Technology

We trained a humanoid with 22-DoF dexterous hands to assemble model cars, operate syringes, sort poker cards, fold/roll shirts, all learned primarily from 20,000+ hours of egocentric human video with no robot in the loop. Humans are the most scalable embodiment on the planet. We discovered a near-perfect log-linear scaling law (R² = 0.998) between human video volume and action prediction loss, and this loss directly predicts real-robot success rate. Humanoid robots will be the end game, because they are the practical form factor with minimal embodiment gap from humans. Call it the Bitter Lesson of robot hardware: the kinematic similarity lets us simply retarget human finger motion onto dexterous robot hand joints. No learned embeddings, no fancy transfer algorithms needed. Relative wrist motion + retargeted 22-DoF finger actions serve as a unified action space that carries through from pre-training to robot execution. Our recipe is called "EgoScale": - Pre-train GR00T N1.5 on 20K hours of human video, mid-train with only 4 hours (!) of robot play data with Sharpa hands. 54% gains over training from scratch across 5 highly dexterous tasks. - Most surprising result: a *single* teleop demo is sufficient to learn a never-before-seen task. Our recipe enables extreme data efficiency. - Although we pre-train in 22-DoF hand joint space, the policy transfers to a Unitree G1 with 7-DoF tri-finger hands. 30%+ gains over training on G1 data alone. The scalable path to robot dexterity was never more robots. It was always us. - Website: https://lnkd.in/gxzgeP-2 - Paper: https://lnkd.in/g7PJdz_8

𝗧𝗵𝗲 𝗽𝗮𝗿𝗮𝗱𝗼𝘅 𝗼𝗳 𝗺𝗼𝗱𝗲𝗿𝗻 𝗵𝗲𝗮𝗹𝘁𝗵 𝘁𝗲𝗰𝗵: 𝗧𝗵𝗲 𝗺𝗼𝗿𝗲 𝘄𝗲 𝗺𝗼𝗻𝗶𝘁𝗼𝗿, 𝘁𝗵𝗲 𝗺𝗼𝗿𝗲 𝗮𝗻𝘅𝗶𝗼𝘂𝘀 𝘄𝗲 𝗯𝗲𝗰𝗼𝗺𝗲. We track our bodies 24/7. Count every calorie. Measure sleep, HRV, glucose, stress. From Apple Watch. To Oura Ring. To the latest “temple” device. Somewhere along the way, awareness turned into obsession. Here’s the paradox no one talks about: We have the best health-tracking tools in history, and some of the worst health outcomes. Something doesn’t add up. 𝗪𝗵𝗮𝘁 𝘁𝗵𝗲 𝗿𝗲𝘀𝗲𝗮𝗿𝗰𝗵 𝗮𝗰𝘁𝘂𝗮𝗹𝗹𝘆 𝘀𝗵𝗼𝘄𝘀 𝗦𝗹𝗲𝗲𝗽 𝘁𝗿𝗮𝗰𝗸𝗶𝗻𝗴 𝗰𝗮𝗻 𝘄𝗼𝗿𝘀𝗲𝗻 𝘀𝗹𝗲𝗲𝗽 Studies on orthosomnia (an obsession with “perfect” sleep metrics) show that people who fixate on sleep scores experience more sleep anxiety, lighter sleep, and poorer recovery—even when objective sleep doesn’t improve. Trying to optimize sleep can literally break it. 𝗛𝗥𝗩 𝗺𝗼𝗻𝗶𝘁𝗼𝗿𝗶𝗻𝗴 𝗶𝗻𝗰𝗿𝗲𝗮𝘀𝗲𝘀 𝘀𝘁𝗿𝗲𝘀𝘀 𝗳𝗼𝗿 𝗺𝗮𝗻𝘆 𝘂𝘀𝗲𝗿𝘀 HRV is a useful trend marker—but daily fluctuations are normal. Research shows that constant HRV checking can heighten health anxiety and perceived stress, especially when users don’t understand variability or context. Ironically, stressing about HRV often lowers HRV. 𝗠𝗼𝗿𝗲 𝗱𝗮𝘁𝗮 ≠ 𝗯𝗲𝘁𝘁𝗲𝗿 𝗵𝗲𝗮𝗹𝘁𝗵 𝗱𝗲𝗰𝗶𝘀𝗶𝗼𝗻𝘀 Behavioral science research consistently finds that excessive self-monitoring leads to hypervigilance, loss of bodily trust, and decision fatigue. When every sensation becomes a data point, people stop listening to internal cues and start deferring to dashboards. In short: 𝗢𝘃𝗲𝗿-𝗺𝗲𝗮𝘀𝘂𝗿𝗲𝗺𝗲𝗻𝘁 𝗿𝗲𝗽𝗹𝗮𝗰𝗲𝘀 𝗮𝘄𝗮𝗿𝗲𝗻𝗲𝘀𝘀 𝘄𝗶𝘁𝗵 𝗮𝗻𝘅𝗶𝗲𝘁𝘆. So what actually creates health? The same fundamentals that worked 5,000 years ago: • Deep, peaceful sleep • Regular sunlight • Real, nourishing food • Daily movement • Time with people you love These don’t need algorithms. They need presence. Use wearables if they serve you—I do, occasionally. But don’t let them become your master. Your life isn’t an algorithm waiting to be optimized. It’s a system meant to be felt, explored, and course-corrected. The best health coach you’ll ever have is already inside you. Trust it.

Every cloud provider faces the same AI infrastructure challenge: chips need to be positioned close together to exchange data quickly, but they generate intense heat, creating unprecedented cooling demands. We needed a strategic solution that allowed us to use our existing air-cooled data centers to do liquid cooling without waiting for new construction. And it needed to be rapidly deployed so we could bring customers these powerful AI capabilities while we transition towards facility-level liquid cooling. Think of a home where only one sunny room needs AC, while the rest stays naturally cool – that’s what we wanted to achieve, allowing us to efficiently land both liquid and air-cooled racks in the same facilities with complete flexibility. The available options weren't great. Either we could wait to build specialized liquid-cooled facilities or adopt off-the-shelf solutions that didn't scale or meet our unique needs. Neither worked for our customers, so we did what we often do at Amazon… we invented our own solution. Our teams designed and delivered our In-Row Heat Exchanger (IRHX), which uses a direct-to-chip approach with a "cold plate" on the chips. The liquid runs through this sealed plate in a closed loop, continuously removing heat without increasing water use. This enables us to support traditional workloads and demanding AI applications in the same facilities. By 2026, our liquid-cooled capacity will grow to over 20% of our ML capacity, which is at multi-gigawatt scale today. While liquid cooling technology itself isn't unique, our approach was. Creating something this effective that could be deployed across our 120 Availability Zones in 38 Regions was significant. Because this solution didn't exist in the market, we developed a system that enables greater liquid cooling capacity with a smaller physical footprint, while maintaining flexibility and efficiency. Our IRHX can support a wide range of racks requiring liquid cooling, uses 9% less water than fully-air cooled sites, and offers a 20% improvement in power efficiency compared to off-the-shelf solutions. And because we invented it in-house, we can deploy it within months in any of our data centers, creating a flexible foundation to serve our customers for decades to come. Reimagining and innovating at scale has been something Amazon has done for a long time and one of the reasons we’ve been the leader in technology infrastructure and data center invention, sustainability, and resilience. We're not done… there's still so much more to invent for customers.

🗞️ A must-have for anyone teaching Russian disinformation tactics. A comprehensive yet highly pedagogical and illustrated catalogue of tactics with concrete examples. 👏🏼Well done @center for countering disinformation with the support of The European Union Advisory Mission Ukraine (#EUAM Ukraine) 🇪🇺 1️⃣ The first part is dedicated to the Mechanisms of destructive information influence: • Bots 🤖 • Fake accounts 🤳🏻 • Anonymous authority 👁️ • Appeal to authority 🔨 • Deepfakes 👾 • Potemkin villages 🤡 • Duplicating websites or accounts 👨🏻💻 • Framing 🖼️ • Information overload 🌧️ • Agenda-setting 📆 • Demonisation • Polarisation 🤯 • Confirmation bias 🧠 • Primacy effect 🪢 • Deceptive sources 🎭 • Information alibi 🥸 2️⃣ The second part offers an overview of the Tactics of destructive information influence. Particularly useful to identifies the perverse rhetorical tricks at play and counter them with the right arguments: • Clickbaiting • Rating • Information sandwich • Lost in translation • Presence effects • Contextomy • Gish gallop • Whataboutism • Conspiracy theories • Talking away • Mundanisation • Doublespeak • Sleeper effect • “Check it if you can” • False analogy • Trolling • False dilemma • Using jokes or memes • Stereotyping 3️⃣ The last part describes the various soft power tools weaponized to leverage influence : Soft power tools: Russia’s influence through… • films 🎦 • e-sports 🎮 • literature 📕 • music 🎶 • sports ⚽️ • churches ⛪️ • cultural centre networks 🤝🏻 • educational programmes and grants 🎓 • historical revisionism 🖊️ • loyal political structures🏰 👐🏻Many thanks to the authors for a reference document which deserves to be widely shared As someone who srudied humanities, I always longed for the ancient “class of rhetorics” which was, until the late 19th century, the penultimate year of secondary education in France before philosophy: students learned the full art of persuasion—finding ideas, structuring them, refining style, memorizing, and delivering speeches—through constant practice and study of classical models. The purpose was to train them in the art of eloquence—to speak and write clearly, elegantly, and persuasively. And to prepare future orators -lawyers, priests, politicians- as well as any educated citizen. Were this classical knowledge more widely shared today, we might be better equipped to resist the tactics outlined in part 2️⃣ as we would more spontaneously recognize the persuasion strategies used against us -even if they come in alluring video forms these days! - and be able to counter them with the tools of logic and structured argument.

AI just helped a couple get pregnant - after 19 years and 15 failed IVF cycles. The breakthrough came with an AI tool built by a team at Columbia University. It’s called STAR - the world’s first AI system trained to find sperm that embryologists can’t. The husband had azoospermia - a condition where no sperm is visible under the microscope. Dozens of attempts, surgeries, and even overseas experts had failed. But the team at Columbia didn’t give up. They spent 5 years building STAR (Sperm Track and Recovery). The system scans 8 million images per hour using a chip and computer vision, then gently isolates viable sperm missed by even the most experienced lab techs. And it worked. ▶︎ STAR found 44 sperm in a sample that had been manually searched for two full days. ▶︎ That one breakthrough led to a pregnancy that had felt impossible for nearly two decades. ▶︎ And it did so without chemicals, donor samples, or invasive extraction methods. For millions of couples dealing with infertility, this is a glimpse of what AI-assisted reproductive medicine could unlock. But more importantly - this shows us what AI in healthtech should be aiming for: Not just more data. Not just smarter models. But real clinical results that change lives. And as a healthtech investor, this is what I look for in AI-driven care: → A clear pain point → A targeted intervention → And a story no one can ignore What’s your take - could AI reshape fertility care the way it’s starting to reshape diagnostics and mental health? #entrepreneurship #healthtech #innovation

We’ve all heard about AI’s potential to boost productivity. But what truly matters to me is whether it’s making work better for the people who show up every day. At Cisco, our People Intelligence team, in collaboration with IT, has been exploring this very topic, and the findings are fascinating. Here are five key insights from our research that leaders should take seriously: 1. Leaders are key to adoption. At Cisco, employees are 2x more likely to use AI if their direct leader uses it. 2. Generic AI training doesn’t work. Role-specific, practical training accelerates AI use. 3. Confidence gaps exist among senior leaders. Directors at Cisco often feel less confident with AI than mid-level employees, underscoring the need for tailored support at all levels. 4. Employee autonomy fuels adoption. Hybrid work environments are powerful accelerators for AI adoption, while mandates can hinder it. Employees who voluntarily go to the office are more likely to use AI, while those who are required to work on-site have lower adoption. 5. AI use is linked to employee well-being, but the relationship is complex, with both benefits and trade-offs that require thoughtful navigation. This is just the beginning. Next, we’re looking at how AI is transforming the way teams operate. For now, one thing is clear, employees who use AI aren’t just more productive. They’re also more engaged, better aligned with company strategy, and empowered to focus on meaningful work. #AIAdoption #EmployeeExperience #FutureOfWork

The New York Times profiled a start-up with 28 employees serving nearly 50 million users. That company is us. The traditional startup playbook: raise massive funding, hire hundreds of employees, and worry about profitability "later." But there's another way. Everyone at Gamma could fit in a small restaurant. We're not just surviving—we've been profitable for 15+ consecutive months, with revenue growing month over month, and lifetime negative net burn (we have more money in the bank than we've raised). This isn't an accident. We've deliberately designed our organization to maximize impact per person. Instead of creating specialist silos, we hire versatile generalists who can solve problems across domains. Rather than building management hierarchies, we find player-coaches who both lead and execute. Our team leverages AI tools throughout our workflow - Claude for data analysis, Cursor for coding efficiency, NotebookLM for customer research synthesis. These aren't just productivity hacks; they're force multipliers. Examples: — When our growth PM needed better analytics, he didn't file a ticket with a data team—he built a self-serve system that anyone can use without SQL knowledge. — When our marketing lead needed to understand our customers better, she fed thousands of interactions into an LLM and created actionable personas that now guide our entire strategy. — When our design team needs to test a hypothesis, we create a rapid prototype and show it to our power users. What we're seeing isn't just about "doing more with less." It's about fundamentally changing what's possible per person. The most valuable employees aren't specialists who excel in narrow domains - they're resourceful problem-solvers who continuously expand their capabilities. This approach creates remarkable resilience. Since everyone understands multiple functions, we don't have single points of failure when someone leaves or moves to another project. If you're building today, the question isn't how quickly you can scale headcount … it's how much impact you can create with the smallest possible team. The future belongs to tiny teams of extraordinary people.

#Batteries are starting to dominate the evening peak in California's grid, charging up with daytime solar then discharging as solar ramps down. On 5th April they set another new record for share of supply, peaking at over 34% at 7pm. This represents a rapid progression - two years ago the record was just 13%. And they remained the largest source of supply on the grid from 6:35pm until 9:40pm. As more and more battery storage enters the mix, batteries will continue to play an increasing role in the state's grid, and continue to break more records. They are flexible and extremely quick to respond. By charging in the middle of the day they are soaking up excess solar and are then putting this to good use later, reducing the need for gas and imports in the nighttime hours. From just 0.5 GW in 2018, by late 2024 California already had over 13 GW of battery storage capacity, with more on the way. While that may sound like a lot, there is still some way to go with the California Energy Commission estimating the state will need around 52 GW of battery storage to meet it's 2045 target of getting all its power from carbon-free sources. Batteries will play an important role in the decarbonised grid of the future. As prices continue to fall we will see more and more batteries deployed, and are certainly seeing this happen in Australia - especially Western Australia. We are just on the cusp of much more widespread adoption. Onwards and upwards! #energy #sustainability #renewables #energytransition

As organizations increasingly adopt hybrid-cloud architectures, understanding the right path and tools is crucial for professionals aiming to deliver resilient, scalable, and efficient applications. Here’s a Cloud Native roadmap breaking down the skills and tools to master across critical domains. Dive in and explore the ecosystem that powers modern applications! 🔴 𝟭. 𝗟𝗶𝗻𝘂𝘅 𝗙𝘂𝗻𝗱𝗮𝗺𝗲𝗻𝘁𝗮𝗹𝘀   Linux remains at the heart of cloud-native systems. Get comfortable with terminal commands, bash scripting, and distributions like Ubuntu and Red Hat for a solid start. 🟢 𝟮. 𝗡𝗲𝘁𝘄𝗼𝗿𝗸𝗶𝗻𝗴 𝗘𝘀𝘀𝗲𝗻𝘁𝗶𝗮𝗹𝘀   Protocols like HTTP, SSL, and SSH form the backbone of connectivity. Tools like Wireshark are invaluable for monitoring and securing network traffic. 🔵 𝟯. 𝗖𝗹𝗼𝘂𝗱 𝗦𝗲𝗿𝘃𝗶𝗰𝗲𝘀   The cloud is non-negotiable! Whether AWS, Azure, or Google Cloud, understanding SaaS, PaaS, and IaaS is key to harnessing the cloud's potential. 🟣 𝟰. 𝗦𝗲𝗰𝘂𝗿𝗶𝘁𝘆   Security is foundational in cloud-native environments. Tools like Open Policy Agent and Prisma provide the framework for enforcing policies and securing applications. 🟡 𝟱. 𝗖𝗼𝗻𝘁𝗮𝗶𝗻𝗲𝗿𝘀 & 𝗢𝗿𝗰𝗵𝗲𝘀𝘁𝗿𝗮𝘁𝗶𝗼𝗻   Containers revolutionized app deployment! Master Docker, Kubernetes, and service meshes like Istio to orchestrate, scale, and manage applications seamlessly. 🟠 𝟲. 𝗜𝗻𝗳𝗿𝗮𝘀𝘁𝗿𝘂𝗰𝘁𝘂𝗿𝗲 𝗮𝘀 𝗖𝗼𝗱𝗲 (𝗜𝗮𝗖)   IaC tools like Terraform, Chef, and Puppet automate infrastructure, ensuring consistency and efficiency across deployments. IaC is a must for scalable cloud-native applications. 🟢 𝟳. 𝗢𝗯𝘀𝗲𝗿𝘃𝗮𝗯𝗶𝗹𝗶𝘁𝘆   With tools like Prometheus, Grafana, and Elastic Stack, observability gives you the visibility needed to monitor, troubleshoot, and optimize performance in real time. 🔵 𝟴. 𝗖𝗜/𝗖𝗗   Continuous Integration and Delivery streamline deployments. GitLab, Jenkins, and GitOps practices (Argo) enable rapid, reliable application delivery. This roadmap covers essential areas for cloud-native development, from Linux fundamentals to CI/CD and observability. But, the cloud-native landscape is vast and rapidly evolving! Did I miss any critical tools or concepts? Whether it's a tool you swear by or an emerging trend you're excited about, drop it in the comments! 👇

After spending three decades in the aerospace industry, I’ve seen firsthand how crucial it is for different sectors to learn from each other. We no longer can afford to stay stuck in our own bubbles. Take the aerospace industry, for example. They’ve been looking at how car manufacturers automate their factories to improve their own processes. And those racing teams? Their ability to prototype quickly and develop at a breakneck pace is something we can all learn from to speed up our product development. It’s all about breaking down those silos and embracing new ideas from wherever we can find them. When I was leading the Scorpion Jet program, our rapid development – less than two years to develop a new aircraft – caught the attention of a company known for razors and electric shavers. They reached out to us, intrigued by our ability to iterate so quickly, telling me "you developed a new jet faster than we can develop new razors..." They wanted to learn how we managed to streamline our processes. It was quite an unexpected and fascinating experience that underscored the value of looking beyond one’s own industry can lead to significant improvements and efficiencies, even in fields as seemingly unrelated as aerospace and consumer electronics. In today’s fast-paced world, it’s more important than ever for industries to break out of their silos and look to other sectors for fresh ideas and processes. This kind of cross-industry learning not only fosters innovation but also helps stay competitive in a rapidly changing market. For instance, the aerospace industry has been taking cues from car manufacturers to improve factory automation. And the automotive companies are adopting aerospace processes for systems engineering. Meanwhile, both sectors are picking up tips from tech giants like Apple and Google to boost their electronics and software development. And at Siemens, we partner with racing teams. Why? Because their knack for rapid prototyping and fast-paced development is something we can all learn from to speed up our product development cycles. This cross-pollination of ideas is crucial as industries evolve and integrate more advanced technologies. By exploring best practices from other industries, companies can find innovative new ways to improve their processes and products. After all, how can someone think outside the box, if they are only looking in the box? If you are interested in learning more, I suggest checking out this article by my colleagues Todd Tuthill and Nand Kochhar where they take a closer look at how cross-industry learning are key to developing advanced air mobility solutions. https://lnkd.in/dK3U6pJf