AI Basics

What is a Model Collapse?

Model collapse occurs when an AI system is trained primarily on data created by other AI models instead of original human content. When a model learns from synthetic output, it begins to repeat and amplify the mistakes or biases present in that data. Over time, the system loses its ability to understand the nuances of human language and logic. This creates a feedback loop where the quality of the output steadily degrades, eventually making the model less accurate, less creative, and less reliable for real-world tasks.

Why this matters to you

As the internet fills with AI-generated content, it becomes harder for developers to find high-quality human data. If future AI systems rely too heavily on this machine-made information, they may become less useful and more prone to errors, which directly impacts the quality of the tools you use in your daily work.

How you might hear this

We need to be careful about how we source our training sets because relying too much on automated content could lead to model collapse and lower the performance of our internal tools.

AI Jargon Buster

Search any AI term, explained in plain English.

Type a term below and search. You will be taken straight to the tool.

AI is changing how you get hired

See how your CV performs against the ATS algorithms that screen candidates before a human ever reads your application.

Try the CV Optimiser

Understand the bigger picture

How AI job displacement actually works, what it means for your career, and what to do about it.

When the Ground Shifts

Job hunting right now?

A 39-page guide covering the whole search: beating the ATS, fixing your CV and LinkedIn, and walking into interviews prepared. Written by a recruiter.

Get the Job Search Playbook

Related terms

Training Data Synthetic Data Generative AI (Gen AI) Hallucination Data Poisoning Large Language Model (LLM)

This tool uses AI to generate your results.