From data
to decisions.
AI systems learn patterns from examples, then use those patterns to respond to new inputs.
Examples become data
Text, images, audio, or measurements provide examples. Their quality, relevance, and representation shape what a system can learn.
Training adjusts the model
A learning algorithm adjusts internal parameters to improve performance on a training objective. Learning patterns is different from memorizing a set of hand-written rules.
New inputs produce outputs
At inference time, the trained model processes a new input and produces a prediction or response. A convincing output can still be incorrect.
Same idea.
Different kinds of input.
Choose an example to see a simplified input → model → output flow. This illustration does not run an AI model.
Predictive or generative?
Predictive systems estimate an outcome, such as likely demand. Generative systems create content, such as a draft paragraph or image. Both depend on learned patterns, and both need evaluation for the task at hand.