Feedback loop
A feedback loop is a process in which observations of outcomes or outputs influence subsequent actions, inputs, or system behavior.
Hidden Influences
A cyclic process in which a system's outputs influence subsequent inputs, causing later outputs to be shaped by earlier ones.
- How it is used
- Explains how recommender systems and users influence each other and why that relationship complicates measuring impact.
- How deep it goes
- Traces recommendations and recorded interactions through a cycle, including evolving preferences and measurement limits, without specifying an implementation.
- Examples
- Illustrates feedback with popular songs, social media feed ranking, and police deployment based on reported crime and arrests.
- What you can do after
- Evaluate how recorded behavior can reflect prior recommendations rather than independent preferences or content value.
Building Reliable AI Systems
A process that uses information about a system's outputs to guide its later behavior or improvement.
- How it is used
- Uses feedback loops to improve deployed AI services and illustrate how repeated training can amplify bias.
- How deep it goes
- Traces collection, outcome measurement, investigation, intervention, and reassessment, while illustrating a harmful loop without specifying an automated implementation.
- Examples
- Uses billing-question escalation and unequal customer-service responses to illustrate beneficial and harmful loops.
- What you can do after
- Evaluate and improve deployed AI behavior using voluntary feedback, task outcomes, and human escalation.
Read how Building Reliable AI Systems explains feedback loop →
Also covered in
- Microservices Patterns, Second EditionIn early access (MEAP)
- The Architecture of ReasoningIn early access (MEAP)
- Financial AI in Practice
- Applied Reinforcement LearningIn early access (MEAP)
- Prompt Engineering in Practice
- Context EngineeringIn early access (MEAP)
What to know first
- hypothesisA hypothesis is a proposed proposition, explanation, prediction, or model whose validity or plausibility can be assessed against evidence.
- fine-tuningFine-tuning is the continued training of a pretrained model on targeted data or objectives to adapt its parameters or behavior to a task, domain, or use.
- customer satisfactionCustomer satisfaction is the degree to which customers or users evaluate a product, service, interaction, or outcome as meeting their needs and expectations.
- hallucinationA hallucination is model-generated content that is false, fabricated, or unsupported by relevant evidence but presented as plausible or factual.
Related concepts
- pheromone deposithas part · Pheromone deposit is an ant colony optimization operation that adds simulated pheromone to paths or edges traversed by artificial ants.
- pheromone evaporationhas part · Pheromone evaporation is the reduction of pheromone strength over time, weakening the influence of previously favored paths.
- pipelinecontrasts with · A pipeline is an ordered sequence of connected processing stages through which data or work flows, with each stage passing or transforming its output for subsequent stages.
Questions about feedback loop
What is feedback loop?
A feedback loop is a process in which observations of outcomes or outputs influence subsequent actions, inputs, or system behavior.
Which Manning books explain feedback loop?
Hidden Influences and Building Reliable AI Systems.
What should I know before learning feedback loop?
Manning books that teach feedback loop usually introduce hypothesis, fine-tuning, customer satisfaction and hallucination first.
What is the difference between feedback loop and pipeline?
A feedback loop is a process in which observations of outcomes or outputs influence subsequent actions, inputs, or system behavior. A pipeline is an ordered sequence of connected processing stages through which data or work flows, with each stage passing or transforming its output for subsequent stages.
What are the parts of feedback loop?
Feedback loop includes pheromone deposit, pheromone evaporation and predictive algorithmic system.
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.

