I love coffee, and I love to improve my coffee-brewing skills.
Today, I asked Chat GPT-4 to guide me through the process of making a pour-over coffee.
The goal was to create the best possible cup of coffee using my Comandante grinder and a Hario V60 dripper.
To begin, I placed my phone near my coffee brewing-setup and started a conversation with Chat GPT-4.
I asked aloud “how can I ‘dial-in’ my Comandante grinder for a light-roast pour-over coffee?”.
GPT-4 audibly responds to my question, suggesting that I set the grinder to about 20-24 clicks from its zero point, which results in a medium grind to a medium-coarse grind.
I also spoke at my phone to ask” what’s the ideal coffee-to-water ratio for 500ml of water“; a voice within my phone recommended using about 31.25g of coffee beans.
With the grind size and coffee dosage determined, I asked my phone for “step-by-step instructions on using the Hario V60 to brew 500ml of coffee“. Here’s the resulting ai-created coffee recipe.
Coffee-Brewing Recipe & Instructions, by GPT-4:
Coffee Ingredients:
- 31.25g of coffee, ground to a medium consistency
- 500ml of hot water
Coffee Brewing (Pour-Over Method):
- Place your V60 over your carafe or mug and insert a filter.
- Rinse the filter with hot water to remove any paper taste and to preheat your V60 and carafe.
- Discard the rinse water, add your ground coffee to the filter, and start your timer.
- Pour hot water to saturate the grounds evenly for the bloom, using about double the weight of the coffee (60-70ml for 31g of coffee).
- Wait for about 30 seconds to let the coffee bloom, releasing the gases.
- After the bloom, continue pouring the rest of the water in a slow, steady spiral motion, pausing when needed to avoid overflow, until you reach 500ml.
- Let the water drain through the grounds completely, aiming for a total brew time of about 2.5 to 3 minutes.
I followed GPT’s advice and recipe, carefully executing each step. I admit, a few times, I wanted to do “my own” method that I’m more familiar with, but I continued doing what the machine told me, in the name of accurate scientific results.
As I watched the “draw-down” phase, I looked forward to trying out the resulting coffee.
I poured the freshly brewed coffee into my mug and took a sip.
To my delight, the coffee was excellent – it was at least as good as my usual brews. The AI had successfully guided me through the process, providing accurate and helpful information along the way.
In conclusion, using GPT-4 to assist in making coffee was a fun and educational experience. The AI’s recommendations for grind size, coffee dosage, and brewing instructions were spot-on, resulting in a delicious cup of pour-over coffee.