A curious maker's laboratory where the analog craft of brewing became a practical domain for understanding AI, teaching others, publishing experiments, and building digital tools.
The Purpose
Between late 2022 and 2024, generative AI moved from an abstract novelty to a wildly capable utility. Rather than just reading about the shifts, I wanted to understand them practically. I needed a domain where I could test these tools against real-world constraints—something physical, complex, and deeply familiar.
I chose homebrewing. With a decade of brewing experience, I possessed the domain expertise required to validate AI outputs. I knew when a recipe hallucinated an impossible mashing temperature, or when a fermentation profile defied biology. Homebrewing.ai became my sandbox for AI literacy, learning design, and creative experimentation.
My Role & The Output
I built and owned the entire ecosystem: the brand identity, the website experience, the content strategy, the publishing cadence, and the audience growth. Over this period, I produced more than 40 articles blending brewing practice with generative AI tooling, developed a custom Python/OpenAI tool using BeerXML, configured a specialized brewing GPT, and studied for my BJCP (Beer Judge Certification Program) exams with AI assistance.
This wasn't about replacing the brewer or building a massive commercial enterprise. It was an exercise in human-in-the-loop experimentation—maintaining a beginner's mindset, testing new workflows, and publishing the lessons learned. The work culminated in practical digital marketing experience and a featured article in Brew Your Own magazine.
The Laboratory
Hands-On Experiments
Recipe Ideation & Refinement
Iterating on complex grain bills and hop schedules, using AI to suggest historical variations while relying on human expertise to ensure the math and biology remained sound.
Python & BeerXML
Building a custom beer-name generator that parsed raw BeerXML recipe data, feeding stylistic and ingredient context into the OpenAI API for highly relevant creative outputs.
Custom Brewing GPT
Designing and launching the Homebrewing.ai GPT—a specialized assistant configured to respect established BJCP guidelines and provide structured brewing guidance.
BJCP Certification Study
Using generative AI as an adaptive tutor to prepare for the rigorous Beer Judge Certification Program, testing its ability to simulate blind tastings and quiz on off-flavor chemistry.
Published Work
The Article Collection
A curated selection of the 40+ articles published during the project, documenting the intersection of traditional craft and emerging technology.
Foundation
Establishing the core philosophy and demystifying AI for the brewing community.
Introducing Homebrewing.ai
The launch piece establishing the project's intent: blending a decade of homebrewing craft with the emerging frontier of generative AI.
An Introduction to AI for Homebrewers
A foundational guide demystifying artificial intelligence for the brewing community, focusing on practical utility over hype.
Brewing Practice
Applying generative tools directly to the craft.
Builder Experiments
Coding and building custom solutions for brewing challenges.
How to Brew Up Creative Beer Names with Python and OpenAI
A hands-on coding project using Python, BeerXML data, and OpenAI to generate context-aware, stylistically appropriate beer names.
Introducing Homebrewing.ai GPT: Your AI Brewing Assistant
The development and release of a custom GPT trained specifically to assist brewers while respecting established brewing standards.
Learning Mindset
Approaching new technology with curiosity and humility.
Education
Studying, learning, and teaching in an AI-accelerated world.
Empowering the Next Generation: The Intersection of AI and STEAM Education
Exploring how the rapid evolution of generative AI necessitates shifts in educational frameworks and critical thinking.
Become a Certified BJCP Beer Judge with the Help of Generative AI
Using AI as a study partner to prepare for the rigorous Beer Judge Certification Program tasting and written exams.
Recognition
Validating the approach through established industry publications.
Human Expertise at the Center
A persistent theme across Homebrewing.ai was that generative AI does not replace the artisan; it accelerates the curious maker. The AI could suggest a beautiful recipe for a Belgian Tripel, but it couldn't taste the astringency of a sparge gone wrong or smell the diacetyl from a stressed fermentation. The human palate, biological realities, and iterative validation remained the ultimate arbiters of success.
Takeaways
- Domain knowledge is the ultimate filter: Generative AI is most powerful when operated by someone who understands the rules well enough to know when the AI is hallucinating or breaking them.
- The power of a public sandbox: Committing to publishing a project forces a higher standard of learning. I didn't just tinker; I had to synthesize, write, and explain concepts to an audience.
- Systems over prompts: Moving beyond simple chat interfaces to build structured tools (like the Python script mapping BeerXML or the customized GPT) revealed how AI will integrate into specialized workflows.
- Tactile application breeds true literacy: Abstract conversations about AI capabilities are less effective than using those tools to solve concrete, domain-specific problems.