1 Understanding Generative AI Basics

 

This chapter covers

  • An introduction to generative AI: what’s really going on under the hood?
  • Distinguishing between the many generative AI models
  • A review of the global trends that brought us to the generative AI revolution

Welcome! As advertised, this book is obsolete. Which means that by the time you got around to opening it, most of what’s written here will either not work, or will be so outdated as to be useless. Now I bet you’re feeling just a bit silly for sinking good money into a product like this. Well I assure you: you don’t feel half as weird for buying this book as I felt writing it.

We will definitely get around to the fun stuff - or, at least stuff that was fun back in the Before Times when I was originally writing this - soon enough. We’ll learn how generative artificial intelligence can be used for far more than just stand-alone ChatGPT prompts. Curious to see whether:

  • AI can read statistical data archives and then derive serious insights?
  • AI can access the live internet, aggregate data from multiple sites, and use that to pick out real-world trends?
  • AI can accurately summarize large bodies of your own text-based content?
  • AI models can be fine-tuned to provide responses that’re a better match to your needs?
  • AI models can be used to generate original video and audio content?

Me too. Let’s find out.

1.1 Stepping into the generative AI world

1.2 Categorizing AI models by function and objective

1.2.1 Understanding usage tokens

1.2.2 GPT-4 models

1.2.3 GPT-3.5 models

1.2.4 GPT-3 models

1.3 Model fine-tuning

1.4 The technologies that make generative AI work

1.5 AI and Data Privacy and Ownership

1.6 AI and Reliability

1.7 What’s still ahead:

1.8 Summary