by Sumeet Pathania
Surely you’ve heard the buzz about AI or artificial intelligence but you know very little about what it is. Are you not able to debate with people that think AI could take over the world? Well, you’re in the right place. This article will teach you the basics of AI and how it works. So let’s start…
What is Artificial Intelligence
Simply put, AI is creating intelligent machines. Machines and programs that can do things similar to humans. For years people have thought computers will never be able to compete with our brains but our exponential growth and curiosity have proven that statement wrong. Before we dive any deep I want to make it clear that AI is a broad field and consistents of many subfields.
The Birth of Artificial Intelligence
Although AI seems like a fairly new technology, the idea of AI dates back to the 1950s when it was first coined by John McCarthy. But achieving an artificially intelligent being wasn’t so simple. After several reports criticizing progress in AI, government funding and interest in the field dropped off. During 1972-1980 progress in AI was at its lowest. It is know known as the “AI Winter”. Luckily the field was later revived in the 1980s when the British government started funding it again to compete with efforts by the Japanese. Now those are some good facts to know!
What can we use Artificial Intelligence for?
Everything! (ok maybe not everything but a lot of things). Believe it or not, AI is being used for a variety of things from making important business decisions to helping you pick your next Netflix show. One company that takes advantage of artificial intelligence to help provide a better user experience is Youtube. We have all spent hours on youtube binging videos. Telling ourselves “ this is the last one then I’ll do some work” but then as soon as the video ends we see “up Next: …” and why would you not want to see that video. Anyways AI is responsible for that.
How Artificial Intelligence Works
AI works by using smart and algorithms combined with huge sets of data. This also the algorithms to find patterns and make decisions based on that data. There are many subfields within AI. For now, we’ll talk about machine learning (letting the machine learn from data).
What is Machine Learning?
Three of the most widely used methods of machine learning are reinforcement learning, supervised learning, and unsupervised learning-but there are also other methods of machine learning. For now, let’s just learn about these three.
1. Reinforcement Learning
This is mainly used for games and robotics. Reinforcement learning is essentially letting the algorithms figure it out through trial and error. After thousands of tests, the machine then improves its actions based on the feedback provided. The goal of reinforcement learning is to achieve the best outcome. The video below is a great example of how reinforcement learning works.
2. Supervised Learning
Supervised learning as the name indicates, means learning when supervised. Basically, in this method, we teach the machine to use labelled data. For example, suppose you are given a basket filled with different kinds of fruits. Now the first step is to train the machine with all different fruits one by one like this:
- if the shape is a sphere and orange, fruit equals orange
- if the shape is a small oval sphere and colour is purple, fruit equals grape
Now that the machine has a labelled set of data, It will first classify the fruit with its shape and colour and output the fruit. Thus, the machine learns the things from training data (basket containing fruits) and then applies the knowledge to test data(new fruit).
Supervised learning generally falls into two categories: Classification and Regression
Classification is what we did above with the fruits. It sorts out the new data into groups. So the apples will go with the apples and the grapes will go with the grapes.
Regression models a target prediction value based on independent variables. It is mostly used for finding out the relationship between variables and forecasting. For example, a person’s age and salary. (the older the person the more they earn).
Linear relationship. However, regression models can get a bit more complicated when there isn’t a linear relationship.
3. Unsupervised Learning
In this machine learning technique, you don’t supervise the model. Instead, you allow the model to figure it out by itself using the unclassified/ unlabeled data. Here the task of machine is to group unsorted information according to similarities and patterns without any prior training of data. Unsupervised learning algorithms can perform more complex processing tasks than supervised learning systems. The only downside is unsupervised learning can be way more unpredictable since there are limited instructions. An example of unsupervised learning is giving an algorithm of pictures of cats and dogs and then having it sort the images on its own.
4 Key Takeaways about Artificial Intelligence
- AI is much faster and more efficient at learning certain tasks compared to -humans
- AI is going to have a huge impact on every industry and will change the way we perceive most things.
- AI is an umbrella term with many subfields such as machine learning
- Machine learning uses supervised, unsupervised, or reinforcement learning to accomplish tasks.