Generative AI & LLMs
Natural language understanding
Computer methods for working out what someone's words mean or what they are asking for.
Example
A system interprets 'move it to Friday' using the conversation's background information.
Why people use it
It helps software work out what a person means, not just which words appear.
What you'll hear
“It needs to understand what 'move it' refers to.”
What this means for you
Test ambiguous phrasing and whether the system asks useful clarifying questions.
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 the name mean the computer has human awareness?
- No. It describes a practical ability to interpret language, rather than proof of human-like experience.
- Why can sarcasm be difficult?
- The intended meaning can differ from the literal words and depend on shared background information.
- Can an unclear request have several reasonable meanings?
- Yes. Sometimes asking a follow-up is more useful than making a confident guess.