 |
| Nutritional Analysis by OAII |
In today's culinary landscape, the fusion of artificial intelligence (AI) and nutritional analysis is paving the way for healthier and more informed cooking experiences. Join us as we explore how AI is empowering individuals to make conscious decisions about their diet and nutrition by examining the nutritional content of recipes.
Sample Prompt: I have a recipe for [Your Dish]. It includes [List of Ingredients and Quantities]. Here are the directions to prepare it [Directions]. This recipe will make [X] servings. Can you help me estimate the nutritional content per serving, including calories, protein, fat, carbohydrates, fiber, and sugar?
AI-powered nutritional analysis opens up a world of possibilities for mindful cooking, allowing you to create dishes that align with your health and wellness goals. To illustrate this, we'll perform a nutritional analysis on a classic dish: The Best Scalloped Potatoes from Food Network Kitchens.
AI-Powered Nutritional Analysis
Now, let's turn our attention to the nutritional analysis of this savory delight. AI, with its ability to process vast amounts of nutritional data, can provide valuable insights into the health profile of our favorite recipes. I checked with four seperate AI chatbots to see what they could generate. In the prompt, I included the text of the ingredients, the recipie, and the number of servings (which you can view from the Food Network Recipe page noted above).
Note: My prompt was not as specific as the one suggested above.
| Measure | Google Bard | FBN's Norm | Microsoft Copilot | ChatGPT 3.5 |
| Calories | 528 | 725 | 1,000 | ChatGPT 3.5 could not generate nutriational value and returned this instead:
"I don't have real-time internet access or the ability to perform external calculations, including nutritional analysis. However, I can guide you on how you might calculate a basic estimate using available resources. You can use online nutritional calculators or databases like the USDA FoodData Central to look up nutritional information for each ingredient. Input the quantity used in the recipe, and these tools can provide approximate values for calories, protein, fat, carbohydrates, fiber, and sugar. Once you have the individual values, you can sum them up to get an estimate per serving. If you'd like assistance with a more detailed or specific analysis, you might consider consulting a nutritionist or using specialized nutritional analysis software. If you have any other questions or need further guidance, feel free to let me know!" |
| Fat | 33.2g (44% of daily value) | 51g | 70g | |
| Carbohydrates | 50.5g (17% of daily value) | 47g | 64g | |
| Protien | 14.2g (28% of daily value) | 20g | 29g | |
| Sodium | 915mg (39% of daily value) | 930mg | 1200mg | |
| Potassium | 617mg (14% of daily value) | - | - | |
| Calcium | 326mg (33% of daily value) | - | - | |
| Vitamin A | 515 IU (15 of daily value) | - | - | |
| Vitamin C | 14mg (7% of daily value> | - | - | |
| Cholesterol | - | 170mg | 230mg | |
| Details | Provided significant notes and images | Provided a short summary | Provided a table and sources |
There are several issues to note in this quick experiment:
- No specific nutritional measure was requested so we see the default response only
- ChatGPT 3.5 was not able to complete the request at all
- Among those that could complete the request, there appears to be significant variance in the results
- Calories: range from 528 to 1000 and show a significant deviation of 472. This indicates a large discrepancy between the lowest and highest values and suggests a considerable lack of consistency in the repeated measurements.
- Fat: The fat estimates also show a significant deviation with a range of 36.8 and IQR of 17.8, indicating a lack of consistency.
- Carbohydrates: Estimates show a significant deviation with a range of 17.0 and IQR of 8.5, indicating a lack of consistency.
- Protien: Estimates show a significant deviation with a range of 49.8 and IQR of 7.0, indicating a lack of consistency.
- Sodium: Estimates show a significant deviation with a range of 285 and IQR of 142.5, indicating a lack of consistency.
However, there still can be great benefit from even using free AI chatbots as they can provide overall direction and support, even if the actual measuremenats are not as accurate as we like. For example, we can conlude from all three analysis that even though the measurements varying and inconsistent--scallped potatoes tend to be high in fat, cholestorol, carbohydrates, and sodium.
Embracing Informed Culinary Choices
As we navigate the intersection of AI and culinary arts, the power of nutritional analysis stands out as a valuable tool for conscious cooking. Whether you're exploring classic recipes or crafting your own culinary creations, AI-driven insights ensure that every meal is a step towards a healthier and more informed lifestyle.
Comments