Follow us on Google News
Artificial intelligence is moving beyond writing software and helping people turn everyday ideas into physical products. A growing group of technology enthusiasts is using chatbots to design components, generate computer-aided design (CAD) files and prepare prototypes for 3D printing or commercial fabrication.
The emerging practice is being called “vibe manufacturing”, a reference to vibe coding, in which users describe the software they want and allow an AI model to generate much of the underlying code. In manufacturing, the same conversational approach is being applied to objects such as metal brackets, jewellery, robot arms, toys and electronic devices.
A November 1 report by Business Insider documented several people using AI tools to move from an idea to a physical prototype. Their experiences suggest that the technology is lowering the entry barrier to hardware experimentation, although the results still depend on repeated testing and human decisions.
From A Spoken Idea To A Manufactured Metal Part
One example involves Jeff Deehan, a real estate developer in Michigan who wanted a small steel component to secure a camper to his truck. He described the required shape and positioning to OpenAI’s GPT-6 Astra through voice mode.
The chatbot generated a design, downloaded it as a DXF file and used the website of metal fabrication company SendCutSend to submit an order. The component cost $156 and arrived around 10 days later.
The process illustrates how AI can connect several stages that traditionally require different skills. A user can explain the purpose of a component, ask a model to create its design file and use a manufacturing service to produce the physical object.
Other enthusiasts are using similar methods for more ambitious projects. Jake Fitzgerald designed a sheet-metal robot arm with a reach of 0.6 metres, a motor and a camera. He paid approximately $200 to have it manufactured.
Fitzgerald had previously faced repeated design failures, including loose screw holes, and said earlier iterations had cost him around $2,000. He reported that newer AI models had improved the design process, although such individual experiences do not establish that the technology can consistently produce engineering-ready components.
Text-To-CAD Is Changing Who Can Design Hardware
The technology behind this trend is commonly known as text-to-CAD. CAD software is used by engineers and product designers to create precise digital models of physical objects before they are manufactured.
Traditionally, producing a useful CAD model requires knowledge of dimensions, geometry, materials and the manufacturing process. Text-to-CAD tools attempt to simplify the initial design stage by allowing users to describe an object in ordinary language.
Amy Zhou and Jake Fitzgerald are developing an open-source text-to-CAD tool that has attracted more than 23,000 stars on GitHub. Zhou, who previously worked on Apple’s iPhone cases team, has also used AI to create jewellery designs from photographs that can subsequently be cast in wax.
Another example comes from London-based maker Sourish Sharma, who used AI to convert MRI data into a printable model of a human brain. He had previously struggled with an open-source tool that required substantial manual editing.
The applications are not limited to decorative objects. AI is also helping hobbyists design electronic devices by suggesting components and guiding assembly. David Plakon, for instance, used GPT-6 Astra to develop an answering machine that allows his children to leave recorded messages for their grandparents without using the internet.
Why AI-Generated Designs Still Need Engineering Checks
The growing interest in AI-assisted manufacturing does not mean that traditional engineering has become unnecessary. Generating a model that looks correct on a computer screen is different from producing a component that fits, performs reliably and can be manufactured safely.
Engineering design involves constraints that a simple visual model may not capture. These include material strength, wall thickness, tolerances, screw-hole positioning, heat resistance and the forces a component must withstand.
An October 30 technical discussion by engineering software company SimScale highlighted these limitations. It described how text-to-CAD systems can produce shapes that appear correct but lack the constraints and design relationships required for reliable manufacturing.
A model may also fail when a user changes one of its dimensions or attempts to fit it into a larger assembly. Complex components can require repeated revisions, physical prototypes and specialist review before they are ready for production.
The distinction is particularly important for products involving electricity, moving machinery, vehicles or children’s safety. An AI-generated prototype should not be treated as a certified or commercially ready product without appropriate testing and qualified engineering assessment.
What This Means For Small Manufacturers And Inventors
Vibe manufacturing could make early-stage product development more accessible to people who have ideas but lack formal CAD training. Instead of hiring a designer for every initial concept, an individual may be able to use AI to create a first draft and approach a fabrication service with a workable file.
Small manufacturers could also see more customers arriving with AI-generated designs. This may create new opportunities for businesses that offer sheet-metal cutting, machining, 3D printing and prototype production.
However, the quality of those orders will matter. Manufacturers may need to spend time correcting dimensions, checking material requirements and identifying designs that cannot be produced using their equipment.
For inventors and hobbyists, the technology could reduce the distance between imagining an object and testing it in the real world. But it does not remove the costs of materials, manufacturing, failed prototypes or safety checks.
Vibe manufacturing is therefore better understood as a new interface between people and engineering tools than as a replacement for engineers. AI can increasingly translate an idea into a physical prototype, but turning that prototype into a reliable product still requires technical judgement and testing.
