Original subtitles

Okay, so you just saw a tour of the areas of AI. And we'll go into these

throughout the course in a little more depth. But really AI is all around you in a way

that doesn't always categorize so neatly.

In general applied AI is automation of some kind. There's some kind of decision problem you want to

automate, something that's too hard for you to do by hand and you want

computer assistance.

So scheduling--airline rallying, military planning--things like that, regularly done

by AI.

Route planning--every time you go get a traffic map, that's AI.

The search algorithms aren't so different than the kinds of things we'll do when we study search.

Medical diagnosis.

You can input your symptoms. There are medical diagnosis systems that will then

try to figure out the underlying diagnosis. These systems actually in many cases

are very good. There are even cases where they can be better than humans.

Web search engines. Some of you may have used a web search engine...

This has got AI everywhere. It's got some natural language in the text,

there's machine learning in figuring out how to rank results using

multiple kinds of features. It's kind of AI everywhere in these things.

Spam classifiers: you may have noticed that the amount of spam you've got your life,

probably,

shot up while you were young,

right, as kind of people learned how to spam,

and then recently I think it's been under much better control and that has to do

with a lot better, both machine learning

and AI techniques for spam classification.

Automated help-desks, that's AI.

Fraud detection on your credit card, product recommendations when

you go to Amazon--all this stuff is AI. So there are a lot of things

that don't really boil down to challenging a grandmaster at chess, but are

equally applicable to industry's problems today, that come from the

techniques that you're gonna learn about in this course.