Video & Creative AI
InstructMesh repairs AI 3D models for real-world use
MIT News AIhigh confidence
Why it matters
- →Bridges the gap between aesthetic AI generation and functional physical fabrication.
- →Enables non-experts to repair structural flaws using natural language prompts.
- →Combines generative geometry with LLM reasoning for intuitive design iteration.
Researchers from MIT CSAIL, Google, and Northeastern University have developed InstructMesh, a system that repairs structurally flawed AI-generated 3D models for immediate fabrication. The tool pairs Microsoft’s TRELLIS generative model with GPT-4 to interpret natural language commands, allowing users to fix design errors without traditional modeling expertise. In testing, novices successfully identified and corrected nearly 90% of structural flaws in pre-existing models. The system enables the creation of functional items like custom mugs and robot enclosures that work as intended, bridging the gap between aesthetic generation and physical utility.
Read originalThe story around this
Earlier coverage that leads up to this article, and what followed. Lines connect each piece to the closest one after it, converging here.
How we got here
- 1Agent Builds 3D Paris Gallery Using Hugging Face Spaces
Hugging Face Blog · June 9, 2026 · Background
- 2Figma Enhances Design Tool with AI and Code Layers
TechCrunch AI · June 24, 2026 · Background
- 3MIT's SceneSmith Enhances Robot Training with AI Agents
MIT News AI · July 13, 2026 · Related
- 4MIT Develops System to Enhance CAD Model Generation
MIT News AI · July 16, 2026 · Related
- 5Voice Features Enable Hands-Free 3D Design with ChatGPT and Claude
Matt Wolfe · August 1, 2026 · Background
- 6Cole Medin's MIT-licensed Claude Code skills
Cole Medin · August 13, 2026 · Background
- 7ChatGPT 6 Astra: Hands-on with Multimodal Capabilities
Skill Leap AI · September 9, 2026 · Background
- 8Skild AI Uses Video to Teach Robots New Tasks
NVIDIA Blog · September 10, 2026 · Background
More from MIT News AI
© MIT News AIInvestment · $2.1M
Market & Regulationbusiness
MIT Transit Lab gets $2.1M for AI transit platform
Google.org is backing MIT’s Public Transit Intelligence Hub with $2.1 million to tackle the fragmented reality of transit control rooms. The project doesn't aim to automate dispatchers but to unify siloed data streams into a single AI-orchestrated interface, blending predictive models with large language model reasoning. This shifts the focus from benchmark-driven AI to institutional trust and operational reality in complex, dynamic environments. It signals a move toward AI that augments human judgment in critical infrastructure rather than replacing it.
MIT News AI
© MIT News AIResearchresearch
Ataraxos AI beats top human Stratego players
A multi-university team has released Ataraxos, an AI that dominates the imperfect-information game Stratego by a massive margin. Unlike previous attempts that burned millions in compute, this system uses efficient self-play and decision-time planning to outperform top humans with less than one-thirtieth of the training effort. It secured a 39-2 record against world-class players, proving that calculated risk assessment can scale beyond poker-like games. This marks a significant leap in handling hidden information without prohibitive computational costs.
MIT News AI
More in Video & Creative AI
© TechCrunch AIVideo & Creative AIimage
ChatGPT launches global virtual try-on and favorites
OpenAI is pushing ChatGPT deeper into e-commerce with a global virtual try-on feature powered by its new Images 2.5 model. Users can now upload selfies or product screenshots to visualize clothing on themselves, moving beyond simple text-based recommendations. This update also introduces a Favorites library for saving discoveries, directly challenging Pinterest and Google’s dominance in visual fashion search. While the underlying image generation improvements are notable, the real shift is ChatGPT becoming a proactive shopping assistant rather than just an information retriever.
TechCrunch AI
© The Verge Creator EconomyVideo & Creative AIvideo
RoseBerry repackages Neighbours as AI microdrama
RoseBerry Media is using an AI-powered workflow to recut classic episodes of Neighbours into vertical clips for the Epis platform, effectively rebranding archival footage as 'microdramas.' This move blurs the line between genuine short-form storytelling and simple content repurposing, raising questions about whether studios are innovating or just cannibalizing existing IP. The strategy mirrors efforts by other startups to monetize legacy media through vertical feeds, but it lacks the narrative innovation that defined the genre's origins in China. By leveraging AI for production efficiency, RoseBerry aims to introduce new audiences to the show while cutting costs, though it risks diluting the format's cultural impact.
The Verge Creator Economy
© TechCrunch AIVideo & Creative AIvideo
Synthesia launches interactive AI avatars for enterprise roleplay
Synthesia is moving beyond static video generation into agentic interaction with its new Roleplay Sessions product. This feature allows enterprises to train digital twins on specific knowledge bases, enabling employees to practice sales pitches or surveys against an avatar that listens and responds in real-time. By combining voice-to-text, LLM reasoning, and Synthesia’s proprietary video rendering, the company is bridging the gap between passive content creation and active simulation. This marks a significant shift for the $4 billion valuation startup, positioning it as a platform for immersive training rather than just marketing asset production.
TechCrunch AI