
AI Words You Keep Hearing – Simple explanations
Quick alphabetical cheat sheet (most important terms first)
- AGI → Artificial General Intelligence = AI that can do any intellectual task a human can
- Alignment → making sure AI wants the same things we want
- ASI → Artificial Superintelligence = much smarter than best humans in every domain
- Diffusion model → current best way to create images (Midjourney, Flux, Stable Diffusion…)
- Fine-tuning → taking big model and teaching it specific behaviour/style
- Generative AI → AI that creates new content (text, image, video, music, code…)
- Hallucination → when AI confidently says something completely wrong
- LLM → Large Language Model = the big text AIs (GPT, Claude, Grok, Llama…)
- Multimodal → AI that understands several types of input (text + image + audio…)
- Prompt → the instruction you give the AI
- Prompt engineering → art & science of writing better instructions
- RLHF → Reinforcement Learning from Human Feedback
- Scaling → making model bigger + more data + more compute → usually gets much better
- Singularity → point where AI improves itself so fast we can't keep up/predict anymore
- Transformer → architecture almost all powerful AIs use since ~2017
Tip: when you see scary/long AI term → just ask any recent LLM: "Explain [term] like I'm 14"
Next step → Part 4 – Generative AI – The Tools Everyone’s Using