Concern 126 · Working draft · editorial review in progress
Uncertainty in generative outputs
Uncertainty in generated answers is the difficulty of expressing how reliable a whole response is when it contains many different claims.
Understanding the concern
A model's likelihood for the next word measures a pattern in text generation, not directly the truth of every sentence. One paragraph can contain both well-supported facts and unsupported details.
A single confidence number may therefore hide important differences inside the answer. Useful communication identifies which claims were checked, which are inferences, and where evidence is missing. Evaluation should match the kind of answer users receive.
An example
A travel summary may correctly describe a city while guessing a restaurant's current opening hours.
Illustrative unless identified as a documented case.
Words in plain English
- Model
- The learned part of an AI system. It uses patterns found during training to produce an answer or prediction.
Evidence and its limits
Classifier calibration supplies a contrast; NIST discusses generative confabulation. Token probability is not presented as factual confidence.
Guo and colleagues · ICML 2017 · Research study
On Calibration of Modern Neural Networks ↗ (opens in a new tab)Studies agreement between classification confidence and observed accuracy; fluent language is not a calibrated probability.
NIST · AI 600-1 · 2024 · Official risk framework
Artificial Intelligence Risk Management Framework: Generative Artificial Intelligence Profile ↗ (opens in a new tab)Describes generative-AI risk categories and management actions; it does not measure the failure rate of every product.
Broader frameworks and background
These resources provide context. They do not establish every claim in this concern.
A question to discuss
Which specific statements are supported, rather than assigning one confidence label to the whole response?
Keep track of different views, what needs checking, and any next step in your own discussion notes.
Open discussion notes →Principles for the conversation
Before you finish · reflect together
Virtues and principles to consider
Virtues are qualities we practice in how we treat people and make choices. They are not scores for judging someone. Use these ideas to explore what a thoughtful response could look like.
Humility
Humility means recognizing the limits of our knowledge and being willing to learn from others. Expertise can be valuable without answering every question.
State where evidence is weak, invite criticism, and listen to people with different experience. Leave room to revise a decision when new information appears.
Ask together: What might we be missing, and who could help us see it?
Prudence
Prudence means using careful judgment before and during action. It balances worthwhile opportunities with the need to avoid harm that could have been anticipated.
Start with a limited trial, set stopping points, and prefer choices that can be undone when uncertainty is high. Review what happens rather than assuming the original plan remains sound.
Ask together: What is a careful next step, and when should we stop or change course?
Truthfulness
Truthfulness means making claims that match the evidence and being clear about what we do not know. It includes correcting mistakes rather than protecting a polished story.
Check sources, label guesses, and distinguish a trial result from a proven benefit. If AI helped produce a claim, someone still needs to take responsibility for checking it.
Ask together: What can we honestly say, and what still needs checking?
Editorial ratings
These are editorial judgments from the register, not measured probabilities or predictions about your situation.
About the register and scoring →More ways to explore
Search for more writing
For more specific or recent material, open a search below or ask for a web summary. AI-generated summaries can contain mistakes; open the original articles to check them.