Close
When GenAI can be helpful and when it is likely to give less successful results.
In this video: Capabilities and limitations of GenAI
However, this is not the case. These models do not have the capacity to understand and cannot differentiate between truth and fiction or right and wrong. As a result, they can generate plausible outputs that are untrue or inaccurate – these are called hallucinations. They can also produce outputs that are biased or offensive, violate the copyright of others or contradict themselves mid-conversation.
Because of this, some people begin to use GenAI cautiously and review its outputs thoroughly in the early stages. Over time, they might develop a level of trust in its outputs for particular tasks. Some people report that GenAI becomes a partner in their content creation, helping with tasks like brainstorming ideas, summarising text or improving writing in a particular style.
If you are representing GenAI outputs as fact, be sure that you independently verify them for truthfulness. This includes any references that it provides.
