users, public sources such as transportation authorities, and a mix of third-party providers. Ars Technica contacted Google to ask about the cause of the problem. A spokesperson said the company "investigated a technical issue that temporarily showed inaccurate road closures on the map" and has "since removed them." With Google Maps drawing information from third parties, the issue could partly have been related to the German Automobile Club's warning that there may be heavy traffic at the start of the holiday. Google also added AI features to Maps recently, and we all know how reliable they can be. There have been plenty of other incidents in which Google Maps got things very wrong. Germany was cursing the service again earlier this month when it showed highway tunnels being closed in part of the country when they were open. In 2023, Google was sued by the family of a North Carolina man who drove his car off a collapsed bridge as he followed directions given by Google Maps. The case is ongoing. Masthead: Bhavik Nasit