
AI Recipes are becoming better and better each day
This is because AI software and the language learning models are becoming even more advanced to provide users with the best possible recipes that most align with their health needs and dietary restrictions. Engines like ChatGPT and Bard are delivering high-quality recipes that continue to become tastier as these engines develop.
ChatGPT's Bolognese Recipe

ChatGPT was prompted with the request to make a recipe using spaghetti, tomatoes, onions, celery, ground beef, and stocked spices and herbs. It suggested a delicious spaghetti Bolognese recipe in less than 5 seconds.
​
Ingredients:
-
Spaghetti noodles
-
Ground beef
-
Tomatoes
-
Onions
-
Celery
-
Garlic (if you have it)
-
Olive oil
-
Salt
-
Pepper
-
Italian seasoning (or a mix of oregano, basil, and thyme)
Instructions:
-
Start by cooking your spaghetti noodles according to the package instructions. Drain and set aside.
-
In a large skillet or pan, heat some olive oil over medium heat.
-
Add chopped onions and celery to the skillet. Sauté until they're softened and translucent.
-
If you have garlic, mince it and add it to the skillet. Sauté for another minute until fragrant.
-
Add the ground beef to the skillet. Break it apart with a spatula and cook until it's browned and cooked through.
-
Dice the tomatoes and add them to the skillet. Stir everything together.
-
Season the mixture with salt, pepper, and Italian seasoning (or a mix of oregano, basil, and thyme). Adjust the seasoning to taste.
-
Let the mixture simmer for about 15-20 minutes, stirring occasionally, until the flavors meld together and the sauce thickens slightly.
-
Serve the Bolognese sauce over the cooked spaghetti noodles. Optionally, you can garnish with some grated Parmesan cheese and fresh herbs like parsley or basil.
​
​

Future Studies & Research
to improve the product and design
01
Focus group testing
To ensure our product is easily navigable and convenient, we would host a focus group session to test a prototype version of the app. Participants would be able to scan available items and navigate the app; they could then share their opinions on what they like and what they don't like.
03
Testing scanning ability
AI has the capability to identify objects with a camera as machine learning progresses. To ensure that these capabilities are accurate for the safety of those with dietary restrictions, future testing of food items needs to be accomplished in order to check scanning validity.
02
AI generation recipe testing
Although AI can create recipes with items you give it, these recipes and their proportions of ingredients still need to be tested. To test the validity of these recipes, a study could be created to measure satisfaction levels of meals created by online recipes, a professional chef, and a language learning model.
04
Implementation with existing Yuka App
This prototype must work in conjunction with the existing app. In order to make the app navigable, further studies need to be conducted in navigability between existing and new features.