Maybe the artificial intelligence conversation makes you feel tired and jaded. Maybe you’re tired of tech talk in general. Maybe looking at everyone on their smartphone on lunch break or talking about their shiny new Apple Watch brings chills sooner than thrills. But before you set your status to “permanently unimpressed,” consider the fact that we literally have self-driving cars. Your ten-year-old self would be beside themselves. Just don’t give up yet: there is still a lot to say about new technology like AI. Some of it just needs a little explanation.
AI is Not Just for Trekkies
Cleanse your mind of 1960’s spandex StarTrek suits. Artificial intelligence is not science fiction or evil robots. AI is already pretty much omnipresent, and it is anything but evil. Millions use it unconsciously every day: Googling their medical symptoms, scrutinizing Amazon for the cheapest HDMI cable, binge-watching Korean dramas on Netflix. The technology behind all of that internet fun time is already miraculous, but the kind of AI that makes headlines nowadays is actually AGI. That stands for artificial general intelligence—AI’s more complex cousin. So what are companies striving for when they delve into AGI? Well, in essence, machines that are indistinguishable from human behavior.
To build anything with AGI, you need some building blocks: machine learning, deep learning, natural language understanding, context awareness, and data privacy. Let’s take a look at each:
- Machine Learning: this is not a synonym for AI! It’s part of the whole. Machine learning means using algorithms to learn from past data to create a future behavior. It’s the initial step in teaching a machine how to decide what to do with new data. For example: Let’s say you show the machine a color palette and tell it which color is which. Once the machine is trained, it can tell you what color it “sees” when presented with something new—but it can only implement the algorithm made under your supervision.
- Deep Learning: this is a more advanced kind of machine learning. It uses artificial neural networks that mimic the neurons of a human brain. It means finding patterns in raw data. As more layers of deep learning are created, the machine can understand increasingly abstract concepts. Facial recognition is an application of deep learning. To be able to do this, the machine needs to look at a few layers at once: skin color, head shape, hair color, and so on. Then it combines all the information. After processing enough data points it can recognize specific features, and eventually be able to ask you if you’d like to tag your friend David in a new Facebook picture.
- Natural Language Processing: this is a way for computers to communicate with us. This one is hard because human language is very complex and filled to the brim with emotion and humor (depending on your conversation partner, that is.) Siri, Alexa, and Google are deep into the natural language processing exploration.
- Context Awareness: this means having access to relevant information. AI is only as good as the information it can access; for example, your artificial assistant can’t get you a table at that new Sushi place without Google maps, your calendar, and your contact list.
Impressed now? The stars have to align for all the information to fit perfectly so that AI can run. We are making huge steps in a breathtakingly daunting process. In the scope of history, it is objectively incredible. And we didn’t even talk about what all this technology could mean for our future. The time we’re living in is fascinating, inspiring, and full of intrigue. It’s far from boring.