Artificial Intelligence Wines 1.0
While the Artificial Intelligence race has been getting a lot of attention, we have quietly trademarked Artificial Intelligence wines, spirits, and cider. Who even knew that we could do such a thing? Von Stiehl Artificial Intelligence wines were introduced as red and white blends in the summer of 2023.
What do Artificial Intelligence and wine have to do with each other? Aside from how smart you are for choosing to drink von Stiehl wines, this label requires a scheme that involves microchips. Here’s how the first experiment worked. Tasters enjoyed our Artificial Intelligence wines, reviewed the wines, and posted pairing notes on our website.
We collected the reviews up to March 20, 2024. Results were compiled and submitted to two separate Artificial Intelligences. (Is Intelligences a word? Perhaps Intelligenci?) We submitted to both ChatGPT and Gemini. The results were given to our winemaking team to utilize for the production of our Artificial Intelligence wines 2.0. Version 2.0 may be the only professionally produced, crowd-sourced wines! Artificial Intelligence 2.0 wines will be released in June, 2024. What then? We’ll do it all again. Purchase and taste Artificial Intelligence wines 2.0 red and white blends, then submit your reviews following the QR codes on the back labels of the wines.
Results from both AI Chatbots are posted below. You can see what they had to say.
RESULTS FROM 1.0 RED
ARTIFICIAL INTELLIGENCE WINES (RED) CHATGPT:
Please summarize this feedback regarding Artificial Intelligence Red wine into a single paragraph:
ChatGPT:
Feedback on the AI Red wine indicates a diverse array of sensory experiences, with descriptors ranging from dark red hues to a variety of aromas including oak, cherry, and berries, and flavors encompassing notes of tannin, berry, cherry, and wood, albeit with some mentioning a spoiled or fizzy aspect. Mouthfeel and finish opinions vary widely, from smooth and long to bitter and sour, with mixed impressions overall on sweetness and depth. Some enjoy it with certain pairings like sharp cheese, while others find it challenging to drink initially but improve with aeration. Impressions of the label are generally positive, with sentiments toward its futuristic and geeky appeal, although some suggest adding more information about the AI involvement. Recommendations for AI 2.0 include adjustments such as reducing sulfites, enhancing body or oakiness, lowering sweetness, and providing more detailed AI information on the label, with preferences for a drier profile or alternative flavor progression from sweet to tart.
The current makeup of our Artificial Intelligence Red wine is:
Chancellor 30%
Marechal Foch 5%
Dechaunac 25%
Chambourcin 25%
Rougeon 15%
Residual Sweetness: 0.6% BRIX
PH (Acidity): 3.48
Alc. By Volume: 12%
How do we change this current makeup to taste more like our summarized feedback results?
ChatGPT:
To align the wine makeup with the summarized feedback, several adjustments can be made based on the preferences expressed:
- Reducing Sweetness: Since feedback mentions concerns about sweetness, reducing the residual sweetness level would be beneficial. You could aim for a residual sweetness level closer to 0.3-0.4% BRIX.
- Enhancing Body and Depth: To address comments about lacking depth, increasing the proportion of Chancellor and Chambourcin, which are known for their fuller body, could help. Adjusting the blend to include more of these varieties, perhaps increasing each to 35%, could enhance the wine’s body and depth.
- Balancing Acidity: While the acidity level seems reasonable, you might slightly adjust it to enhance balance. This could involve either tweaking the blend proportions or considering other winemaking techniques such as malolactic fermentation to soften acidity without compromising structure.
- Adjusting Oakiness: Some feedback mentions oak aromas and flavors, so you might consider incorporating oak aging or oak alternatives during winemaking. This could enhance the wine’s complexity and add desired oak characteristics.
- Addressing Fizziness: If some bottles exhibit a fizzy quality, ensuring proper degassing during bottling and storage can help minimize this issue. Additionally, evaluating the closure and bottling conditions could prevent unwanted carbonation.
- Experimenting with Yeast Strains: To refine the wine’s sensory profile, experimenting with different yeast strains during fermentation could yield desired aroma and flavor characteristics, such as reducing any perceived off-notes mentioned in the feedback.
- Labeling and Information: As suggested, updating the back label with more information about the AI involvement and possibly adding a QR code for further details can enhance consumer understanding and engagement.
By making these adjustments, you can create a wine that better aligns with the feedback’s preferences while maintaining the unique character of the AI Red blend.
ARTIFICIAL INTELLIGENCE WINES (RED) GOOGLE GEMINI:
Please summarize this feedback regarding Artificial Intelligence Red wine into a single paragraph:
Google Gemini:
AI Red received mixed reviews. People liked the deep color and some enjoyed the fruity flavors, but others found it too sweet or unbalanced. The oak and tannins were noticeable, with some preferring them and others finding them too strong. The mouthfeel ranged from smooth to bitter, and the finish varied as well. Opinions on the label were split, with some loving the futuristic design and others wanting more information about the AI process. Recommendations for AI2.0 included reducing sweetness, making it drier, offering a fuller body, and providing more information about the use of AI.
The current makeup of our Artificial Intelligence Red wine is:
Chancellor 30%
Marechal Foch 5%
Dechaunac 25%
Chambourcin 25%
Rougeon 15%
Residual Sweetness: 0.6% BRIX
PH (Acidity): 3.48
Alc. By Volume: 12%
How do we change this current makeup to taste more like our summarized feedback results below?
Google Gemini:
Reduce Sweetness:
- Lower residual sugar: Since your current residual sugar is already quite low (0.6% BRIX), further significant reduction might be difficult. However, focusing on grape varieties with naturally lower sugar content can help. Consider:
- Increasing Chambourcin: Chambourcin is known for its lower sugar content compared to Chancellor and DeChaunac. Increase Chambourcin by 5-10% (up to 30-35%) and decrease Chancellor or DeChaunac by a proportional amount.
- Introducing new grapes: Explore grapes like Dornfelder, Gamaret, or Regent, known for lower sugar and higher acidity. Consider replacing 5-10% of the blend with one of these.
Balance Sweetness and Acidity:
- Increase acidity: The feedback suggests some tasters prefer a more balanced wine. Grapes with higher natural acidity can help achieve this.
- Increase Marechal Foch: This grape has higher acidity than some others in the blend. Consider increasing Marechal Foch by 5% (up to 10%).
- Introduce grapes with high acidity: Consider adding grapes like St. Laurent or Carménère, known for their high acidity, in small quantities (around 5%).
Tannin and Body:
- Maintain or slightly increase tannins: The feedback seems divided on tannins. Maintain the current level or consider a slight increase by:
- Keeping Chancellor: Chancellor contributes some tannins to the blend.
- Introducing grapes with high tannins: A small addition (around 5%) of grapes like Tannat or Petit Verdot could add some structure. However, be cautious not to overpower the wine.
Fruity Flavor Profile:
- Maintain or adjust fruit character: The feedback mentions enjoying the fruity flavors. Maintain the current blend for a balanced fruit profile, or consider adjustments depending on the desired fruit character:
- More black fruit: Increase Chambourcin or introduce grapes like Dornfelder with a bolder black fruit character.
- More red fruit: Maintain or increase DeChaunac for a red fruit emphasis.
Remember: These are suggestions. Always conduct pilot trials and tastings to determine the optimal blend for AI Red 2.0.
Additional Tips:
- Consult a winemaker to finalize the new blend composition and fermentation techniques that can further influence sweetness and mouthfeel.
- Consider offering both a drier and a slightly sweeter version of AI Red to cater to diverse preferences.
RESULTS FROM 1.0 WHITE
ARTIFICIAL INTELLIGENCE WINES (WHITE) CHATGPT:
Please summarize this feedback regarding Artificial Intelligence White wine into a single paragraph:
ChatGPT:
Feedback on the AI White wine showcases a diverse range of sensory experiences, with descriptors encompassing light yellow hues and a plethora of aromas from citrus to pear and floral notes, along with flavors spanning from sweet to citrusy and slightly sour. Mouthfeel impressions vary from soft and smooth to cloying, while finishes range from tangy and sweet to fairly short, with opinions on sweetness being a central theme. Preferences on the wine itself are divided, with some enjoying its sweetness and finding it delicious and chuggable, while others express a preference for drier styles and find it unbalanced or uninteresting. Impressions of the label vary from fun and intriguing to potentially needing more information about AI involvement, with suggestions for a QR code for additional details. Recommendations for AI 2.0 include reducing residual sugar, making the wine less sweet, updating the back label with more AI information and a QR code, and potentially adjusting the finish to be less sour.
The current makeup of our Artificial Intelligence White wine is:
Pinot Grigio % 17
Voignier % 2.5
Chenin Blanc % 2.5
Riesling % 33
Gewurztraminer % 10
Muscat Canelli % 35
Residual Sweetness: 2% BRIX
PH (Acidity): 3.25
Alcohol By Volume: 11%
How do we change this current makeup to taste more like our summarized feedback?
ChatGPT:
To align the AI White wine with the summarized feedback, several adjustments can be made:
- Reducing Sweetness: Given the feedback’s emphasis on sweetness, reducing the residual sweetness level is crucial. Aim for a residual sweetness level closer to 0.8-1.2% BRIX to make the wine less sweet.
- Balancing Varietals: Adjust the blend to include more varietals known for their drier profile and balanced acidity. Consider reducing the proportion of Muscat Canelli, which tends to contribute to sweetness, and increasing varietals like Pinot Grigio and Chenin Blanc, known for their crispness and acidity.
- Enhancing Acidity: While the current acidity level seems reasonable, enhancing it slightly could help balance the wine, especially if reducing sweetness. This can be achieved by increasing the proportion of varietals like Riesling and Chenin Blanc, known for their higher acidity levels.
- Adjusting Aromatic Profile: To address comments about overripe fruit and sweetness, consider reducing the percentage of Muscat Canelli and Gewurztraminer, which contribute heavily to fruity and floral aromas. Increasing varietals like Pinot Grigio and Voignier, which offer more subtle aromatics, can help achieve a balanced profile.
- Experimenting with Fermentation Techniques: Experimenting with fermentation techniques, such as cooler fermentation temperatures or extended lees contact, can help enhance mouthfeel and complexity while reducing perceived sweetness.
- Labeling and Information: Update the back label with more information about AI involvement and possibly add a QR code for further details, as suggested by feedback, to enhance consumer understanding and engagement.
By making these adjustments, you can create a white wine that better aligns with the preferences expressed in the summarized feedback, offering a drier profile with balanced acidity and a more nuanced aromatic and flavor profile.
ARTIFICIAL INTELLIGENCE WINES (WHITE) GOOGLE GEMINI:
Please summarize this feedback regarding Artificial Intelligence Red wine into a single paragraph:
Google Gemini:
AI White received mixed reviews, with some enjoying its sweetness and fruity flavors while others found it too sweet and lacking in balance. The aromas were generally pleasant, with mentions of citrus, pear, and stone fruit. Mouthfeel ranged from smooth and viscous to crisp and clean. The finish was also divisive, with some preferring a tangy or dry ending and others enjoying the sweetness. While some liked the current style, the dominant recommendation for AI White 2.0 is to reduce the residual sugar for a drier wine.
The current makeup of our Artificial Intelligence White wine is:
Pinot Grigio % 17
Voignier % 2.5
Chenin Blanc % 2.5
Riesling % 33
Gewurztraminer % 10
Muscat Canelli % 35
Residual Sweetness: 2% BRIX
PH (Acidity): 3.25
Alcohol By Volume: 11%
How do we change this current makeup to taste more like our summarized feedback results?
Google Gemini:
Based on the feedback for a drier AI White 2.0, here’s how you can adjust the grape composition:
Reduce Sweetness:
- Lower residual sugar grapes: Your current residual sugar is 2% BRIX, which is already somewhat sweet. Focus on grapes with naturally lower sugar:
- Reduce Muscat Canelli: Muscat Canelli is known for its high sugar content. Reduce it by 15-20% (down to 15-20%).
- Increase Pinot Grigio and/or Chenin Blanc: These grapes typically have lower sugar levels compared to others in the blend. Increase Pinot Grigio by 10-15% (up to 27-32%) and/or Chenin Blanc by 5-10% (up to 7.5-12.5%).
Maintain Acidity and Balance:
- Maintain current acidity: The feedback doesn’t suggest a major need for acidity adjustment. Maintain the current blend for a balanced acidity profile.
Adjust Fruit Character (Optional):
- Maintain or adjust fruit character: The feedback mentions enjoying the fruity flavors. Maintain the current blend for a balanced fruit profile, or consider adjustments depending on the desired fruit character:
- More citrus: Increase Pinot Grigio or Riesling for a more prominent citrus character.
- Less floral: Reduce Gewurztraminer slightly (by 2-5%) if the floral notes are overpowering.
Remember: These are suggestions. Always conduct pilot trials and tastings to determine the optimal blend for AI White 2.0.
Additional Tips:
- Consult a winemaker to finalize the new blend composition and fermentation techniques that can further influence sweetness and mouthfeel.
- Consider offering both a drier and a slightly sweeter version of AI White to cater to diverse preferences.
