Chapter 6: Types of AI

Not all AI systems are the same. Understanding the different types of AI helps you make sense of what today’s tools can and cannot do, and where the technology is heading.

Most AI you interact with today falls into two categories: narrow AI and generative AI. A third category, general AI, remains theoretical but is the subject of active research.

Narrow AI — sometimes called “weak AI” — refers to systems designed to perform a single specific task extremely well.

Examples include:

• Facial recognition systems
• Spam filters
• Navigation and routing algorithms
• Recommendation engines on streaming platforms
• Speech recognition for voice assistants
• Image classification models

Narrow AI can outperform humans within its domain but is useless outside of it. A chess AI cannot drive a car; a driving AI cannot summarize a document.

Generative AI represents the most recent leap forward. These systems can create entirely new content — text, images, audio, video, code, and more — by learning patterns from large datasets.

Examples include:

• ChatGPT, Claude, Gemini (text generation)
• Midjourney, DALL·E, Stable Diffusion (image generation)
• Luma, Runway, Pika (video generation)
• GitHub Copilot and similar coding assistants

Generative AI is flexible, creative, and widely accessible. It has become a productivity booster for writing, research, design, education, and software development.

However, it can also produce incorrect or biased outputs and must be used with discernment.

General AI — often referred to as AGI — is a hypothetical form of artificial intelligence capable of understanding, learning, and reasoning across any domain at human-level ability.

AGI would not just perform tasks; it would generalize knowledge, adapt to new situations, and solve problems it has never encountered before.

No system like this exists today. Leading AI labs, researchers, and governments are exploring paths toward AGI, but there is no consensus on when or whether it will be achieved.

The concept of AGI raises significant questions around safety, ethics, and global impact, making it one of the most debated topics in technology.

A recent development is the rise of small language models (SLMs) — compact AI systems that run efficiently on laptops, mobile devices, or even offline.

While not as powerful as frontier LLMs, SLMs offer advantages:

• Faster and more private
• Lower energy use
• Great for on-device tasks like summarization or translation

SLMs are shaping a future where AI is not only cloud-based, but also personal and portable.

Most AI today falls into two categories: narrow AI, which performs specific tasks, and generative AI, which creates new content based on learned patterns.

General AI remains a future possibility, not a present reality — while small language models signal a growing shift toward fast, personal, and privacy-friendly AI.

Knowing the differences helps you understand the tools you use and set realistic expectations for what AI can currently do.