The following diagram provides a high-level overview of the relationship between AI, ML, deep learning, GenAI, and LMs. Each layer represents a subset of the previous, showcasing the evolution of AI technology. Starting with AI in 1956, ML in 1997, and deep learning in 2017, the diagram also highlights how LMs and GenAI, which emerged more recently, fit into this broader context. Further details on these technologies are discussed in the following section.

Figure 1.1 – Brief AI history

Let’s dive deeper.

AI

While AI refers to the broader goal of simulating human intelligence, ML is one of the core methods used to achieve it. ML provides the statistical techniques and models that enable AI to learn from data.

AI is like a smart assistant that can perform tasks that typically require human intelligence, such as understanding language, recognizing images, making decisions...

Azure provides a comprehensive suite of AI services designed to accelerate the development of intelligent applications. These services cover a broad range of capabilities, including vision, language, speech, search, and GenAI. With prebuilt models, APIs, and customization options, developers can quickly integrate advanced AI features into their solutions without needing deep ML expertise.

At the core of this ecosystem is the Azure AI Foundry platform (discussed in detail in the AI Foundry section of Chapter 2)—a unified environment for building, deploying, and managing AI applications. It streamlines the development process by combining model training, data integration, and deployment workflows with enterprise-grade security and compliance features. Azure AI Foundry empowers teams to collaborate efficiently while scaling AI solutions across the organization.

Figure 1.4 – Overview of Azure AI services

The following...

In this chapter, we explored the fundamental concepts of AI and ML, delving into how they power modern innovations across industries. By understanding key types of learning—supervised, unsupervised, and reinforcement learning—we gained insights into how AI systems are trained to make predictions and decisions. Additionally, we touched on deep learning and GenAI, which extend AI’s capabilities by enabling the creation of original content such as text and images.

These skills are essential for building AI-powered applications that automate tasks, analyze data, and generate useful content, enhancing productivity and innovation. By understanding how LMs and neural networks function, you’re equipped to develop smarter, more efficient solutions. The knowledge gained here is foundational for leveraging AI in real-world scenarios.

In the next chapter, we will explore how to plan AI solutions that meet responsible AI principles. We’ll also cover...

Answer the following questions to test your knowledge of this chapter:

1. Which of the following best describes the primary function of ML?
   1. ML is used to manually program computers to perform specific tasks
   2. ML focuses on training models to make predictions based on data without explicit programming for each task
   3. ML is primarily used for data storage and retrieval
   4. ML is a subset of NLP that deals with speech recognition

   Correct answer: B

2. What is the key difference between supervised learning and unsupervised learning in ML?
   1. Supervised learning uses labeled data for training, while unsupervised learning does not use labels
   2. Supervised learning is used for speech recognition, while unsupervised learning is used for image recognition
   3. Supervised learning involves reinforcement, while unsupervised learning involves deep learning
   4. Supervised learning is always more accurate than unsupervised learning

   Correct answer: A

3. Which AI technology is specifically designed to generate...

To learn more about the topics that were covered in this chapter, look at the following resources:

• AI architecture design: https://learn.microsoft.com/en-us/azure/architecture/ai-ml/
• Deep learning vs. machine learning in Azure Machine Learning: https://learn.microsoft.com/en-us/azure/machine-learning/concept-deep-learning-vs-machine-learning?view=azureml-api-2
• AI and machine learning on Databricks: https://learn.microsoft.com/en-us/azure/databricks/generative-ai/generative-ai
• Grounding LLMs: https://techcommunity.microsoft.com/t5/fasttrack-for-azure/grounding-llms/ba-p/3843857
• Understand embeddings in Azure OpenAI Service: https://learn.microsoft.com/en-us/azure/ai-services/openai/concepts/understand-embeddings
• Strategies for Optimizing High-Volume Token Usage with Azure OpenAI: https://techcommunity.microsoft.com/t5/fasttrack-for-azure/strategies-for-optimizing-high-volume-token-usage-with-azure/ba-p/4007751
• What’s Azure AI Search...