In an Era of Rapid Technological Advancement.The use of Artificial Intelligence (AI) is no longer limited to just work or entertainment–it can also assist in health management and weight loss. AI is a powerful tool that makes weight control more effective and tailored to individual lifestyles. Here, we’ll explore how AI can help you lose weight efficiently.
1. Using AI to Collect Data
The most important aspect of leveraging AI is data collection to train the AI. For example, if you want to lose weight, you first need to understand: “How does weight loss work?”
For me, who had purchased personal trainers multiple times but saw little progress, the problem seemed to stem from food intake. Therefore, when the goal was set to track daily calorie consumption, the challenge became finding a way to record it without feeling bored or annoyed.
The method I used was:
Take pictures of each meal, send them via Line > then upload to Chat GPT 4o > Google Sheet to log daily food intake.
Create a personal database (Database) to analyze calories by designing a table format that specifies what data to track.
2. Analyzing Food with AI
I used Chat GPT 4o to analyze food by providing clear instructions for accurate analysis.
Example Prompt:
“You’re a nutritionist expert in Bangkok. You will help me identify the dish name and calories. You’ll also provide a short analysis of micronutrients such as protein, carbs, fat, saturated fat, sugar, and sodium contents in the food.
If there’s more than one dish, summarize everything including total calories in the picture, which might be combined for different dishes or estimated calories per 1 serving size/meal.”
Reason for specifying Bangkok: To give the AI context that we are in Thailand, so it interprets unfamiliar dishes as Thai food first. This worked well, as the AI successfully identified Thai dishes like “Yam Naem” (fermented sausage salad) and “Kuay Jab” (noodle soup).
3. Converting Data into JSON
After analyzing the food, I used AI to convert the data into JSON format for use in Google Sheets.
Example Command:
“You are the data parser. Extract the data and replace the value for the JSON below. Return only the JSON with exact formatting without any preceding words (calories should be numbers only). If there’s more than one food recorded, summarize it into one entry.”
After collecting daily food data, I used Qwen to analyze the information, such as:
What foods do we eat regularly?
How should we adjust our eating habits to lose weight?
Advantages of Qwen:
Reads and analyzes data effectively.
Can upload files and perform long analyses without requiring payment.
Disadvantages of Qwen: Cannot pull data from Line of automate tasks like Chat GPT 4o.
Thus, I used Chat GPT 4o for data collection and Qwen for analysis.
5. Solving the Eternal Question: “What Should I Eat Today?”
When creating tables in Google Sheets, I also collected additional data such as restaurant names, locations, and ratings to help decide what to eat on indecisive days. This made life much easier.
Solution: A simple Streamlit web app was created to randomly recommend food based on weighted probabilities, such as:
If not eaten in a while, increase the likelihood.
If the review score is high, increase the likelihood.
If the food is healthy, increase the likelihood.
In Summary: AI is a tool that makes weight control easier and more efficient, whether it’s planning nutrition, collecting data, analyzing results, or providing motivation. However, using AI should be part of a balanced approach, combined with natural health practices.
Precautions: > Use trusted apps or tools. > Consult a doctor or expert if you have doubts about your weight loss plan.
Finally, try following these methods and start managing your weight effectively today! For more information, please contact us @davoy.tech .
