Evaluation & Quality
Epistemic uncertainty
Uncertainty caused by gaps in what an AI knows or limitations in how it works.
Example
An AI system is uncertain about a kind of case rarely represented in training.
Why people use it
It highlights uncertainty caused by what the system does not yet know.
What you'll hear
“We don't have enough examples of this situation.”
What this means for you
Identify gaps in coverage rather than interpreting every uncertainty as random noise.
Can you control it?
No
No direct control. This describes a wider issue, concept or result rather than something you can simply switch on or off in a tool.
Common questions
- Can an AI sound certain in an unfamiliar situation?
- Yes. Its confidence may fail to reflect important gaps in what it has learned.
- Can new information reduce it?
- Often. Better examples or better assumptions can reduce gaps in what is known.
- Is it the same as unavoidable randomness?
- No. Lack of knowledge and variation that remains despite good knowledge are different sources of uncertainty.