Iterating with Intelligence: AI-Assisted Glaze Development
Potters often face frustration when a glaze looks good in a bucket but fails in the kiln. It might crawl, craze, or look dull. Guessing your way through recipes wastes time, materials, and kiln space. You can use AI to turn intuition into a repeatable loop. This method helps you find the exact surface you want while managing safety and costs.
The Glaze Design Brief
The Glaze Design Brief is a tool to capture every decision before you touch a calculator. This tool forces you to define your needs clearly. You must list functional requirements like food safety and clay fit. You also list material constraints to avoid toxic or expensive ingredients. Finally, you set a target surface. For example, you might aim for a smooth satin finish with 60% reflectance.
Once you complete this brief, the AI uses your trusted base recipe as a chemical anchor. It then changes only one variable at a time. If you increase flux by 1%, 2%, or 3% in different columns, the data shows the exact effect of that change. This makes it easy for the AI to suggest the next step toward your target.
Scenario
You start with a matte base that is too dull. Your brief asks for a satin surface with 60% reflectance. The AI suggests a small flux increase. Your first test hits 48% reflectance. The AI sees the trend and suggests a second increase. The next test hits 61% reflectance, meeting your goal perfectly.
How to Implement
Define and Record: Fill out the Glaze Design Brief with all constraints and targets. Input your proven base recipe.
Generate and Fire: Let the AI propose a single-variable test matrix. Mix the batches, fire the tiles, and log the results.
Analyze and Iterate: Feed the measurements back into the model. The AI updates the brief and recommends the next adjustment.
Summary
A Glaze Design Brief turns vague wishes into measurable goals.
Testing one material at a time isolates cause and effect.
Looping your tests with AI helps you achieve consistent results with less waste.
Optional learning community: https://t.me/GyaanSetuAi
