What Amazon’s Astro Taught Me About Giving Robots a Soul

<img src="https://spectrum.ieee.org/media-library/cute-wheeled-home-robot-with-a-tablet-face-set-against-a-blue-heart-patterned-background.jpg?id=66906422&width=1200&height=400&coordinates=0%2C417%2C0%2C417"/><br/><br/><p>In 2018, Amazon brought me in as the lead UX Sound Designer for <a href="https://spectrum.ieee.org/amazon-astro-robot" target="_blank">Astro, their first consumer home robot</a>. Astro used cameras and other sensors to map and navigate your <a href="https://spectrum.ieee.org/ai-robots" target="_blank">home and workplace</a>, and could proactively patrol, check up on loved ones, and transport small items using its built-in cargo bin. While there was a well-defined feature set and form factor, initially there was no character direction. In fact, even before <a href="https://www.amazon.com/Introducing-Amazon-Astro/dp/B078NSDFSB" target="_blank">Astro</a> had a name, there were two main questions—was it simply Alexa on wheels, or was it a robot with its own character?</p><p>The Astro team was divided. One option was to focus on Alexa, and treat the mobile robot simply as an added utility. I argued for Astro to not focus on Alexa, along with the majority of the UX team. Our belief was that a thing that moves through your home and turns toward you with intent can never be just an appliance. People would ascribe character to whether we wanted them to or not, and so the only question was whether we shaped that character or let it happen by accident.</p><p>Ultimately, <a href="https://www.aboutamazon.com/news/devices/meet-astro-a-home-robot-unlike-any-other" target="_blank">Astro became Astro rather than Alexa</a>, and user testing backed up our decision. People <em><em>didn’t</em></em> see the robot as Alexa. They saw it as its own character, and that’s what they wanted it to be. Alexa on the device felt somewhat strange and creepy, but building Astro its own voice was too slow and expensive in 2018. So, we settled on Alexa as a supporting character that handled any actual talking, while Astro was the main character, communicating as much as it could without words, through sound, motion, and facial expressions.</p><p>I had been brought on to the Astro team to define the robot’s sound design language and voice. But there was no one to flesh out the robot’s actual character. You cannot make a single real decision about a character without defining it first. Every choice about how Astro moved, sounded, paused, or reacted was a character choice, and those choices required all disciplines working together. As Sound Lead, I was weaving together sound, motion, and character, and how they played together inside each story moment. The animators, who programmed Astro’s motion and facial expressions, were extraordinary at what they did, but the emotional arc they were animating came from the sound (and therefore character) work first. So I stepped into that role, which is where my real work started. What I learned about building character for robots applies to nearly everything being built in embodied AI right now.</p><h2>Character Is a Design System</h2><p>Developing a character for Astro meant answering questions that had never been asked about a product at Amazon: What is the emotional range of this robot’s baseline state? How does this robot communicate uncertainty without eroding trust? Where is the line between being expressive and annoying? What are the vulnerabilities of this device’s character?</p><p>These are design questions. They have real answers, and every team working on the product has to build from them. For example, Astro’s emotional range was designed to be relatively small at first. We never wanted Astro to get too sad or too angry. It could play sad, but would snap out of it quickly and end the reaction on a high note to keep things positive.</p><p class="shortcode-media shortcode-media-youtube"> <span class="rm-shortcode" data-rm-shortcode-id="5ace7686175eb510c58a3b79ecc7f5e3" style="display:block;position:relative;padding-top:56.25%;"><iframe frameborder="0" height="auto" lazy-loadable="true" scrolling="no" src="https://www.youtube.com/embed/r1eS3TitrHc?rel=0" style="position:absolute;top:0;left:0;width:100%;height:100%;" width="100%"></iframe></span></p><p>Character leaks out of every seam and can create a disjointed experience if not defined correctly. Even if it’s just animation timing that’s slightly off, or a response that’s technically correct but contextually tone-deaf, users feel every one of these inconsistencies, even if they can’t name them. Watch what happens at the beginning and end of this Sing sequence:</p><p class="shortcode-media shortcode-media-youtube"> <span class="rm-shortcode" data-rm-shortcode-id="24123281b2c3cce6b288876b59fed097" style="display:block;position:relative;padding-top:56.25%;"><iframe frameborder="0" height="auto" lazy-loadable="true" scrolling="no" src="https://www.youtube.com/embed/HtePtQyiTDs?rel=0" style="position:absolute;top:0;left:0;width:100%;height:100%;" width="100%"></iframe></span></p><p>Astro goes from nothing, into the emotional moment, and then lands back on nothing. No build up, no cool down, no sense that the feeling came from somewhere or had anywhere to go. I pushed hard for better character stitching, the transitions in and out of expressive moments that make a performance feel continuous rather than assembled, but it never got implemented. The moment itself works. But without the stitching, it reads as a clip playing on a robot rather than coming from within the robot character itself.</p><h2>Story and Sound at the Beginning</h2><p>We had decided that Astro would have no spoken dialogue, but it had something that functioned the same way: a vocabulary of sounds, tones, and rhythms that acted as its voice. This vocabulary became the leading output of the character’s personality. The robot’s motion and facial expressions were built around it.</p><p>Astro’s wake-up sequence is a great example. Waking wasn’t just a boot animation on the screen; it was an entire performance. Slow and humble at first, the robot oriented itself quietly, then stretched its screen, checked its wheels, and finally, with an upward gesture toward its telescoping mast, it popped it up slightly, and did a little dance of joy. Sound, motion, and eyes hit every beat<em> </em>together in full choreography.</p><p class="shortcode-media shortcode-media-youtube"> <span class="rm-shortcode" data-rm-shortcode-id="3f2f54b4b3d6b267224490a3eaf3d339" style="display:block;position:relative;padding-top:56.25%;"><iframe frameborder="0" height="auto" lazy-loadable="true" scrolling="no" src="https://www.youtube.com/embed/coPva7ltAgM?rel=0&start=261" style="position:absolute;top:0;left:0;width:100%;height:100%;" width="100%"></iframe></span></p><p>The character’s output in that sequence was first written as a story. Astro is waking up in its new home for the first time. Its main aspiration is to be part of a family, so this is the moment it has been waiting for, this is its purpose. Being the responsible character that it is, it wants to make sure everything is good to go before it introduces itself and starts learning its new home.</p><p>This narrative came first because it drove every other decision that we made. After the story was written, sound gave that story a metaphorical voice: the excited tones, the pacing as it checked its wheels, and the bright melodic phrase as Astro looked up at its new family for the first time and introduced itself. Once the sound was laid down, animation did their thing with motion and facial expressions, taking cues from the emotional arc the sound had established. Motion didn’t lead—it followed the feeling of the story and the sounds, the same way an animator follows a recorded vocal take.</p><p>That wake up sequence became one of the most-discussed moments in early user testing. People described it as “alive.” What they were responding to wasn’t any single element. It was all three channels (sound, motion, and facial expressions) expressing the same defined character in harmony.</p><h2>Context Is Where Character Becomes Real</h2><p>The most compelling characters are defined not by a fixed disposition but by how they respond to their environments and the people in them. They’re still recognizably themselves even as they adapt. This is what I call contextual character. A robot living in a home doesn’t occupy a single emotional state. It moves through rooms with different energy, encounters people in different moods, operates at different times of day, and responds to an endless range of social situations it was never explicitly designed for.</p><p>We got close to a contextual character output with Astro’s sound. When a specific piece of environmental context was fed in, the system adapted beautifully, and Astro felt completely alive. But every state like this was still a prediction we made by hand—a situation we had to imagine in advance and design a response for. A random home throws more situations at a robot than anyone can possibly predict, so there was always a longer tail of moments the system was never prepared for.</p><p>The difference between a product people describe as “smart” and one they describe as “aware” often comes down to this. Smartness is capability. Awareness is context. Presence is character. And character is always in reaction to the people around it, to its environment, to its own evolving state. That’s what makes it feel like something is emotionally present with you.</p><p>This is where AI changes the game for character design in ways that go well beyond what was possible with Astro. AI-driven adaptation doesn’t require the contextual predictions that we relied on. It learns the specific rhythms, preferences, and emotional context of the people it lives and works with. The character doesn’t just respond to context. It <em><em>grows</em></em> into it.</p><h2>What Industry Is Missing</h2><p>The character and soul of the impending wave of embodied AI products appears to almost always be an afterthought. And character defined late is character defined by default. It becomes the sum of a thousand small decisions made by different people thinking about anything but character. People project character onto devices whether you plan for it or not, especially if those devices move—a robot that moves is <em><em>already</em></em> a character. If nobody has designed this character, the result will be products that feel like nothing, or worse, feel confusing and not trustworthy. Technically impressive, but lifeless.</p><p>We did not get this fully right with Astro. So many things were moving in parallel that character was rarely treated as a utility, and it made sense why. When you are building a first-of-its-kind product, the things that are the loudest are the ones that break, the deadlines, the costs, the features a customer can point to on a box. Character is quieter than all of that. It’s easy to assume it can come later. On a team as large as the Amazon Astro team, it’s lucky to get any idea onto the roadmap when it is competing with a hundred others that all feel more urgent in the moment. None of this came from people not caring. It came from character being the kind of thing that is hard to prioritize until you see what its absence costs you.</p><h2>My Asks to Product Leaders</h2><p>If you are building a product that will share physical or conversational space with people, three things are worth considering:</p><p><strong>Define character before you define interactions.</strong> You need a defensible character with enough emotional logic to answer hard questions consistently. Find answers to character questions early, and have every discipline build from the same foundation.</p><p><strong>Build story and sound into the character pipeline, not the production pipeline.</strong> Story and sound developed alongside character definition has the chance to inform motion, expression, and interaction logic. This requires a different kind of collaboration, and a different kind of hire.</p><p><strong>Design for adaptation, not just consistency.</strong> A consistent character is necessary, but the products that will matter most in people’s lives are the ones that deepen through use. The infrastructure to support that is more and more accessible, but the design thinking to take advantage of it is still rare.</p><div class="horizontal-rule"></div><p><em><em>An unabridged version of this story can be read on <a href="https://medium.com/@mikeforstmusic/what-amazons-astro-taught-me-about-giving-ai-a-soul-989fcd9c45f4" target="_blank">Medium</a>.</em></em></p> Reference: https://ift.tt/SUGdfWR

Microsoft discovers new lightweight backdoor that steals cryptocurrency

<p>Microsoft says it has detected new self-propagating malware that spreads through USB drives in search of cryptocurrency credentials, which it then sends to attacker-controlled servers.</p> <p>The company named the worm Crypto Clipper because it monitors the contents of device clipboards for patterns consistent with wallet addresses or seed phrases. When found, the malware also takes five screenshots over a 10-second period. Both the credentials and the screenshots are then sent to the attacker through Tor, a network protocol that provides anonymous routing by sending traffic through redundant nodes so logs can’t capture both the sending and receiving IP addresses. Crypto Clipper establishes the Tor connection by using a SOCKS5 proxy, a network protocol that sends traffic through a proxy server, which then forwards it to its final destination.</p> <h2>A lightweight backdoor</h2> <p>“The execution of this clipper is notable because it does not depend on a traditional installer or exposed IP-based C2 infrastructure,” Microsoft <a href="https://www.microsoft.com/en-us/security/blog/2026/06/17/crypto-clipper-uses-tor-worm-like-propagation-for-persistence-control/">said</a> Thursday. “Instead, it deploys a portable Tor client, routes traffic through a local SOCKS5 proxy, and blends data theft with remote code execution, turning a financially motivated stealer into a lightweight backdoor.”</p><p><a href="https://arstechnica.com/security/2026/06/microsoft-spots-new-self-propagating-malware-for-stealing-cryptocurrency/">Read full article</a></p> <p><a href="https://arstechnica.com/security/2026/06/microsoft-spots-new-self-propagating-malware-for-stealing-cryptocurrency/#comments">Comments</a></p> Reference : https://ift.tt/tXMK0SZ

Apple patches high-severity eavesdropping vulnerability in Beats Studio Buds

<p>Apple has updated its Beats Studio Buds wireless earbuds to patch a high-severity vulnerability that could be exploited by nearby hackers to eavesdrop on users.</p> <p>The vulnerability, <a href="https://www.cve.org/CVERecord?id=CVE-2025-20701">CVE-2025-20701</a>, allowed improper authentication in the firmware running on the Bluetooth-related chips, which made it possible for people within signal range to impersonate devices that had previously been paired with the earbuds. The researchers demonstrated this in a series of end-to-end attacks that allowed them to eavesdrop on conversations or sounds within earshot of the phone microphone.</p> <h2>Apple joins the patch party</h2> <p>“Impact: An attacker within Bluetooth range may be able to listen through the microphone of a device which is not yet paired and actively seeking pair requests,” Apple said in a Tuesday security <a href="https://support.apple.com/en-us/127557">advisory</a>. The fix is contained in Beats Firmware Update 1B211, which is delivered automatically while headphones are paired with and within Bluetooth range of a user’s iPhone, iPad, or Mac. Users can check their firmware version by going to Settings on their device, navigating to Bluetooth, and tapping the info button next to the headphones.</p><p><a href="https://arstechnica.com/apple/2026/06/apple-patches-high-severity-eavesdropping-vulnerability-in-beats-studio-buds/">Read full article</a></p> <p><a href="https://arstechnica.com/apple/2026/06/apple-patches-high-severity-eavesdropping-vulnerability-in-beats-studio-buds/#comments">Comments</a></p> Reference : https://ift.tt/CoadXKT

Before SpaceX IPO, investors in China secretly acquired stakes

<p>A businessman with ties to Chinese military contractors was among the overseas investors who acquired stakes in SpaceX while it was still a private company. An entity linked to the Qatari royal family also took a stake.</p> <p>The new details come from a <a href="https://www.documentcloud.org/documents/28232877-jx-537-r/">private investor list</a> obtained by ProPublica that sheds light on a particularly delicate issue for Elon Musk’s rocket company: which people in countries like China bought into the company, and how. SpaceX built its business off sensitive US government work like making spy satellites for the Pentagon. While there is no ban on Chinese investment in US military contractors, such investment is heavily regulated.</p> <p>In a sign of its sensitivity to the concerns, SpaceX barred investors from China and Hong Kong from buying shares in its initial public offering last week due to “regulatory and compliance risks,” <a href="https://www.bloomberg.com/news/articles/2026-06-05/chinese-hk-investors-banned-from-spacex-ipo-on-security-grounds">Bloomberg reported</a>. The US government alleges that China has a strategy of using investments in sensitive industries for espionage and to get access to cutting-edge technology.</p><p><a href="https://arstechnica.com/information-technology/2026/06/before-spacex-ipo-investors-in-china-secretly-acquired-stakes/">Read full article</a></p> <p><a href="https://arstechnica.com/information-technology/2026/06/before-spacex-ipo-investors-in-china-secretly-acquired-stakes/#comments">Comments</a></p> Reference : https://ift.tt/TOsxqRX

Tesco moving 40,000 server workloads off VMware amid Broadcom's “abusive conduct”

<p>Tesco, a retail conglomerate headquartered in the United Kingdom, is moving 40,000 server workloads off of VMware amid "abusive conduct" from Broadcom, recent legal filings claim.</p> <p>Tesco filed a lawsuit in the UK’s High Court against Broadcom alleging breach of contract last year. According to a September report from <a href="https://www.theregister.com/software/2025/09/03/supermarket-giant-tesco-sues-vmware-for-breach-of-contract/1420651">The Register</a>, the lawsuit claimed that in January 2021, Tesco bought perpetual licenses for VMware’s vSphere Foundation and Cloud Foundation, a subscription to VMware Tanzu, plus support services until 2026, with the option to extend support for four additional years.</p> <p>But when <a href="https://arstechnica.com/information-technology/2022/05/broadcom-will-pay-61-billion-to-become-the-latest-company-to-acquire-vmware/">Broadcom took over VMware</a> in November 2023, it would not honor the deal and instead tried to get Tesco to pay “excessive and inflated prices for virtualization software for which Tesco has already paid” and would not allow it to buy support services for its perpetually licensed software without buying “duplicative subscription-based licenses for those same Software products," the initial complaint read, <a href="https://www.theregister.com/software/2025/09/03/supermarket-giant-tesco-sues-vmware-for-breach-of-contract/1420651">The Register reported</a> at the time.</p><p><a href="https://arstechnica.com/information-technology/2026/06/tesco-moving-40000-server-workloads-off-vmware-amid-broadcoms-abusive-conduct/">Read full article</a></p> <p><a href="https://arstechnica.com/information-technology/2026/06/tesco-moving-40000-server-workloads-off-vmware-amid-broadcoms-abusive-conduct/#comments">Comments</a></p> Reference : https://ift.tt/2XABILx

How Musicians Can Get Paid for Training AI

<img src="https://spectrum.ieee.org/media-library/conceptual-illustration-of-two-quarter-note-stems-going-through-an-s-resembling-a-dollar-sign.jpg?id=66750724&width=1200&height=400&coordinates=0%2C417%2C0%2C417"/><br/><br/><p>Musicians are accustomed to getting paid each time their creative work is used. Across vinyl/CD sales, streams, radio, cover versions, and those numerous niches like karaoke, there are agreements in place about what “use” means. Underlying this is a simple economic principle: The more something is used, the more money it makes.</p><p><span>Generative AI has <a href="https://spectrum.ieee.org/ai-art-generator" target="_blank">complicated the definition of use</a>. On the one hand, you could argue that the use of a piece of musical training data happens just once, at the point of training. On the other hand, creators would be right to complain that the creative essence of their work lives on in the structure of the model, used every time the model produces an output.</span></p><p><span></span><span>Now, companies like Sureel and SoundVerse are working to re-create the essential economic principle that motivates creativity in an era of AI. Such initiatives aim to turn the generative AI industry from one guilty of “the biggest act of copyright theft in history” into one that coexists harmoniously with hardworking artists.</span></p><h2>Music Royalties for the AI era </h2><p><a href="https://www.sureel.ai/" target="_blank">Sureel</a>, a startup Warner Music Group just <a href="https://www.musicbusinessworldwide.com/warner-music-group-acquires-sureel-ai-the-attribution-startup-that-traces-how-ai-models-use-artists-work/" target="_blank">acquired</a>, has partnered with the Swedish copyright agency <a href="https://www.stim.se/" rel="noopener noreferrer" target="_blank">STIM</a> to explore the potential for<a href="https://www.stim.se/en/news/stim-launches-the-worlds-first-ai-license-for-music" rel="noopener noreferrer" target="_blank"> music creators to get paid when their music is used to train generative AI tools</a>. Sureel’s software labels online media, such as a music file, with instructions determined by the owner. The instructions specify whether an AI company may use the media freely in training, limit its influence in any given training set, or avoid it altogether. The software then tracks how the AI company uses the media in training and sets licensing fees accordingly. </p><p>Meanwhile, the founders of the AI music company SoundVerse “[reject] one-time royalty buyouts as insufficient and [advocate] for ongoing participation of artists in the AI lifecycle,” they wrote in a <a href="https://www.soundverse.ai/whitepaper.pdf" rel="noopener noreferrer" target="_blank">2025 white paper</a>. They argue that each time a generative AI system produces an output, certain pieces of training data play a greater role than others. If the system outputs music resembling jazz, the jazz in the training set has arguably contributed more than, say, the folk music. You can therefore differentially reward each piece of training data for each output.</p><p> Sureel’s Co-President Benji Rogers told me, “Attribution isn’t about re-creating the old economics. It’s about measuring, for the first time, the thing the old economics only approximated.”</p><p>Such influence attribution needs to do more than superficially measure how similar a training data point is to the AI output. The challenge is to attribute causality, or a relationship between the training data and the trained AI, Sureel CEO Tamay Aykut says. </p><p> Even if the AI industry achieved that, however, it might encourage people to create music designed to maximize training-data royalties. While all creative markets lead to new incentives (music streaming, for example, has driven songs to have shorter intros), the industry could do without another economic structure that is easily gamed, in which someone’s reverse-engineered pastiche diverts royalties away from original works of creative expression.</p><p class="ieee-inbody-related">RELATED: <a href="https://spectrum.ieee.org/midjourney-copyright" target="_self">Generative AI Has a Visual Plagiarism Problem</a></p><p>Inferring the influence of a particular piece of music on a generated piece of music, if a well-defined problem at all, may involve more advanced information theoretic principles, or modelling the actual historical role and impact of individual works. Aykut proposes that in carefully designed attribution systems, more unusual and unpolished musical works could even have more inherent value than radio standards.</p><p> Simon Gozzi, Head of Business Development at STIM, says the company is in the process of seeing how Sureel’s attribution reports could underlie licensing agreements between musicians and AI companies. Could generative AI attribution strategies not only sustain the economic logic that “popularity pays,” but also motivate musical experimentation and diversity? It’s a compelling concept when public sentiment rightly fears generative AI’s threat to cultural vibrancy, pushing power towards tech companies, deskilling creative workers, shrinking revenue in the creative sector, and filling the internet with slop. “Attribution is one of the few credible tools we have,” Rogers says.</p><p class="pull-quote"> There’s a window of opportunity to debate and establish approaches to paying for AI training data that serve a vibrant and sustainable creative sector.</p><p>The technical problem of training data attribution is both complex and ill-defined. Just as a simplistic attribution strategy based on measuring similarity might motivate people to reverse-engineer the canonical works of a genre to capture royalties, a more complex attribution strategy based on some information theory of originality might be easily gamed or fail to reward human cultural production. </p><p> For creative workers, there’s good reason to fear that even with the best intentions, AI attribution will only compound the baroque and opaque arms races that they are already weary of navigating. Some voices within the music AI sector are also skeptical. Drew Silverstein, president of SourceAudio, says, “Attribution would seem to be the obvious answer, but it’s flawed in AI, so we have to look at other models.” He advocates simple negotiated agreements with an agreed or annually recurring price at the point of training.</p><p>Meanwhile, the copyright lawsuits that have dominated the generative AI revolution are beginning to give way to an increasing number of privately negotiated agreements, such as those between <a href="https://www.theverge.com/news/790405/warner-universal-music-ai-deals" rel="noopener noreferrer" target="_blank">Universal, Warner, and major AI companies</a> to work together on training models with copyright consent. Although <a href="https://www.musicbusinessworldwide.com/sunos-licensing-talks-with-major-labels-in-limbo-with-no-path-forward-report/" rel="noopener noreferrer" target="_blank">little is certain</a>, these agreements may have considerable influence over the industry norms that arise. </p><p>Right now, there’s a window of opportunity to debate and establish approaches that pay for AI training data while also sustaining a vibrant creative sector. Sophisticated engineering solutions will have a role to play, but they need to take into account the cultural complexity of the challenge, and enable fairness and transparency through good design. </p><h2>Making AI training pay off </h2><p> It remains to be seen whether monolithic generative models such as Suno actually have as much credibility as first touted. In many creative applications of AI, there’s a renewed focus on smaller customized models that are tailored for specific human creative expressive needs such as <a href="https://forum.ircam.fr/projects/detail/rave/" rel="noopener noreferrer" target="_blank">IRCAM’s RAVE</a> model or <a href="https://www.jenmusic.ai/stylefilters" rel="noopener noreferrer" target="_blank">Jen’s Style Filters</a>. Meanwhile, more mainstream “end user” creative applications may be shifting towards a focus on fan engagement. <a href="https://www.nytimes.com/2026/03/24/technology/openai-shutting-down-sora.html" rel="noopener noreferrer" target="_blank">OpenAI’s sudden dropping of Sora</a>, despite being in negotiations with Disney and <a href="https://www.youtube.com/watch?v=-XZQx4PFqvs" rel="noopener noreferrer" target="_blank">Suno’s recent emphasis on building fan engagement experiences that draw directly on the work of artists</a>, following its deal with Universal, both point to teething troubles in the creative AI sector. </p><p> A move to smaller, more targeted models and applications would give more room for creator alliances. For example, collectives of musicians might band together to provide the training data for a smaller custom model, for which revenue splits might be egalitarian or based on other principles of fairness.</p><p>The same may possibly be true of hybrid model architectures and structured training regimes where different data sources are used at different points in the training process, as well as retrieval augmented generation, which mixes context-specific information with training data to improve results. An approach that produces worse results but enables fairer or more transparent paths of attribution may be more successful if it brings creators on board with more lucrative royalty flows and even clear credits.</p><p> Also, no matter how sophisticated an attribution algorithm is, it will always be grounded in human decisions, ranging from the wise and the fair to the arbitrary and corrupt. Ask a music industry insider to explain how the percentage split between recording and songwriting royalties is determined, and you’re in for a long answer. At best, the machinery of training data attribution will enable open and informed discussion about what makes our creative and cultural sectors fair and vibrant. At worst, it will conceal already opaque private agreements in complex black boxes.</p><p> This is where national policies are vital. Attribution must be “multi-layered and auditable, open to expert and regulatory scrutiny,” Rogers says. Crafting such policies will take expertise from computer science, musicology, law, and economics. AI-competitive governments will be able to boost their cultural and creative sectors by supporting institutions that fulfil this purpose. </p><p> Even the most neoliberal economies look beyond markets to sustain cultural expression, whether through public arts funding or measures like local music quotas for radio. As the economic impact of generative AI in the creative sector takes form, taxation, redistribution, and active support of cultural infrastructures may still be the most effective way to support positive social outcomes. Taxing big AI and redistributing that revenue back to the creative workers that contributed to the industry’s wealth is, after all, another “AI attribution strategy.” </p> Reference: https://ift.tt/zKUyo3j

The Secret to Marathon-Winning Humanoid Robots

<img src="https://spectrum.ieee.org/media-library/a-red-and-black-humanoid-runs-alone-through-a-marathon-course.jpg?id=66940897&width=1200&height=400&coordinates=0%2C295%2C0%2C296"/><br/><br/><p>On April 19, 2026, the <a href="https://www.cnn.com/2026/04/19/china/china-robot-half-marathon-intl-hnk" rel="noopener noreferrer" target="_blank">Honor Lightning humanoid robot ran a half-marathon in 50 minutes and 26 seconds</a>, beating the human world record by 7 minutes and the best robot time from 2025 by almost two hours.</p><p>How did they do it? Is there some magical technology or technique that unlocked this performance? How did they beat the significantly better-known Unitree (who reportedly had to supply an ice backpack to try and complete the race without overheating)? My doctoral thesis involved <a href="https://www.avikde.me/p/phd-defense" rel="noopener noreferrer" target="_blank">building and controlling hopping and running robots</a>, and <a href="https://www.avikde.me/p/ghost-robotics-minitaur" rel="noopener noreferrer" target="_blank">since then I’ve tried to design and build efficient commercial legged robots</a>, giving me a decent idea of the constraints involved. In this article, we take a look at the fundamental underlying constraints to try and answer these questions.</p><hr/><h3>The Physics of Running</h3><p><a href="https://spectrum.ieee.org/ai-institute" target="_blank">Running</a> consists of alternating phases of a leg pushing against the ground (“stance phase”) and the body flying through the air (“aerial phase”). In the aerial phase, the body falls due to gravity, losing vertical momentum. The leg in stance phase pushes against the ground to redirect the vertical momentum upward, while the other leg swings forward to reposition for the next foothold.</p><p><a href="https://spectrum.ieee.org/ev-motor" target="_blank">Electric motors</a> use energy to produce torque- the higher the torque, the more energy lost as heat. Adding a geartrain after the motor amplifies its torque and reduces its speed. A large reduction helps with torque production, but since the rotor of the motor itself has to spin faster, it becomes very sluggish at accelerating its output. This is obviously bad for the swing phase described above. These competing effects mean that for a particular motor, there is usually a sweet spot for the gear ratio:</p><p class="shortcode-media shortcode-media-rebelmouse-image"> <img alt="A graph showing the relationship between gearing and motor efficiency, with an optimal gearing ratio in the relationship between stance and swing." class="rm-shortcode" data-rm-shortcode-id="4c2224acc293d6b3ce8b8b6553aa30f5" data-rm-shortcode-name="rebelmouse-image" id="10bd7" loading="lazy" src="https://spectrum.ieee.org/media-library/a-graph-showing-the-relationship-between-gearing-and-motor-efficiency-with-an-optimal-gearing-ratio-in-the-relationship-between.jpg?id=66940901&width=980"/> <small class="image-media media-caption" placeholder="Add Photo Caption...">The power consumed by a robot leg is minimized at an optimal gear ratio (30:1 in this example).</small><small class="image-media media-photo-credit" placeholder="Add Photo Credit...">Avik De/Datawrapper</small></p><h3>How Honor Did It</h3><p>While the Lightning’s motor specifications are not published, the hip and knee motors roughly have a 110-150mm outer diameter. For an approximate set of motor parameters, I looked to the <a href="https://www.tq-group.com/en/products/tq-robodrive/servo-kits/ilm115x25/" target="_blank">ILM115x25 motor</a> due to its relevant size and detailed specifications.</p><p>We can use a simple physics model to estimate the power consumption for running at 7 m/s (the Lightning’s average half marathon speed) as gear ratio varies:</p><p class="shortcode-media shortcode-media-rebelmouse-image"> <img alt="A graph showing that optimal gearing for a robot\u2019s motor dissipates the amount of heat that the motor generates." class="rm-shortcode" data-rm-shortcode-id="e04f969907417a25696dd3127e090008" data-rm-shortcode-name="rebelmouse-image" id="185f3" loading="lazy" src="https://spectrum.ieee.org/media-library/a-graph-showing-that-optimal-gearing-for-a-robot-u2019s-motor-dissipates-the-amount-of-heat-that-the-motor-generates.jpg?id=66940912&width=980"/> <small class="image-media media-caption" placeholder="Add Photo Caption...">The light blue curve shows how to pick the optimal gearing (45:1). The dark blue curve shows how much heat will be produced in the knee motor, ~150W for the optimal gearing.</small><small class="image-media media-photo-credit" placeholder="Add Photo Credit...">Avik De/Datawrapper</small></p><p>We see that the drivetrain is not magical: with a gear ratio <em><em>chosen for this task</em></em> (we’ll return to this below), the approximate robot power consumption would be a very reasonable 400W.</p><p>However, the dissipated knee power ( typically the main thermal limiting factor) is ~150W. This is almost an unavoidable consequence — running at human speeds with a humanoid-sized robot will inevitably generate this amount of heat! Over a prolonged period, keeping the motor from overheating would be a challenge, but the Lightning has a <a href="https://eu.36kr.com/en/p/3775418378027520" target="_blank">trick up its sleeve</a>:</p><blockquote>According to Honor, the liquid - cooling pipes penetrate deep into the motors like capillaries. The high - power liquid pump has a heat - exchange flow rate of more than 4 liters per minute. Each of the four drive motors in the lower limbs is equipped with an independent liquid - cooling circuit.</blockquote><p>Liquid cooling is not new, but it’s definitely not a commodity. It has shown up in research periodically, and on the commercial side <a href="https://apptronik.com/news-collection/apptronik-readies-its-humanoid-robot-for-a-summer-unveil" rel="noopener noreferrer" target="_blank">Apptronik tried it for a few of their prototypes</a> but (to my knowledge) does not use it on their main <a href="https://apptronik.com/apollo" target="_blank">Apollo</a> platform. Basic air convection-based cooling would not continuously be able to extract 150W out of the knee motor, and so the cooling technology is a key enabler of this type of performance.</p><h3>Why Others Couldn’t Compete</h3><p>Why did Honor’s competitors, including more <a href="https://www.forbes.com/sites/johnkoetsier/2026/01/09/top-10-humanoid-robot-companies-by-shipments-revealed/" rel="noopener noreferrer" target="_blank">established and widely-shipped humanoids</a> such as from <a href="https://www.unitree.com/g1" target="_blank">Unitree</a> or <a href="https://www.agibot.com/" target="_blank">Agibot</a>, not compete as well?</p><p>We can use the same model to generate an equivalent energetics plot for walking at 1.5 m/s, a much more modest but potentially more common activity for a commercial humanoid robot:</p><p class="shortcode-media shortcode-media-rebelmouse-image"> <img alt="A graph showing that robots with gear ratios optimized for running or walking are inefficient when walking or running respectively." class="rm-shortcode" data-rm-shortcode-id="5bbe64af17f8581b4106547f468728a4" data-rm-shortcode-name="rebelmouse-image" id="616f5" loading="lazy" src="https://spectrum.ieee.org/media-library/a-graph-showing-that-robots-with-gear-ratios-optimized-for-running-or-walking-are-inefficient-when-walking-or-running-respective.jpg?id=66940939&width=980"/> <small class="image-media media-caption" placeholder="Add Photo Caption...">The solid and dashed light blue lines show a running-optimized design, while green lines show a walking-optimized design. The optimal ratio for walking is much lower (30:1 vs 45:1). However, the power dissipated in the knee motor while running (dark blue) is much higher at 30:1 vs 45:1—the price to pay for running with a walking-optimized design.</small><small class="image-media media-photo-credit" placeholder="Add Photo Credit...">Avik De/Datawrapper</small></p><p>The plot adds a new green curve for the walking power, and the optimal gearing is significantly different!</p><p>Let’s say you design your robot to excel at the normal walking task and choose the green design with 30:1 gearing. The knee motor power to run a half marathon is over 300W (red arrow), more than 2x what we had with the running-optimized design. It wouldn’t be so surprising to need ice packs!</p><p>Conversely, visually following the green curve shows that the running-optimized robot wastes more power for walking. Using larger motors sized for running increases the weight of the robot and wastes power when it is standing or walking. The larger motors also pose practical issues like bumping into objects while operating in homes or factories.</p><h3>Closing Thoughts</h3><p>Honor’s half marathon performance was an impressive engineering effort and result. It didn’t need any magical leaps in technology, but the deployment of the capillary motor cooling solution is a notable advance without which this running pace would have been unsustainable. The cooling, weight optimization, and robustness advances may well be useful for more practical purposes like carrying heavy payloads down the line.</p><p class="shortcode-media shortcode-media-rebelmouse-image"> <img alt="A comparison showing two similar humanoid robots, but one has significantly smaller motors on its hips." class="rm-shortcode" data-rm-shortcode-id="3ef7dc89b86a70493190325135f1f20f" data-rm-shortcode-name="rebelmouse-image" id="19121" loading="lazy" src="https://spectrum.ieee.org/media-library/a-comparison-showing-two-similar-humanoid-robots-but-one-has-significantly-smaller-motors-on-its-hips.jpg?id=66941011&width=980"/> <small class="image-media media-caption" placeholder="Add Photo Caption...">The Honor Lighting robot [right] has much larger motors driving its legs than the Unitree H1 robot [left], making it a more efficient runner but a less efficient walker.</small><small class="image-media media-photo-credit" placeholder="Add Photo Credit...">Left: Wei Zhiyang/Zhejiang Daily Press Group/VCG/Getty Images; Right: VCG/Getty Images</small></p><p>However, the Lightning is not as well-suited to other tasks as a robot designed for greater versatility. Engineering is always characterized by tradeoffs, and making the correct ones separates good products from great ones. With consistently improving AI language models, this very human skill is becoming the most valuable one an engineer can have.</p><p>The news coverage seemed to overly focus on the fact that the human half-marathon record had been broken by a robot. Machines and humans have very different capabilities and constraints, so why should we ever have expected the half marathon time for a robot and human to be related? As in <a href="https://en.wikipedia.org/wiki/Deep_Blue_versus_Garry_Kasparov" rel="noopener noreferrer" target="_blank">Deep Blue’s 1997 defeat of Garry Kasparov in chess</a>, where it couldn’t physically move the pieces, the Honor robot’s capabilities are much narrower than a human running elbow-to-elbow with other runners while visually navigating the course without GPS. Comparing the robot runner to a human runner is just an apples-to-oranges comparison, and only risks diminishing Honor’s engineering achievement on one hand, and human athletic achievement on the other.</p> Reference: https://ift.tt/MZjHF4n