You’ve certainly heard of Artificial Intelligence (AI) – in fact, you probably even have a go-to image in your head. Maybe it’s evil robots… maybe it’s lines of code from the Matrix swirling around… or maybe it’s just Haley Joel Osment.
Whatever comes to mind… it’s probably NOT accurate!
There’s so much misinformation about modern technology, and Artificial Intelligence may be the most misunderstood of all. So let’s break it down as simply as possible!
What Is Artificial Intelligence?
According to Andrew Moore of Carnegie Mellon University, “Artificial intelligence is the science and engineering of making computers behave in ways that, until recently, we thought required human intelligence.”
In other words, Artificial Intelligence refers to any attempt to improve computers so they can think like humans.
The “artificial” part refers to the fact that it is a program (not a living organism), and the “intelligence” part refers to the fact that it is capable of learning over time.
This means “AI” is a very broad term. There are countless ways Artificial Intelligence is being applied in technology – but one of them stands far above the rest: Machine Learning.
In fact, you might be surprised to learn that the concept you think is called Artificial Intelligence… is actually something called Machine Learning!
Huh? What’s Machine Learning?
Machine Learning is a specific type of AI that focuses purely on teaching computer software to learn via algorithms.
This means Machine Learning is included within the broader definition of AI… but calling something “AI” when it really uses Machine Learning would be inaccurate (though not totally off the mark).
Then there’s Deep Learning within Machine Learning… but we’ll save that for another day.
Now, Machine Learning itself is also a pretty broad term. Machine Learning refers to any computer program that is capable of learning – they get better over time because they get feedback, which changes how they operate.
This is unlike most computer programs, which can only perform tasks in the way they were originally designed. (That doesn’t make them “dumb” though – for example, your calculator is quite “smart”, but it can only perform math equations that you ask it to perform.)
I Don’t Get It. Can I Have An Example?
Yes! To illustrate this, let’s consider two different computer programs for translating English into Spanish. One program was designed with Machine Learning and the other wasn’t.
The program without Machine Learning will probably be able to tell you everything in the Spanish dictionary. However, if it makes a mistake, it will continue making that mistake (at least until the program is updated).
On the other hand, the program with Machine Learning doesn’t just give you translations – it also asks you for feedback.
If it makes a mistake, and you correct it, then it will pay attention – and next time, it won’t make that mistake!
Obviously, this is a very simplified way to think about it – it takes more than just one response for a program to “learn” the correct answer, and there’s lots of complicated math going on.
All you really need to know is that over time, Machine Learning has the chance to give us better information than any human can – not to mention all the amazing applications for making our lives easier.
Let’s check some of those out, shall we?
What Are Some Uses For AI (Erm, Machine Learning)?
You’ve surely benefited from Machine Learning – perhaps even in the time it’s taken you to read this article!
Machine Learning can be added to a huge variety of computer programs – anything that can theoretically get better over time is a perfect candidate.
For example, have you noticed your Netflix recommendations getting better? Or maybe you’ve asked Siri for advice – and noticed she’s giving better answers these days.
In most cases, you don’t need to give feedback yourself; the program can do its own testing. For example, Netflix knows you’re there to watch movies – and if you watch a movie all the way through, you probably like it. It “learns” what you like based on the movies you choose and even how long you watch them for.
That’s why they even use different images for the same movie on different users – the program is also learning what types of pictures you are more likely to click on!
These are all examples of Machine Learning at work! Of course, Machine Learning can also use other AI technologies to do its job, like neural networks (we’ll cover that some other day).
Whew, What A Lesson – Let’s Review All That, Shall We?
The bottom line is that most people are actually thinking of “Machine Learning” when they think of Artificial Intelligence.
Machine Learning refers to any computer program that improves over time thanks to feedback, while AI refers to the general field of making computers more like humans.
AI is a constantly evolving field – 50 years ago, we considered a chess video game to be “AI”, because it was unheard of that a computer could match a human at chess. Nowadays, that’s standard and even old-school… so it’s time for the next AI challenge!
What do you think? Is AI a cool way to improve our lives, or does it make you feel a little creeped out?? Let us know in the comments below!