{"slug": "why-can-ai-generate-super-mario-but-not-a-wedge-ramp-for-my-robot-vacuum", "title": "Why can AI generate Super Mario but not a wedge ramp for my robot vacuum?", "summary": "A developer reports that AI generation tools struggle to produce functional 3D parts like a wedge ramp for a robot vacuum, while excelling at aesthetic objects such as a Super Mario figurine. The user, who bought a Bambu P2S printer, found that text-to-model AIs and agent-based Python geometry generation failed, but succeeded by decomposing the part into ordered steps and executing them in Blender via blender-mcp. The developer shared the approach in a GitHub repository (spec-3d-model) and questions whether the issue stems from training data, representation, or evaluation, and whether converting 3D modeling to code is the right strategy.", "body_md": "I've been puzzled by something: AI generation can produce an elaborate figurine, a cartoon character, even a convincing Super Mario — yet it can't reliably make a simple wedge ramp so my robot vacuum can climb a step.\n\n```\n  For context: I bought a Bambu P2S but can't model. I tried the \"describe\n  it and get a model\" AIs — the output is unusable, you can't adjust it,\n  it's never quite what I meant. I tried having an agent write Python to\n  build geometry directly — it tops out at simple primitives.\n \n  What finally worked: geometric decomposition. I break a complex part into\n  ordered, grouped steps, describe each as a small spec, and let an agent\n  execute them in Blender (via blender-mcp). That process turned out to\n  abstract into a small engine — the key insight being it converts the 3D\n  spatial reasoning LLMs are bad at, into the structured code they're good\n  at. I wrote it up here: https://github.com/zhuchaokn/spec-3d-model\n \n  My questions:\n  - Why is \"functional part\" generation so much weaker than\n  \"figurine/aesthetic\" generation? Is it data (no parametrized-CAD training\n  sets), representation (mesh vs B-rep), or evaluation (nobody benchmarks\n  \"does it print / is it watertight\")?\n  - Is \"turn 3D modeling into code for an LLM\" the right framing, or am I\n  missing something better?\n```\n\nComments URL: [https://news.ycombinator.com/item?id=49405520](https://news.ycombinator.com/item?id=49405520)\n\nPoints: 2\n\n# Comments: 0", "url": "https://wpnews.pro/news/why-can-ai-generate-super-mario-but-not-a-wedge-ramp-for-my-robot-vacuum", "canonical_source": "https://news.ycombinator.com/item?id=49405520", "published_at": "2026-08-23 01:45:17+00:00", "updated_at": "2026-08-23 02:13:24.660467+00:00", "lang": "en", "topics": ["generative-ai", "ai-tools", "ai-agents"], "entities": ["Bambu P2S", "Blender", "blender-mcp", "spec-3d-model", "Super Mario"], "alternates": {"html": "https://wpnews.pro/news/why-can-ai-generate-super-mario-but-not-a-wedge-ramp-for-my-robot-vacuum", "markdown": "https://wpnews.pro/news/why-can-ai-generate-super-mario-but-not-a-wedge-ramp-for-my-robot-vacuum.md", "text": "https://wpnews.pro/news/why-can-ai-generate-super-mario-but-not-a-wedge-ramp-for-my-robot-vacuum.txt", "jsonld": "https://wpnews.pro/news/why-can-ai-generate-super-mario-but-not-a-wedge-ramp-for-my-robot-vacuum.jsonld"}}