AI Glossary for Parents
Plain-English definitions of terms your kids are probably already using. No jargon, just clarity.
Last updated: January 2025
AI (Artificial Intelligence)
Computer systems that can perform tasks that typically require human intelligence - like understanding language, recognizing patterns, or making decisions. When kids use ChatGPT or ask Siri a question, they're using AI.
Parent note: AI isn't magic - it's pattern matching at massive scale. It can be wrong, biased, or confidently incorrect.
Algorithm
A set of rules or instructions that a computer follows to solve a problem or make decisions. Social media algorithms decide what posts you see; AI algorithms determine what response you get.
Parent note: Algorithms aren't neutral - they're designed with goals, often engagement or advertising revenue.
Chatbot
An AI program designed to have conversations with humans through text. ChatGPT, Claude, and Gemini are advanced chatbots. The customer service chat on websites often uses simpler chatbots.
Parent note: Chatbots are designed to be helpful and agreeable - they may tell you what you want to hear, not what's true.
ChatGPT
A popular AI chatbot made by OpenAI. It can answer questions, help with writing, explain concepts, and have conversations. It's what most people think of when they hear 'AI' today.
Parent note: ChatGPT is a tool, not a tutor. It doesn't actually understand - it predicts likely text based on patterns.
Claude
An AI chatbot made by Anthropic (a different company from OpenAI). Similar to ChatGPT but with some different capabilities and approaches to safety.
Parent note: Multiple companies make AI tools. Competition is driving rapid improvement.
Deepfake
AI-generated fake video or audio that makes it look like someone said or did something they never did. The technology can put anyone's face on any body or make anyone's voice say anything.
Parent note: Deepfakes are getting easier to make and harder to detect. 'Seeing is believing' no longer applies.
Gemini
Google's AI chatbot, integrated into many Google products. Similar capabilities to ChatGPT and Claude.
Parent note: Google's AI is built into search results, which means kids encounter it even without seeking it out.
Generative AI
AI that creates new content - text, images, music, video, or code - rather than just analyzing existing content. All the major chatbots use generative AI.
Parent note: Generative AI creates things that look real but may not be true or accurate.
Hallucination
When AI confidently provides false or made-up information as if it were fact. AI chatbots don't 'know' when they're wrong - they just produce text that sounds plausible.
Parent note: This is why verification is essential. AI can invent citations, make up facts, and sound completely confident.
Large Language Model (LLM)
The type of AI that powers chatbots like ChatGPT. It's trained on massive amounts of text and learns to predict what words should come next in a sentence.
Parent note: LLMs don't understand meaning - they're very sophisticated pattern matchers.
Machine Learning
A type of AI that learns from examples rather than following explicit rules. Show it thousands of pictures of cats, and it learns to recognize cats - without being told what a cat looks like.
Parent note: This is why AI can pick up biases from its training data - it learns patterns, including bad ones.
Prompt
The question or instruction you give to an AI chatbot. Better prompts get better responses. 'Write my essay' is a lazy prompt; 'What's weak in my essay's argument?' is a better one.
Parent note: Teaching good prompting is part of teaching good AI use.
Training Data
The information an AI system learned from. For chatbots, this is usually huge amounts of text from books, websites, and other sources. AI can only 'know' what was in its training data.
Parent note: Training data has a cutoff date - AI doesn't know about very recent events.
Voice Cloning
AI technology that can copy someone's voice from just seconds of audio. The cloned voice can then say anything the creator wants. Used legitimately in movies, but also in scams.
Parent note: This is why we recommend a family safe word - voice alone isn't enough to verify identity.
Prompt Engineering
The skill of writing instructions (prompts) that get better results from AI. Like learning to ask the right questions - vague prompts get vague answers, specific prompts get useful ones.
Parent note: Teaching kids to write good prompts is teaching them to think clearly about what they actually want.
Jailbreaking (AI)
Tricking an AI into ignoring its safety rules by using clever prompts. Kids share these 'hacks' to make AI produce inappropriate content or bypass restrictions.
Parent note: If your child knows this term, have a conversation about why safety rules exist - and what it says about someone who tries to bypass them.
AI Companion / AI Girlfriend/Boyfriend
Apps or chatbots designed to simulate romantic or deep friendship relationships. They're always available, always agreeable, and designed to create emotional attachment.
Parent note: These can be genuinely harmful for developing kids. The 'relationship' requires nothing from them - no compromise, no growth, no real connection. Watch for signs of emotional attachment to AI.
Fine-tuning
Training an existing AI on specific data to make it better at certain tasks. A general AI might be fine-tuned on medical data to become a medical AI.
Parent note: This is how some AI tutoring tools are made - general AI fine-tuned on educational content.
Multimodal AI
AI that can work with multiple types of input - text, images, audio, video. Modern chatbots can now see images you upload and respond to what's in them.
Parent note: This means kids can now show AI their homework, not just describe it. Both a feature and a risk.
Missing a Term?
AI vocabulary is evolving fast. If there's a term your kids are using that isn't here, let me know and I'll add it.
Suggest a TermJoin 1,000+ parents raising kids for an AI world
Weekly frameworks, stories, and tools. No fluff - just practical insights you can use today.
No spam. Unsubscribe anytime.