LLMs are terrible at building real things and everyone in AI knows it. Ask one for a CAD model and you get a pretty 3D blob an engineer can't use. SKY-001 just beat GPT-6 Astra on CAD and BIM: → 51-125x faster → 150-210x cheaper → Higher precision than every frontier model they tested It breaks an object down to the screws, sketches every part, then rebuilds it in order so it opens straight in CAD software. The team behind Maluuba built this. World models are coming for manufacturing next.
TL;DR · AI 摘要
Alex Prompter on X: "LLMs are terrible at building real things and everyone in AI knows it. Ask one for a CAD model and...
核心要点
- 主题聚焦：LLMs are terrible at building real things and ev
- 来源：God of Prompt(@godofprompt)，建议结合原文判断细节。
- AI 分析暂不可用，本条为保底评分与摘要。
Alex Prompter on X: "LLMs are terrible at building real things and everyone in AI knows it. Ask one for a CAD model and you get a pretty 3D blob an engineer can't use. SKY-001 just beat GPT-6 Astra on CAD and BIM: → 51-125x faster → 150-210x cheaper → Higher precision than every frontier model th… / X
Alex Prompter
@alex_prompter
LLMs are terrible at building real things and everyone in AI knows it. Ask one for a CAD model and you get a pretty 3D blob an engineer can't use. SKY-001 just beat GPT-6 Astra on CAD and BIM: → 51-125x faster → 150-210x cheaper → Higher precision than every frontier model they tested It breaks an object down to the screws, sketches every part, then rebuilds it in order so it opens straight in CAD software. The team behind Maluuba built this. World models are coming for manufacturing next.
Sam Pasupalak
@spisallyouneed
7h
Introducing SKY-001, the first Engineering World Model. On CAD and BIM tasks, it beats GPT-6 Astra and every frontier model we tested: -> 51-125x faster - > 150-210x cheaper -> Higher IOU than all frontier models SKY-001 lets you dream a physical object before you build it. 18
Show more
$
00:00
/$
Paid partnership
7:51 PM · Oct 19, 2026
·
5,313
Views
3
7
27
17