What Is Generative AI?

What is generative AI?

In the previous courses, you learned what AI is and how to use it in everyday life. Now we're diving into one of its most exciting subfields: generative AI, meaning artificial intelligence that creates new content. Texts, images, videos, music, code, 3D models: generative AI can produce all of these, and in recent years the quality has reached a level that was unthinkable until recently.

But what exactly does "generative" mean? And how does generative AI differ from other forms of artificial intelligence? In this lesson, you'll get the overview you need to fully understand the following lessons of this course.

What does "generative" mean? The word "generative" comes from the Latin "generare": to produce, to create. Generative AI refers to AI systems that can produce new, original content that did not exist in this form in the training data. Unlike analytical AI, which classifies existing data or recognizes patterns, generative AI creates something new. When you ask ChatGPT to write a story or have Midjourney generate an image, you are working with generative AI.

Generative vs. discriminative models

To really understand generative AI, it helps to draw a distinction. AI research distinguishes between two basic types of models:

Discriminative models

Analyze and classify existing data. They answer the question: "What is this?"

Examples:

Spam filter: Is this email spam or not?
Image recognition: Is there a cat or a dog in the photo?
Credit check: Is this application high-risk?
Medical diagnostics: Does the scan show an anomaly?

Discriminative models make decisions about existing data but do not produce new content.

Generative models

Produce new data that resembles the training data. They answer the question: "What could exist?"

Examples:

ChatGPT/Claude: Produces new texts that sound human
ChatGPT/Midjourney: Produces new images from text descriptions
Veo/Kling: Produces new videos from prompts
Suno: Produces new pieces of music

Generative models learn the structure and patterns of the training data and can use them to create something new.

What content can generative AI produce?

The range of generative AI is enormous today. Here are the most important content types that AI systems can produce today:

Text generation

Large language models (LLMs) such as ChatGPT, Claude, Gemini, or Llama produce all kinds of texts: articles, emails, summaries, translations, creative writing, analyses. The quality is now so high that AI-generated texts often can no longer be distinguished from texts written by humans.

As of 2026: Texts of up to several thousand words in consistent quality, support for more than 100 languages, contextual memory across long conversations.

Image generation

Tools such as ChatGPT, Midjourney, Nano Banana (Google), Stable Diffusion, and FLUX produce photorealistic images, illustrations, art, and designs from text descriptions. From product photos to artistic paintings. The quality now surpasses many professional stock photos.

As of 2026: Photorealism at a professional level, consistent characters, reliable rendering of text in images, resolutions up to 4K+.

Video generation

Text-to-video models such as Veo (Google), Runway, Kling, and Pika can produce short video clips from text descriptions. Development is rapid: as recently as 2023 the results were barely usable, and in 2026 they are production-ready for many use cases.

As of 2026: Clips of usually 5 to 15 seconds in HD, often with matching sound, improved physical consistency, camera control, scenes made of several shots.

Audio and music generation

AI can produce speech, music, and sound effects. Text-to-speech models (TTS) such as ElevenLabs produce natural-sounding voices. Music generators such as Suno create complete songs with vocals in various genres.

As of 2026: TTS that can hardly be distinguished from real voices, personalized voice clones, complete songs with arrangements and vocals in studio quality.

Code generation

AI assistants such as GitHub Copilot, Claude Code, and Cursor write working code in dozens of programming languages. They can generate entire functions, tests, documentation, and even complete small applications.

As of 2026: Reliable generation of functions and classes, context-aware suggestions, automated bug fixing, agent-based development of entire features.

3D generation

Newer models can generate 3D objects and scenes from text or images. A game changer for game development, architecture, and product design. Examples: Meshy, Tripo, Luma AI.

As of 2026: Usable 3D models from single images or text descriptions, automatic texturing, first progress with animated 3D characters.

How does generation work? (simplified)

Without going too deep into the mathematics: during training, generative models learn the statistical structure of their training data. They build an internal model, a so-called latent space, which represents the essential features and relationships of the data in compressed form.

When you then enter a prompt, the model navigates through this latent space and samples (selects) new data points that match your request. For language models, this means that the next token is chosen token by token, usually one of the most probable ones. For image models, an image is constructed step by step from noise. The result is slightly different every time. That's why you get different results from the same prompts.

Milestones of generative AI

Development has been exponentially fast. Here are the most important milestones:

2020: GPT-3. OpenAI releases GPT-3 with 175 billion parameters. For the first time, machines can generate coherent, convincing texts. The world is amazed, but access is limited.
2021: DALL-E & Codex. For the first time, AI generates usable images from text. Codex (the basis for GitHub Copilot) shows that AI can also program.
2022: The breakthrough year. Stable Diffusion makes image generation open source. Midjourney delivers artistic images. DALL-E 2 becomes public. And in November 2022, ChatGPT is released and changes everything. An estimated 100 million users in 2 months.
2023: GPT-4 & multimodality. GPT-4 shows massively improved capabilities. Claude 2 and Gemini enter the field. Models become multimodal (text + image).
2024: Video & audio. Sora is announced, and music generators such as Suno become popular. Open models (Llama, Mixtral) catch up with the leaders.
2025–2026: Agents & integration. AI agents can complete multi-step tasks on their own. Generative AI is integrated into operating systems, office suites, and industry software. Video models produce matching sound at the same time (e.g., Google Veo, 2025), and open models such as DeepSeek R1 show that top performance can also be achieved more cheaply.

Economic significance

Generative AI is not a toy: it is changing entire industries. In 2023, McKinsey estimated the potential economic value of generative AI at 2.6 to 4.4 trillion USD per year. Particularly affected: marketing and the creative industries, software development, customer service, education, healthcare, and legal advice. Companies that use generative AI early and wisely gain a measurable competitive advantage.

Documented: According to OpenAI CEO Sam Altman, ChatGPT reached one million users after about five days (December 2022).

What is the key difference between generative and discriminative AI models?

Correct! Generative models produce new data (texts, images, videos), while discriminative models analyze and classify existing data (e.g., detecting spam, categorizing images).

Not quite. The key difference lies in the function: generative models create new content (texts, images, videos), while discriminative models analyze existing data and sort it into categories. Both can use deep learning and exist in different sizes.

Key Takeaways:

Generative AI produces new, original content: texts, images, videos, audio, code, and 3D models.
Unlike discriminative models (which classify data), generative models create something new from learned patterns.
Generation works through a latent space, in which the model samples new data points that match your prompt.
Development has been explosive: from GPT-3 (2020) and ChatGPT (2022) to AI agents and integration into everyday software (2026).
The economic impact is enormous: generative AI is fundamentally changing industries such as marketing, software, education, and healthcare.

This lesson is part of a course with final exam and certificate. View the course "Generative AI"