Artificial Intelligence Concepts for Architects, Developers, and Executives
Learn how Artificial Intelligence, Machine Learning, and Deep Learning are being used to teach computers to understand the world around us. In business they’re used to understand customers, forecast financial results, automated processes, and more. Lucid Thoughts explains AI concepts in simple terms that architects, developers, and executives alike will understand.
Subscribe to the Lucid Thoughts channel and be sure to leave your questions and comments on each video.
Admit it, you keep hearing about AI in every meeting, right? So what do you do? You go to your computer, you Google it, and it leads you right to me.
I’m Brynn. And this is a judgment-free place full of information about artificial intelligence.
AI is the practice of using math to teach computers how to think like humans. A machine learning, deep learning, linear regression, and natural language processing. And all those things are really just putting the computers through the same lessons you had as a kid.
Let’s start with machine learning.
Think of machine learning as that classic childhood game, “One of these things is not like the others.” Just like you learn the rules as a game as a kid, the computer also learns the rules. But in AI, we call those algorithms. They teach the computer to recognize patterns in data using one or two methods: Supervised learning and unsupervised learning.
If you use supervised learning method, your machine is trained with data that’s already labeled by humans. As more data is added into the system, the computer learns to identify and classify it. Oh! Remember that. It’ll come up later, trust me. The computer learns to classify all future data.
Linear regression is one form of supervised learning. It’s like that day in school where you got that fresh sheet of graph paper and you learned to plot all the points. You were learning to make projections. And that’s what linear regression is teaching computers, how to plot and project, so we can predict the possible outcomes.
Then, there’s unsupervised learning.
By giving the machine unlabeled data, it plots like with like and identifies similarities. I mean you don’t need a training set of humans really do this. Pro tip! Clustering also comes back later.
So, how do you know which to use? I mean, when do you use supervised learning, when do you use unsupervised learning? Well, it’s like when you use a hammer versus a bulldozer, I mean, it depends on your situation and goals.
Still with me? All right, good.
Let’s dive a little deeper.
Deep learning is pushing machines to new levels. Now, with some crazy math we’re able to create an artificial neural network, basically a model of the human brain. Then the machine can do things like recognize images and faces and noises.
Speaking of recognition, natural language processing teaches computers to recognize human speech patterns. Kind of like the hours you spent trying to get your nephew to say your name, machine and deep learning techniques help devices to understand and talk back.
Prime example, “Siri, when is national pizza day?”
Siri: “Today is national pizza day.”
My lucky day. Oh, wait, you didn’t catch all that? No worries. You get the rest of the season to get a little deeper into each of these topics. Until then. See you later, homeslice.