1
00:00:04,306 --> 00:00:05,972
ERIK DEMAINE: I mean,
there are a couple

2
00:00:05,972 --> 00:00:08,160
of different approaches
to teaching this class.

3
00:00:08,160 --> 00:00:10,860
One is to make it fun.

4
00:00:10,860 --> 00:00:15,240
I really like to make
material as engaging

5
00:00:15,240 --> 00:00:17,209
and as accessible as possible.

6
00:00:17,209 --> 00:00:19,500
So, especially if you have
relatively little experience

7
00:00:19,500 --> 00:00:21,010
in algorithms--
you've just taken

8
00:00:21,010 --> 00:00:25,290
6006-- and some of
these algorithms

9
00:00:25,290 --> 00:00:26,700
get really complicated.

10
00:00:26,700 --> 00:00:31,630
And they're hard to
understand and it

11
00:00:31,630 --> 00:00:35,730
makes you wonder, how did
they ever come up with that?

12
00:00:35,730 --> 00:00:38,230
So I like to talk about that.

13
00:00:38,230 --> 00:00:41,390
In some of the
lectures, I'll kind of

14
00:00:41,390 --> 00:00:43,564
give an idea of how they
might have gotten there.

15
00:00:43,564 --> 00:00:45,980
It's like, oh we could start
with this really simple step.

16
00:00:45,980 --> 00:00:48,230
It's really inefficient, but
at least it solves a problem.

17
00:00:48,230 --> 00:00:50,521
But then we have this extra
thing and this extra thing,

18
00:00:50,521 --> 00:00:52,930
and eventually you get to
the really fast solution,

19
00:00:52,930 --> 00:00:55,820
so it's less of the mystery
of how you got there,

20
00:00:55,820 --> 00:00:58,320
or how the original
researchers got there.

21
00:00:58,320 --> 00:01:02,120
And I also try to
remind students,

22
00:01:02,120 --> 00:01:04,800
especially in the harder topics,
that this is not easy stuff

23
00:01:04,800 --> 00:01:07,510
and you should think
about it a little more.

24
00:01:07,510 --> 00:01:12,320
Do the problem sets and you'll
learn it-- figure it out

25
00:01:12,320 --> 00:01:13,070
at a deeper level.

26
00:01:13,070 --> 00:01:14,700
But a lot of this
material is hard,

27
00:01:14,700 --> 00:01:19,520
and I think being honest about
that makes it more reasonable.

28
00:01:19,520 --> 00:01:21,020
Students don't feel
bad for thinking

29
00:01:21,020 --> 00:01:23,186
that it's hard, because the
first time you learn it,

30
00:01:23,186 --> 00:01:24,590
it is hard.

31
00:01:24,590 --> 00:01:27,020
But the whole algorithmic
way of thinking

32
00:01:27,020 --> 00:01:29,961
of setting up a problem-- what
should the input and the output

33
00:01:29,961 --> 00:01:32,460
be-- and then thinking about
all different solutions for how

34
00:01:32,460 --> 00:01:36,770
to fill in the middle I think is
a really powerful perspective,

35
00:01:36,770 --> 00:01:39,600
and it can shape pretty
much your entire life,

36
00:01:39,600 --> 00:01:42,760
but especially if you're
doing computer science.

37
00:01:42,760 --> 00:01:45,030
It's a really useful
way of thinking,

38
00:01:45,030 --> 00:01:47,580
and I've gotten lots of
letters later from students

39
00:01:47,580 --> 00:01:51,950
where they're doing some job,
some computer thing and they

40
00:01:51,950 --> 00:01:54,350
ran this problem and then
they discover-- and then they

41
00:01:54,350 --> 00:01:55,849
looked back at their
notes and said,

42
00:01:55,849 --> 00:01:59,000
oh, this is the right data
structure I should use now.

43
00:01:59,000 --> 00:02:01,150
And they solve it, and
they tell me about it.

44
00:02:01,150 --> 00:02:07,084
And it's kind of exciting to see
years later the way of thinking

45
00:02:07,084 --> 00:02:08,750
and lots of the
solutions and techniques

46
00:02:08,750 --> 00:02:11,840
you see in this class really are
useful out there in the world,

47
00:02:11,840 --> 00:02:15,670
so it's a useful class to take.

48
00:02:15,670 --> 00:02:17,300
It's a relatively hard class.

49
00:02:17,300 --> 00:02:19,630
The exams are challenging.

50
00:02:19,630 --> 00:02:22,170
We try to make them
not super challenging,

51
00:02:22,170 --> 00:02:25,060
but it's hard,
especially whenever

52
00:02:25,060 --> 00:02:27,610
creativity is involved.

53
00:02:27,610 --> 00:02:31,800
It's hard to do that
live in an exam.

54
00:02:31,800 --> 00:02:34,830
So students may not perform
super well my class,

55
00:02:34,830 --> 00:02:37,780
but still they get the
idea that this is a way

56
00:02:37,780 --> 00:02:40,680
to approach problems,
and they understand

57
00:02:40,680 --> 00:02:42,950
some of the techniques.

58
00:02:42,950 --> 00:02:46,220
And over time, they can
fill in whatever holes they

59
00:02:46,220 --> 00:02:48,524
left behind, and
say, you know, it

60
00:02:48,524 --> 00:02:50,690
seems like a problem where
dynamic programming would

61
00:02:50,690 --> 00:02:51,440
be really useful.

62
00:02:51,440 --> 00:02:54,280
It's time I really learned
dynamic programming.

63
00:02:54,280 --> 00:02:57,000
Dynamic programming is one that
we actually teach both in 006

64
00:02:57,000 --> 00:03:01,135
and 046, because it's
really conceptually weird.

65
00:03:03,689 --> 00:03:05,730
It's actually really easy
once you understand it,

66
00:03:05,730 --> 00:03:08,188
but getting to that point of
understanding is really tough.

67
00:03:08,188 --> 00:03:10,340
So we spend a lot
of time on it so

68
00:03:10,340 --> 00:03:12,930
that the ideas-- and more
exposure you get to it,

69
00:03:12,930 --> 00:03:15,020
the clearer it will become.

70
00:03:15,020 --> 00:03:17,550
I think by the end of 6046,
they have a pretty good idea.

71
00:03:17,550 --> 00:03:20,360
But if not, years later,
if they come back it,

72
00:03:20,360 --> 00:03:24,670
I'm sure it'll be a lot
easier this third time around.

73
00:03:24,670 --> 00:03:27,229
So a lot of these
concepts are-- I mean,

74
00:03:27,229 --> 00:03:28,770
the ideas are actually
really simple.

75
00:03:28,770 --> 00:03:35,390
They're just hard to shift your
mind into that kind of mindset,

76
00:03:35,390 --> 00:03:38,030
but once you do, things
become really clear.

77
00:03:38,030 --> 00:03:43,010
So hope is, by the end of
the semester, they get there.