Garbage in, garbage out
Garbage in, garbage out is the principle that poor-quality, unsuitable, or unsupported inputs tend to produce unreliable or poor-quality computational outputs.
Software Testing with Generative AI
The principle that poor or insufficient input tends to produce poor or insufficient output from a computational process.
- How it is used
- Improves LLM-assisted software development and test planning by focusing attention on input quality and human review.
- How deep it goes
- Explains the principle through training-data and prompt examples, then stops short of formal model internals.
- Examples
- Uses flawed training data, Copilot suggestions, and vague versus contextual file-upload test prompts.
- What you can do after
- Evaluate and improve LLM inputs while verifying generated code and testing suggestions before using them.
Read how Software Testing with Generative AI explains garbage in, garbage out →
Starting Data Analytics with Generative AI and Python
Garbage in, garbage out is the principle that poor-quality inputs can produce unreliable outputs even when the processing or analysis is sophisticated.
- How it is used
- Guides data-quality checks before analytics or AI-supported decisions.
- How deep it goes
- Explains reliability through business analytics and a coin-flip analogy, but defers a universal data-quality solution.
- Examples
- Uses coin flips, business analytics, and skewed prompt data to illustrate unreliable or biased outputs.
- What you can do after
- Evaluate analysis reliability by inspecting inputs, interpreting variables in context, and accounting for data compromises.
Read how Starting Data Analytics with Generative AI and Python explains garbage in, garbage out →
Related concepts
- language modelused by · A language model is a computational model that represents patterns in linguistic tokens to estimate sequence probabilities and predict or generate language from context.
These guides are generated from the text of Manning books by Manning's knowledge graph and link to the sections they draw on. The full books are available on liveBook and at manning.com.