- Jul 19, 2026
- admin
- 0
AI Terminology for Testers
If you’re beginning your AI testing journey, you’ll encounter many new terms. Understanding these concepts will help you work effectively with AI-powered testing tools and modern automation frameworks.
- Artificial Intelligence (AI)
Definition
Artificial Intelligence is the ability of computers to perform tasks that normally require human intelligence.
Examples
- Chatbots
- Voice assistants
- Recommendation systems
- AI-powered testing tools
QA Example
An AI tool automatically generates test cases from user requirements.
- Machine Learning (ML)
Definition
Machine Learning is a branch of AI where systems learn from data instead of being explicitly programmed for every scenario.
QA Example
A testing tool learns which tests fail most often and prioritizes them during regression testing.
- Deep Learning
Definition
Deep Learning is a subset of Machine Learning that uses neural networks with many layers to solve complex problems.
Examples
- Image recognition
- Speech recognition
- Face detection
- Generative AI
Definition
Generative AI creates new content such as:
- Text
- Code
- Images
- Videos
- Audio
QA Example
Generate Playwright automation scripts from a plain English prompt.
- Large Language Model (LLM)
Definition
An LLM is an AI model trained on vast amounts of text to understand and generate human language.
Examples
- ChatGPT
- Claude
- Gemini
QA Example
Generate test cases, API tests, SQL queries, or automation code.
- Prompt
Definition
A prompt is the instruction or question you provide to an AI model.
Example
Generate Playwright TypeScript test cases for a login page.
The quality of the prompt often influences the quality of the output.
- Prompt Engineering
Definition
Prompt Engineering is the practice of writing clear, structured prompts to get better AI responses.
Example
Instead of:
Write login tests.
Use:
Generate Playwright TypeScript login automation scripts including positive, negative, boundary, and security test scenarios.
- Token
Definition
Tokens are the small pieces of text that AI models process.
For example:
“Playwright is awesome”
may be split into several tokens internally.
Why It Matters
- Longer conversations consume more tokens.
- Token limits affect how much context the model can process.
- Context Window
Definition
The context window is the maximum amount of information an AI model can consider at one time.
QA Example
You can provide requirements, API specifications, and automation code together if they fit within the model’s context window.
- Hallucination
Definition
A hallucination occurs when an AI model generates information that is incorrect, fabricated, or unsupported.
QA Example
An AI assistant invents an API endpoint that does not exist.
Best Practice
Always verify AI-generated outputs.
- Embeddings
Definition
Embeddings convert text into numerical representations that capture meaning.
Uses
- Semantic search
- Similarity matching
- Knowledge retrieval
- Retrieval-Augmented Generation (RAG)
Definition
RAG combines an LLM with external knowledge sources, allowing it to retrieve relevant information before generating a response.
QA Example
A chatbot answers questions using your organization’s test documentation instead of relying only on its training data.
- Fine-Tuning
Definition
Fine-tuning adapts an existing AI model using additional domain-specific data.
Example
Training a model on your company’s testing standards and terminology.
- AI Agent
Definition
An AI agent can perform tasks autonomously by planning, making decisions, and using tools.
QA Example
An AI agent:
- Generates test cases
- Executes tests
- Analyzes failures
- Creates reports
- Model Context Protocol (MCP)
Definition
Model Context Protocol (MCP) is an open standard that enables AI assistants to securely connect to external tools, applications, and data sources.
QA Example
An AI assistant uses an MCP server to inspect a Playwright test, retrieve test results, or interact with browser automation tools.
- Self-Healing
Definition
Self-healing allows AI-powered automation tools to recover from certain UI changes, such as updated element locators.
Benefit
Reduces automation maintenance effort.
- Computer Vision
Definition
Computer Vision enables AI to interpret images and visual interfaces.
QA Example
Visual regression testing that detects layout differences between application versions.
- Natural Language Processing (NLP)
Definition
NLP enables computers to understand and process human language.
QA Example
Converting plain English test descriptions into automation scripts.
- Synthetic Test Data
Definition
Artificially generated data that resembles real-world data while avoiding the use of sensitive information.
Benefits
- Better privacy
- Improved test coverage
- Easier data management
- AI-Assisted Testing
Definition
Using AI to enhance software testing activities rather than replacing testers.
Examples
- Test case generation
- Automation code generation
- Failure analysis
- Test prioritization
- Documentation
Quick Reference Table
| Term | Simple Meaning | QA Example |
| AI | Machines performing intelligent tasks | AI-generated test cases |
| Machine Learning | Systems learn from data | Test prioritization |
| Deep Learning | Advanced neural networks | Image recognition |
| Generative AI | Creates new content | Automation scripts |
| LLM | Language-focused AI model | ChatGPT generating code |
| Prompt | Instruction given to AI | “Generate login tests” |
| Prompt Engineering | Writing better prompts | Detailed automation requests |
| Token | Small unit of text | Input/output size |
| Context Window | Maximum information AI can process | Large requirement documents |
| Hallucination | Incorrect AI-generated information | Fake API endpoint |
| Embeddings | Numeric representation of meaning | Semantic search |
| RAG | LLM + external knowledge | Company documentation chatbot |
| Fine-Tuning | Adapting a model with domain data | Organization-specific AI |
| AI Agent | Autonomous AI assistant | Automated test execution |
| MCP | Standard for connecting AI to tools | AI interacting with Playwright |
| Self-Healing | Automatic locator recovery | UI change adaptation |
| Computer Vision | AI understands images | Visual testing |
| NLP | AI processes human language | Natural language test generation |
| Synthetic Test Data | Artificial but realistic data | Privacy-safe testing |
| AI-Assisted Testing | AI supports testing activities | Faster automation development |
