Page 4 - AI Vol 1: Foundations of AI
P. 4
02 WHAT IS AI?
02A. INTRODUCTION TO AI
In the last decade, AI has evolved from a concept process in the human brain. Deep Learning
in science fiction to an increasingly important utilizes neural networks, algorithms inspired by
part of our professional and personal lives. Public the structure and function of the human brain.
agencies will certainly not be immune from what These neural networks are composed of layers of
promises to be a rapid and transformational period interconnected nodes, much like neurons in the
in technological development. As we explore AI human brain, which allow the system to recognize
and its implications in this document, we begin complex patterns in large amounts of data.
by addressing some of the fundamentals of AI
and Machine Learning (“ML”). Although AI may seem like a recent phenomenon,
elements of AI and ML have long been
AI is the broad term used to describe the concept incorporated into many everyday products and
of machines being able to carry out tasks in a applications. Take voice assistants like Siri,
way that we would consider “smart.” That is, Alexa, and Google Assistant, for example. These
computers with a human-like “intelligence” which systems utilize AI to process audio data, recognize
is “artificial.” In popular parlance, the term AI has voices, and translate spoken words into written
become synonymous with Large Language Models instructions. They then match these instructions
(“LLM”) and other forms of generative AI, but it is to pre-programmed commands for execution.
important to recognize that the term AI refers to a Importantly, they learn from each interaction,
long standing and very general field of study. gradually improving their ability to understand and
respond to requests. Similarly, services like Google
ML, a significant branch of AI, is about teaching Photos use AI vision models to recognize people
machines to learn and make decisions from data. and objects in our photos, simplifying organization
Instead of being programmed with specific rules, and retrieval. Navigation systems such as Google
like traditional programs, ML algorithms use Maps and Apple Maps exemplify AI’s real-time
data to learn and make predictions or decisions. data processing capabilities. They analyze vast
The heart of ML lies in its ability to adapt and amounts of information, like traffic patterns and
improve from experience, uncovering insights road conditions, and continually adapt based on
and identifying patterns without being explicitly new data to provide the most efficient routes. These
programmed to do so. To do this, ML relies on examples highlight that AI and ML have already
deep learning (“DL”) which involves training become familiar aspects of our lives, long before
algorithms on large data sets to identify patterns the current wave of AI enthusiasm.
and make decisions, much like the learning
4 | VOLUME 1 FOUNDATIONS OF AI | LOZANOSMITH.COM
