Student Builds Nuclear Fusion Prototype Using AI Assistant Claude

Hudhayfa Nazoordeen, a 20-year-old mathematics student at the University of Waterloo, has successfully built a nuclear fusion prototype in just four weeks, spending approximately $2,000 on parts ordered primarily online. The hand-sized device generates a glowing orb of plasma—the same state of matter that powers our sun—and represents a remarkable achievement for someone with no prior electrical engineering experience.

The prototype, while unable to produce actual nuclear fusion, successfully achieves plasma generation, which is a critical step in the fusion process. Nuclear fusion occurs when atomic nuclei merge together, producing enormous amounts of energy—nearly four million times more than burning oil or coal. Scientists worldwide have spent decades attempting to build devices capable of sustaining nuclear fusion for extended periods.

What makes Nazoordeen’s project particularly noteworthy is his innovative use of AI technology throughout the construction process. He relied heavily on Claude, an AI assistant, to ensure safety and provide technical guidance. “I’d ask Claude, ‘Should I do this dumb thing where I put this inside the plug point?’ And it’d be like absolutely not,” Nazoordeen explained, emphasizing the potentially lethal nature of working with high-voltage equipment.

The student sourced components from eBay, Amazon, and specialized suppliers, with one of the most challenging pieces being a high-voltage neon sign transformer. After unsuccessful online purchases, he called every neon sign store in his city and took an Uber an hour away to secure the right component. The device worked the same day.

Carlos Paz-Soldan, an associate professor of applied physics at Columbia University, praised Nazoordeen’s capabilities, noting that “You wouldn’t find your average student being able to do this.” While these prototypes have limited real-world applications, they serve as excellent educational tools, teaching students about high voltage, vacuum systems, and other essential fusion-related skills.

Nazoordeen’s next ambitious goal is to build a device capable of generating actual nuclear fusion, which he estimates will cost around $10,000. Despite juggling university studies, employment, and an internship at language model provider Cohere, he remains enthusiastic about pursuing this next challenge.

Key Quotes

I’d ask Claude, ‘Should I do this dumb thing where I put this inside the plug point?’ And it’d be like absolutely not

Hudhayfa Nazoordeen explained how he used the AI assistant Claude throughout his nuclear fusion prototype project, particularly for safety guidance when working with potentially lethal high-voltage equipment. This demonstrates AI’s practical application in preventing dangerous mistakes during complex technical projects.

If something wasn’t grounded properly and you mess it up, you would die. So even when we were building it, we were always just standing back and using a stick to make sure things were completely grounded.

Nazoordeen emphasized the serious safety risks involved in building the high-voltage nuclear fusion prototype, highlighting why AI assistance was crucial for someone without prior electrical engineering experience attempting such a dangerous project.

That student is very capable. You wouldn’t find your average student being able to do this.

Carlos Paz-Soldan, an associate professor of applied physics at Columbia University, praised Nazoordeen’s achievement, acknowledging the exceptional skill required to build such a device, even with AI assistance and online resources.

Our Take

This story represents a fascinating case study in AI-enabled innovation. What’s particularly striking is how Nazoordeen leveraged Claude not just for technical information, but as a real-time safety advisor—essentially an AI guardian preventing potentially fatal errors. This use case demonstrates AI’s evolution from information retrieval to active risk management.

The timing is also significant: Nazoordeen interns at Cohere, a company building the very type of AI technology he used for his project. This creates an interesting feedback loop where AI developers are also AI power users, potentially informing better product development.

Moreover, this story challenges assumptions about expertise barriers. With AI assistance and online communities, a mathematics student with zero electrical experience built a plasma-generating device in four weeks for $2,000. This suggests we’re entering an era where AI democratizes access to specialized knowledge, potentially accelerating grassroots innovation across multiple fields. The question becomes: what other previously inaccessible domains will AI unlock for determined amateurs?

Why This Matters

This story highlights the democratizing power of AI technology in complex scientific endeavors. Nazoordeen’s successful use of Claude as a safety advisor and technical consultant demonstrates how AI assistants are becoming essential tools for ambitious DIY projects that would have been nearly impossible for individuals to attempt safely just years ago. The intersection of AI guidance and accessible online resources is enabling a new generation of makers and innovators to tackle sophisticated engineering challenges.

The broader implications extend beyond this single project. As AI tools become more sophisticated, they’re lowering barriers to entry in traditionally specialized fields like nuclear physics and electrical engineering. This could accelerate innovation by empowering more individuals to experiment with advanced technologies. Additionally, Nazoordeen’s internship at Cohere, a language model provider, while simultaneously using AI tools like Claude, illustrates the symbiotic relationship between AI development and AI application. This story also underscores the importance of AI safety features—Claude’s ability to prevent potentially fatal mistakes demonstrates the critical role AI can play in risk mitigation for high-stakes projects.

For those interested in learning more about artificial intelligence, machine learning, and effective AI communication, here are some excellent resources:

Source: https://www.businessinsider.com/student-builds-nuclear-fusion-prototype-using-ai-glowing-plasma-orb-2024-9