Ke Ding

The Immersive KIND — drones x digital bodies x SAM 3D 🧠✨ Exploring how the physical body can be captured, segmented, and reconstructed into living 3D forms — ready to exist across worlds. From movement → to mesh → to machine intelligence. Tools used: 💥SAM 3 / SAM 3D 💥Midjourney 💥Kling 💥Krea 💥Canva #SAM3D #DigitalBodies #CreativeAI #SpatialComputing #AIArt #FutureOfMovement #TheImmersiveKIND #KadineJames #NewFormats Meta The Immersive KIND special thanks Campbell Orme

10

Ok! Another round up of neat things I have found within AI and design tooling this week. 1️⃣ MagicPath releases v2 The new version of Magic Path is fully agent-enabled now. Make one request and a team of AI agents fans out across your canvas, generating multiple screens and variations at once. Plug in Codex, Claude Code, or Cursor, connect your repo and the agents build alongside your teammates in real time. It also runs in the cloud. So your agents keep working when you close your laptop. Send a request from your phone on a walk, come back to a built prototype on your canvas. 2️⃣ Jitter I posted about this yesterday but worth the wrap up. It's an AI Motion design tool which should stop the After Effects headaches. Now you type what you want and Jitter builds it. The neat announcement yesterday was that you can now generate your own UI controls for what you're creating. 3️⃣ QuiverAI Quiver is the industry leading SVG generation tool. Their Arrow model generates and edits production-quality SVGs from text or image input. They recently announced integration with Paper so if you use Paper to design your UIs, icon generation got a whole lot easier. 4️⃣ Dessn Whether it's Figma, Pencil, Paper or Claude Design, they're all an abstraction layer on top of the code. Dessn aims to solve that by being the first AI design tool that allows you to build and design within your actual production codebase. Some really interesting ideas here, including abstracting the whole localhost problem for designers. They've just raised $6M. 5️⃣ AtlasGrid I love this tool. Coming soon. AtlasGrid lets you point any iOS app and it maps every screen and user path automatically. The designer use case is so useful: run it against your own shipped app and see every state, edge case, and flow your team has built, even the ones nobody documented. It's a complete audit of what your product actually looks like in production, not what it looks like in your Figma file. I've been waiting for something like this for some time. 6️⃣ Ramp's AI-First Design Culture Ramp has 1,500 people and 99% of them are using AI tools. Their design team runs Claude Code for specialised workflows and Anthropic even visited the office because Ramp was using Claude more than Anthropic itself was! Designer Tommy Geoco recently did an interview with Ramps' lead designer Diego Zaks and there are some great takeaways. It's a must-watch. Link in the comments. We're continuing to see experimentation in workflows. The closer I work with designers with this tech, the more I realise that AI is not going to replace human craft in this space at all. What it will do is accelerate our creativity and take away the monotonous. These tools are a great example of this, from allowing you to build directly in the codebase, to mapping out app flows in seconds, to generating your own custom UI for your tooling. The future for design looks bright. Links to all of these in the comments below. 

20

4 Comments

A timely post, given the new reporting from mobilegamer.biz, which highlighted that since the layoffs, Microsoft has allegedly mandated a 100% daily usage of AI on general tasks. Despite the mandate, AI adoption is low, apart from ChatGPT, possibly because staff and leadership are skeptical that the tools are actually helpful or more efficient.

7

2 Comments

🙌 Dropping a new tutorial on my new MiniMax Remover BMO node that harnesses the power of Minimax Object Remover code base which can now be installed directly from the ComfyUI Manager 1️⃣ My first attempt to bring a repo into comfy for a specific use case and man was it fun and a great challenge 🥵 In 2020 severely injured my left hand 🖐 in an accident when a building spike pierced through my median nerve. I lost the ability to feel a lot of sensation and pressure in the palm and fingers. This caused me to really have trouble with my traditional role as a VFX artist using Wacom pens and keyboards simultaneously and destroyed my ambidextrous abilities. 15 years of training and suddenly I was slower and at times unable to do what I used to do on an expert level 😢 Since then I have been learning AI as a way to recreate some of my previously learned crafts and exploring a new career in AI development techniques to help augment tasks I couldn't do as well or as fast. Here is a great example 👇 ⚫ The node allows fast inpainting results for complex shots and solves some of the age old issues with manual cleanup traditionally done in Flame or Nuke and spanning many days. 🎉 Using cursor and my growing python knowledge I was able to code this node in 2 days with the help of my new friend Gemini 2.5 and Claude 4.5 Sonnet 💻 I hope this tutorial helps users understand the power of AI outside the flashy videos we see as a practical application of traditional VFX use cases. Next stop is flame and Nuke integration! 🥂 #comfyui #ai #python #llm

394

28 Comments

#AGI isn’t arriving - it’s already manifesting in real-world workflows. Over three blog instalments, Genspark CEO Eric Jing shares how using hundreds of #AI tools daily convinced him: AGI is real and happening now. What You’ll Discover Across the Series: Part 1: Voting with Your Usage From 300+ AI tools tested to daily hands-on use, AGI reveals itself when you move from demo mode to co-creation mode. Part 2: What To Do Next A playbook: build personal projects, AI-native workflows, and iterate in tools like AI Sheets, Slides, and Deep Research. Part 3: Build AI-Native Teams Your critical edge in 2025? Teams that collaborate with AI, not just consume it. Disruptive intelligence demands disruptive workflows. Why This Matters for Leaders & Builders: 1. AGI is not magic—it’s consistent complexity. It emerges when models coordinate, reflect, and self-supervise across tasks. 2. Genspark’s Mixture‑of‑Agents (MoA) system already runs windows of domain-specialised agents collaborating on search, design, and finance. That’s real agentic intelligence at scale. 3. This isn’t a future recruiter prompt; it’s your competitive mindset upgrade today. The future isn’t about smarter prompts. It’s about building agentic systems that think, adapt, and evolve alongside you. Leaders who adopt this approach in 2025 won’t just survive; they’ll outpace the AI curve.

2

🚀 New Feature Update – True 3D Expansion in Contour Editing At EduCase we’re excited to share a major update in our contour editing toolkit: the introduction of a TPS-style 3D Expansion algorithm. Traditionally, margin expansions in 2D slice-by-slice editing often missed the true geometry of irregular volumes in 3D. This could lead to under-coverage in superior/inferior directions or inconsistencies compared to TPS-generated contours. Our new approach: ✅ Performs true 3D isotropic expansion in voxel space ✅ Uses direction-specific margins (e.g., 5 mm posterior, 10 mm anterior) ✅ Matches TPS behavior by including all slices intersecting the expansion volume ✅ Provides real-time feedback with progress indicators so users know exactly what’s happening What this means for clinicians and trainees: * More consistent contour expansions across all planes * A workflow that mirrors the accuracy and conservativeness of TPS systems * Better alignment between training tools and clinical planning environments This update was inspired directly by feedback from practicing doctors, and we’re thrilled to bring the tool closer to real-world clinical behavior. 🔬 Building tools that bridge the gap between education, research, and clinical practice is always our focus. This is another step forward in that journey! #EduCase #RadiationOncology #MedTech #HealthcareInnovation #3DExpansion #TPS

5

Bringing a coffee shop/fitness studio to life with AR ☕📱🎨 I’ve been experimenting with a new augmented reality setup at Lagree Core Strength + Espresso in Saskatoon. Grab a coffee, scan a QR code, and the artwork wakes up in the room. It’s simple for the viewer, but the effect transforms the space. This one’s especially fun because I got to bring some of my old graffiti sketches into a new dimension—literally. The textures, depth, and movement give it a gritty, immersive feel that fits the energy of the place. How it’s built (for the curious): •Runs fully in the browser — no app downloads needed •Built with open-source tools and Model Viewer •Optimized for indoor lighting and mixed surfaces •Lightweight assets for quick mobile loading Why it matters: AR like this bridges art, design, and technology in an approachable way. Businesses can refresh a space without changing the walls — it’s living artwork that evolves over time. If you’re in Saskatoon, stop by Lagree Core Strength + Espresso and see it in person. It looks even better in the space. ⸻ #augmentedreality #webar #spatialdesign #publicart #digitalart #YXE #Saskatoon #experientialdesign #modelviewer

5

2 Comments

Toronto producer Ahkilo (credits with Drake and Roy Woods) says when inspiration stalls, he turns to AI as a studio partner. Using Staccato, a Canadian AI platform that suggests melodies and chords from a musician’s own material, he calls it “like a friend in the studio — you still make the magic.” Across Canada, artists are experimenting with tools like Staccato and Grimes’s Elf.Tech, blurring the line between inspiration and automation. But as musician David Usher notes, “When there’s no manipulation between the prompt and the product, that’s where it gets tricky.” Ethicists like Jada Watson warn that many AI models rely on unconsented training data — yet she also sees how AI opens creative doors for people with disabilities. The real revolution isn’t AI replacing musicians — it’s AI expanding who gets to be one. https://lnkd.in/gETurUQZ

4

When they say “creative technology” — take it literally. I just experienced Signals a Creative Tech Expo by DigiBC - The Creative Technology Association of British Columbia , The Greater Vancouver International Film Festival Society & IM4 Media Lab and my mind is officially blown. 🤯 For ten days, Vancouver transformed into a playground for the future of storytelling, and I was completely immersed in it. This wasn't just an exhibition; it was a journey into the collective dreams of over 30 artists pushing the boundaries of art, AI, and immersive experience. I walked away forever changed by powerful works like: ✨ Pegah Tabassinejad "Entropic Fields of Displacement" – an eight-channel video installation that turns the simple act of walking into a profound meditation on migration and resilience. I cried. This was a very relatable art, as a migrant, who is navigating a new city and has had to sacrifice some things to come scale a startup; I felt every bit of the emotion this piece tapped into. ✨ Zachery Cameron Longboy’s “SHEMANSHE"** – a deeply personal, layered encounter with land, memory, and spirit that stayed with me long after I left the room. A true legend and a pioneer figure in the emergence and development of contemporary Indigenous performance-based video installation. ✨ Pinyao Liu & Keon Ju Maverick Lee ‘s "ReVerie"** – where whispered dreams morph into a 3D landscape you can explore. It felt like stepping inside someone’s subconscious. I got to contribute my imagination to the digital dreamscape. It was sublime. ✨ The incredible IM4 Media Lab Collective, who brought youth and Elders together in "Thunderbird Dreams," creating a living, interactive mural that celebrated Indigenous creativity. It was a powerful reminder of how technology can strengthen, not fragment, our communities and culture. And special thanks to tge phenomenal project assistant Sussan Yáñez for the work you do and the beautiful art you create. ✨🤍✨ We explored, we wondered, we connected Div Wang ✨🦋✨ The conversations were as boundary-pushing as the art itself – no need for any "buzzword" translations here! The community just *gets* it. A massive thanks to Jessica Vaira and the rest of the team at DigiBC, VIFF, and all the visionary artists and storytellers. You’ve created more than an event; you’ve built a welcoming space where art and technology collide to spark real human connection. Final thanks to Angel Wen for bringing this to my attention. #SignalsExpo #CreativeTech #VancouverTech #ImmersiveArt #AIArt #FutureOfStorytelling #VIFF #DigiBC

13

1 Comment

#unrealengine is becoming the standard for Image Generator for Flight Simulators. When people think of Unreal Engine, they often think Games, Architecture and Automotive Visualization or Media & Entertainment but Epic Games's vision has always been bigger. Today, Unreal Engine is rapidly becoming the image generator of choice for training and simulation across aviation and defense. From FAA / EASA certified flight simulators to mission rehearsal systems, we're seeing global leaders adopt Unreal for its photorealism, real-time performance, large world coordinate support, and flexibility in complex environments. CAE has achieved Level D qualification using its Unreal-powered Prodigy image generator entrol simulators is redefining flight simulation with a cutting-edge LED dome and Unreal integration Sogitec is showcasing next-gen simulation fidelity at Paris Air Show 2025 FlightSafety International has integrated Unreal into its new VITAL visual system These are more than partnerships, they signal a fundamental industry shift: Unreal Engine is becoming the standard for high-fidelity Image Generators, where every pixel counts, and every frame brings training closer to reality. https://lnkd.in/gvkKJTZZ https://lnkd.in/ghiUK6vk https://lnkd.in/gMmYg_jK https://lnkd.in/gbymgS6M

164

3 Comments

🤖 AI Up-Skilling Journey 🧪 Experiment 5: Telegram → n8n → Notion 🎯 Goal: Learn about n8n by building a minimal workflow that captures user messages from Telegram, enrich and structure them via an LLM within n8n, then adds each formatted entry as a Notion page. 🛠️ Tools - OpenAI ChatGPT - n8n - Telegram - Notion 🗺️ Architecture user → Telegram bot → n8n trigger → OpenAI for JSON structuring → Notion page creation. It’s a compact pattern for turning unstructured chat into structured, queryable knowledge. 🧠 Key Learnings - n8n is a strong no-code/low-code orchestrator for multi-tool agentic workflows; - the main friction is enforcing strict JSON from the LLM and mapping it cleanly into typed Notion properties. 🔗 https://n8n.io/ #n8n #NoCode #Automation #OpenAI #Notion #Telegram

6

Here is a quick video Nancy Estrada filmed of me explaining one of the ways I use our new Luma Agents app to work on modify video. Luma AI Dream Lab LA .

108

11 Comments

Our Sigma Software Design team has released a new motion explainer for WhiteBridge.ai Network. It’s a dynamic look at AI Guardian Agent technology — where data moves like a living system and every frame carries intent. This project is an excellent example of how our design team works with complex tech concepts and turns them into clear, visual stories that feel alive and modern. If you’re looking for motion design that can explain, engage, and leave an impression, the team is open to new collaborations. 👉 Watch the full shot on Dribbble: https://lnkd.in/deBhzZza #SigmaSoftware #Design #MotionDesign #AI #CreativeTech #ExplainerVideo

21

A lot of XR work focuses on education and simulation, but another powerful use of virtual reality is the ability to explore things at impossible scales. AGO BRISTOL 1775: From Warship to Prison Hulk a project with the Art Gallery of Ontario and Priam Givord XR Studio gave us the opportunity to do exactly that, transforming highly detailed historical ship models into immersive spaces visitors could move through in both desktop and VR. One of the most interesting parts of the project was the digitization process. We worked with physical ship models captured through CT scanning and photogrammetry, then brought them into Unreal Engine 5 and scaled them up into explorable environments. The asset complexity was substantial: the Bristol CT scan was around 40 million triangles, the ship model reached 90 million triangles with six 8K textures, and the prisoner-of-war model came in at 30 million triangles with six 4K textures. A big part of what made the project unusual was using Nanite in VR, allowing that level of fidelity to hold up in real time in a way most XR projects do not attempt. Created for the Art Gallery of Ontario, the experience was built as a multiplayer application for both desktop and VR, allowing visitors to explore the AGO’s 1775 model of the Bristol freely or through a guided narrative tracing its transformation from British warship to prison hulk. As the lead developer I worked on the guided system, the multiplayer infrastructure and real-time rendering of the scanned assets in Unreal Engine. I had the pleasure of working with a great team at the AGO and Priam GIVORD, an amazing XR Interface Designer and the project lead. Feel free to share and repost! #VR #VirtualReality #UnrealEngine #Nanite #DigitalHeritage #MuseumInnovation #3DScanning #Photogrammetry #ImmersiveExperience #ArtGalleryOfOntario

177

13 Comments

Wen S. — Co-Founder & COO of Genspark — on Venture with Grace, sharing a bold vision for the AI-powered future. Wen envisions a world with thousands of AI models, each solving different problems. Their approach? Combine the best-of-breed models to meet real user needs — not just for developers, but for billions of knowledge workers creating slides, memos, reports, and more. She emphasized building tools that empower people, increase productivity, and ultimately make work better and more meaningful. Episode is live now — a forward-looking conversation on AI, productivity, and shaping the future of work. #AI #FutureOfWork #Productivity #KnowledgeWorkers #TechInnovation

12

2 Comments