| 🏠 Back to Exam Syllabus | 📺 RooCloud on YouTube | 🌐 RooCloud Practice Exams |
Types of AI: ANI, AGI, ASI, Generative, Agentic and Machine Learning
This episode of the ISACA Advanced in AI Security Management (AAISM) exam prep series opens Domain 3 by building the mental map of the AI field that the rest of the technical controls work depends on. It frames how the broad concepts of AI, machine learning, and deep learning relate, the functional and capability classifications you are expected to know, the major model families that show up in the enterprise, the way machines actually learn, and the vocabulary every security manager needs.
What this episode covers
- How AI, machine learning, deep learning, and generative AI nest together as a set of progressively narrower circles.
- The four functional types of AI — reactive, limited memory, theory of mind, and self-aware.
- The three capability tiers — artificial narrow, general, and super intelligence.
- The major model families including generative, agentic, and predictive models, and their distinctive risks.
- The three machine learning paradigms — supervised, unsupervised, and reinforcement learning — and where each fits.
- Neural networks, deep learning, and algorithms, including layers, hyperparameters, and ensemble modeling.
- Essential vocabulary including large language models, transformers, prompts, drift, overfitting, underfitting, and AI as a service.
Watch the full episode above for the worked examples and detailed explanations of each concept.
Frequently Asked Questions
How do AI, machine learning, deep learning, and generative AI relate to each other?
Picture a set of nested circles. The widest is artificial intelligence, the broad goal of machines doing tasks that normally need human intelligence. Inside it sits machine learning, where systems learn patterns from data. Inside that sits deep learning, which uses many-layered networks to learn abstract features. And generative AI lives within deep learning, creating brand-new content from the patterns it has absorbed.
What are ANI, AGI, and ASI?
Artificial narrow intelligence (ANI), sometimes called weak AI, is limited to one specific domain like translating languages, and is essentially all the AI in use today. Artificial general intelligence (AGI) could handle any intellectual task a human can but does not yet exist. Artificial superintelligence (ASI) is a hypothetical stage where machine intelligence would surpass the best human minds across every field, and remains purely a concept.
What is the difference between generative, agentic, and predictive AI?
Generative models, such as generative adversarial networks, create new content like images or audio by pitting two networks against each other. Agentic AI makes autonomous decisions and takes actions to achieve a goal, with fewer hallucinations than ordinary generative AI but still needing human oversight. Predictive models use historical data to forecast future events such as equipment failures.
What are the three main machine learning paradigms?
Supervised learning trains on labeled data and covers regression for continuous numbers and classification for categories. Unsupervised learning works on raw, unlabeled data and performs clustering, association rules, and dimensionality reduction. Reinforcement learning has an agent learn by trial and error through rewards and penalties, suiting autonomous control.
What are overfitting and underfitting in machine learning?
Underfitting means a model is too simple to capture the patterns in the data. Overfitting means a model has memorized the training data, including the noise, and fails on anything new. Both are central concepts that every security professional should understand when evaluating model reliability.
📚 Master the ISACA AAISM Exam!
Ready to test your knowledge? Access chapter-specific Multiple Choice Questions (MCQs) and full-length practice exams for the ISACA AAISM certification at RooCloud.com. Solve the chapter-wise questions to reinforce this lesson before moving to the next episode.
Reference: This article is based on concepts discussed in Types of AI: ANI, AGI, ASI, Generative, Agentic & Machine Learning.