IntroductionHello! My name is Denys. I was born and raised in Nigeria, where I'm currently based. I began my journey into 3D art in March 2022, teaching myself through online resources, starting, of course, with the iconic donut tutorial on YouTube. Since then, I've continued to grow my skills independently, and now I'm working toward a career in 3D generalism, with a particular interest in environment art.I originally got into Blender because SketchUp wasn't free, and I could not keep up with the subscriptions. While searching for alternatives, I came across Blender. That's when I realized I had installed it once years ago, but back then, the interface completely intimidated me, and I gave up on it. This time, though, I decided to stick with it – and I'm glad I did.I started out creating simple models. One of my first big projects was modeling the entire SpongeBob crew. That led to my first animation, and eventually, the first four episodes of a short animated series (though it's still incomplete). As I grew more confident, I began participating in online 3D competitions, like cgandwe, where I focused on designing realistic environments. Those experiences have played a huge role in getting me to where I am today.Getting Started Before starting any scene, I always look for references. It might not be the most original approach, but it's what works best for me. One piece that inspired me was a beautiful artwork by Calder Moore. I bookmarked it as soon as I saw it back in 2023, and luckily, I finally found the time to bring it to life last month.BlockoutThe goal was to match the original camera angle and roughly model the main frame of the structures. It wasn't perfect, but modeling and placing the lower docks helped me get the perspective right. Then I moved on to modeling and positioning the major structures in the scene.I gave myself two weeks to complete this project. And as much as I enjoy modeling, I also enjoy not modeling, so I turned to asset kits and free models to help speed things up. I came across an awesome paid kit by Bigmediumsmall and instantly knew it would fit perfectly into my scene.I also downloaded a few models from Sketchfab, including a lamp, desk console, freighter controls, and a robotic arm, which I later took apart to add extra detail. Another incredibly helpful tool was the Random Flow add-on by BlenderGuppy, which made adding sci-fi elements much easier. Lastly, I pulled in some models from my older sci-fi and cyberpunk projects to round things out.Kitbashing Once I had the overall shape I was aiming for, I moved on to kitbashing to pack in as much detail as possible. There wasn't any strict method to the madness; I simply picked assets I liked, whether it was a set of pipes, vents, or even a random shape that just worked in the sci-fi context. I focused first on kitbashing the front structure, and used the Random Flow add-on to fill in areas where I didn't kitbash manually. Then I moved on to the other collections, following the same process.The freighter was the final piece of the puzzle, and I knew it was going to be a challenge. Part of me wanted to model it entirely from scratch, but the more practical side knew I could save a lot of time by sticking with my usual method. So I modeled the main shapes myself, then kitbashed the details to bring it to life. I also grabbed some crates from Sketchfab to fill out the scene.Texturing This part was easily my favorite, and there was no shortcut here. I had to meticulously create each material myself. Well, I did use PBR materials downloaded from CGAmbient as a base, but I spent a lot of time tweaking and editing them to get everything just right.Texturing has always been my favorite stage when building scenes like this. Many artists prefer external tools like Substance 3D Painter (which I did use for some of the models), but I've learned so much about procedural texturing, especially from RyanKingArt, that I couldn't let it go. It's such a flexible and rewarding approach, and I love pushing it as far as I can.I wanted most of the colors in the scene to be dark, but I did keep the original color of the pipes and the pillars, just to add a little bit of vibrance to the scene. I also wanted the overall texture to be very rough and grungy. One of the biggest helps in achieving this was using the Grunge Maps from Substance 3D Painter. I found a way to extract them into Blender, and it helped.A major tool during the texturing phase was Jsplacement, which I used to procedurally generate sci-fi grids and plates. This was the icing on the cake for adding intricate details. Whenever an area felt too flat, I applied bump maps with these grids and panels to bring the materials to life. For example, both the lamp pole and the entire black metal material feature these Jsplacement Maps.Lighting For this, I didn't do anything fancy. I knew the scene was in a high altitude, so I looked for HDRI with a cloudless sky, and I boosted the saturation up a little to give it that high altitude look.Post-Production The rendering phase was challenging since I was working on a low-end laptop. I couldn't render the entire scene all at once, so I broke it down by collections and rendered them as separate layers. Then, I composited the layers together in post-production. I'm not big on heavy post-work, so I kept it simple, mostly tweaking brightness and saturation on my phone. That's about it for the post-production process.Conclusion The entire project took me 10 days to complete, working at least four hours each day. Although I've expressed my love for texturing, my favorite part of this project was the detailing and kitbashing. I really enjoyed piecing all the small details together. The most challenging part was deciding which assets to use and where to place them. I had a lot of greebles to choose from, but I'm happy with the ones I selected; they felt like a perfect fit for the scene.I know kitbashing sometimes gets a negative reputation in the 3D community, but I found it incredibly relieving. Honestly, this project wouldn't have come together without it, so I fully embraced the process.I'm excited to keep making projects like this. The world of 3D art is truly an endless and vast realm, and I encourage every artist like me to keep exploring it, one project at a time.Denys Molokwu, 3D Artist

News Health & Medicine A genetic test may predict which weight loss drugs work best for patients Clinical trials show people with different “genetic scores” lose more weight on specific drugs Genetics tests may help predict whether GLP-1 drugs, such as semaglutide and liraglutide, or a different type of medication may work better for some people. © Obesity Action Coalition By Tina Hesman Saey June 13, 2025 at 9:00 am People trying to lose weight often count calories, carbs, steps and reps and watch the scales. Soon, they may have another number to consider: a genetic score indicating how many calories a person needs to feel full during a meal. This score may help predict whether someone will lose more weight on the drugs liraglutide or phentermine-topiramate, researchers report June 6 in Cell Metabolism. A separate study, posted to medRXiv.org in November, suggests that individuals with a higher genetic propensity for obesity benefit less from semaglutide compared to those with a lower genetic predisposition. Sign up for our newsletter We summarize the week's scientific breakthroughs every Thursday.

IntroductionHi, my name is Pankaj Kholiya, and I am a Senior 3D Character Artist. I've been working in the game industry for the past 8 years. I worked on titles like Call of Duty: Black Ops 6, That Christmas, Ghost of Tsushima Director's Cut, Star Wars: Outlaws, Alan Wake 2, Street Fighter 6, and many more. Currently, I'm working as a freelancer for the gaming and cinematics industry.Since my last interview, I made a few personal works, was a part of a Netflix movie, That Christmas, and worked with Platige on Star Wars: Outlaws and Call of Duty: Black Ops 6 cinematic.The Breakdancing Clothing ProjectIt all started when I witnessed a dance battle that a friend organized. It was like watching Step Up live. There, I got the inspiration to create a break dancer. I started by gathering different references from the internet. I found one particular image on Pinterest and decided to recreate it in 3D.At first, the idea was to create the outfit in one pose, but along the way, I also decided to create a dancing version of the character and explore Unreal Engine. Here is the ref I used for the dancing version:Getting StartedFor the upcoming talents, I'll try to describe my process in a few points. Even before starting Marvelous Designer, I made sure to have my base character ready for animation and simulation. This time, I decided to use the MetaHuman creator for the base due to its high-quality textures and materials. My primary focus was on the clothing, so using MetaHuman saved a lot of time.After I was satisfied with how my MetaHuman looked, I took it to Mixamo to get some animations. I was really impressed by how good the animations worked on the MetaHuman. Once I had the animations, I took the animation into Marvelous Designer and simulated the clothes.For the posed character, I adjusted the rig to match the pose like the reference and used the same method as in this tutorial to pose the character:ClothingFor this particular project, I didn't focus on the topology as it was just for a single render. I just packed the UVs in Marvelous Designer, exported the quad mesh from Marvelous Designer, subdivided it a few times, and started working on the detailing part in ZBrush.For the texture, I used the low-division mesh from the ZBrush file, as I already had the UVs on it. I then baked the normal and other maps on it and took it to Substance 3D Painter.AnimationThere are multiple ways to animate the metahuman character. For this project, I've used Mixamo. I imported my character into Mixamo, selected the animation I liked, and exported it. After that, I just imported it to Marvelous Designer and hit the simulation button. You can check my previous breakdown for the Mixamo pipeline.Once happy with the result, I exported the simulated cloth as an Alembic to Unreal Engine. Tutorial for importing clothes into Unreal Engine:Lighting & RenderingThe main target was to match the lighting closely to the reference. This was my first project in Unreal Engine, so I wanted to explore the lighting and see how far I could go with it. Being new to the Unreal Engine, I went through a lot of tutorials. Here are the lights I've used for the posed version:For the dancing version, I've created a stage like the ref from the Step Up movie: Some tips I found useful for the rendering are in the video below:ConclusionAt first, I had a clear direction for this project and was confident in my skills to tackle the art aspect of it. But things changed when I dived into Unreal Engine for my presentation. More than half the time on this project went into learning and getting used to Unreal Engine. I don't regret a single second I invested in Unreal, as it was a new experience. It took around 15 days to wrap this one up.The lesson I learned is that upgrading your knowledge and learning new things will help you grow as an artist in the long run. Approaching how you make an artwork has changed a lot ever since I started 3D, and adapting to the changing art environment is a good thing. Here are some recommendations if you are interested in learning Unreal Engine.Pankaj Kholiya, Senior 3D Character ArtistInterview conducted by Amber Rutherford

8 Sage Green Color Palettes You’ve Got to Experience In this article:See more ▼Post may contain affiliate links which give us commissions at no cost to you.There’s something undeniably calming about sage green that makes it one of my absolute favorite colors to work with as a designer. This muted, earthy hue has this incredible ability to ground a space while still feeling fresh and contemporary. Whether you’re working on a branding project, designing an interior space, or creating digital content, sage green offers a versatility that few colors can match. What I love most about sage green is how it bridges the gap between trendy and timeless. It’s not going anywhere anytime soon, and honestly, I don’t think it ever should. This sophisticated color has been quietly revolutionizing design palettes across every industry, and today I’m excited to share eight of my favorite sage green color combinations that will elevate your next project. 👋 Psst... Did you know you can get unlimited downloads of 59,000+ fonts and millions of other creative assets for just $16.95/mo? Learn more »The 8 Most Inspiring Sage Green Color Palettes 1. Garden Fresh #D2E5C4 #B2C69E #95B07B #79955D #5A743C Download this color palette 735×1102 Pinterest image 2160×3840 Vertical wallpaper 900×900 Square 3840×2160 4K Wallpaper This monochromatic sage palette is pure perfection for anyone wanting to create depth without complexity. I use this combination constantly in botanical-themed projects because it captures every shade of green you’d find in a thriving garden. The progression from light to dark creates natural hierarchy, making it incredibly functional for both print and digital work. 2. Misty Morning #BDC9BB #ACBAA1 #B2C1A2 #A4B1A0 #ADC3B7 Download this color palette 735×1102 Pinterest image 2160×3840 Vertical wallpaper 900×900 Square 3840×2160 4K Wallpaper When I need something soft and ethereal, this is my go-to palette. These gentle sage tones remind me of early morning fog rolling over hills. It’s perfect for wellness brands, spa environments, or any project that needs to evoke tranquility and peace. The subtle variations create interest without ever feeling overwhelming. 3. Harvest Moon #9AAB89 #647056 #D6C388 #F8C565 Download this color palette 735×1102 Pinterest image 2160×3840 Vertical wallpaper 900×900 Square 3840×2160 4K Wallpaper Get 300+ Fonts for FREEEnter your email to download our 100% free "Font Lover's Bundle". For commercial & personal use. No royalties. No fees. No attribution. 100% free to use anywhere. The combination of sage green with warm golds creates magic every single time. This palette captures that perfect autumn moment when the light hits everything just right. I love using this for brands that want to feel both grounded and optimistic – it’s earthy sophistication with a sunny disposition. 4. Moody Botanical #4D5D42 #6A894B #8DA67E #9B999A #C6B5DF Download this color palette 735×1102 Pinterest image 2160×3840 Vertical wallpaper 900×900 Square 3840×2160 4K Wallpaper For projects that need a bit more drama, this palette delivers beautifully. The deeper sage tones paired with that unexpected lavender create intrigue without losing the calming essence of green. I find this combination works wonderfully for upscale restaurants or luxury lifestyle brands that want to feel approachable yet refined. 5. Countryside Charm #A3AC9A #8A9A5B #93A395 #748B74 #827D67 Download this color palette 735×1102 Pinterest image 2160×3840 Vertical wallpaper 900×900 Square 3840×2160 4K Wallpaper This palette feels like a walk through the English countryside – all rolling hills and weathered stone walls. The mix of sage greens with those earthy undertones creates incredible depth. I use this combination for projects that need to feel established and trustworthy, like financial services or heritage brands. 6. Industrial Farmhouse Zen #CED3D2 #3F5054 #6F675E #9CAB86 #C8CAB5 Download this color palette 735×1102 Pinterest image 2160×3840 Vertical wallpaper 900×900 Square 3840×2160 4K Wallpaper The marriage of sage green with industrial grays might seem unexpected, but it creates this incredibly sophisticated modern aesthetic. This palette is perfect for tech companies or architectural firms that want to feel innovative yet grounded. The sage adds warmth to what could otherwise be cold, sterile colors. 7. Desert Sage #9AAB89 #B2AC88 #A06464 #8C909C #C9AD99 Download this color palette 735×1102 Pinterest image 2160×3840 Vertical wallpaper 900×900 Square 3840×2160 4K Wallpaper Inspired by the American Southwest, this palette combines sage with dusty terra cottas and warm beiges. There’s something so comforting about these colors together – they feel like sunset in the desert. I love using this for hospitality brands or any project that wants to evoke adventure and warmth. 8. Forest Floor #B2C69E #ACB6A6 #5B7553 #745000 #462800 Download this color palette 735×1102 Pinterest image 2160×3840 Vertical wallpaper 900×900 Square 3840×2160 4K Wallpaper This rich, earthy combination takes sage green into deeper territory with those gorgeous chocolate browns. It reminds me of walking through an old-growth forest where the light filters through layers of leaves. Perfect for organic brands, outdoor companies, or any project that wants to feel authentic and connected to nature. Why Sage Green Is Having Its Moment As someone who’s been watching color trends for years, I can tell you that sage green’s popularity isn’t just a passing fad. This color speaks to our collective desire for calm in an increasingly chaotic world. It’s the visual equivalent of taking a deep breath – immediately soothing and centering. The rise of biophilic design has also played a huge role in sage green’s dominance. As we spend more time indoors, we’re craving those connections to nature, and sage green delivers that botanical feeling without being overly literal. It’s nature-inspired design at its most sophisticated. What makes sage green particularly special is its incredible adaptability. Unlike brighter greens that can feel overwhelming or dated, sage green has this chameleon-like quality that allows it to work in virtually any context. Pair it with warm woods and it feels rustic; combine it with metallics and it becomes luxurious; add some crisp whites and suddenly it’s Scandinavian minimalism. Mastering Sage Green in Your Design Work The key to working with sage green successfully is understanding its undertones. Some sage greens lean more yellow, others more blue or gray. Recognizing these subtle differences will help you create more cohesive palettes and avoid color clashes that can make your work feel off. I always recommend testing your sage green palettes in different lighting conditions. What looks perfect on your computer screen might feel completely different in natural light or under warm artificial lighting. This is especially crucial for interior design projects or any work that will be viewed in physical spaces. When building palettes around sage green, I like to think about the mood I’m trying to create. For calm, peaceful vibes, I’ll pair it with other muted tones and plenty of white space. For something more energetic, I might add unexpected pops of coral or sunny yellow. The beauty of sage green is that it’s such a diplomatic color – it plays well with almost everything. Sage Green Across Different Design Applications Branding and Logo Design In branding work, sage green communicates reliability, growth, and environmental consciousness without hitting people over the head with it. I love using it for wellness companies, sustainable brands, and professional services that want to feel approachable. The key is pairing it with typography that reinforces your brand personality – clean sans serifs for modern feels, or elegant serifs for more traditional approaches. Interior Spaces Sage green walls have become incredibly popular, and for good reason. The color creates an instant sense of calm while still feeling current. I particularly love using darker sage greens in dining rooms or bedrooms where you want that cozy, enveloping feeling. Lighter sages work beautifully in kitchens and bathrooms where you want freshness without the sterility of pure white. Digital Design For websites and apps, sage green offers a refreshing alternative to the blues and grays that dominate digital design. It’s easy on the eyes, which makes it perfect for apps focused on wellness, meditation, or any platform where users will spend extended time. Just be mindful of accessibility – always test your sage green backgrounds with various text colors to ensure proper contrast ratios. Product Design The natural, organic feeling of sage green makes it perfect for product packaging, especially in the beauty, food, and wellness sectors. It communicates quality and naturalness without feeling overly earthy or crunchy. I’ve seen it work beautifully on everything from skincare packaging to high-end kitchen appliances. The Psychology Behind Sage Green’s Appeal Color psychology tells us that green represents growth, harmony, and balance – all things we desperately need in our modern lives. But sage green takes these positive associations and adds sophistication. It’s green without the intensity, nature without the rawness. There’s also something inherently honest about sage green. It doesn’t try too hard or demand attention the way brighter colors do. This authenticity resonates with consumers who are increasingly skeptical of brands that feel forced or overly polished. Sage green whispers where other colors shout, and sometimes that’s exactly what your message needs. Looking Forward: Sage Green’s Staying Power While I can’t predict the future, I’m confident that sage green will remain relevant for years to come. It hits all the right notes for contemporary design – it’s calming without being boring, natural without being literal, and sophisticated without being pretentious. The color also photographs beautifully, which matters more than ever in our Instagram-driven world. Whether it’s a sage green accent wall or a product shot featuring sage packaging, this color translates perfectly to social media, helping brands create that coveted “aesthetic” that drives engagement. As we continue to prioritize wellness and sustainability in design, sage green offers the perfect visual shorthand for these values. It’s a color that makes people feel good, and in a world that often doesn’t, that’s incredibly powerful. Bringing It All Together These eight sage green palettes represent just the beginning of what’s possible with this incredible color. Whether you’re drawn to the monochromatic serenity of Garden Fresh or the unexpected sophistication of Industrial Zen, there’s a sage green palette that can elevate your next project. The secret to success with sage green is trusting its natural elegance. Don’t feel like you need to overstyling or complicate things – sage green’s beauty lies in its understated sophistication. Let it be the calm, confident foundation that allows other elements of your design to shine. So go ahead and embrace the sage green revolution. Your designs (and your stress levels) will thank you for it. After all, in a world full of visual noise, sometimes the most powerful statement you can make is a quiet one. Riley Morgan Riley Morgan is a globe-trotting graphic designer with a sharp eye for color, typography, and intuitive design. They are a color lover and blend creativity with culture, drawing inspiration from cities, landscapes, and stories around the world. When they’re not designing sleek visuals for clients, they’re blogging about trends, tools, and the art of making design feel like home—wherever that may be. 8 Stunning Sunset Color PalettesThere’s something absolutely magical about watching the sun dip below the horizon, painting the sky in breathtaking hues that seem...10 Warm Color Palettes That’ll Brighten Your DayThere’s nothing quite like the embracing quality of warm colors to make a design feel inviting and alive. As someone...10 Luxurious Jewel Tone Color PalettesAs a designer who’s always searching for color combinations that exude sophistication and richness, I find myself constantly returning to...

Meta's CTO, Palantir's CTO, and OpenAI's chief product officer are being appointed as lieutenant colonels in America's Army Reserve, reports The Register. (Along with OpenAI's former chief revenue officer). They've all signed up for Detachment 201: Executive Innovation Corps, "an effort to recruit senior tech executives to serve part-time in the Army Reserve as senior advisors," according to the official statement. "In this role they will work on targeted projects to help guide rapid and scalable tech solutions to complex problems..." "Our primary role will be to serve as technical experts advising the Army's modernization efforts," [Meta CTO Andrew Bosworth] said on X... As for Open AI's involvement, the company has been building its ties with the military-technology complex for some years now. Like Meta, OpenAI is working with Anduril on military ideas and last year scandalized some by watering down its past commitment to developing non-military products only. The Army wasn't answering questions on Friday but an article referenced by [OpenAI Chief Product Officer Kevin] Weil indicated that the four will have to serve a minimum of 120 hours a year, can work remotely, and won't have to pass basic training... "America wins when we unite the dynamism of American innovation with the military's vital missions," [Palantir CTO Shyam] Sankar said on X. "This was the key to our triumphs in the 20th century. It can help us win again. I'm humbled by this new opportunity to serve my country, my home, America." Read more of this story at Slashdot.

Coolers News Scythe Solvency Update, "Scycopter" Liquid Cooler, New $45 Air CoolersJune 10, 2025Last Updated: 2025-06-10We looked at Scythe’s Scycopter liquid cooler, Magoroku air cooler, Big Shuriken 4, and moreThe HighlightsScythe showed off its liquid cooler, which is currently going by the working name “Scycopter”The Magoruku is a $50 CPU cooler that’s supposed to be relatively high performing with 6x6mm heat pipes coupled with a nickel-plated copper cold plateWe talked to Scythe about the news of its European branch closing downTable of ContentsAutoTOC Grab a GN Tear-Down Toolkit to support our AD-FREE reviews and IN-DEPTH testing while also getting a high-quality, highly portable 10-piece toolkit that was custom designed for use with video cards for repasting and water block installation. Includes a portable roll bag, hook hangers for pegboards, a storage compartment, and instructional GPU disassembly cards.IntroWe visited Scythe’s booth at Computex 2025 and the company showed off several new coolers, including a mockup of a liquid cooler. Our visit comes off the heels of the news that Scythe will be closing its European branch, which we discussed with the company.Editor's note: This was originally published on May 22, 2025 as a video. This content has been adapted to written format for this article and is unchanged from the original publication.CreditsHostSteve BurkeCamera, Video EditingMike GaglioneVitalii MakhnovetsWriting, Web EditingJimmy ThangScythe Liquid CoolerTo our knowledge, we saw Scythe’s first liquid cooler at the show. We spoke with Kitagawa-san, lead designer at Scythe, who told us that he spent about the last year studying liquid coolers. The company also showed us a 3D-printed prototype peg with a piece of tape underneath it, which allows you to essentially stick it to any fan you want. A fan can then socket on top of the cooler and be angled to shoot air down toward the VRM or RAM, etc. The working name of the liquid cooler is the “Scycopter,” which is really cool and is a combination of Scythe and helicopter. Currently, the radiator thickness is pretty standard at 27mm, but that might change. The standard pump block will have an option that will allow you to install a fan on top of it. For the fins, the pitch is .1mm. That makes them pretty close together. Scythe also tells us that the total height of the copper coldplate is 1.6mm. Magoruku Grab a GN15 Large Anti-Static Modmat to celebrate our 15th Anniversary and for a high-quality PC building work surface. The Modmat features useful PC building diagrams and is anti-static conductive. Purchases directly fund our work! (or consider a direct donation or a Patreon contribution!)We showed Scythe’s Magoruku CPU cooler at last year’s Computex, but it’s coming out now. It’s supposed to be $50, but the company tells us that it might be able to bring it down to $44 in the US depending on market conditions. The Magoruku is supposed to be a relatively high-performing, mid-range/budget cooler. Scythe is going with a flat nickel-plated copper for its cold plate coupled with 6x6mm heat pipes. The company is using 2x120mm “Wonder Tornado” fans as Scythe calls them. They are 25mm-thick fans and use metal brackets to adjust the fan height. Mugen 6 TUFThe Mugen 6 TUF is an ASUS-themed version of the CPU cooler. Big Shuriken 4Scythe also showed off its Big Shuriken 4 CPU cooler, which the company also showed last year, but is now about final. It has cut-outs on the side of the fan, which Scythe says helps with performance as it allows air to escape from the sides. One of the things that Scythe is trying to figure out with the Big Shuriken 4 is whether to make it all black or ARGB. Scythe Closing Its European Branch Visit our Patreon page to contribute a few dollars toward this website's operation (or consider a direct donation or buying something from our GN Store!) Additionally, when you purchase through links to retailers on our site, we may earn a small affiliate commission.In regards to Scythe’s closed European branch, it sounds like the company is restructuring and moving operations to Taiwan. Scythe tells us it will still ship and sell to European customers.

AI researchers have recently been asking themselves a version of the question, “Is that really Zuck?”As first reported by Bloomberg, the Meta CEO has been personally asking top AI talent to join his new “superintelligence” AI lab and reboot Llama. His recruiting process typically goes like this: a cold outreach via email or WhatsApp that cites the recruit’s work history and requests a 15-minute chat. Dozens of researchers have gotten these kinds of messages at Google alone. For those who do agree to hear his pitch (amazingly, not all of them do), Zuckerberg highlights the latitude they’ll have to make risky bets, the scale of Meta’s products, and the money he’s prepared to invest in the infrastructure to support them. He makes clear that this new team will be empowered and sit with him at Meta’s headquarters, where I’m told the desks have already been rearranged for the incoming team.Most of the headlines so far have focused on the eye-popping compensation packages Zuckerberg is offering, some of which are well into the eight-figure range. As I’ve covered before, hiring the best AI researcher is like hiring a star basketball player: there are very few of them, and you have to pay up. Case in point: Zuckerberg basically just paid 14 Instagrams to hire away Scale AI CEO Alexandr Wang. It’s easily the most expensive hire of all time, dwarfing the billions that Google spent to rehire Noam Shazeer and his core team from Character.AI (a deal Zuckerberg passed on). “Opportunities of this magnitude often come at a cost,” Wang wrote in his note to employees this week. “In this instance, that cost is my departure.”Zuckerberg’s recruiting spree is already starting to rattle his competitors. The day before his offer deadline for some senior OpenAI employees, Sam Altman dropped an essay proclaiming that “before anything else, we are a superintelligence research company.” And after Zuckerberg tried to hire DeepMind CTO Koray Kavukcuoglu, he was given a larger SVP title and now reports directly to Google CEO Sundar Pichai. I expect Wang to have the title of “chief AI officer” at Meta when the new lab is announced. Jack Rae, a principal researcher from DeepMind who has signed on, will lead pre-training. Meta certainly needs a reset. According to my sources, Llama has fallen so far behind that Meta’s product teams have recently discussed using AI models from other companies (although that is highly unlikely to happen). Meta’s internal coding tool for engineers, however, is already using Claude. While Meta’s existing AI researchers have good reason to be looking over their shoulders, Zuckerberg’s $14.3 billion investment in Scale is making many longtime employees, or Scaliens, quite wealthy. They were popping champagne in the office this morning. Then, Wang held his last all-hands meeting to say goodbye and cried. He didn’t mention what he would be doing at Meta. I expect his new team will be unveiled within the next few weeks after Zuckerberg gets a critical number of members to officially sign on. Tim Cook. Getty Images / The VergeApple’s AI problemApple is accustomed to being on top of the tech industry, and for good reason: the company has enjoyed a nearly unrivaled run of dominance. After spending time at Apple HQ this week for WWDC, I’m not sure that its leaders appreciate the meteorite that is heading their way. The hubris they display suggests they don’t understand how AI is fundamentally changing how people use and build software.Heading into the keynote on Monday, everyone knew not to expect the revamped Siri that had been promised the previous year. Apple, to its credit, acknowledged that it dropped the ball there, and it sounds like a large language model rebuild of Siri is very much underway and coming in 2026.The AI industry moves much faster than Apple’s release schedule, though. By the time Siri is perhaps good enough to keep pace, it will have to contend with the lock-in that OpenAI and others are building through their memory features. Apple and OpenAI are currently partners, but both companies want to ultimately control the interface for interacting with AI, which puts them on a collision course. Apple’s decision to let developers use its own, on-device foundational models for free in their apps sounds strategically smart, but unfortunately, the models look far from leading. Apple ran its own benchmarks, which aren’t impressive, and has confirmed a measly context window of 4,096 tokens. It’s also saying that the models will be updated alongside its operating systems — a snail’s pace compared to how quickly AI companies move. I’d be surprised if any serious developers use these Apple models, although I can see them being helpful to indie devs who are just getting started and don’t want to spend on the leading cloud models. I don’t think most people care about the privacy angle that Apple is claiming as a differentiator; they are already sharing their darkest secrets with ChatGPT and other assistants. Some of the new Apple Intelligence features I demoed this week were impressive, such as live language translation for calls. Mostly, I came away with the impression that the company is heavily leaning on its ChatGPT partnership as a stopgap until Apple Intelligence and Siri are both where they need to be. AI probably isn’t a near-term risk to Apple’s business. No one has shipped anything close to the contextually aware Siri that was demoed at last year’s WWDC. People will continue to buy Apple hardware for a long time, even after Sam Altman and Jony Ive announce their first AI device for ChatGPT next year. AR glasses aren’t going mainstream anytime soon either, although we can expect to see more eyewear from Meta, Google, and Snap over the coming year. In aggregate, these AI-powered devices could begin to siphon away engagement from the iPhone, but I don’t see people fully replacing their smartphones for a long time. The bigger question after this week is whether Apple has what it takes to rise to the occasion and culturally reset itself for the AI era. I would have loved to hear Tim Cook address this issue directly, but the only interview he did for WWDC was a cover story in Variety about the company’s new F1 movie.ElsewhereAI agents are coming. I recently caught up with Databricks CEO Ali Ghodsi ahead of his company’s annual developer conference this week in San Francisco. Given Databricks’ position, he has a unique, bird’s-eye view of where things are headed for AI. He doesn’t envision a near-term future where AI agents completely automate real-world tasks, but he does predict a wave of startups over the next year that will come close to completing actions in areas such as travel booking. He thinks humans will need (and want) to approve what an agent does before it goes off and completes a task. “We have most of the airplanes flying automated, and we still want pilots in there.”Buyouts are the new normal at Google. That much is clear after this week’s rollout of the “voluntary exit program” in core engineering, the Search organization, and some other divisions. In his internal memo, Search SVP Nick Fox was clear that management thinks buyouts have been successful in other parts of the company that have tried them. In a separate memo I saw, engineering exec Jen Fitzpatrick called the buyouts an “opportunity to create internal mobility and fresh growth opportunities.” Google appears to be attempting a cultural reset, which will be a challenging task for a company of its size. We’ll see if it can pull it off. Evan Spiegel wants help with AR glasses. I doubt that his announcement that consumer glasses are coming next year was solely aimed at AR developers. Telegraphing the plan and announcing that Snap has spent $3 billion on hardware to date feels more aimed at potential partners that want to make a bigger glasses play, such as Google. A strategic investment could help insulate Snap from the pain of the stock market. A full acquisition may not be off the table, either. When he was recently asked if he’d be open to a sale, Spiegel didn’t shut it down like he always has, but instead said he’d “consider anything” that helps the company “create the next computing platform.”Link listMore to click on:If you haven’t already, don’t forget to subscribe to The Verge, which includes unlimited access to Command Line and all of our reporting.As always, I welcome your feedback, especially if you’re an AI researcher fielding a juicy job offer. You can respond here or ping me securely on Signal.Thanks for subscribing.See More:

Join the event trusted by enterprise leaders for nearly two decades. VB Transform brings together the people building real enterprise AI strategy. Learn more When DeepSeek released its R1 model this January, it wasn’t just another AI announcement. It was a watershed moment that sent shockwaves through the tech industry, forcing industry leaders to reconsider their fundamental approaches to AI development. What makes DeepSeek’s accomplishment remarkable isn’t that the company developed novel capabilities; rather, it was how it achieved comparable results to those delivered by tech heavyweights at a fraction of the cost. In reality, DeepSeek didn’t do anything that hadn’t been done before; its innovation stemmed from pursuing different priorities. As a result, we are now experiencing rapid-fire development along two parallel tracks: efficiency and compute.  As DeepSeek prepares to release its R2 model, and as it concurrently faces the potential of even greater chip restrictions from the U.S., it’s important to look at how it captured so much attention. Engineering around constraints DeepSeek’s arrival, as sudden and dramatic as it was, captivated us all because it showcased the capacity for innovation to thrive even under significant constraints. Faced with U.S. export controls limiting access to cutting-edge AI chips, DeepSeek was forced to find alternative pathways to AI advancement. While U.S. companies pursued performance gains through more powerful hardware, bigger models and better data, DeepSeek focused on optimizing what was available. It implemented known ideas with remarkable execution — and there is novelty in executing what’s known and doing it well. This efficiency-first mindset yielded incredibly impressive results. DeepSeek’s R1 model reportedly matches OpenAI’s capabilities at just 5 to 10% of the operating cost. According to reports, the final training run for DeepSeek’s V3 predecessor cost a mere $6 million — which was described by former Tesla AI scientist Andrej Karpathy as “a joke of a budget” compared to the tens or hundreds of millions spent by U.S. competitors. More strikingly, while OpenAI reportedly spent $500 million training its recent “Orion” model, DeepSeek achieved superior benchmark results for just $5.6 million — less than 1.2% of OpenAI’s investment. If you get starry eyed believing these incredible results were achieved even as DeepSeek was at a severe disadvantage based on its inability to access advanced AI chips, I hate to tell you, but that narrative isn’t entirely accurate (even though it makes a good story). Initial U.S. export controls focused primarily on compute capabilities, not on memory and networking — two crucial components for AI development. That means that the chips DeepSeek had access to were not poor quality chips; their networking and memory capabilities allowed DeepSeek to parallelize operations across many units, a key strategy for running their large model efficiently. This, combined with China’s national push toward controlling the entire vertical stack of AI infrastructure, resulted in accelerated innovation that many Western observers didn’t anticipate. DeepSeek’s advancements were an inevitable part of AI development, but they brought known advancements forward a few years earlier than would have been possible otherwise, and that’s pretty amazing. Pragmatism over process Beyond hardware optimization, DeepSeek’s approach to training data represents another departure from conventional Western practices. Rather than relying solely on web-scraped content, DeepSeek reportedly leveraged significant amounts of synthetic data and outputs from other proprietary models. This is a classic example of model distillation, or the ability to learn from really powerful models. Such an approach, however, raises questions about data privacy and governance that might concern Western enterprise customers. Still, it underscores DeepSeek’s overall pragmatic focus on results over process. The effective use of synthetic data is a key differentiator. Synthetic data can be very effective when it comes to training large models, but you have to be careful; some model architectures handle synthetic data better than others. For instance, transformer-based models with mixture of experts (MoE) architectures like DeepSeek’s tend to be more robust when incorporating synthetic data, while more traditional dense architectures like those used in early Llama models can experience performance degradation or even “model collapse” when trained on too much synthetic content. This architectural sensitivity matters because synthetic data introduces different patterns and distributions compared to real-world data. When a model architecture doesn’t handle synthetic data well, it may learn shortcuts or biases present in the synthetic data generation process rather than generalizable knowledge. This can lead to reduced performance on real-world tasks, increased hallucinations or brittleness when facing novel situations.  Still, DeepSeek’s engineering teams reportedly designed their model architecture specifically with synthetic data integration in mind from the earliest planning stages. This allowed the company to leverage the cost benefits of synthetic data without sacrificing performance. Market reverberations Why does all of this matter? Stock market aside, DeepSeek’s emergence has triggered substantive strategic shifts among industry leaders. Case in point: OpenAI. Sam Altman recently announced plans to release the company’s first “open-weight” language model since 2019. This is a pretty notable pivot for a company that built its business on proprietary systems. It seems DeepSeek’s rise, on top of Llama’s success, has hit OpenAI’s leader hard. Just a month after DeepSeek arrived on the scene, Altman admitted that OpenAI had been “on the wrong side of history” regarding open-source AI.  With OpenAI reportedly spending $7 to 8 billion annually on operations, the economic pressure from efficient alternatives like DeepSeek has become impossible to ignore. As AI scholar Kai-Fu Lee bluntly put it: “You’re spending $7 billion or $8 billion a year, making a massive loss, and here you have a competitor coming in with an open-source model that’s for free.” This necessitates change. This economic reality prompted OpenAI to pursue a massive $40 billion funding round that valued the company at an unprecedented $300 billion. But even with a war chest of funds at its disposal, the fundamental challenge remains: OpenAI’s approach is dramatically more resource-intensive than DeepSeek’s. Beyond model training Another significant trend accelerated by DeepSeek is the shift toward “test-time compute” (TTC). As major AI labs have now trained their models on much of the available public data on the internet, data scarcity is slowing further improvements in pre-training. To get around this, DeepSeek announced a collaboration with Tsinghua University to enable “self-principled critique tuning” (SPCT). This approach trains AI to develop its own rules for judging content and then uses those rules to provide detailed critiques. The system includes a built-in “judge” that evaluates the AI’s answers in real-time, comparing responses against core rules and quality standards. The development is part of a movement towards autonomous self-evaluation and improvement in AI systems in which models use inference time to improve results, rather than simply making models larger during training. DeepSeek calls its system “DeepSeek-GRM” (generalist reward modeling). But, as with its model distillation approach, this could be considered a mix of promise and risk. For example, if the AI develops its own judging criteria, there’s a risk those principles diverge from human values, ethics or context. The rules could end up being overly rigid or biased, optimizing for style over substance, and/or reinforce incorrect assumptions or hallucinations. Additionally, without a human in the loop, issues could arise if the “judge” is flawed or misaligned. It’s a kind of AI talking to itself, without robust external grounding. On top of this, users and developers may not understand why the AI reached a certain conclusion — which feeds into a bigger concern: Should an AI be allowed to decide what is “good” or “correct” based solely on its own logic? These risks shouldn’t be discounted. At the same time, this approach is gaining traction, as again DeepSeek builds on the body of work of others (think OpenAI’s “critique and revise” methods, Anthropic’s constitutional AI or research on self-rewarding agents) to create what is likely the first full-stack application of SPCT in a commercial effort. This could mark a powerful shift in AI autonomy, but there still is a need for rigorous auditing, transparency and safeguards. It’s not just about models getting smarter, but that they remain aligned, interpretable, and trustworthy as they begin critiquing themselves without human guardrails. Moving into the future So, taking all of this into account, the rise of DeepSeek signals a broader shift in the AI industry toward parallel innovation tracks. While companies continue building more powerful compute clusters for next-generation capabilities, there will also be intense focus on finding efficiency gains through software engineering and model architecture improvements to offset the challenges of AI energy consumption, which far outpaces power generation capacity.  Companies are taking note. Microsoft, for example, has halted data center development in multiple regions globally, recalibrating toward a more distributed, efficient infrastructure approach. While still planning to invest approximately $80 billion in AI infrastructure this fiscal year, the company is reallocating resources in response to the efficiency gains DeepSeek introduced to the market. Meta has also responded, With so much movement in such a short time, it becomes somewhat ironic that the U.S. sanctions designed to maintain American AI dominance may have instead accelerated the very innovation they sought to contain. By constraining access to materials, DeepSeek was forced to blaze a new trail. Moving forward, as the industry continues to evolve globally, adaptability for all players will be key. Policies, people and market reactions will continue to shift the ground rules — whether it’s eliminating the AI diffusion rule, a new ban on technology purchases or something else entirely. It’s what we learn from one another and how we respond that will be worth watching. Jae Lee is CEO and co-founder of TwelveLabs. Daily insights on business use cases with VB Daily If you want to impress your boss, VB Daily has you covered. We give you the inside scoop on what companies are doing with generative AI, from regulatory shifts to practical deployments, so you can share insights for maximum ROI. Read our Privacy Policy Thanks for subscribing. Check out more VB newsletters here. An error occured.

In Brief Posted: 11:46 AM PDT · June 14, 2025 Image Credits:Matthias Balk/picture alliance / Getty Images Google reportedly plans to cut ties with Scale AI Meta’s big investment in Scale AI may be giving some of the startup’s customers pause. Reuters reports that Google had planned to pay Scale $200 million this year but is now having conversations with its competitors and planning to cut ties. Microsoft is also reportedly looking to pull back, and OpenAI supposedly made a similar decision months ago, although its CFO said the company will continue working with Scale as one of many vendors. Scale’s customers include self-driving car companies and the U.S. government, but Reuters says its biggest clients are generative AI companies seeking access to workers with specialized knowledge who can annotate data to train models. Google declined to comment on the report. A Scale spokesperson declined to comment on the company’s relationship with Google, but he told TechCrunch that Scale’s business remains strong, and that it will continue to operate as an independent company that safeguards its customers’ data. Earlier reports suggest that Meta invested $14.3 billion in Scale for a 49% stake in the company, with Scale CEO Alexandr Wang joining Meta to lead the company’s efforts to develop “superintelligence.” Topics AI, Google, Meta, Scale AI