1
00:00:19,360 --> 00:00:22,336
How many of you have used
an electronic spreadsheet,

2
00:00:22,360 --> 00:00:23,840
like Microsoft Excel?

3
00:00:24,840 --> 00:00:26,096
Very good.

4
00:00:26,120 --> 00:00:30,696
Now, how many of you have run a business
with a spreadsheet by hand,

5
00:00:30,720 --> 00:00:33,800
like my dad did for his small
printing business in Philadelphia?

6
00:00:34,760 --> 00:00:35,920
A lot less.

7
00:00:36,620 --> 00:00:39,220
Well, that's the way it was done
for hundreds of years.

8
00:00:40,180 --> 00:00:43,196
In early 1978,
I started working on an idea

9
00:00:43,220 --> 00:00:45,620
that eventually became VisiCalc.

10
00:00:46,180 --> 00:00:47,836
And the next year it shipped,

11
00:00:47,860 --> 00:00:51,139
running on something new
called an Apple II personal computer.

12
00:00:51,940 --> 00:00:55,996
You could tell that things
had really changed when, six years later,

13
00:00:56,020 --> 00:00:58,196
the Wall Street Journal ran an editorial

14
00:00:58,220 --> 00:01:01,500
that assumed you knew what VisiCalc was
and maybe even were using it.

15
00:01:02,460 --> 00:01:05,836
Steve Jobs back in 1990

16
00:01:05,860 --> 00:01:09,100
said that "spreadsheets
propelled the industry forward."

17
00:01:09,580 --> 00:01:13,500
"VisiCalc propelled the success of Apple
more than any other single event."

18
00:01:14,140 --> 00:01:15,300
On a more personal note,

19
00:01:16,300 --> 00:01:19,436
Steve said, "If VisiCalc had been written
for some other computer,

20
00:01:19,460 --> 00:01:21,660
you'd be interviewing
somebody else right now."

21
00:01:22,060 --> 00:01:27,647
So, VisiCalc was instrumental in getting
personal computers on business desks.

22
00:01:27,960 --> 00:01:29,120
How did it come about?

23
00:01:29,880 --> 00:01:32,800
What was it? What did I go through
to make it be what it was?

24
00:01:34,120 --> 00:01:39,456
Well, I first learned to program
back in 1966, when I was 15 --

25
00:01:39,480 --> 00:01:41,720
just a couple months
after this photo was taken.

26
00:01:42,400 --> 00:01:45,360
Few high schoolers had access
to computers in those days.

27
00:01:45,760 --> 00:01:49,556
But through luck
and an awful lot of perseverance,

28
00:01:49,580 --> 00:01:51,820
I was able to get
computer time around the city.

29
00:01:52,700 --> 00:01:57,898
After sleeping in the mud at Woodstock,
I went off to MIT to go to college,

30
00:01:58,020 --> 00:02:01,100
where to make money,
I worked on the Multics Project.

31
00:02:01,820 --> 00:02:06,383
Multics was a trailblazing
interactive time-sharing system.

32
00:02:06,860 --> 00:02:10,036
Have you heard of the Linux
and Unix operating systems?

33
00:02:10,060 --> 00:02:11,300
They came from Multics.

34
00:02:11,820 --> 00:02:14,116
I worked on the Multics versions

35
00:02:14,140 --> 00:02:17,956
of what are known
as interpreted computer languages,

36
00:02:17,980 --> 00:02:20,436
that are used by people
in noncomputer fields

37
00:02:20,460 --> 00:02:23,652
to do their calculations
while seated at a computer terminal.

38
00:02:24,860 --> 00:02:27,116
After I graduated from MIT,

39
00:02:27,140 --> 00:02:30,414
I went to work for
Digital Equipment Corporation.

40
00:02:31,320 --> 00:02:34,821
At DEC, I worked on software

41
00:02:34,845 --> 00:02:37,939
for the new area
of computerized typesetting.

42
00:02:38,300 --> 00:02:42,636
I helped newspapers
replace their reporters' typewriters

43
00:02:43,060 --> 00:02:44,260
with computer terminals.

44
00:02:44,780 --> 00:02:45,996
I'd write software

45
00:02:46,020 --> 00:02:49,716
and then I'd go out in the field
to places like the Kansas City Star,

46
00:02:49,740 --> 00:02:52,156
where I would train users
and get feedback.

47
00:02:52,180 --> 00:02:54,036
This was real-world experience

48
00:02:54,060 --> 00:02:57,300
that is quite different
than what I saw in the lab at MIT.

49
00:02:58,780 --> 00:03:01,396
After that, I was project leader

50
00:03:01,420 --> 00:03:06,220
of the software for DEC's first
word processor, again a new field.

51
00:03:06,620 --> 00:03:12,116
Like with typesetting, the important thing
was crafting a user interface

52
00:03:12,140 --> 00:03:16,220
that was both natural and efficient
for noncomputer people to use.

53
00:03:18,260 --> 00:03:22,236
After I was at DEC, I went
to work for a small company

54
00:03:22,260 --> 00:03:28,380
that made microprocessor-based electronic
cash registers for the fast-food industry.

55
00:03:29,260 --> 00:03:32,956
But I had always wanted to start
a company with my friend Bob Frankston

56
00:03:32,980 --> 00:03:35,356
that I met on the Multics project at MIT.

57
00:03:35,380 --> 00:03:39,076
So I decided to go back to school to learn
as much as I could about business.

58
00:03:39,100 --> 00:03:41,876
And in the fall of 1977,

59
00:03:41,900 --> 00:03:44,900
I entered the MBA program
at Harvard Business School.

60
00:03:45,740 --> 00:03:48,236
I was one of the few
percentage of students

61
00:03:48,260 --> 00:03:51,100
who had a background
in computer programming.

62
00:03:51,980 --> 00:03:55,156
There's a picture of me from the yearbook
sitting in the front row.

63
00:03:55,180 --> 00:03:56,196
(Laughter)

64
00:03:56,220 --> 00:03:58,836
Now, at Harvard,
we learned by the case method.

65
00:03:58,860 --> 00:04:00,796
We'd do about three cases a day.

66
00:04:00,820 --> 00:04:04,813
Cases consist of up to a few dozen pages

67
00:04:04,837 --> 00:04:07,144
describing a particular
business situation.

68
00:04:08,140 --> 00:04:13,171
They often have exhibits,
and exhibits often have words and numbers

69
00:04:13,195 --> 00:04:16,355
laid out in ways that make sense
for the particular situation.

70
00:04:16,860 --> 00:04:18,796
They're usually all somewhat different.

71
00:04:18,820 --> 00:04:20,036
Here's my homework.

72
00:04:20,060 --> 00:04:23,236
Again, numbers, words,
laid out in ways that made sense.

73
00:04:23,260 --> 00:04:27,340
Lots of calculations --
we got really close to our calculators.

74
00:04:27,740 --> 00:04:29,620
In fact, here's my calculator.

75
00:04:31,400 --> 00:04:33,976
For Halloween, I went
dressed up as a calculator.

76
00:04:34,000 --> 00:04:35,808
(Laughter)

77
00:04:36,985 --> 00:04:40,176
At the beginning of each class,
the professor would call on somebody

78
00:04:40,200 --> 00:04:41,479
to present the case.

79
00:04:42,279 --> 00:04:45,056
What they would do is
they would explain what was going on

80
00:04:45,080 --> 00:04:48,816
and then dictate information
that the professor would transcribe

81
00:04:48,840 --> 00:04:51,745
onto the many motorized blackboards
in the front of the class,

82
00:04:51,769 --> 00:04:53,296
and then we'd have a discussion.

83
00:04:53,320 --> 00:04:57,776
One of the really frustrating things
is when you've done all your homework,

84
00:04:57,800 --> 00:05:00,816
you come in the next day
only to find out that you made an error

85
00:05:00,840 --> 00:05:03,376
and all of the other numbers
you did were wrong.

86
00:05:03,400 --> 00:05:05,176
And you couldn't participate as well.

87
00:05:05,200 --> 00:05:07,440
And we were marked by class participation.

88
00:05:08,520 --> 00:05:13,440
So, sitting there with 87 other people
in the class, I got to daydream a lot.

89
00:05:14,480 --> 00:05:18,336
Most programmers in those days
worked on mainframes,

90
00:05:18,360 --> 00:05:24,696
building things like inventory systems,
payroll systems and bill-paying systems.

91
00:05:24,720 --> 00:05:27,216
But I had worked
on interactive word processing

92
00:05:27,240 --> 00:05:29,336
and on-demand personal computation.

93
00:05:29,360 --> 00:05:34,056
Instead of thinking
about paper printouts and punch cards,

94
00:05:34,080 --> 00:05:36,856
I imagined a magic blackboard

95
00:05:36,880 --> 00:05:40,256
that if you erased one number
and wrote a new thing in,

96
00:05:40,280 --> 00:05:43,136
all of the other numbers
would automatically change,

97
00:05:43,160 --> 00:05:44,920
like word processing with numbers.

98
00:05:45,800 --> 00:05:50,040
I imagined that my calculator
had mouse hardware on the bottom of it

99
00:05:50,440 --> 00:05:53,400
and a head-up display,
like in a fighter plane.

100
00:05:53,920 --> 00:05:58,056
And I could type some numbers in,
and circle it, and press the sum button.

101
00:05:58,080 --> 00:06:01,816
And right in the middle of a negotiation
I'd be able to get the answer.

102
00:06:01,840 --> 00:06:04,840
Now I just had to take my fantasy
and turn it into reality.

103
00:06:05,840 --> 00:06:08,280
My father taught me about prototyping.

104
00:06:08,760 --> 00:06:10,376
He showed me mock-ups

105
00:06:10,400 --> 00:06:13,736
that he'd make to figure out
the placement on the page

106
00:06:13,760 --> 00:06:16,296
for the things for brochures
that he was printing.

107
00:06:16,320 --> 00:06:19,256
And he'd use it
to get feedback from customers

108
00:06:19,280 --> 00:06:22,880
and OKs before he sent the job
off to the presses.

109
00:06:23,640 --> 00:06:28,736
The act of making a simple, working
version of what you're trying to build

110
00:06:28,760 --> 00:06:31,080
forces you to uncover key problems.

111
00:06:31,920 --> 00:06:36,320
And it lets you find solutions
to those problems much less expensively.

112
00:06:36,920 --> 00:06:39,160
So I decided to build a prototype.

113
00:06:39,960 --> 00:06:44,536
I went to a video terminal
connected to Harvard's time-sharing system

114
00:06:44,560 --> 00:06:46,056
and got to work.

115
00:06:46,080 --> 00:06:49,336
One of the first problems
that I ran into was:

116
00:06:49,360 --> 00:06:52,040
How do you represent values in formulas?

117
00:06:52,536 --> 00:06:53,896
Let me show you what I mean.

118
00:06:55,200 --> 00:06:58,376
I thought that you would point somewhere,

119
00:06:58,400 --> 00:07:01,496
type in some words,
then type in some somewhere else,

120
00:07:01,520 --> 00:07:05,044
put in some numbers and some more numbers,
point where you want the answer.

121
00:07:05,432 --> 00:07:08,656
And then point to the first, press minus,
point to the second,

122
00:07:08,680 --> 00:07:09,880
and get the result.

123
00:07:10,560 --> 00:07:13,661
The problem was:
What should I put in the formula?

124
00:07:14,000 --> 00:07:16,620
It had to be something
the computer knew what to put in.

125
00:07:16,644 --> 00:07:18,336
And if you looked at the formula,

126
00:07:18,360 --> 00:07:21,360
you needed to know
where on the screen it referred to.

127
00:07:22,040 --> 00:07:25,136
So the first thing I thought
was the programmer way of doing it.

128
00:07:25,160 --> 00:07:27,066
The first time you pointed to somewhere,

129
00:07:27,090 --> 00:07:29,770
the computer would ask you
to type in a unique name.

130
00:07:30,960 --> 00:07:34,736
It became pretty clear pretty fast
that that was going to be too tedious.

131
00:07:34,760 --> 00:07:38,000
The computer had to automatically
make up the name and put it inside.

132
00:07:38,800 --> 00:07:43,160
So I thought, why not make it be
the order in which you create them?

133
00:07:43,680 --> 00:07:46,176
I tried that. Value 1, value 2.

134
00:07:46,200 --> 00:07:48,962
Pretty quickly I saw
that if you had more than a few values

135
00:07:48,986 --> 00:07:51,496
you'd never remember
on the screen where things were.

136
00:07:51,520 --> 00:07:56,816
Then I said, why not instead
of allowing you to put values anywhere,

137
00:07:56,840 --> 00:07:58,440
I'll restrict you to a grid?

138
00:07:58,920 --> 00:08:01,318
Then when you pointed to a cell,

139
00:08:01,342 --> 00:08:04,062
the computer could put
the row and column in as a name.

140
00:08:05,040 --> 00:08:11,216
And, if I did it like a map and put ABC
across the top and numbers along the side,

141
00:08:11,240 --> 00:08:14,136
if you saw B7 in a formula,

142
00:08:14,160 --> 00:08:16,360
you'd know exactly
where it was on the screen.

143
00:08:17,040 --> 00:08:20,464
And if you had to type the formula
in yourself, you'd know what to do.

144
00:08:21,110 --> 00:08:24,190
Restricting you to a grid
helped solve my problem.

145
00:08:24,870 --> 00:08:30,470
It also opened up new capabilities,
like the ability to have ranges of cells.

146
00:08:31,030 --> 00:08:32,606
But it wasn't too restrictive --

147
00:08:32,630 --> 00:08:36,950
you could still put any value,
any formula, in any cell.

148
00:08:37,789 --> 00:08:41,630
And that's the way we do it to this day,
almost 40 years later.

149
00:08:42,779 --> 00:08:46,446
My friend Bob and I decided that we were
going to build this product together.

150
00:08:46,890 --> 00:08:51,306
I did more work figuring out exactly
how the program was supposed to behave.

151
00:08:51,530 --> 00:08:54,786
I wrote a reference card
to act as documentation.

152
00:08:54,810 --> 00:08:59,586
It also helped me ensure
that the user interface I was defining

153
00:08:59,610 --> 00:09:03,089
could be explained concisely
and clearly to regular people.

154
00:09:03,810 --> 00:09:09,226
Bob worked in the attic of the apartment
he rented in Arlington, Massachusetts.

155
00:09:09,250 --> 00:09:10,970
This is the inside of the attic.

156
00:09:12,530 --> 00:09:15,306
Bob bought time on the MIT Multics System

157
00:09:15,330 --> 00:09:18,050
to write computer code
on a terminal like this.

158
00:09:18,690 --> 00:09:22,186
And then he would download test versions
to a borrowed Apple II

159
00:09:22,210 --> 00:09:25,466
over a phone line
using an acoustic coupler,

160
00:09:25,490 --> 00:09:26,650
and then we would test.

161
00:09:27,210 --> 00:09:31,930
For one of these tests I prepared
for this case about the Pepsi Challenge.

162
00:09:33,010 --> 00:09:35,930
Print wasn't working yet,
so I had to copy everything down.

163
00:09:36,770 --> 00:09:39,186
Save wasn't working,
so every time it crashed,

164
00:09:39,210 --> 00:09:42,210
I had to type in all of the formulas
again, over and over again.

165
00:09:42,530 --> 00:09:46,506
The next day in class, I raised my hand;
I got called on, and I presented the case.

166
00:09:46,530 --> 00:09:50,106
I did five-year projections.
I did all sorts of different scenarios.

167
00:09:50,130 --> 00:09:53,810
I aced the case.
VisiCalc was already useful.

168
00:09:54,450 --> 00:09:57,050
The professor said, "How did you do it?"

169
00:09:57,759 --> 00:10:00,426
Well, I didn't want to tell him
about our secret program.

170
00:10:00,450 --> 00:10:01,970
(Laughter)

171
00:10:02,490 --> 00:10:04,346
So I said, "I took this and added this

172
00:10:04,370 --> 00:10:06,466
and multiplied by this
and subtracted that."

173
00:10:06,490 --> 00:10:08,586
He said, "Well,
why didn't you use a ratio?"

174
00:10:08,610 --> 00:10:11,626
I said, "Hah! A ratio --
that wouldn't have been as exact!"

175
00:10:11,650 --> 00:10:14,226
What I didn't say was,
"Divide isn't working yet."

176
00:10:14,250 --> 00:10:17,346
(Laughter)

177
00:10:17,370 --> 00:10:21,226
Eventually, though,
we did finish enough of VisiCalc

178
00:10:21,250 --> 00:10:23,090
to be able to show it to the public.

179
00:10:23,730 --> 00:10:25,706
My dad printed up a sample reference card

180
00:10:25,730 --> 00:10:27,730
that we could use as marketing material.

181
00:10:28,570 --> 00:10:34,346
In June of 1979, our publisher
announced VisiCalc to the world,

182
00:10:34,370 --> 00:10:38,490
in a small booth at the giant National
Computer Conference in New York City.

183
00:10:39,090 --> 00:10:43,250
The New York Times had
a humorous article about the conference.

184
00:10:43,770 --> 00:10:46,226
"The machines perform
what seem religious rites ...

185
00:10:46,250 --> 00:10:49,811
Even as the believers gather,

186
00:10:49,835 --> 00:10:52,931
the painters in the Coliseum sign room
are adding to the pantheon,

187
00:10:52,955 --> 00:10:56,126
carefully lettering 'VISICALC'
in giant black on yellow.

188
00:10:56,150 --> 00:10:57,966
All hail VISICALC!"

189
00:10:57,990 --> 00:11:01,566
(Gasp) New York Times:
"All hail VISICALC."

190
00:11:01,590 --> 00:11:03,086
(Laughter)

191
00:11:03,110 --> 00:11:07,206
That was the last mention
of the electronic spreadsheet

192
00:11:07,230 --> 00:11:11,127
in the popular business press
for about two years.

193
00:11:11,151 --> 00:11:12,881
Most people didn't get it yet.

194
00:11:13,370 --> 00:11:14,570
But some did.

195
00:11:15,178 --> 00:11:19,378
In October of 1979, we shipped VisiCalc.

196
00:11:20,250 --> 00:11:22,488
It came in packaging
that looked like this.

197
00:11:23,090 --> 00:11:25,570
And it looked like this,
running on the Apple II.

198
00:11:26,370 --> 00:11:28,330
And the rest, as they say, is history.

199
00:11:28,730 --> 00:11:30,826
Now, there's an awful lot
more to this story,

200
00:11:30,850 --> 00:11:32,906
but that'll have to wait for another day.

201
00:11:32,930 --> 00:11:35,170
One thing, though, Harvard remembers.

202
00:11:35,770 --> 00:11:36,970
Here's that classroom.

203
00:11:37,610 --> 00:11:41,050
They put up a plaque
to commemorate what happened there.

204
00:11:42,010 --> 00:11:49,010
(Applause)

205
00:11:50,770 --> 00:11:53,706
But it also serves as a reminder

206
00:11:53,730 --> 00:12:00,426
that you, too, should take
your unique backgrounds, skills and needs

207
00:12:00,450 --> 00:12:05,706
and build prototypes to discover
and work out the key problems,

208
00:12:05,730 --> 00:12:07,850
and through that, change the world.

209
00:12:08,610 --> 00:12:09,826
Thank you.

210
00:12:09,850 --> 00:12:14,611
(Applause)