cloudnative wiki

AI vs ML vs DL vs GenAI vs LLMs — What's the Difference?

2 min read

AI vs ML vs DL vs GenAI vs LLMs — What’s the Difference?

graph TD
    AI[Artificial Intelligence - AI<br/>Simulating human intelligence] --> ML[Machine Learning - ML<br/>Learning from data without explicit rules]
    ML --> DL[Deep Learning - DL<br/>Multi-layer neural networks]
    DL --> GenAI[Generative AI - GenAI<br/>Creating novel content: text, code, images]
    GenAI --> LLMs[Large Language Models - LLMs<br/>Transformer architectures trained on massive text]

    AI --> NLP[Natural Language Processing - NLP]
    NLP -.-> LLMs

The landscape of Artificial Intelligence (AI) encompasses various nested subfields, techniques, and applications. Here is an architectural overview of how they fit together:

1. Artificial Intelligence (AI)

The broadest umbrella term, AI refers to the development of computational systems that perform tasks typically requiring human intelligence:

  1. Reasoning and problem-solving
  2. Knowledge representation and management
  3. Planning and decision-making
  4. Perception and sensing
  5. Natural Language Processing (NLP)
  6. Robotics and computer vision

2. Machine Learning (ML)

A subset of AI, ML develops algorithms and statistical models that learn patterns from data rather than relying on hardcoded procedural rules:

  1. Supervised learning: Regression, classification (labeled datasets).
  2. Unsupervised learning: Clustering, dimensionality reduction (unlabeled datasets).
  3. Reinforcement learning: Agent optimization via reward signals and policy gradients.

3. Deep Learning (DL)

A subfield of ML based on artificial neural networks with multiple representation layers:

  1. Convolutional Neural Networks (CNNs): Spatial pattern processing for computer vision.
  2. Recurrent Neural Networks (RNNs) / LSTMs: Sequential processing for time-series data.
  3. Transformers: Self-attention architectures that revolutionized language and multimodal modeling.

4. Generative AI (GenAI)

A subset of Deep Learning focused on generating novel, synthetic data artifacts rather than merely classifying or predicting values:

  • Text and code generation (Transformers)
  • Image and video generation (Diffusion models, GANs)
  • Audio and speech synthesis

5. Large Language Models (LLMs)

A specific class of foundation models trained on massive text corpora using the Transformer architecture:

  • Autoregressive next-token prediction
  • Zero-shot and few-shot reasoning
  • Instruction fine-tuning (RLHF / DPO) for aligned dialogue execution

6. Summary Taxonomy

  • AI is the overarching field of intelligent machines.
  • ML is learning patterns from data.
  • DL is learning patterns using deep neural networks.
  • GenAI is generating new content using deep generative models.
  • LLMs are language-specialized generative foundation models.

Graph View

Backlinks