From code generation to device building
The latest AI interface experiment is moving off the screen and onto the workbench. Wired reports that Schematik, created by Amsterdam-based founder Samuel Beek, is positioning itself as a “Cursor for Hardware,” a tool meant to help users describe a physical device they want to build and then receive guidance on components, sourcing, and assembly.
The pitch is easy to understand because it comes from a very concrete failure. Beek told Wired that he once blew every fuse in his house after relying on ChatGPT-generated wiring guidance for an electric door opener. He said the problem pushed him to build AI that “deeply understands what it’s talking about” in hardware contexts, where mistakes are not just annoying but potentially destructive.
Why hardware is a harder AI problem
Software “vibe coding” has become a shorthand for prompting AI systems into producing working code quickly. Hardware is less forgiving. A faulty software output may crash an app. A faulty hardware instruction can short a connection, damage equipment, or create safety risks. Wired’s reporting uses exactly that tension as the backdrop for Schematik’s emergence.
According to the article, the product lets users specify what they want to build, after which the system suggests the necessary wires and components, provides links to buy them, and acts as a guide for putting everything together. That pushes the AI interface beyond ideation and into a more operational role: selecting real-world parts and shaping the process of assembly.
The promise is obvious. Someone without deep hardware training can move from idea to object more quickly. The danger is obvious too. If the model’s judgment is wrong, the physical result can fail in ways that are more consequential than a buggy web app. Schematik’s appeal therefore rests on whether it can reduce that gap between creative ambition and reliable execution.
Early traction is already visible
Wired says Beek posted about the idea on X in February and drew strong interest from people willing to test it out. One of them, N8N branding lead Marc Vermeeren, said he used Schematik to build several devices, including an MP3 player and a Tamagotchi-style bot called Clawy to help manage Claude coding sessions. The article describes other users creating their own variations as well.
That matters because maker tools often live or die by community enthusiasm before they mature into polished businesses. In this case, the startup appears to have both user experimentation and investor backing. Wired reports that Schematik has raised $4.6 million from Lightspeed Venture Partners and that Beek plans to build a business around it.
Anthropic’s role is not investment, but enablement
The article’s title suggests Anthropic wants in, and the body clarifies what that means. Anthropic engineer Felix Rieseberg posted on X that the company has enabled a Bluetooth API for makers and developers. In context, that looks like platform support for the kind of hardware-building workflows tools like Schematik want to unlock.
That is an important distinction. Based on the supplied text, Anthropic is not described here as an investor in Schematik. What Wired shows is a growing alignment between frontier AI models and maker-oriented tooling. If large model providers expose interfaces useful for devices, hobby electronics, and connected products, the boundary between coding assistant and hardware assistant starts to erode.
The bigger shift behind the story
Schematik is interesting not just because it helps people assemble gadgets, but because it extends a broader pattern in AI product design. Users increasingly expect models to be agents across workflows, not just answer engines. In software, that expectation has already become normal. In hardware, it remains experimental, partly because the cost of being wrong is higher and the relevant knowledge is more grounded in parts, tolerances, connections, and constraints.
That is why Schematik’s description as a “Cursor for Hardware” resonates. It translates a familiar software metaphor into a more difficult domain. Whether the comparison fully holds is still an open question. But the ambition is clear: reduce the distance between a prompted idea and a functioning physical artifact.
Why this could matter beyond hobbyists
If these tools improve, their relevance will not stop at weekend tinkering. Faster iteration on prototypes could matter for education, product design, internal tooling, and small manufacturing teams. The core advantage is not magic. It is compression. A system that can recommend parts, suggest assembly steps, and stay context-aware across an entire build process can lower the activation energy for making something real.
That said, Wired’s framing keeps the central caution in view. Hardware is where vague AI confidence can produce burned components, wasted time, or worse. The real test for this category is not whether it feels creative. It is whether it can be trusted when the wires are real.
What to watch
- Whether maker communities continue adopting AI-native hardware tools in public build workflows.
- How far model providers like Anthropic go in exposing interfaces aimed at devices and peripherals.
- Whether reliability and safety become the main differentiators in AI-assisted hardware design.
Schematik captures a real frontier in applied AI: the move from generating software to orchestrating physical creation. The opportunity is large. So is the penalty for getting it wrong.
This article is based on reporting by Wired. Read the original article.
Originally published on wired.com
