Placeholder Image

字幕表 動画を再生する

  • SPEAKER 1: This is CS50, and this is week 1.

  • And by the end of the day, you will know how

  • to create programs that look like this.

  • So this, of course, is binary.

  • This is the only language that machines ultimately understand.

  • But thankfully, per last week, there's so many abstractions

  • and there are so many humans that have come before us that we don't actually

  • have to write anything at this level.

  • We can abstract way above it like we did with Scratch already

  • and like we will starting today with C.

  • But does anyone nonetheless want to take a guess

  • at what that program, when fed to your Mac or PC, actually does?

  • Anyone recognize?

  • Anyone want to hazard a guess?

  • It's perhaps the simplest program you could write.

  • Indeed, it does, when fed to the brain of your computer,

  • the so-called CPU simply prints that.

  • So how do we actually get to that point?

  • Well, recall where we started this conversation last time

  • talking about computer science more generally

  • and problem solving-- we proposed could be distilled really is this.

  • You've got some inputs.

  • You want some outputs.

  • And somewhere in the middle, you need to do something with those inputs.

  • And to get to that point, though, we had to represent those inputs and outputs.

  • We just had to decide as humans, how are we going to represent all of the inputs

  • to our problem when it comes time to have a computer actually process them.

  • And at the end of the day, all of the phones and the computers that we're all

  • using only at the end of the day plug into the wall

  • to get their physical resource, electricity,

  • and they might store that temporarily in a battery.

  • But that really is our only input.

  • It's either plugged in or it's not.

  • It's either a 1 or a 0, true or false.

  • So the world really reduces to those two states, so to speak.

  • And so you can think of those states then as just being like a light bulb,

  • on or off.

  • I pulled up my cell phone last time to turn the flashlight on or off,

  • 1 or 0, true or false.

  • Now of course, if you only have one light bulb,

  • you can only count from 0 to 1.

  • But if you start to have a bunch of them back to back to back to back,

  • you can permute them like I did my finger--

  • 0, 1, 2, 3, and so forth.

  • And so we started talking about binary more generally.

  • And so here for instance were three sequences

  • of 0's and 1's and each of those represented something,

  • but we don't need to think about the world at that level.

  • We can abstract on top of that.

  • All of us are so much more familiar with decimal of course,

  • and indeed recall that this was just 72, 73, and 33, which if anyone recalls,

  • when you use ASCII--

  • which is this global standard for mapping numbers

  • to letters-- we got what message?

  • Yeah, it was just high, capital H capital I exclamation point.

  • And so that's an abstraction on top of those otherwise binary numbers.

  • But we don't have to model just text using numbers.

  • At the end of the day, our only resource is still that electricity,

  • and the only way we think about it digitally is still zeros and ones.

  • But if we take the same value--

  • 72, 73, 33-- and treat them in the context of Photoshop

  • or a photo program or a graphics program,

  • we can instead interpret them as like some amount

  • of red, some amount of green, some amount of blue,

  • which gave us last time, recall, this yellowish color.

  • So now we had another abstraction on top of binary colors,

  • and this is just one pixel.

  • What can you do once you have more than one pixel?

  • What can you represent next?

  • Yeah, right, images.

  • So we're continuing the conversation up and up and up,

  • and we could represents something like a graphical emoji on the screen, which

  • has more than just one yellow dot.

  • It's got a whole bunch of yellow dots and other colors as well.

  • And recall that, if we want to animate things,

  • whether it's through silly things like animojis on a phone or just more proper

  • videos and movies, well, those are just sequences

  • of images flying past your human eyes really quite quickly.

  • So that's where we kind of left off last time starting at the base level

  • and abstracting away so that we could stipulate thereafter we

  • can represent inputs, and we can represent

  • outputs, whatever those happen to be.

  • And here on out, we don't need to think at that level.

  • We can just assume we all know how to do this.

  • And even if it eventually becomes kind of a distant memory,

  • we know that someone can indeed do this.

  • And that's the value of abstraction.

  • But inside of this black box are so-called algorithms,

  • the secret sauce--

  • this is where the problems are actually solved.

  • And we not only talked about what these algorithms are, but for instance,

  • how efficient they were.

  • So recall that this red line represented a very simple algorithm just turning

  • the phone book page by page one at a time.

  • And the reason that it's a straight line is because there's a one to one

  • correspondence between how many pages there are in the book

  • and how many page turns there are-- one page, one more page, one more

  • turn, and so forth.

  • If I fly through it at twice the speed--

  • 2, 4, 6, 8--

  • I can do better.

  • And so that yellow line now, recall, was lower on the graph.

  • If you just look at any two points, yellow and red,

  • yellow is below red, saying it takes less time.

  • But it was not quite correct.

  • There was one bug when I was looking for Mike two pages at a time.

  • What was that issue?

  • Yeah, I might miss him.

  • He might accidentally get sandwiched in between two pages-- not a huge deal

  • because I could fix it, but I have to fix it.

  • I have to apply that additional logic and double back

  • at least a page if I go too fast.

  • But of course the final algorithm-- and frankly all of our initial intuition

  • probably-- was the dividing and conquer, open it roughly to the middle,

  • look down, and then go left, and go right,

  • and just repeat that process as the problem gets

  • this big to this big to this big to this big to just one page left.

  • So that was all about efficiency.

  • But to get to that point we needed to express ourselves more precisely.

  • And so we introduced pseudo code.

  • There's no formal definition.

  • It can be English, English like.

  • It's just meant to be succinct and get the point across.

  • And recall that, along the way, we introduced a whole bunch of concepts,

  • many of which you probably experimented with Scratch, like loops

  • and conditions, Boolean expressions, variables, and so forth.

  • And those were building blocks that came out of this kind of demonstration here.

  • But honestly, even in this demonstration, in this pseudo code,

  • there were a whole bunch of assumptions.

  • If you read these instructions one at a time

  • and you're holding the phone book yourself,

  • odds are you can execute this pseudocode, this algorithm.

  • But what does it really mean to, say, open to the middle of the phone book?

  • All of us have an intuitive understanding of what that means.

  • But honestly, if you were explaining that

  • to a kid or someone who's learning English or whatever language

  • for the first time, open to the middle of the phone book,

  • you should probably sets forth some assumptions.

  • OK, this thing in front of you has 1,000 pages, pieces of paper.

  • Turn to the 500th page, and let's call that the middle.

  • This would very quickly get tedious if all of us humans

  • are talking at that level of detail.

  • And so we abstract away with more sweeping

  • statements like open to the middle of the phone book,

  • but that's an abstraction.

  • And it's not quite as precise as is probably ideal,

  • especially feeding this algorithm to a newbie or to a robot or a computer.

  • But it's useful because we can then make a 12 step program instead of a 20 step

  • program by elaborating too much.

  • And for instance, throughout here too we had our loops and conditions

  • and so forth, but even call Mike.

  • What does that mean?

  • Well, if you imagine that the human knows how to use the phone,

  • then it goes without saying.

  • But if he or she also needs to be programmed to use the phone,

  • you've got to explain-- pick it up, hit this button, type

  • this sequence of buttons, and so forth.

  • So call Mike is also an abstraction.

  • So these abstractions are useful, but they can sometimes get in the way,

  • especially if you're not precise enough to program the computer correctly.

  • And to paint this picture, I thought we could begin a little heartedly here.

  • I brought some breakfast, if you didn't quite make it next door or beyond.

  • Just need a couple of volunteers if you're

  • comfortable appearing on stage and on the internet here.

  • Let me kind of-- there's a lot of lights here.

  • How about over there on the left and over here in the front?

  • Yeah, right there.

  • I think your hand was up.