Artificial Intelligence is a topic that has been getting a lot of attention, mostly because of the rapid improvement that this field has seen since the turn of the 21st century. Amazing innovations are laying the foundation for ongoing breakthrough achievements. In this article, I’m going to focus on three specific topics:
- What Is Artificial Intelligence?—I’ll discuss and explain what AI is, what it’s being used for and some examples of its use in the future.
- What Is Machine Learning?—I’ll explain what this strange topic is and its connection to AI. Additionally, I’ll give a small example of machine learning in action.
- Flying Cars (Easter Egg)—I’ll discuss a possible implementation of AI on the horizon: flying cars.
What Is Artificial Intelligence (AI)?
In the 1950s, AI pioneers Minsky and McCarthy described artificial intelligence as any task performed by a program or a machine that, was it performed by a human, would have required that human to apply intelligence to accomplish the task.
This is a fairly broad description. Nowadays, all tasks associated with human intelligence are described as AI when performed by a computer. This includes planning, learning, reasoning, problem-solving, knowledge representation, perception, motion, manipulation and, to a lesser extent, social intelligence and creativity.
“Artificial intelligence is defined as the branch of science and technology that [is] concerned with the study of software and hardware to provide machines the ability to learn insights from data and [the] environment and the ability to adapt in changing situation[s] with high precision, accuracy and speed.”―Amit Ray, Compassionate Artificial Superintelligence AI 5.0 — AI with Blockchain, BMI, Drone, IOT and Biometric Technologies
What Is AI used for?
Now that we know what AI actually means, let’s find out what it’s used for!
While surfing the web, have you ever wondered how most ads are related to your interests? That’s a representation of AI, more specifically, machine learning. However, AI is more commonly associated with robots, such as the ability of a robot to think on its own and the potential for computer consciousness. While these would be astounding achievements, they involve highly complex algorithms which we still can’t produce today.
What Is Machine Learning?
Machine learning is a big part of AI, and it might be the key reason for this field’s meteoric rise. It’s based on the principle of trial and error; every time we try to solve a problem, like a maze, we’re going to fail at least once. However, failing is a good thing in machine learning, because it enables the program to learn new information. That information is stored as data, and each time an AI goes down a specific path, it will reference the data from prior trials to see which one will work best this time.
To expand on the above example, I’m going to teach you one of the first AI algorithms (often used to solve mazes), the A* algorithm.
To understand this algorithm, let’s visualize our maze as a chess board with inaccessible regions (like a maze) that we’ll call nodes.
- Let’s define our beginning node and final node as the top left and bottom right squares, respectively.
- From here, we will give every node of the board a value, determined by the sum of the distance from the beginning and the distance to the end node.
- Now, the program is going to pick a node based on this sum, which provides an arrow path to the end! If the program reaches a node it can’t pass, it comes back to the last node that was picked before and tries again.
This is a fun example of AI in action, since flying cars would be reliant on AI to function properly. In the future, scientists believe we’re going to have autonomous cars that transport people to their desired destinations. This involves cars having some sort of artificial intelligence, more specifically, machine learning, because they need to always find the best possible course to the destination, not crash into buildings and respect other vehicles. A very basic implementation of this, although extremely ineffective and slow, could be the A* algorithm, where buildings represent inaccessible nodes. However, some good alternatives exist that we didn’t review in detail due to their high levels of complexity:
- Neural networks – Layered systems that loosely mimic the biological structure of the brain. They can be “trained” to recognize patterns in
inputs,and then estimate a possible output, often with surprising reliability.
- Genetic Programming – Treats programs as the parameters. For example, you would be breeding pathfinding algorithms instead of paths, and your fitness function would rate each algorithm based on how well it does.
This article was written to provide a fun introduction to AI and to show its potential for future technologies. More than ever, it’s crucial to know the principles of artificial intelligence since it will be so important in the future.
We need to constantly be open to new ideas and approaches, such as artificial intelligence (AI), and be willing to challenge assumptions of what this technology can achieve.