Building Agentic AI: Workflows, Fine-Tuning, Optimization, and Deployment

7

Reasoning LLMs and Computer Use

Introduction

I would argue that the most interesting case studies in this book revolve around techniques that can be employed to make AI applications “smarter” in some way, rather than techniques that make systems cheaper or faster. We tend to want AI systems that can actually reason, remember, learn, perceive, communicate, and perhaps even set their own goals, and in many cases we are willing to wait and/or pay for these smarter systems. This chapter lays the groundwork for the final part of this book’s main mission: to bring AI applications to the forefront of usefulness.

We will begin by walking through the practical and philosophical steps that will move us toward artificial general intelligence (AGI). ...