What makes generative AI different from other deep learning applications?
Most deep learning applications before ~2020 were discriminative — classify this image, predict this value, detect this object. Generative AI produces new content resembling its training data: new text, new images, new audio. The architectural shift that enabled this at scale was the transformer, which is why 'generative AI' and 'transformer-based models' arrived together in most people's vocabulary.
Answered in
AI vs Machine Learning vs Deep Learning vs Generative AI: The Actual HierarchyAI, ML, deep learning, and generative AI aren't four competing terms — they're four nested circles. Here's exactly what's inside what, and why it matters.
Read the full analysisOther questions this article answers
More learning & thinking with ai questions
- What's the actual difference between machine learning and deep learning?
- Do I need to understand neural networks to use LLMs well?
- What's the difference between fine-tuning and prompt engineering, and when do I use each?
- What makes something an AI agent instead of just an LLM call?
- Why do vector databases matter for AI applications?
- Is machine learning the same thing as AI?
- Is deep learning always better than other machine learning approaches?
- Is ChatGPT machine learning, deep learning, or generative AI?
Every answer on Crashtech is written by the editor of the article it comes from — never auto-summarised. Browse all answers or the Learning & Thinking with AI beat.