Model Training & Adaptation
Online learning
Updating an AI gradually as new examples arrive.
Example
A predictor updates from a continuing stream of labeled examples.
Why people use it
It lets AI adapt as new examples arrive instead of waiting for a full rebuilding cycle.
What you'll hear
“Will new examples change it as we go?”
What this means for you
Ask whether and when incoming data changes the AI system.
Can you control it?
Developer-only
The people building or running the AI choose this setup. An everyday user generally needs their help to change how this part works.
Common questions
- Does every online AI service use online learning?
- No. Being accessible online is different from updating the AI system continuously.
- Can bad new examples quickly cause harm?
- Yes. Frequent updates can spread mistakes unless incoming information is checked.
- Can it forget older patterns?
- It can become less effective on older situations if recent examples receive too much influence.