Artificial intelligence (AI) is one of the most powerful innovation tools of our century. 

However, there are different types with diverse and complementary functions. 
So today we'll analyze the differences between generative AI (GPT Chat) and predictive AI (Google Maps).
 

1) Main objective
Generative AI learns patterns from large volumes of data and then generates new output, such as text, images, audio, video, or code.
Predictive AI uses historical data to estimate future results or assign a category: for example, forecasting sales, detecting fraud, estimating customer churn, or deciding if an email is spam.
 

2) Output type
The output of generative AI is usually something "created": an email draft, an image, a summary, a conversational response, or a piece of code. The output of predictive/traditional AI is usually a probability, label, score, recommendation, or decision: “high risk,” “customer with an 82% probability of canceling,” “approved/rejected,” “product recommended.”
 

3) How they “think”
In simple terms:
 

• Predictive AI tries to answer: “What will happen?” or “What category does this belong to?”
• Generative AI tries to answer: “What would a new example similar to what has been learned look like?”
 

4) Data and training
Generative AI is usually trained with very large and varied datasets because it needs to learn the structure of the content to be able to produce new, convincing outputs.
Predictive AI can work with more focused and labeled data for a specific task, such as default (yes/no), churn (yes/no), demand per week, or diagnosis by class. 5) 

Everyday Examples.
 

Generative
• Spam filter.
• Credit scoring.
• Fraud detection.
• Sales forecasting.
• Recommendation tools.
• Document or image classification.
 

Predictive 
• Write an email.
• Create an image.
• Summarize a report.
• Generate code.
• Create a transcript or a conversation response.
 

6)  An easy way to remember it
• Generative: write the report.
• Predictive: calculate the probability of something happening.