AI glossary / Safety and trust
Alignment
The work of making sure an AI actually does what people intend and behaves in line with human values.
More words in Safety and trust
- AnthropomorphismOur habit of treating AI as if it were a person with feelings and intentions, because it talks like one.
- BiasWhen an AI gives unfair or skewed results because the data it learned from was unfair or skewed.
- Black boxA system where you can see what goes in and what comes out, but not clearly why it decided what it did.
- Constitutional AIA training approach where a model is taught to follow a written set of principles, its constitution.
- Content credentialsA digital label attached to a photo or video recording where it came from and whether AI was involved.