Understanding Artificial Intelligence: A Beginner's Guide

Artificial Intelligence (AI) is reshaping industries, economies, and our daily lives. But what exactly is AI? In our beginner’s guide we will demystify AI, explaining what it is, how it works, and why it matters.  

What is Artificial Intelligence?

Artificial Intelligence is an overlap of lots of fields (statistics, computer science, math, engineering, etc.), but generally focuses on creating systems that can perform tasks that mimic human intelligence. We can draw endless amounts of concentric circles and Venn diagrams to show what it is and isn’t, but we can leave that to another time. For now, we’ll just say that common AI tasks include learning, reasoning, problem-solving, understanding natural language, recognizing patterns, and even perceiving emotions (and for now we’ll ignore areas of AI that include robotics).

How Does AI Work?

At its core, AI works by processing large amounts of data, identifying patterns within that data, and using these patterns to make predictions or decisions. We can create AI systems with traditional software engineering approaches with rules and logic, like early autonomous breaking systems in cars, for example. Those systems would take sensor data of distance (or change of distance) from one object (your car) to another object in front of you. A simple program can then be created that takes that one-dimensional measurement and says that if the distance between objects is approaching the known minimum breaking distance, then apply the break to stop the car. This system would seem intelligent to humans but is still just rules and logic.  

Most systems are much more complex than this with significantly more data needed to make judgements, so we use Machine Learning (ML) to create models to make judgements. Therefore, we say that ML is an efficient way to create an AI system.

Here’s a breakdown of how AI typically operates:

1. Data Collection: Most ML models used today have learn from data, these are called supervised ML models. This data can be anything from images, text, or numbers to complex data sets like stock market trends or user behavior.

2. Training Models: Once the data is collected, ML algorithms (such as neural networks) are trained on this data. During training, the ML learns to recognize patterns, make predictions, and improve its accuracy over time. The resulting object of this is called an ML model.  

3. Making Predictions: After the model is trained, it can be used to make predictions or decisions based on new data. For example, a trained ML model can predict whether an email is spam or not, or it can suggest a movie you might like based on your past viewing history or detect if you’re driving over a lane line and attempt to correct you.  

4. Continuous Improvement: Just about all AI systems will start off with a static model, meaning it’s not continuously updating on new data. However, in production systems, we often want the ML model to continuously learn and improve as they are exposed to better data or changing behavior of the input system, making them more accurate and effective over time. This last piece is a significantly complex topic that we’ll dive into much more in future posts, along with various types of feedback or reinforcement, both in global and local variety.  

Applications of AI

AI is already embedded in many aspects of our daily lives. Some common applications include:

• Personal Assistants: Voice-activated assistants like Siri, Alexa, and Google Assistant rely on AI to understand commands and perform tasks.

• Healthcare: AI helps doctors diagnose diseases, predict patient outcomes, and even assist in surgery through robotic systems.

• Finance: AI is used in fraud detection, algorithmic trading, and customer service in banks.

• Entertainment: AI powers recommendation systems on platforms like Netflix and Spotify, suggesting content based on your past behavior.

• Transportation: Self-driving cars use AI to navigate roads, avoid obstacles, and make driving decisions.

Why Does AI Matter?

AI is significant because it has the potential to revolutionize virtually every industry. It enhances productivity, improves decision-making, and enables the creation of new products and services that were previously unimaginable.

Stay tuned! We’re going to expand in all these areas in upcoming posts, both in the technical areas (for people interested in more complex stuff!) and in the applications of how AI will change the world. Just think...will a parking lot at a suburb department store be necessary if everyone has self-driving cars that just drop you off and then become a taxi, earning income, until it picks you back up for your journey home?

‍