Okay, let's restart.

Remember we're in the part where we want to generally use structured representations

as our world models, which means rather than having factored representations, which basically

tells us what the descriptions look like, we're basically completely free to design

descriptions of the world as we please.

And we're going to see a couple of them.

We're going to see a couple of description languages of different expressiveness, where

we can say different things about the world.

Where the inference procedures are going to be different, where we have different notions

of semantics and so on.

One of the main, and this field is called logic and inference, automated reasoning.

Those are names for this field.

And one of the big misconceptions that people are teaching, many people are teaching, is

that there's only one logic.

Well, maybe two.

And it's called first order logic.

Well, and maybe propositional logic.

That's not true.

On the contrary, for artificial intelligence, one of the things an agent designer must do

is design a suitable language that is capable of succinctly capturing the structures in the

world.

And depending on what kind of an environment we're in, we need different languages.

And that means different logics.

Logics are tools that are designed, that are engineered.

And so my goal in this part of the course is less giving you an introduction to propositional

logic, or first order logic, or ALC.

All of this we'll learn, but what I want you to understand is the principles by which you

can then start engineering your own logics.

Logics make very good pets.

Everybody should have one or more of them.

Okay, that being said, we're going to look at propositional logic after all.

But I want you to keep your eyes open, and I'll just rub your nose in.

The commonalities.

So we'll start with propositional logic.

And what we'll see time and again is that we have a formal language which is kind of

made up by layers.

From simple things.

That's the alphabet.

Here our alphabet consists of essentially what we call propositional variables.

In almost all logics we're going to have variables.

And I'm sure you're good friends with variables, right?

They are things that you've learned in high school latest.

Variables are things that are intended to be substituted by something suitable.

And in high school that's usually some kind of a number.

Here we're going to substitute them or replace them by true and false.

Same thing.

Just funny numbers.

So we're going to have variables.

They're going to have names.

That's the important thing about variables.