Learn AI · Level 1 · Foundations · step 3 of 17 · word 5 of 9
What is predictive AI?
Predictive AI uses past data to forecast or score something, such as whether a payment is fraud or how many umbrellas a shop will sell next week.
In plain English
Predictive AI answers questions like “how likely?” or “how many?”. It learns patterns from past examples, then gives a number, a score or a yes/no for new cases.
Generative AI creates something new, such as text or images. Predictive AI estimates. Most AI in daily life before ChatGPT was predictive, and much of it still is.
An example
Your bank flags a card payment as suspicious. A model compared it with millions of past payments and gave it a high fraud score. Film suggestions, delivery-time estimates and spam filters work in a similar way.
Watch out
A prediction is only as fair as the data behind it. If past decisions were biased, for example in hiring or lending, the model can repeat that bias at scale. See AI bias.
Learn AI path · Level 1 · Foundations · step 3 of 17
Next wordLarge language model (LLM) →See all of level 1