Loading video...

Video Failed to Load

There was a problem loading this video. This could be due to a temporary network issue or the video might be unavailable.

FrankenMotion: Part-level human motion generation; - enables compositional synthesis by combining atomic body-part movements; - DiT + FrankenAgent LLM. Indie game devs/animators can make complex character animations via text without manual keyframing.

Wildminder

10,367 subscribers

33,890 views • 8 months ago •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

0 Comments

No comments available

Comments from the original post will appear here

Related Videos

Multi-Track Timeline Control for Text-Driven 3D Human Motion Generation paper page: Recent advances in generative modeling have led to promising progress on synthesizing 3D human motion from text, with methods that can generate character animations from short prompts and specified durations. However, using a single text prompt as input lacks the fine-grained control needed by animators, such as composing multiple actions and defining precise durations for parts of the motion. To address this, we introduce the new problem of timeline control for text-driven motion synthesis, which provides an intuitive, yet fine-grained, input interface for users. Instead of a single prompt, users can specify a multi-track timeline of multiple prompts organized in temporal intervals that may overlap. This enables specifying the exact timings of each action and composing multiple actions in sequence or at overlapping intervals. To generate composite animations from a multi-track timeline, we propose a new test-time denoising method. This method can be integrated with any pre-trained motion diffusion model to synthesize realistic motions that accurately reflect the timeline. At every step of denoising, our method processes each timeline interval (text prompt) individually, subsequently aggregating the predictions with consideration for the specific body parts engaged in each action. Experimental comparisons and ablations validate that our method produces realistic motions that respect the semantics and timing of given text prompts.

Multi-Track Timeline Control for Text-Driven 3D Human Motion Generation paper page: Recent advances in generative modeling have led to promising progress on synthesizing 3D human motion from text, with methods that can generate character animations from short prompts and specified durations. However, using a single text prompt as input lacks the fine-grained control needed by animators, such as composing multiple actions and defining precise durations for parts of the motion. To address this, we introduce the new problem of timeline control for text-driven motion synthesis, which provides an intuitive, yet fine-grained, input interface for users. Instead of a single prompt, users can specify a multi-track timeline of multiple prompts organized in temporal intervals that may overlap. This enables specifying the exact timings of each action and composing multiple actions in sequence or at overlapping intervals. To generate composite animations from a multi-track timeline, we propose a new test-time denoising method. This method can be integrated with any pre-trained motion diffusion model to synthesize realistic motions that accurately reflect the timeline. At every step of denoising, our method processes each timeline interval (text prompt) individually, subsequently aggregating the predictions with consideration for the specific body parts engaged in each action. Experimental comparisons and ablations validate that our method produces realistic motions that respect the semantics and timing of given text prompts.

AK

126,654 views • 2 years ago

Tencent's Hunyuan 3D Studio is LIVE! 🎨 AI-powered 3D creation for professionals, slashing workflows from days to minutes.⚡️ Key Features: 🔹Text-to-3D: Generate controllable geometry (multi-view, multi-style, A pose, bbox-control) from text/images. 🔹Auto Part Split: Break models into 50+ editable components (hats, clothes, etc.) 🔹1-Min UV Unwrap: AI-driven, professional-grade UVs, no more manual fixes. 🔹PBR Textures: Precisely edit global or local textures with text or image prompts & professional material sphere generation. 🔹Auto Rigging: Quickly rig various characters, with specialized motion templates for human models. Hunyuan3D Studio is built to empower game developers, animators, and designers to create faster and smarter. 📄Tech Report: 👉🏻Try it now:

Tencent's Hunyuan 3D Studio is LIVE! 🎨 AI-powered 3D creation for professionals, slashing workflows from days to minutes.⚡️ Key Features: 🔹Text-to-3D: Generate controllable geometry (multi-view, multi-style, A pose, bbox-control) from text/images. 🔹Auto Part Split: Break models into 50+ editable components (hats, clothes, etc.) 🔹1-Min UV Unwrap: AI-driven, professional-grade UVs, no more manual fixes. 🔹PBR Textures: Precisely edit global or local textures with text or image prompts & professional material sphere generation. 🔹Auto Rigging: Quickly rig various characters, with specialized motion templates for human models. Hunyuan3D Studio is built to empower game developers, animators, and designers to create faster and smarter. 📄Tech Report: 👉🏻Try it now:

Tencent Hy

46,998 views • 1 year ago

✨We are excited to open-source Tencent HY-Motion 1.0, a billion-parameter text-to-motion model built on the Diffusion Transformer (DiT) architecture and flow matching. Tencent HY-Motion 1.0 empowers developers and individual creators alike by transforming natural language into high-fidelity, fluid, and diverse 3D character animations, delivering exceptional instruction-following capabilities across a broad range of categories. The generated 3D animation assets can be seamlessly integrated into typical 3D animation pipelines.🎮🎥 Highlights: 🔹Billion-Scale DiT: Successfully scaled flow-matching DiT to 1B+ parameters, setting a new ceiling for instruction-following capability and generated motion quality. 🔹Full-Stage Training Strategy: The industry’s first motion generation model featuring a complete Pre-training → SFT → RL loop to optimize physical plausibility and semantic accuracy. 🔹Comprehensive Category Coverage: Features 200+ motion categories across 6 major classes—the most comprehensive in the industry, curated via a meticulous data pipeline. 🌐Project Page: 🔗Github: 🤗Hugging Face: 📄Technical report:

✨We are excited to open-source Tencent HY-Motion 1.0, a billion-parameter text-to-motion model built on the Diffusion Transformer (DiT) architecture and flow matching. Tencent HY-Motion 1.0 empowers developers and individual creators alike by transforming natural language into high-fidelity, fluid, and diverse 3D character animations, delivering exceptional instruction-following capabilities across a broad range of categories. The generated 3D animation assets can be seamlessly integrated into typical 3D animation pipelines.🎮🎥 Highlights: 🔹Billion-Scale DiT: Successfully scaled flow-matching DiT to 1B+ parameters, setting a new ceiling for instruction-following capability and generated motion quality. 🔹Full-Stage Training Strategy: The industry’s first motion generation model featuring a complete Pre-training → SFT → RL loop to optimize physical plausibility and semantic accuracy. 🔹Comprehensive Category Coverage: Features 200+ motion categories across 6 major classes—the most comprehensive in the industry, curated via a meticulous data pipeline. 🌐Project Page: 🔗Github: 🤗Hugging Face: 📄Technical report:

Tencent Hy

329,750 views • 9 months ago

How do you give a humanoid the general motion capability? Not just single motions, but all motion? Introducing SONIC, our new work on supersizing motion tracking for natural humanoid control. We argue that motion tracking is the scalable foundation task for humanoids. So we "supersized" it: 9k+ GPU hours and 100M+ motion frames. But tracking alone is not enough; we show how to make a useful control system out of it: - Universal Kinematic Planner: Enables game-like gamepad control and high-level teleoperation, just like controlling a character in a game. - VR Full-Body Teleop: Direct, real-time whole-body control by a human wearing a VR headset. - VR Keypoint Teleop: Control the upper body (hands/head) while our planner handles robust locomotion automatically. - VLA Integration: We connect this motion tracker to autonomous Visual-Language-Action (VLA) models for autonomous task execution! We use a Universal Token Space to UNIFY this command space, turning our robust tracker into a general-purpose, programmable humanoid brain. This is the generalist "System 1" for humanoids. 🚀 Project: #Humanoids #Robotics #AI #FoundationModels #NVIDIAResearch 🧠🔥

How do you give a humanoid the general motion capability? Not just single motions, but all motion? Introducing SONIC, our new work on supersizing motion tracking for natural humanoid control. We argue that motion tracking is the scalable foundation task for humanoids. So we "supersized" it: 9k+ GPU hours and 100M+ motion frames. But tracking alone is not enough; we show how to make a useful control system out of it: - Universal Kinematic Planner: Enables game-like gamepad control and high-level teleoperation, just like controlling a character in a game. - VR Full-Body Teleop: Direct, real-time whole-body control by a human wearing a VR headset. - VR Keypoint Teleop: Control the upper body (hands/head) while our planner handles robust locomotion automatically. - VLA Integration: We connect this motion tracker to autonomous Visual-Language-Action (VLA) models for autonomous task execution! We use a Universal Token Space to UNIFY this command space, turning our robust tracker into a general-purpose, programmable humanoid brain. This is the generalist "System 1" for humanoids. 🚀 Project: #Humanoids #Robotics #AI #FoundationModels #NVIDIAResearch 🧠🔥

Zhengyi “Zen” Luo

62,782 views • 11 months ago

Alibaba presents MIMO Controllable Character Video Synthesis with Spatial Decomposed Modeling Character video synthesis aims to produce realistic videos of animatable characters within lifelike scenes. As a fundamental problem in the computer vision and graphics community, 3D works typically require multi-view captures for per-case training, which severely limits their applicability of modeling arbitrary characters in a short time. Recent 2D methods break this limitation via pre-trained diffusion models, but they struggle for pose generality and scene interaction. To this end, we propose MIMO, a novel framework which can not only synthesize character videos with controllable attributes (i.e., character, motion and scene) provided by simple user inputs, but also simultaneously achieve advanced scalability to arbitrary characters, generality to novel 3D motions, and applicability to interactive real-world scenes in a unified framework. The core idea is to encode the 2D video to compact spatial codes, considering the inherent 3D nature of video occurrence. Concretely, we lift the 2D frame pixels into 3D using monocular depth estimators, and decompose the video clip to three spatial components (i.e., main human, underlying scene, and floating occlusion) in hierarchical layers based on the 3D depth. These components are further encoded to canonical identity code, structured motion code and full scene code, which are utilized as control signals of synthesis process. The design of spatial decomposed modeling enables flexible user control, complex motion expression, as well as 3D-aware synthesis for scene interactions. Experimental results demonstrate effectiveness and robustness of the proposed method.

Alibaba presents MIMO Controllable Character Video Synthesis with Spatial Decomposed Modeling Character video synthesis aims to produce realistic videos of animatable characters within lifelike scenes. As a fundamental problem in the computer vision and graphics community, 3D works typically require multi-view captures for per-case training, which severely limits their applicability of modeling arbitrary characters in a short time. Recent 2D methods break this limitation via pre-trained diffusion models, but they struggle for pose generality and scene interaction. To this end, we propose MIMO, a novel framework which can not only synthesize character videos with controllable attributes (i.e., character, motion and scene) provided by simple user inputs, but also simultaneously achieve advanced scalability to arbitrary characters, generality to novel 3D motions, and applicability to interactive real-world scenes in a unified framework. The core idea is to encode the 2D video to compact spatial codes, considering the inherent 3D nature of video occurrence. Concretely, we lift the 2D frame pixels into 3D using monocular depth estimators, and decompose the video clip to three spatial components (i.e., main human, underlying scene, and floating occlusion) in hierarchical layers based on the 3D depth. These components are further encoded to canonical identity code, structured motion code and full scene code, which are utilized as control signals of synthesis process. The design of spatial decomposed modeling enables flexible user control, complex motion expression, as well as 3D-aware synthesis for scene interactions. Experimental results demonstrate effectiveness and robustness of the proposed method.

AK

149,079 views • 2 years ago

Can reinforcement learning from AI feedback unlock new capabilities in AI agents? Introducing Motif, an LLM-powered method for intrinsic motivation from AI feedback. Motif extracts reward functions from Llama 2's preferences and uses them to train agents with reinforcement learning. On the complex NetHack game, Motif solves previously unsolved tasks without needing any expert demonstrations. Surprisingly, Motif's reward leads to better game score than the one obtained by using the score itself as a reward. Given access to an event captioning mechanism, a few properties make Motif a general method: • it is entirely based on open models • the LLM doesn't need direct access to the environment dynamics (e.g., its source code) • the LLM doesn't need to understand observation and action spaces The best part? You can start using Motif right now, even on a small compute budget: the whole pipeline can take less than two GPU-days. Feel free to read our paper and try our code out. Paper: Code: Blog post: Work co-lead by Martin Klissarov and myself, with Shagun Sodhani Roberta Raileanu Pierre-Luc Bacon Pascal Vincent @yayitsamyzhang Mikael Henaff Learn more in the thread 🧵

Can reinforcement learning from AI feedback unlock new capabilities in AI agents? Introducing Motif, an LLM-powered method for intrinsic motivation from AI feedback. Motif extracts reward functions from Llama 2's preferences and uses them to train agents with reinforcement learning. On the complex NetHack game, Motif solves previously unsolved tasks without needing any expert demonstrations. Surprisingly, Motif's reward leads to better game score than the one obtained by using the score itself as a reward. Given access to an event captioning mechanism, a few properties make Motif a general method: • it is entirely based on open models • the LLM doesn't need direct access to the environment dynamics (e.g., its source code) • the LLM doesn't need to understand observation and action spaces The best part? You can start using Motif right now, even on a small compute budget: the whole pipeline can take less than two GPU-days. Feel free to read our paper and try our code out. Paper: Code: Blog post: Work co-lead by Martin Klissarov and myself, with Shagun Sodhani Roberta Raileanu Pierre-Luc Bacon Pascal Vincent @yayitsamyzhang Mikael Henaff Learn more in the thread 🧵

Pierluca D'Oro

312,048 views • 3 years ago

Thrilled to announce Kingnet AI V2 is now officially live ! We have officially deployed on the BNB Chain first ! Whether you're an enthusiast or a professional game developer, come and try it out now： Each generated asset costs approximately $3 and supports export in professional game-editing formats. We will soon support exporting assets in NFT on-chain formats, empowering Web3 users and partners with seamless integration. Jump down more rabbit holes next.👇 📔 Product Introduction： By conversing naturally with agent Joi, users can achieve a complete automated game development cycle - from requirement proposal to finished product delivery. Users simply need to describe their game concepts and design requirements in natural language, and Joi will automatically utilize built-in generator including: • Animation Generator： AI-driven motion generation with auto-rigging technology for instant character animation • Map Generator： Procedural map generation with built-in logic validation for consistent world-building • Numerical Generator： Automated game economy tuning for fair yet challenging gameplay systems • Editable Code Generator： Generates clean, maintainable game logic code with multi-platform/multi-language support • Interface Generator： Intelligent layout engine that optimizes user experience and interaction flow Joi intelligently generates all necessary game components, performs multi-dimensional feasibility checks, and ultimately completes game synthesis, packaging and deployment. Users can directly click to try the game on the chat interface, or download the complete editable code package to achieve rapid iteration and secondary development. 🎯 Core Architecture： 1/ Natural Language Understanding & Multimodal Intent Parsing: Utilizing advanced deep learning NLP models (e.g., Transformer-based language understanding models), Joi precisely interprets user natural language inputs and extracts core game design intents and parameters. Through semantic segmentation and entity recognition, complex requirements are decomposed into specific tasks for animation, map, numerical systems, UI, and code modules. 2/ Modular Editor System & API Integration: Joi employs a unified API framework to enable seamless collaboration between editor modules, ensuring high compatibility in data formats and workflows. 3/ Intelligent Validation & Quality Assurance: The system incorporates multi-dimensional verification mechanisms including animation continuity checks, map pathfinding and physical logic validation, game balance analysis, UI interaction consistency verification, and static/dynamic code security testing. Automated testing and feedback loops ensure outputs meet high-standard game design specifications. 4/ Automatic Synthesis, Packaging & Instant Deployment: Verified resources are automatically integrated to complete game compilation, packaging and deployment. Supports one-click generation of playable online links and downloadable complete code packages for immediate testing or deep customization/iterative development. 5/ Interactive Chat Interface & Seamless UX: The entire workflow is completed within the chat interface, significantly reducing traditional game development's communication and operational barriers. Users accomplish complex game design and development through conversation while receiving real-time feedback and adjustment suggestions, democratizing game creation. 6/ Industry-Disrupting Value: Transforms traditional manual development into AI-driven automated pipelines.

Thrilled to announce Kingnet AI V2 is now officially live ! We have officially deployed on the BNB Chain first ! Whether you're an enthusiast or a professional game developer, come and try it out now： Each generated asset costs approximately $3 and supports export in professional game-editing formats. We will soon support exporting assets in NFT on-chain formats, empowering Web3 users and partners with seamless integration. Jump down more rabbit holes next.👇 📔 Product Introduction： By conversing naturally with agent Joi, users can achieve a complete automated game development cycle - from requirement proposal to finished product delivery. Users simply need to describe their game concepts and design requirements in natural language, and Joi will automatically utilize built-in generator including: • Animation Generator： AI-driven motion generation with auto-rigging technology for instant character animation • Map Generator： Procedural map generation with built-in logic validation for consistent world-building • Numerical Generator： Automated game economy tuning for fair yet challenging gameplay systems • Editable Code Generator： Generates clean, maintainable game logic code with multi-platform/multi-language support • Interface Generator： Intelligent layout engine that optimizes user experience and interaction flow Joi intelligently generates all necessary game components, performs multi-dimensional feasibility checks, and ultimately completes game synthesis, packaging and deployment. Users can directly click to try the game on the chat interface, or download the complete editable code package to achieve rapid iteration and secondary development. 🎯 Core Architecture： 1/ Natural Language Understanding & Multimodal Intent Parsing: Utilizing advanced deep learning NLP models (e.g., Transformer-based language understanding models), Joi precisely interprets user natural language inputs and extracts core game design intents and parameters. Through semantic segmentation and entity recognition, complex requirements are decomposed into specific tasks for animation, map, numerical systems, UI, and code modules. 2/ Modular Editor System & API Integration: Joi employs a unified API framework to enable seamless collaboration between editor modules, ensuring high compatibility in data formats and workflows. 3/ Intelligent Validation & Quality Assurance: The system incorporates multi-dimensional verification mechanisms including animation continuity checks, map pathfinding and physical logic validation, game balance analysis, UI interaction consistency verification, and static/dynamic code security testing. Automated testing and feedback loops ensure outputs meet high-standard game design specifications. 4/ Automatic Synthesis, Packaging & Instant Deployment: Verified resources are automatically integrated to complete game compilation, packaging and deployment. Supports one-click generation of playable online links and downloadable complete code packages for immediate testing or deep customization/iterative development. 5/ Interactive Chat Interface & Seamless UX: The entire workflow is completed within the chat interface, significantly reducing traditional game development's communication and operational barriers. Users accomplish complex game design and development through conversation while receiving real-time feedback and adjustment suggestions, democratizing game creation. 6/ Industry-Disrupting Value: Transforms traditional manual development into AI-driven automated pipelines.

Kingnet AI

46,437 views • 1 year ago

Generated with seedance 2 on PixVerse Prompt :- Create a 21-second vertical 9:16 cinematic stop-motion animation. A tiny humanoid figure made entirely from naturally shaped smooth beach stones comes to life in a shallow tidal pool on a rocky seashore. Each body part is constructed from individual stones—large oval stone as the head, stacked rounded stones for the torso, smaller stones forming arms and legs. The scene is filmed through a slightly out-of-focus chain-link fence, creating a natural foreground frame. Behind the character is a calm shoreline, shallow seawater, barnacle-covered rocks, seaweed, and a softly blurred forest under an overcast sky. Photorealistic textures, natural muted colors, shallow depth of field, realistic water reflections. Animation: The stone figure slowly balances itself and begins moving like a playful little human. It shifts its weight, raises one leg, bends its knees, swings its stone arms, and performs a quirky little dance while carefully maintaining balance on the wet surface. Tiny ripples form beneath every footstep, and its reflection moves naturally in the water. The movements should feel hand-crafted stop-motion, slightly imperfect but believable, with realistic stone physics and weight. Toward the final seconds, the figure loses its balance, stumbles, and collapses naturally, with the individual stones separating and tumbling onto the shallow wet ground. The character completely disassembles into ordinary stones. End with the camera holding on the scattered stones and their reflections. Camera: locked-off smartphone-style vertical composition, subtle natural camera movement, medium-full shot, low angle close to water level, strong foreground bokeh from the fence, cinematic depth of field. Lighting: soft diffused daylight, overcast coastal atmosphere, realistic reflections and wet stone highlights. Style: ultra-realistic live-action environment + whimsical photorealistic stone stop-motion character, tactile stone textures, physically believable movement, cinematic macro photography, no CGI-looking surfaces, no text, no humans. Negative prompt: cartoon, plastic-looking stones, exaggerated facial features, smooth CGI character, floating objects, unrealistic physics, extra limbs, changing environment, camera cuts, text, watermark, oversaturated colors.

Generated with seedance 2 on PixVerse Prompt :- Create a 21-second vertical 9:16 cinematic stop-motion animation. A tiny humanoid figure made entirely from naturally shaped smooth beach stones comes to life in a shallow tidal pool on a rocky seashore. Each body part is constructed from individual stones—large oval stone as the head, stacked rounded stones for the torso, smaller stones forming arms and legs. The scene is filmed through a slightly out-of-focus chain-link fence, creating a natural foreground frame. Behind the character is a calm shoreline, shallow seawater, barnacle-covered rocks, seaweed, and a softly blurred forest under an overcast sky. Photorealistic textures, natural muted colors, shallow depth of field, realistic water reflections. Animation: The stone figure slowly balances itself and begins moving like a playful little human. It shifts its weight, raises one leg, bends its knees, swings its stone arms, and performs a quirky little dance while carefully maintaining balance on the wet surface. Tiny ripples form beneath every footstep, and its reflection moves naturally in the water. The movements should feel hand-crafted stop-motion, slightly imperfect but believable, with realistic stone physics and weight. Toward the final seconds, the figure loses its balance, stumbles, and collapses naturally, with the individual stones separating and tumbling onto the shallow wet ground. The character completely disassembles into ordinary stones. End with the camera holding on the scattered stones and their reflections. Camera: locked-off smartphone-style vertical composition, subtle natural camera movement, medium-full shot, low angle close to water level, strong foreground bokeh from the fence, cinematic depth of field. Lighting: soft diffused daylight, overcast coastal atmosphere, realistic reflections and wet stone highlights. Style: ultra-realistic live-action environment + whimsical photorealistic stone stop-motion character, tactile stone textures, physically believable movement, cinematic macro photography, no CGI-looking surfaces, no text, no humans. Negative prompt: cartoon, plastic-looking stones, exaggerated facial features, smooth CGI character, floating objects, unrealistic physics, extra limbs, changing environment, camera cuts, text, watermark, oversaturated colors.

Marcos

15,852 views • 1 month ago

This will retire 90% of RAG systems with dignity (and a sad song playlist). Powered by DSPy: If you're still building "text in, text out" chatbots that only perform blind vector and text searches, you're not gonna make it! My team just dropped Elysia, and it's not just an incremental successor to Verba… It's a whole rethink of how we interact with our data using AI. 𝗪𝗵𝗮𝘁 𝗶𝘀 𝗘𝗹𝘆𝗶𝘀𝗮? An open-source platform for building agentic RAG architectures. It learns from your preferences, intelligently categorizes, labels, and searches through your data, and provides complete transparency into its decision-making process. The long & exciting feature list: • 𝗧𝗿𝗮𝗻𝘀𝗽𝗮𝗿𝗲𝗻𝘁 𝗗𝗲𝗰𝗶𝘀𝗶𝗼𝗻-𝗧𝗿𝗲𝗲 𝗔𝗴𝗲𝗻𝘁𝘀: Elysia’s core is a customizable decision tree, and it visualizes its entire reasoning process, showing you why it chooses a specific tool or path. It enables advanced error handling, self-healing from failed queries, and prevents infinite loops. You can also add custom tools and branches to build complex, state-aware workflows. • 𝗗𝗮𝘁𝗮 𝗔𝘄𝗮𝗿𝗲𝗻𝗲𝘀𝘀: Before it even attempts a query, Elysia performs a full analysis of your data collections. This eliminates the blind search problem plaguing most RAG systems and allows for far more complex and accurate query generation. • 𝗗𝘆𝗻𝗮𝗺𝗶𝗰 𝗗𝗮𝘁𝗮 𝗗𝗶𝘀𝗽𝗹𝗮𝘆𝘀: Your RAG pipeline shouldn't be limited to text, right? That’s why Elysia analyzes each query's results and chooses the best way to display them, from tables and charts to product cards and GitHub tickets. It also features a comprehensive data explorer with search, sorting, and filtering capabilities. • 𝗛𝘆𝗽𝗲𝗿-𝗣𝗲𝗿𝘀𝗼𝗻𝗮𝗹𝗶𝘇𝗮𝘁𝗶𝗼𝗻 𝘃𝗶𝗮 𝗙𝗲𝗲𝗱𝗯𝗮𝗰𝗸: It uses your positively-rated queries as few-shot examples to improve future responses. This allows you to use smaller, faster models that perform like larger ones over time, cutting costs without sacrificing quality for most use cases. • 𝗖𝗵𝘂𝗻𝗸-𝗢𝗻-𝗗𝗲𝗺𝗮𝗻𝗱: Elysia chunks documents at query time. It performs initial searches on document-level vectors and only chunks relevant documents on the fly, storing them in a parallel quantized collection with cross references for future use. 𝗧𝗵𝗲 𝗦𝘁𝗮𝗰𝗸 Elysia is built from scratch on Weaviate, using its native features like named vectors, a variety of search types, filters, cross references, quantization, etc. It uses DSPy for LLM interactions and is delivered as a production-ready application via FastAPI, serving a NextJS frontend as static HTML. Also available as a Python package via pip: 𝗽𝗶𝗽 𝗶𝗻𝘀𝘁𝗮𝗹𝗹 𝗲𝗹𝘆𝘀𝗶𝗮-𝗮𝗶 Type: 𝗲𝗹𝘆𝘀𝗶𝗮 𝘀𝘁𝗮𝗿𝘁 Connect your Weaviate cluster and go explore what’s possible.

This will retire 90% of RAG systems with dignity (and a sad song playlist). Powered by DSPy: If you're still building "text in, text out" chatbots that only perform blind vector and text searches, you're not gonna make it! My team just dropped Elysia, and it's not just an incremental successor to Verba… It's a whole rethink of how we interact with our data using AI. 𝗪𝗵𝗮𝘁 𝗶𝘀 𝗘𝗹𝘆𝗶𝘀𝗮? An open-source platform for building agentic RAG architectures. It learns from your preferences, intelligently categorizes, labels, and searches through your data, and provides complete transparency into its decision-making process. The long & exciting feature list: • 𝗧𝗿𝗮𝗻𝘀𝗽𝗮𝗿𝗲𝗻𝘁 𝗗𝗲𝗰𝗶𝘀𝗶𝗼𝗻-𝗧𝗿𝗲𝗲 𝗔𝗴𝗲𝗻𝘁𝘀: Elysia’s core is a customizable decision tree, and it visualizes its entire reasoning process, showing you why it chooses a specific tool or path. It enables advanced error handling, self-healing from failed queries, and prevents infinite loops. You can also add custom tools and branches to build complex, state-aware workflows. • 𝗗𝗮𝘁𝗮 𝗔𝘄𝗮𝗿𝗲𝗻𝗲𝘀𝘀: Before it even attempts a query, Elysia performs a full analysis of your data collections. This eliminates the blind search problem plaguing most RAG systems and allows for far more complex and accurate query generation. • 𝗗𝘆𝗻𝗮𝗺𝗶𝗰 𝗗𝗮𝘁𝗮 𝗗𝗶𝘀𝗽𝗹𝗮𝘆𝘀: Your RAG pipeline shouldn't be limited to text, right? That’s why Elysia analyzes each query's results and chooses the best way to display them, from tables and charts to product cards and GitHub tickets. It also features a comprehensive data explorer with search, sorting, and filtering capabilities. • 𝗛𝘆𝗽𝗲𝗿-𝗣𝗲𝗿𝘀𝗼𝗻𝗮𝗹𝗶𝘇𝗮𝘁𝗶𝗼𝗻 𝘃𝗶𝗮 𝗙𝗲𝗲𝗱𝗯𝗮𝗰𝗸: It uses your positively-rated queries as few-shot examples to improve future responses. This allows you to use smaller, faster models that perform like larger ones over time, cutting costs without sacrificing quality for most use cases. • 𝗖𝗵𝘂𝗻𝗸-𝗢𝗻-𝗗𝗲𝗺𝗮𝗻𝗱: Elysia chunks documents at query time. It performs initial searches on document-level vectors and only chunks relevant documents on the fly, storing them in a parallel quantized collection with cross references for future use. 𝗧𝗵𝗲 𝗦𝘁𝗮𝗰𝗸 Elysia is built from scratch on Weaviate, using its native features like named vectors, a variety of search types, filters, cross references, quantization, etc. It uses DSPy for LLM interactions and is delivered as a production-ready application via FastAPI, serving a NextJS frontend as static HTML. Also available as a Python package via pip: 𝗽𝗶𝗽 𝗶𝗻𝘀𝘁𝗮𝗹𝗹 𝗲𝗹𝘆𝘀𝗶𝗮-𝗮𝗶 Type: 𝗲𝗹𝘆𝘀𝗶𝗮 𝘀𝘁𝗮𝗿𝘁 Connect your Weaviate cluster and go explore what’s possible.

Philip Vollet

93,694 views • 1 year ago

Announcing a new Coursera course: Retrieval Augmented Generation (RAG) You'll learn to build high performance, production-ready RAG systems in this hands-on, in-depth course created by and taught by , experienced AI and ML engineer, researcher, and educator. RAG is a critical component today of many LLM-based applications in customer support, internal company Q&A systems, even many of the leading chatbots that use web search to answer your questions. This course teaches you in-depth how to make RAG work well. LLMs can produce generic or outdated responses, especially when asked specialized questions not covered in its training data. RAG is the most widely used technique for addressing this. It brings in data from new data sources, such as internal documents or recent news, to give the LLM the relevant context to private, recent, or specialized information. This lets it generate more grounded and accurate responses. In this course, you’ll learn to design and implement every part of a RAG system, from retrievers to vector databases to generation to evals. You’ll learn about the fundamental principles behind RAG and how to optimize it at both the component and whole-system levels. As AI evolves, RAG is evolving too. New models can handle longer context windows, reason more effectively, and can be parts of complex agentic workflows. One exciting growth area is Agentic RAG, in which an AI agent at runtime (rather than it being hardcoded at development time) autonomously decides what data to retrieve, and when/how to go deeper. Even with this evolution, access to high-quality data at runtime is essential, which is why RAG is a key part of so many applications. You'll learn via hands-on experiences to: - Build a RAG system with retrieval and prompt augmentation - Compare retrieval methods like BM25, semantic search, and Reciprocal Rank Fusion - Chunk, index, and retrieve documents using a Weaviate vector database and a news dataset - Develop a chatbot, using open-source LLMs hosted by Together AI, for a fictional store that answers product and FAQ questions - Use evals to drive improving reliability, and incorporate multi-modal data RAG is an important foundational technique. Become good at it through this course! Please sign up here:

Announcing a new Coursera course: Retrieval Augmented Generation (RAG) You'll learn to build high performance, production-ready RAG systems in this hands-on, in-depth course created by and taught by , experienced AI and ML engineer, researcher, and educator. RAG is a critical component today of many LLM-based applications in customer support, internal company Q&A systems, even many of the leading chatbots that use web search to answer your questions. This course teaches you in-depth how to make RAG work well. LLMs can produce generic or outdated responses, especially when asked specialized questions not covered in its training data. RAG is the most widely used technique for addressing this. It brings in data from new data sources, such as internal documents or recent news, to give the LLM the relevant context to private, recent, or specialized information. This lets it generate more grounded and accurate responses. In this course, you’ll learn to design and implement every part of a RAG system, from retrievers to vector databases to generation to evals. You’ll learn about the fundamental principles behind RAG and how to optimize it at both the component and whole-system levels. As AI evolves, RAG is evolving too. New models can handle longer context windows, reason more effectively, and can be parts of complex agentic workflows. One exciting growth area is Agentic RAG, in which an AI agent at runtime (rather than it being hardcoded at development time) autonomously decides what data to retrieve, and when/how to go deeper. Even with this evolution, access to high-quality data at runtime is essential, which is why RAG is a key part of so many applications. You'll learn via hands-on experiences to: - Build a RAG system with retrieval and prompt augmentation - Compare retrieval methods like BM25, semantic search, and Reciprocal Rank Fusion - Chunk, index, and retrieve documents using a Weaviate vector database and a news dataset - Develop a chatbot, using open-source LLMs hosted by Together AI, for a fictional store that answers product and FAQ questions - Use evals to drive improving reliability, and incorporate multi-modal data RAG is an important foundational technique. Become good at it through this course! Please sign up here:

Andrew Ng

124,656 views • 1 year ago

Hollywood has a dirty secret. That perfectly clean green screen shot in your favorite Marvel movie? A human being sat in a dark room for 6 hours fixing it frame by frame. The AI keyer got the body. A person painted every strand of hair. By hand. At 2 AM. For 400 frames. The software costs $5,000 a year. And it still cannot key hair. Nuke: $4,988/year. Cannot key hair in motion blur. After Effects: $264/year. Cannot key transparent glass. Boris FX: $1,865. Cannot key fine edges without haloing. The industry's solution for 30 years has been the same: pay for expensive software, then pay a human to fix what the software couldn't. The YouTubers behind Corridor Crew looked at this and asked a different question. What if the AI didn't try to remove the green? What if it figured out what color was actually there before the green contaminated it? They trained a neural network on synthetic 3D data. Not scraped footage. Not stolen clips. Perfectly rendered scenes where every pixel's true color was already known. Hair strands. Motion blur. Transparent glass. All simulated with known ground truth. Then they fed it real green screen footage. It worked. They called it CorridorKey. Then they open sourced it. → Feed it raw green screen footage → AI reconstructs the true foreground color for every pixel → Hair stays perfect. Every strand. → Motion blur stays intact. Every frame. → Transparent glass stays transparent. → 16-bit and 32-bit EXR output. Nuke-ready. Resolve-ready. → Handles 4K natively → Runs on consumer GPUs. 6 GB VRAM minimum. → Runs on Apple Silicon via MLX → Auto-detects green or blue screen → Removes tracking markers automatically → DaVinci Resolve plugin live → Standalone GUI for non-technical users → One-click installer. No Python setup. Here's the wildest part: Within 2.5 months: 13,000+ GitHub stars. 787 forks. Active Discord. Community built a cloud render farm so you can process footage without owning a GPU. DaVinci Resolve plugin shipped. Nuke and After Effects plugins in development. VFX freelancers are already offering CorridorKey-powered services to production companies. One person. One GPU. Hollywood-quality keys. A business built on free software. Nuke: $4,988/year. Still needs manual cleanup. After Effects: $264/year. Still needs manual cleanup. Rotoscope artist: $50 to $150/hour. For the cleanup. CorridorKey: $0. No cleanup needed. One honest flag: the license is CC BY-NC-SA 4.0. Free forever for personal projects, students, indie films, and learning. Commercial use requires permission from Corridor Digital. They did not pretend it was MIT. A problem that plagued Hollywood for 30 years. Solved by YouTubers. Open sourced for free. 13,000+ stars. 787 forks. CC BY-NC-SA 4.0. Your footage. Your keys. No rotoscoping.

Hollywood has a dirty secret. That perfectly clean green screen shot in your favorite Marvel movie? A human being sat in a dark room for 6 hours fixing it frame by frame. The AI keyer got the body. A person painted every strand of hair. By hand. At 2 AM. For 400 frames. The software costs $5,000 a year. And it still cannot key hair. Nuke: $4,988/year. Cannot key hair in motion blur. After Effects: $264/year. Cannot key transparent glass. Boris FX: $1,865. Cannot key fine edges without haloing. The industry's solution for 30 years has been the same: pay for expensive software, then pay a human to fix what the software couldn't. The YouTubers behind Corridor Crew looked at this and asked a different question. What if the AI didn't try to remove the green? What if it figured out what color was actually there before the green contaminated it? They trained a neural network on synthetic 3D data. Not scraped footage. Not stolen clips. Perfectly rendered scenes where every pixel's true color was already known. Hair strands. Motion blur. Transparent glass. All simulated with known ground truth. Then they fed it real green screen footage. It worked. They called it CorridorKey. Then they open sourced it. → Feed it raw green screen footage → AI reconstructs the true foreground color for every pixel → Hair stays perfect. Every strand. → Motion blur stays intact. Every frame. → Transparent glass stays transparent. → 16-bit and 32-bit EXR output. Nuke-ready. Resolve-ready. → Handles 4K natively → Runs on consumer GPUs. 6 GB VRAM minimum. → Runs on Apple Silicon via MLX → Auto-detects green or blue screen → Removes tracking markers automatically → DaVinci Resolve plugin live → Standalone GUI for non-technical users → One-click installer. No Python setup. Here's the wildest part: Within 2.5 months: 13,000+ GitHub stars. 787 forks. Active Discord. Community built a cloud render farm so you can process footage without owning a GPU. DaVinci Resolve plugin shipped. Nuke and After Effects plugins in development. VFX freelancers are already offering CorridorKey-powered services to production companies. One person. One GPU. Hollywood-quality keys. A business built on free software. Nuke: $4,988/year. Still needs manual cleanup. After Effects: $264/year. Still needs manual cleanup. Rotoscope artist: $50 to $150/hour. For the cleanup. CorridorKey: $0. No cleanup needed. One honest flag: the license is CC BY-NC-SA 4.0. Free forever for personal projects, students, indie films, and learning. Commercial use requires permission from Corridor Digital. They did not pretend it was MIT. A problem that plagued Hollywood for 30 years. Solved by YouTubers. Open sourced for free. 13,000+ stars. 787 forks. CC BY-NC-SA 4.0. Your footage. Your keys. No rotoscoping.

Nav Toor

296,112 views • 5 months ago

Day 14. Building the game I've dreamed of for 15 years with AI. I saved the knight on a white horse for last. I was convinced he'd be a nightmare to animate, so I kept putting him off. Once I started, it was much more manageable than I'd imagined. Guys, don't be like me. Getting started was the hardest part. The model took the least time. I used Tripo P2.0 with the new Smart UV, and I'm so happy with how the horse turned out. I prepared the armor and kept the rider separate so I could rig everything properly. Big thanks to Tripo for this one. GPT-6 Astra helped me research the animation setup. I found a $19 horse-riding pack for Unreal, even though my game is in Unity. It had no combat, but it gave us a good foundation. I did some of the transfer manually. Astra helped bring the animations across and sort out the skeleton, then retarget and adapt the movement to my character. For combat, we took arm swings from Mixamo warriors and mixed in mocap I'd recorded myself. A bit of a Frankenstein, and it worked. This has become my favorite way to build animations. Almost every clip needs adapting. I had soldier animations that were about 90% right, except the footwork looked awful in my game. With Astra, I redid the legs and kept the arm and sword movement. Now I review a clip, point out what needs changing, and choose which movements to combine. I even had Astra make a little Blender tool for pose feedback. I can move an arm wider at a particular frame and save the pose with a comment. It can then use that exact pose to correct the animation. Much easier than trying to explain it all in words. The first characters took a lot of figuring out. With this workflow in place, the knight's full animation set took just a few hours. I finally have the complete roster for the first playable level, with all their animations and even some VFX. Sound effects and a bit more game logic are still missing. Seeing all these units together makes me ridiculously happy.

Day 14. Building the game I've dreamed of for 15 years with AI. I saved the knight on a white horse for last. I was convinced he'd be a nightmare to animate, so I kept putting him off. Once I started, it was much more manageable than I'd imagined. Guys, don't be like me. Getting started was the hardest part. The model took the least time. I used Tripo P2.0 with the new Smart UV, and I'm so happy with how the horse turned out. I prepared the armor and kept the rider separate so I could rig everything properly. Big thanks to Tripo for this one. GPT-6 Astra helped me research the animation setup. I found a $19 horse-riding pack for Unreal, even though my game is in Unity. It had no combat, but it gave us a good foundation. I did some of the transfer manually. Astra helped bring the animations across and sort out the skeleton, then retarget and adapt the movement to my character. For combat, we took arm swings from Mixamo warriors and mixed in mocap I'd recorded myself. A bit of a Frankenstein, and it worked. This has become my favorite way to build animations. Almost every clip needs adapting. I had soldier animations that were about 90% right, except the footwork looked awful in my game. With Astra, I redid the legs and kept the arm and sword movement. Now I review a clip, point out what needs changing, and choose which movements to combine. I even had Astra make a little Blender tool for pose feedback. I can move an arm wider at a particular frame and save the pose with a comment. It can then use that exact pose to correct the animation. Much easier than trying to explain it all in words. The first characters took a lot of figuring out. With this workflow in place, the knight's full animation set took just a few hours. I finally have the complete roster for the first playable level, with all their animations and even some VFX. Sound effects and a bit more game logic are still missing. Seeing all these units together makes me ridiculously happy.

Stefan 3D AI

47,908 views • 24 days ago

🇫🇷 PRX alecks: "It felt like today we were desperate to win. I think the emotional part of the finals is the thing that is difficult for us to overcome because we are an emotionally driven team and sometimes it gets the better of us. Today is one of those days where we forget everything that we have practiced and everyone is trying to win the game for the team. It snowballs into terrible mistakes that they feel really bad for. It seems to be that in grand finals, the team that plays tighter, in a sense, because we play a very loose style, and this thing backfires on us when it matters the most sometimes. Unfortunately, it's a live by the sword, die by the sword kind of thing for us. We'll have to figure out how to make sure that when we arrive in grand finals, we can still keep the same play style without having the weight of expectation on our own shoulders. This time, we lost to our own emotions, not taking anything away from LEVIATÁN. I think we have another level that we can get to, and I think it'll be a nicer grand final, or at least a nicer final map." #VALORANTMasters via zhobiii @ Champs Shanghai

🇫🇷 PRX alecks: "It felt like today we were desperate to win. I think the emotional part of the finals is the thing that is difficult for us to overcome because we are an emotionally driven team and sometimes it gets the better of us. Today is one of those days where we forget everything that we have practiced and everyone is trying to win the game for the team. It snowballs into terrible mistakes that they feel really bad for. It seems to be that in grand finals, the team that plays tighter, in a sense, because we play a very loose style, and this thing backfires on us when it matters the most sometimes. Unfortunately, it's a live by the sword, die by the sword kind of thing for us. We'll have to figure out how to make sure that when we arrive in grand finals, we can still keep the same play style without having the weight of expectation on our own shoulders. This time, we lost to our own emotions, not taking anything away from LEVIATÁN. I think we have another level that we can get to, and I think it'll be a nicer grand final, or at least a nicer final map." #VALORANTMasters via zhobiii @ Champs Shanghai

VALO2ASIA

249,838 views • 3 months ago

I got this custom scorpion rigged, animated and running in my game in one full day with AI. GPT-6 built the entire rig. My part was preparing the model and motion references, then giving feedback. After trying a few approaches, I found a workflow that worked: 1. Give it a model that makes sense. I generated the model with Tripo P2.0, then cleaned it up myself. Low-poly, optimized, split into logical parts. I think that preparation mattered a lot. GPT-6 works on the mesh through code. A manageable vertex count and sensible parts give it a clearer starting point for bone placement and skinning. 2. Show it the movement you want. I took a reference image of the character and used Seedance to generate side-view animation videos. I decided which animations the game needed and selected the takes with the movements I liked. Then I gave GPT-6 the model, the videos and that animation list. There was no existing rig to start from. 3. Compare the animation frame by frame. This was the idea I wanted to test: ask GPT-6 to compare its Blender animation against the reference video frame by frame, from the same view. That gave it concrete poses and timing to work toward. I could point to where the feet should be planted or when the tail should strike. 4. Expect to give feedback. A generated run can look like hopping. An attack can start with a jump you never wanted. Choosing the references matters, and a side view still leaves depth and hidden legs to figure out. It took several iterations. I explained what looked wrong and how I wanted it to move, and GPT-6 kept refining the animation. Once the walk worked, I also had it create left and right strafe animations procedurally from that walk. Those didn't need their own reference videos. By the end of the day, the creature was in Unity. I've recorded the full workflow, including the Blender setup and feedback process:

I got this custom scorpion rigged, animated and running in my game in one full day with AI. GPT-6 built the entire rig. My part was preparing the model and motion references, then giving feedback. After trying a few approaches, I found a workflow that worked: 1. Give it a model that makes sense. I generated the model with Tripo P2.0, then cleaned it up myself. Low-poly, optimized, split into logical parts. I think that preparation mattered a lot. GPT-6 works on the mesh through code. A manageable vertex count and sensible parts give it a clearer starting point for bone placement and skinning. 2. Show it the movement you want. I took a reference image of the character and used Seedance to generate side-view animation videos. I decided which animations the game needed and selected the takes with the movements I liked. Then I gave GPT-6 the model, the videos and that animation list. There was no existing rig to start from. 3. Compare the animation frame by frame. This was the idea I wanted to test: ask GPT-6 to compare its Blender animation against the reference video frame by frame, from the same view. That gave it concrete poses and timing to work toward. I could point to where the feet should be planted or when the tail should strike. 4. Expect to give feedback. A generated run can look like hopping. An attack can start with a jump you never wanted. Choosing the references matters, and a side view still leaves depth and hidden legs to figure out. It took several iterations. I explained what looked wrong and how I wanted it to move, and GPT-6 kept refining the animation. Once the walk worked, I also had it create left and right strafe animations procedurally from that walk. Those didn't need their own reference videos. By the end of the day, the creature was in Unity. I've recorded the full workflow, including the Blender setup and feedback process:

Stefan 3D AI

12,563 views • 1 month ago

Why the character movement in my custom game engine felt janky and how I fixed it. In a game engine, most often, a character moves using the physics engine. Meaning, the player is not just a coordinate in space but a physical body. It has velocity, it handles collisions, and it interacts with the world. Now, as you might know, physics engines need stability. If you run them at variable framerates, things start breaking. Objects phase through walls or fly off into space because the math becomes unpredictable. This is why most game engines lock their physics loop to a 60Hz fixed rate. But here’s the problem: If you have a high-end system, you don't want to limit it at 60 FPS. That's a waste of good hardware. Now, that said, if the GPU is rendering at 144 FPS but the player's position (physics driven) only updates 60 times a second, it creates a micro-stutter that ruins the "smooth" feel of the game. A good way to fix this is to treat the character as two separate things: 1. The Physics Body (Invisible part): This is the "real" character. It lives in the 60Hz physics world, it moves the player and handles collisions. 2. The Visual Model and Camera (Visible part): This is what the player actually sees. It doesn't care about collisions, its only job is to look nice and smooth at whatever framerate the GPU is pushing. Once you have this separation, you can use interpolation to keep them in sync. Every time the physics clock ticks, you save the previous position of the invisible body before moving it to the new one. Between those ticks, calculate how far we are between the last physics update and the next one. By using this to drive the visible parts of the game, the stutters disappear. The physics loop stays fixed behind the scenes, while the visuals slide smoothly between the snapshots. Example: - Right after a tick: blend_weight= 0.0 (The visual model stays at the old physics position). - Halfway to the next: blend_weight= 0.5 (The visual model slides to the middle point). - Just before the next: blend_weight= 0.9 (The visual model is almost at the new physics position). Pro-Tip A critical mistake I made initially, and one many devs make, is parenting the camera and visible parts directly to the player body. If you do this, the camera inherits the discrete 60Hz physics movement by default. In that setup, interpolation won't work because the camera is "stuck" to the physics clock. For this fix to work you must decouple the camera and visuals from the body and move them separately. Player movement processing in Detis Engine: - fixed_process: Physics runs at 60Hz. Handles collisions and raw movement. - process: Variable rate. Mainly used for player input caching in the player case. - late_process: Variable rate. Handles interpolated camera movement after physics and everything else is done being processed. - render. Submits the final interpolated transforms to the GPU. The test environment in the video is running on an old 2070-based laptop. Hopefully the video compression won't introduce any stutter... I’m sharing this in hopes it helps a fellow dev. Cheers.

Why the character movement in my custom game engine felt janky and how I fixed it. In a game engine, most often, a character moves using the physics engine. Meaning, the player is not just a coordinate in space but a physical body. It has velocity, it handles collisions, and it interacts with the world. Now, as you might know, physics engines need stability. If you run them at variable framerates, things start breaking. Objects phase through walls or fly off into space because the math becomes unpredictable. This is why most game engines lock their physics loop to a 60Hz fixed rate. But here’s the problem: If you have a high-end system, you don't want to limit it at 60 FPS. That's a waste of good hardware. Now, that said, if the GPU is rendering at 144 FPS but the player's position (physics driven) only updates 60 times a second, it creates a micro-stutter that ruins the "smooth" feel of the game. A good way to fix this is to treat the character as two separate things: 1. The Physics Body (Invisible part): This is the "real" character. It lives in the 60Hz physics world, it moves the player and handles collisions. 2. The Visual Model and Camera (Visible part): This is what the player actually sees. It doesn't care about collisions, its only job is to look nice and smooth at whatever framerate the GPU is pushing. Once you have this separation, you can use interpolation to keep them in sync. Every time the physics clock ticks, you save the previous position of the invisible body before moving it to the new one. Between those ticks, calculate how far we are between the last physics update and the next one. By using this to drive the visible parts of the game, the stutters disappear. The physics loop stays fixed behind the scenes, while the visuals slide smoothly between the snapshots. Example: - Right after a tick: blend_weight= 0.0 (The visual model stays at the old physics position). - Halfway to the next: blend_weight= 0.5 (The visual model slides to the middle point). - Just before the next: blend_weight= 0.9 (The visual model is almost at the new physics position). Pro-Tip A critical mistake I made initially, and one many devs make, is parenting the camera and visible parts directly to the player body. If you do this, the camera inherits the discrete 60Hz physics movement by default. In that setup, interpolation won't work because the camera is "stuck" to the physics clock. For this fix to work you must decouple the camera and visuals from the body and move them separately. Player movement processing in Detis Engine: - fixed_process: Physics runs at 60Hz. Handles collisions and raw movement. - process: Variable rate. Mainly used for player input caching in the player case. - late_process: Variable rate. Handles interpolated camera movement after physics and everything else is done being processed. - render. Submits the final interpolated transforms to the GPU. The test environment in the video is running on an old 2070-based laptop. Hopefully the video compression won't introduce any stutter... I’m sharing this in hopes it helps a fellow dev. Cheers.

Ioannis Koukourakis

48,703 views • 9 months ago

This Chinese mathematician earned $10,000 a month inventing the hardest problems to train Neural Networks through Scale AI. Today his income dropped to zero. All the solutions are now generated by the model itself. He used to just hold the problem in his head and spell it out in plain text. His work is pure intellect. An expert in higher mathematics, he made his money hand-crafting the trickiest puzzles to test and train neural networks via RLHF. The bastion of "human" logic rested entirely on him, on people with PhDs who knew how to invent the problem. The collapse is simple. The shift to RLAIF and synthetic data. The model plays against itself, builds trees of logical inference, and solves deeper than a human can even invent the problem. No PhD data engineers, no hand-written prompt-completion examples, no manual grading. Just the model, search algorithms, and Chain of Thought. Ready-made "smart human-time" still sells on the market for many times more. His old rate was $50–100 per problem. The internal "mini-app" was written by the model too. Inside there's no pretty shell, just bare logic with exact steps: input: the problem statement inference tree: thousands of branches per second check: every step verifies itself output: a proof a human never had time to invent And here is what the whole setup looked like. He no longer needs to write an example by hand. He gave the model a direct instruction in human words, without a single formal term: "solve the problem yourself and grade yourself yourself" That's it. After that the algorithm found the solution, checked it, and trained on its own result, with no human. → the contractor got $50–100 per problem written → from 5,000 to 10,000 a month → now that income is annulled → a query to a math LLM costs 1–5 cents → a quant or an actuary runs 150,000–250,000 a year → the margin for whoever packages this into an agent is nearly 100% In the author's own words: "I'm no longer able to invent a problem the machine can't solve. The examiner became dumber than the one he's examining." But honestly, he admits the crude mistake himself, and it's not in the math, it's in the positioning. He tied his income to selling "smart human-time", to crafting formulas by hand. As long as he sells formulas, he's left behind. The machine computes faster than he can invent the problem. He names the right move himself: the role shifts from "intellectual craftsman" to "systems architect." Then he doesn't sell his time, he manages compute, packaging that same LLM into an autonomous agent that runs 24/7. Out of everything I've seen this year about the disappearance of intellectual professions, this is the most honest example: $50 per problem zeroed out to 1 cent per query, a doctor of science losing to a search algorithm, one problem stated in human words instead of a hand-written dataset, and right away an out-loud admission of the wrong business model. The barrier to entry in higher mathematics just dropped to the level of "describe the task in words." The only question is who'll be the first to stop selling their time and start managing the machine's compute.

This Chinese mathematician earned $10,000 a month inventing the hardest problems to train Neural Networks through Scale AI. Today his income dropped to zero. All the solutions are now generated by the model itself. He used to just hold the problem in his head and spell it out in plain text. His work is pure intellect. An expert in higher mathematics, he made his money hand-crafting the trickiest puzzles to test and train neural networks via RLHF. The bastion of "human" logic rested entirely on him, on people with PhDs who knew how to invent the problem. The collapse is simple. The shift to RLAIF and synthetic data. The model plays against itself, builds trees of logical inference, and solves deeper than a human can even invent the problem. No PhD data engineers, no hand-written prompt-completion examples, no manual grading. Just the model, search algorithms, and Chain of Thought. Ready-made "smart human-time" still sells on the market for many times more. His old rate was $50–100 per problem. The internal "mini-app" was written by the model too. Inside there's no pretty shell, just bare logic with exact steps: input: the problem statement inference tree: thousands of branches per second check: every step verifies itself output: a proof a human never had time to invent And here is what the whole setup looked like. He no longer needs to write an example by hand. He gave the model a direct instruction in human words, without a single formal term: "solve the problem yourself and grade yourself yourself" That's it. After that the algorithm found the solution, checked it, and trained on its own result, with no human. → the contractor got $50–100 per problem written → from 5,000 to 10,000 a month → now that income is annulled → a query to a math LLM costs 1–5 cents → a quant or an actuary runs 150,000–250,000 a year → the margin for whoever packages this into an agent is nearly 100% In the author's own words: "I'm no longer able to invent a problem the machine can't solve. The examiner became dumber than the one he's examining." But honestly, he admits the crude mistake himself, and it's not in the math, it's in the positioning. He tied his income to selling "smart human-time", to crafting formulas by hand. As long as he sells formulas, he's left behind. The machine computes faster than he can invent the problem. He names the right move himself: the role shifts from "intellectual craftsman" to "systems architect." Then he doesn't sell his time, he manages compute, packaging that same LLM into an autonomous agent that runs 24/7. Out of everything I've seen this year about the disappearance of intellectual professions, this is the most honest example: $50 per problem zeroed out to 1 cent per query, a doctor of science losing to a search algorithm, one problem stated in human words instead of a hand-written dataset, and right away an out-loud admission of the wrong business model. The barrier to entry in higher mathematics just dropped to the level of "describe the task in words." The only question is who'll be the first to stop selling their time and start managing the machine's compute.

Blaze

49,109 views • 4 months ago

Everyone is sleeping on Meta's SAM 3 release. But it's actually a big deal. Here's why: Companies spend millions paying humans to label images and videos frame by frame. A single autonomous driving dataset? Months of work, hundreds of annotators, millions in cost. Without labeled data, you can't train custom models. Without custom models, you're stuck with generic solutions. This is why most companies never move past pilots. SAM 3 breaks this cycle. First let's look at the evolution: SAM 1 segmented objects when you clicked on them. Revolutionary, but one object at a time. SAM 2 added video tracking with memory. Game-changing, but you still manually prompted every object. SAM 3 changes everything with text prompts. Type "yellow school bus" and it finds ALL of them in your image or video. Not just one. Every instance across thousands of frames. Now here's where people get confused: "Can't I just use GPT-5 or Gemini for this?" No, and here's why that's a terrible approach. Large multimodal LLMs are great for reasoning, but they're slow and expensive for production visual tasks. You're paying API costs per image, waiting seconds for responses, getting inconsistent results. SAM 3 runs in 30 milliseconds on a single GPU for 100+ objects. That's 100x faster, and you own the infrastructure. More importantly, SAM 3 gives you precise pixel-level masks, not descriptions. Try asking an LLM to segment every defective part on a manufacturing line in real-time. It won't work. SAM 3 does this effortlessly. The real breakthrough is their data engine. Meta built an AI-human hybrid system that's 5x faster for complex annotations. They trained SAM 3 on 4 million unique visual concepts - 50x more than existing benchmarks like LVIS. SAM 3 is trained on 4 million unique visual concepts, it handles everything: - Text-based concept search - Interactive refinement with clicks - Video tracking across frames - Zero-shot detection of new concepts The model is open source. Weights, code, and benchmarks are on GitHub. If you're building computer vision applications, this is the foundation model to evaluate. The annotation time savings alone will pay for integration costs within weeks. Find the relevant links in the next tweet!

Everyone is sleeping on Meta's SAM 3 release. But it's actually a big deal. Here's why: Companies spend millions paying humans to label images and videos frame by frame. A single autonomous driving dataset? Months of work, hundreds of annotators, millions in cost. Without labeled data, you can't train custom models. Without custom models, you're stuck with generic solutions. This is why most companies never move past pilots. SAM 3 breaks this cycle. First let's look at the evolution: SAM 1 segmented objects when you clicked on them. Revolutionary, but one object at a time. SAM 2 added video tracking with memory. Game-changing, but you still manually prompted every object. SAM 3 changes everything with text prompts. Type "yellow school bus" and it finds ALL of them in your image or video. Not just one. Every instance across thousands of frames. Now here's where people get confused: "Can't I just use GPT-5 or Gemini for this?" No, and here's why that's a terrible approach. Large multimodal LLMs are great for reasoning, but they're slow and expensive for production visual tasks. You're paying API costs per image, waiting seconds for responses, getting inconsistent results. SAM 3 runs in 30 milliseconds on a single GPU for 100+ objects. That's 100x faster, and you own the infrastructure. More importantly, SAM 3 gives you precise pixel-level masks, not descriptions. Try asking an LLM to segment every defective part on a manufacturing line in real-time. It won't work. SAM 3 does this effortlessly. The real breakthrough is their data engine. Meta built an AI-human hybrid system that's 5x faster for complex annotations. They trained SAM 3 on 4 million unique visual concepts - 50x more than existing benchmarks like LVIS. SAM 3 is trained on 4 million unique visual concepts, it handles everything: - Text-based concept search - Interactive refinement with clicks - Video tracking across frames - Zero-shot detection of new concepts The model is open source. Weights, code, and benchmarks are on GitHub. If you're building computer vision applications, this is the foundation model to evaluate. The annotation time savings alone will pay for integration costs within weeks. Find the relevant links in the next tweet!

Akshay 🚀

46,438 views • 10 months ago