0:00

What were you listening to before we hopped on?

0:02

It's your walkout music. New Beyonce. Oh,

0:05

new Beyonce. I haven't heard it yet. How is it? I

0:07

really like it. Nice. I think

0:09

it is reductionist to call it country. Ah.

0:13

I was appropriately enough listening to Start Me

0:15

Up through the ages. Of course you were.

0:18

But I kind of feel like David

0:20

Rosenthal move is that you might've

0:23

been listening to Start Me Up whether we

0:25

were doing Microsoft or not. That's a very

0:27

squarely in your genre song. The Stones though,

0:29

man. It's crazy. They're

0:32

in their 70s, 80s. Amazing. Man.

0:35

I hope we're in our 70s and 80s

0:37

dancing on stage. Season 126. Yeah.

0:41

All right. Let's do it. Let's

0:43

do it. Who got the truth? Is

0:47

it you? Is it you? Is it you? Who

0:50

got the truth now? Is

0:52

it you? Is it you? Is

0:54

it you? Sit it down. Say

0:57

it straight. Another story on the

0:59

way. Welcome

1:01

to season 14, episode four of

1:03

Acquired. It's a podcast about great

1:05

companies and the stories and playbooks

1:07

behind them. I'm Ben Gilbert. I'm

1:09

David Rosenthal. And we are your

1:11

hosts. We often remark

1:13

that selling software is the best

1:16

business model of all time. Well

1:18

today, finally, we tell

1:20

the story of the company

1:22

that created that business, Microsoft.

1:26

Finally. We're like 10 years into Acquired.

1:28

We're finally doing it. It's been daunting.

1:31

We wanted to do it for a while,

1:33

but it takes some chutzpah to tackle Microsoft.

1:35

I'm so fired up. We're ready. It's

1:37

time. Yep. Well listeners, Microsoft

1:40

today is sprawling and massive.

1:42

It is the world's most

1:44

valuable company worth over $3

1:47

trillion. They have

1:49

49 years of history making

1:51

software for consumers and enterprises,

1:54

making hardware, gaming systems, gaming

1:56

studios, Windows apps, iPad apps,

1:58

Mac apps, operating systems, mobile

2:01

operating systems, MP3 players, search engines,

2:03

cloud computing services, on cloud computing,

2:05

programming languages, development environments, the list

2:08

goes on. But it

2:11

did not start out that way. Today,

2:13

we will tell the story of the

2:16

desktop software company. Before

2:18

the enterprise, before IT, before

2:20

the internet, before being

2:22

a trusted partner to governments around

2:24

the free world. And

2:26

really before people even knew what

2:28

to do with personal computers, this

2:31

is the story of a bunch of ragtag

2:33

geniuses in their 20s pushing

2:35

what was possible. Welcome

2:38

to Microsoft, the PC era.

2:41

Well listeners, if you wanna know every

2:43

time an episode drops, you can get

2:45

hints at the next topic and follow

2:47

up, you can sign up at acquired.fm

2:49

slash email. Come talk about this episode

2:51

with the community at acquired.fm slash Slack.

2:54

If you want more from David and I, you should check

2:56

out our second show, ACQ2, where

2:58

we interview founders, investors, and experts often

3:00

as a deeper dive into topics we

3:03

cover on the main show. And

3:05

before we dive in, we wanna briefly

3:07

thank our presenting sponsor, JP Morgan Payments.

3:09

Yes, just like how we say every

3:11

company has a story, every company's story

3:13

is powered by payments and JP Morgan

3:16

Payments is a part of so many

3:18

journeys from seed to IPO and beyond.

3:20

So with that, this show is not investment advice.

3:22

David and I may have investments in the companies

3:25

we discuss and likely all of you if you

3:27

hold any index funds and the

3:29

show is for informational and entertainment purposes

3:31

only. David,

3:34

where on earth do we start the Microsoft

3:36

story? We're right down the middle

3:38

on this one. We're gonna start in 1955 in Seattle, Washington with

3:42

the birth of Bill Gates III or

3:44

Trey as he's known. Yes.

3:48

Growing up, it's so confusing because his dad

3:50

is the second but he goes by senior

3:52

and Bill is junior slash the third slash

3:54

Trey. So Bill in 1955 is

3:56

born as the second of three children

3:59

to Bill and and Mary Gates. Now,

4:01

Bill Gates, senior,

4:04

his father, is from Bremerton, the

4:06

Navy town just across the Sound from

4:08

Seattle, where he grows up in a

4:11

family that owns and runs a furniture

4:13

store there, a long way from

4:16

the software king of the world here. Now,

4:18

Bill Gates, senior slash the second, after high

4:20

school, he joins the army during World War

4:22

Two, serves during World War Two. And then

4:25

he goes, I presume on the GI Bill

4:27

to the University of Washington, where he's the

4:29

first member of his family to go to

4:31

college. And there he gets an undergrad and

4:34

a law degree in four years, and

4:36

then decides to stay in Seattle with

4:38

his new family and become a practicing

4:41

attorney. Now, I say family because

4:43

at the UW, he meets and marries

4:45

one Mary Maxwell.

4:48

And Mary, I don't know how

4:51

to put it other than that she is a force. Yes.

4:54

So Mary's family had founded

4:56

National City Bank. And

4:58

her father was a senior executive at First

5:01

Interstate Bank, which later became a big part

5:03

of Wells Fargo. Now, Mary,

5:05

despite being the daughter of a

5:07

successful business family in that day and

5:10

age was not cast aside like

5:12

so many other daughters we've talked about on

5:14

the show, New York Times, Hermes, where

5:16

it was sort of passed to the son in

5:18

law to continue to run the business. That was

5:20

not the case with Mary Maxwell. No.

5:23

So after she graduates from the UW,

5:25

she becomes first

5:27

the president of the Seattle Junior League.

5:29

And she starts joining nonprofit boards in

5:31

Seattle as a very young woman, she

5:34

joins the Seattle Symphony Board, the Chamber

5:36

of Commerce, the Children's Hospital, the King

5:38

County United Way. And she's

5:40

such a force on these boards that

5:42

she starts getting asked by her fellow

5:44

board members to join their company's

5:46

boards to like the corporate boards, they're so

5:49

impressed with her. So first, she

5:51

joins the board of First Interstate Bank, the bank

5:53

that her family's a part of. Then

5:55

she joins the board of Cairo Television

5:57

in Seattle, she even ends up joining

6:00

the Pacific Northwest Bell Telephone Board.

6:02

That's right, part of the AT&T

6:04

breakup. Yeah, eventually she joins the

6:06

Board of Regents of the University

6:08

of Washington and the whole entire

6:11

National United Way Board. So she

6:13

never works full-time in a corporate

6:15

setting, but it is not an

6:17

overstatement at all to say that Mary Gates became

6:20

one of the most powerful business

6:22

people in the Pacific Northwest, period.

6:25

Absolutely, and Bill Gates Sr. was

6:27

the prominent attorney in the region,

6:29

and so it's quite the power couple.

6:32

Ben, it's like you're reading my script here. Yes, we

6:34

don't want to give Bill Sr. the short shrift

6:36

here either. He becomes a

6:38

superstar lawyer, and he becomes a co-founding

6:41

partner of the firm Preston, Gates and

6:43

Ellis, which today, I didn't even

6:45

realize this till I dug in the research.

6:47

That is K&L Gates today, one of the

6:49

largest law firms in the world. Yep.

6:52

And another fun fact that you probably know

6:54

about Bill Gates Sr., but we got this

6:56

being acquired, we'll talk about, do

6:58

you know what corporate board he joined later

7:00

in life? I do not. Costco.

7:03

Ah ha ha, of course. Bill

7:06

Gates Sr., we should say, too, basically

7:08

galvanized the entrepreneurial community

7:11

in Seattle. He started

7:13

the Tech Alliance. He was a huge

7:15

angel investor. He really did organize, you

7:17

know, angel investors, people who want to put

7:19

high-risk capital to work into startups, and

7:22

his heart was there, obviously, through his

7:24

law practice, long before Bill

7:27

Gates III became the prodigy he

7:29

became. Totally, and that's the point

7:31

we want to land here, is for young Bill,

7:33

Trey, growing up here, he

7:35

is growing up in like a pretty unique

7:38

household. He would later talk about being like nine

7:40

or 10 years old, and most

7:43

nights at dinner at his house, there would

7:45

be a CEO or a senator or a

7:47

governor or somebody who's just over

7:50

for dinner, and Bill would sit there and

7:52

absorb the business conversation. It's

7:54

like the Hermes family, the Dumas family

7:56

that we talked about on that episode. This

7:59

whole thing, it makes me think. of Paul's main character

8:01

in the Dune movies in the book.

8:03

He's kind of bred from birth to

8:05

be this incredible

8:08

business mind. I mean, at age 13,

8:11

with his best friend, who we will

8:13

talk about very soon, he brought up

8:15

the idea, I wonder what company I

8:17

will be the CEO of when I

8:19

grow up. What industry will I go

8:21

after? What problems will I tackle? It

8:23

wasn't a question of if, but which.

8:26

Yes. And it just turned out that he

8:28

would be the CEO and founder of the biggest

8:30

company in the biggest industry ever to

8:32

exist. The other thing that we've

8:34

got to say about Bill growing up, he

8:37

is insanely competitive. So he did

8:39

not and does not like to

8:41

lose at anything. And that is

8:44

putting it mildly, whether it's sports,

8:47

or swimming, or computers, or school, or

8:49

the classroom. There's a quote in one

8:51

of the books we read from a

8:53

childhood friend of his who says, everything

8:56

Bill did, he did competitively and never

8:58

simply to relax. I

9:00

think this used to be more than today. There's kind

9:02

of this image of Bill Gates that he was a

9:04

computer nerd, that he was like this shy little skinny

9:07

kid. And the way he looks doesn't help this. But

9:09

that is not the case at all. This

9:11

guy had a competitive fire in him,

9:13

I'm sure still does, like

9:16

none other. Both things can be true.

9:18

He was the number one math student

9:20

in the state of Washington. He was

9:22

a nerdy kid and a brilliant kid,

9:24

and also fiercely competitive. His childhood friend

9:26

and co-founder of Microsoft, Paul Allen, would

9:28

say about him, you could tell three

9:30

things about Bill Gates pretty quickly. He

9:32

was really smart. He was really

9:35

competitive. He wanted to show you

9:37

how smart he was. And he was

9:39

really, really persistent. That sounds

9:41

about right. So famously, speaking

9:44

of Paul and where Bill and Paul

9:46

meet, when Bill is

9:48

in seventh grade, his parents enroll

9:50

him at the Lakeside School, Which

9:53

now I think is internationally famous because

9:55

of Bill. But It is a super

9:57

rigorous college prep school, middle school, and

9:59

high school. An Bill ends

10:01

up writing the scheduling software

10:03

for class scheduling that he

10:05

puts himself in the classes

10:07

with all the girls like

10:09

this. Like

10:13

this bill: Thirteen years old,

10:16

seventh grade lakeside. This.

10:18

Is when it starts I missing Microsoft

10:20

as instead but during that year. Those.

10:22

Thirteen years old. But. Lakeside

10:24

Mothers Club. Raises.

10:27

Money. To. Buy the school

10:29

a teletype. And connected

10:31

up and rent complete or time

10:33

from a deck Pdp ten. That.

10:36

Is located in downtown Seattle and

10:38

owned by the branch office their

10:40

of General Electric. Now.

10:43

Probably a bunch of the relic. I

10:45

have no idea what any of those

10:47

nerds me and so we gotta set

10:49

some contexts. This is Nineteen, Sixty Eight

10:51

Banks, Sixty Eight of the Beatles, Vietnam,

10:53

The Summer Of Love. This is Not

10:55

the computer. Age Two thousand and one

10:57

A Space Odyssey had just come out.

10:59

Nolan Bushnell has not sounded Atari yet.

11:02

Bob Noise and Gordon Moore are only

11:04

just leaving Fairchild Semiconductor to start Intel.

11:06

Silicon Valley Still dominated by Law T.

11:08

there's no such thing as a microprocessor.

11:10

The United States would land on the

11:12

moon. One year later, totally.

11:15

The Wake Computing. Works.

11:17

Back then it was basically still be

11:19

any act days like a computer met

11:21

two things in either meant. A

11:24

massive room size. Must seek

11:26

The had a bout the

11:28

computing power of a calculator.

11:31

Or. It meant a human. People. Talked

11:33

about computers as humans. Do.

11:35

You ever see the movie Hidden Figures about

11:38

this Black women who did the calculation. Those

11:40

women were that computers. They were called the

11:42

computers. Yes, because they would sit there and

11:44

compute. This. Was a totally different

11:47

era. So the idea that a thirteen

11:49

year old kid to say school this

11:51

middle school would get access to. Share.

11:54

Computer time. I. Can

11:56

imagine there are many other secondary schools in the

11:58

country that we're doing as. This

12:00

is a very early place to make the

12:02

plenty. Microsoft. Is the

12:05

result of tremendous intelligence,

12:07

brilliant strategy, fierce competition,

12:09

And an unbelievable amount of luck.

12:12

Bill Gates was born in Nineteen

12:14

Fifty Five, the same year as

12:17

Steve Jobs to come into adulthood

12:19

just as the Personal Computer Ways

12:21

is starting. And. The.

12:24

Fact that he was at a middle

12:26

school and had this much privilege where

12:28

he could get access to a Pdp

12:31

ten at this point in his life

12:33

to help him understand how important computers

12:35

would be. I mean. There.

12:37

Are. Dozens. Of people

12:39

in America who were as well

12:41

situated as Bill is and that

12:44

might be overly generous, He and

12:46

Paul got a sneak peek into

12:48

the future there at Lakeside. Now

12:50

it's funny You said the personal computer era.

12:52

We are not that far away from the

12:54

personal computer hill. I mean we got us

12:56

of the state what is computing I mentioned

12:58

any act in this room size things. Computers

13:01

did not have screens. You know, curses

13:03

is never late sooner. pixels. Everything

13:06

was done on a teletype,

13:08

the kind of like typewriters

13:10

and. They. Were wired

13:12

up remote. we either in the same

13:14

facility or like went Lakeside is doing.

13:16

You could be remote m and it's

13:19

almost like the cloud today and it

13:21

cold over a phone line. but I

13:23

was the tele type. Yeah exactly. Got

13:25

wired over the phone line, hooked up

13:27

to these mainframes and see you typed

13:29

commands into this teletype and then the

13:31

response team back over the phone line

13:33

or of or whatever table. From.

13:35

The mainframe. And. It got printed

13:37

out on a spool of Keep on The Teletype.

13:40

But. This. Is Power Live

13:42

Normal thirteen year olds don't come

13:45

anywhere near access. Yeah. What?

13:47

Is the computing market at the time? It

13:50

is. Pretty much more compact the

13:52

pretty much in a minute. a

13:55

hundred percent dominated by i b

13:57

m oh yes i b m

13:59

big blue, you know, big iron is

14:01

what it was referred to, like the products

14:03

that they would produce. They

14:06

were the industry. Yes,

14:09

Ben Thompson has a fantastic quote on

14:11

this. He has an article called What

14:13

is a tech company? And here's his

14:15

comment. 50 years ago,

14:17

what is a tech company was an easy

14:19

question to answer. IBM was the tech company,

14:22

and everybody else was IBM customers. That may

14:24

be a slight exaggeration, but not by much.

14:26

IBM built the hardware at that time, the

14:28

system 360. They wrote the

14:31

software, including the operating system

14:33

and the application and provided

14:35

services, including training, ongoing maintenance

14:37

and custom line of business

14:39

software. Yeah, system 360

14:42

was a line of solutions, I

14:44

would say offered by IBM. And it consisted

14:46

of the thing, the room

14:48

size thing, the mainframe, and

14:50

the software, which was system 360, and

14:52

the consulting and the implementation. You know,

14:55

you couldn't just call up UPS and forklift

14:57

one of these things into a company and

14:59

expect it to work. No, you got to

15:01

operate this thing to totally like ASML machines,

15:03

you don't just go ship them off to

15:05

TSMC and say, Good luck making semiconductors. Exactly.

15:07

It's a full solution, full service thing. But

15:09

an important thing that was also happening this

15:11

year 1968, was

15:13

that IBM was undergoing

15:16

some antitrust scrutiny over that huge

15:18

bundle that I just told you

15:20

about. I mean, doesn't it smell

15:22

like antitrust? They do everything

15:24

from the hardware to the software, the

15:27

operating system, the service, the support, they

15:29

are the whole market, they're starting to

15:31

get concerned. And so proactively, they

15:33

unbundled hardware, software and services, and

15:36

they started selling those separately for

15:38

the first time, which was not

15:40

a problem at first. But what

15:43

it did was it cracked the door for customers

15:45

to say, Oh, I can buy

15:47

hardware from IBM and software from someone else.

15:50

And other people were not explaining this, but

15:52

it was possible. Yeah, interesting.

15:55

It cracks the door for Microsoft like 15 years later, 20

15:57

years later. Yeah. But

16:00

this is where the seeds are sown

16:02

of what is the exploitable opportunity when

16:04

Bill Gates is ready to do something.

16:06

Interesting. Now, back to the timing

16:09

thing for Bill and Paul and Microsoft. I

16:11

mentioned when we were setting this up that there's something

16:13

else to talk about here. IBM

16:16

was facing a disruptive force at

16:18

this moment. I think probably for

16:20

the first time in its history,

16:22

certainly in the computing era of

16:24

IBM's history, and that was

16:26

the Digital Equipment Corporation or DEC.

16:29

Just when we said earlier that

16:31

Lakeside is renting computer time from

16:34

the General Electric Computer in downtown

16:36

Seattle, I said it's a

16:38

DEC PDP-10. It's not an

16:41

IBM product. So what's DEC? They're

16:44

the mini computer company. Mini

16:46

in quotes. Yeah, mini, quote unquote. Mini

16:48

meaning it was the size of a closet, not

16:50

the size of a room. It's

16:53

all relative. DEC had been started

16:55

by this guy, Ken Olson from MIT. What

17:00

they did, he had this brilliant insight that would

17:02

play out over and over and over again in

17:04

technology. I'm not going to

17:06

go compete with IBM head on. I'm not going

17:08

to make any DEC. I'm not going to make

17:10

mainframes. But computing

17:13

has advanced enough that

17:16

there's an opportunity to make

17:18

something smaller, less powerful, more

17:21

toy-like. And

17:23

there's enough demand out there that I think I can

17:25

find some new markets for people

17:27

who will buy those types of

17:30

computers. And it's smaller businesses, but

17:32

in particular, it's like branch offices

17:34

of the big company. So I

17:36

am sure General Electric bought

17:39

lots and lots and lots of

17:41

IBM mainframes and products at their

17:43

headquarters. But the GE

17:46

field office in Seattle, they're

17:48

Not going to truck in a mainframe..

17:50

And In 1968, Seattle, I mean, this

17:52

is like a provincial little town. I'm

17:54

standing here right now. It's a major

17:56

city and a huge economy in the

17:58

United States. But at the time, kind

18:01

of.. a podunk, forgotten, sleepy, faraway place.

18:03

Totally. At. Microsoft would go

18:05

a long way to changing that overtime.

18:08

And we should say this is the classic

18:10

low and disruption playbook and in this is

18:12

what Clayton Christensen was talking about. Going from

18:14

mainframe to many computers, I'm in to make

18:17

something that's worse for most things, but better

18:19

for some new things that new customers and

18:21

new markets are gonna care a lot about

18:23

an I B M gonna look at it

18:26

and gulps that can't do any of the

18:28

things that are important to our customers. And

18:30

that's exactly why it works. and specifically, why

18:32

are they not going to care about. I

18:35

don't know exactly what An I B

18:37

M mainframe system system. Three sixty, three,

18:39

seventy whatever costs. I'm imagining tens of

18:41

millions of dollars all in total cost.

18:43

Maybe hundreds of millions of total dollars

18:45

to. Run. A system like that? I'm by

18:47

it. So. The First Deck

18:50

Machine when it comes out the Pdp

18:52

one a few years before this time.

18:55

It was priced at one hundred and twenty

18:57

thousand dollars so like an order of magnitude

18:59

may be to below a system that you

19:02

had by from I B M. Obviously

19:04

still a long, long, long way from the personal computer.

19:06

People are not buying these things for their houses. But.

19:09

You know yet? Zeal? Buy one

19:12

for the field office in Seattle

19:14

or universities will buy them for

19:16

research for their students, for their

19:18

professors. and so deck. China

19:21

creates a new market

19:23

for computing. And. Bill.

19:26

He so studied in business history. The.

19:28

Founder: Ten Olson. Ten his

19:30

bills hero T totally looks up.

19:33

To. Deck and What Deaths Done. Now.

19:36

The deck and mini computer you eyes is

19:38

still the same as the mainframe. You're still

19:40

using a teletype. There's no innovation in terms

19:43

of what the computing is or how use

19:45

it's It's just cheaper and more people have

19:47

access to it. So. Back

19:49

to Leak Side and the Mothers Club raising

19:51

money for. This. Access here for the

19:54

school. Bell. Armor he started

19:56

he's in seventh grade and Lakeside is in

19:58

middle school high school. The. is actually

20:00

in a separate building. The

20:02

computer room that gets installed with the

20:04

teletypes, that's over in the high school.

20:07

But Billy, he doesn't care. He gets exposed to it, I

20:09

think, in a math class one day and he's like, oh,

20:12

I'm hooked. So he goes

20:14

over, he's hanging out with the high schoolers,

20:17

teaches himself how to program. And

20:19

pretty quickly, he becomes known as one of

20:21

the very best programmers there.

20:23

And he and three other kids

20:26

form what they call themselves the

20:28

Lakeside Programmers Group. And

20:30

one of his buddies who he forms it with is,

20:32

of course, the high schooler,

20:34

I think the 10th grader at the time,

20:36

Paul Allen. There is a

20:38

fantastic photo listeners that we will tweet

20:41

of Bill and Paul sitting in the

20:43

computer room at Lakeside. And Bill, I

20:46

think he's like 13, 14,

20:48

he looks like he's about eight. I think on

20:50

the wall, there's this, almost like, printed

20:52

out magazine thing that says the bugs

20:54

slayer that they've hung up over the

20:57

wall. It's amazing. So these

20:59

high school kids, they start the Lakeside

21:01

Programmers Group. They call it

21:03

the Programmers Group because they are programmers.

21:05

This is another super important thing to learn

21:07

to use a computer at this

21:09

time meant to be a programmer. Right.

21:11

There was no packaged software that you

21:14

bought. The software that IBM was selling

21:16

was the operating system to

21:18

make the machines actually function. And

21:20

it was the programming languages

21:22

that you could then program

21:25

on. But you weren't clicking around and using

21:27

Excel or like, you know, pulling up apps.

21:30

Everybody who used the computer wrote their

21:32

own software. Right. There was not this

21:34

multi sided network of you've got developers

21:36

making applications, and then you've got users

21:38

of those applications. No, everybody who used

21:40

the computer was a programmer. So

21:42

the goal of the Lakeside Programmers Group,

21:45

remember, Bill is this business prodigy, is

21:47

to use their very valuable and

21:50

very rare skills as programmers at this

21:52

time to, you know, make

21:54

money. Yep. Do a business. So

21:57

turns out, at the same time, I mean,

21:59

they coincidentally. There

22:02

is a local startup

22:04

coming out of the University of

22:07

Washington called the Computer Center Corporation

22:09

or C-Cubed. The

22:11

business plan behind C-Cubed was that they were

22:13

going to get a bunch of decks, a

22:16

bunch of PDPs, 8s, 10s, 11s, whatever, and

22:19

they were going to be like AWS.

22:21

They were going to just be a

22:23

computer time-sharing company. C-Cubed

22:26

hires the Lakeside

22:28

Programmers Group, these kids, to

22:30

come in and find and

22:32

document bugs in the system,

22:35

and they're going to pay them directly in computer time. So

22:38

when they come into C-Cubed, they

22:40

learn Fortran, they learn Lisp, they

22:42

learn machine language for the PDPs.

22:45

Back at Lakeside, they were just using BASIC,

22:47

the programming language. Which is

22:49

reasonably high level in terms of how abstract

22:51

it is. You're not writing machine language. You

22:54

don't have to know how to

22:56

address memory and registers and all that.

22:59

It reads kind of like English. You know how to

23:01

add numbers together. It's not an elegant

23:03

language, and it's a very verbose language, but

23:05

if you sort of look at it with

23:07

your eyes as a person who speaks English

23:10

and knows BASIC math, you're like, I kind

23:12

of understand what this program does. So there's

23:14

a meaningful amount of translation done by a

23:16

BASIC interpreter that takes you from the BASIC

23:18

code you have to write to what is

23:20

actually running on the machine. Yes,

23:23

but BASIC, we don't want to give

23:25

the impression that it is just BASIC

23:27

or just for kids. No, it's widely

23:29

used. It's going to become hugely, hugely

23:31

important. It is both

23:33

the gateway programming language for everybody, but it's a

23:35

real programming language and a lot of stuff is

23:37

done in it. It's sort of the Python of

23:39

its day. I think the way Python is now

23:42

where you sort of joke that Python

23:44

is so flexible, you can accidentally write

23:46

a program by writing English, and it can kind

23:48

of forgive a lot of mistakes, and it reads

23:50

kind of like English. It's a reasonable

23:52

parallel to draw it way back when with BASIC

23:54

where you say, look, you can

23:56

understand it as a layman, but also it's used

23:58

in a broad set of business applications.

24:01

Totally. When Bill and

24:03

Paul and their buddies come into C cubed,

24:06

they're now getting access to

24:08

learn real hardcore systems programming

24:11

languages, including machine code

24:13

for the PDP-10.

24:16

They're becoming pretty prolific engineers

24:19

here. They're handwriting assembly

24:21

code. Yes. And

24:23

they're getting mentored. One

24:25

of the executives at C cubed

24:28

is a guy named Steve Russell. Did you find

24:30

this, Ben? Do you know about this? No. Uh-uh.

24:32

This is amazing. You're going to die. Steve

24:35

Russell was the guy

24:38

who wrote Space War when he

24:40

was at MIT on the first

24:42

PDP, the PDP-1. He's

24:44

like a computer science legend. Nolan Bushnell told

24:46

us about that. Yes. Space War was the

24:48

first video game, first computer

24:50

game ever written. It was written as

24:53

a fun side project by some MIT

24:55

engineers in the early days of DEC.

24:57

And then that became Nolan Bushnell's inspiration

24:59

for starting Atari and Pong and Space.

25:01

Yeah, Steve Russell, that guy. He mentored

25:04

Bill Gates. And he was here at

25:06

the University of Washington? He

25:08

had come out to the University of Washington and then

25:10

left and was part of one of the execs starting

25:12

this company. C cubed? Wow. That's like a mile from

25:14

my house. Right? Crazy. Wow.

25:18

So, after a little while at C cubed, all of

25:20

this real expertise that these kids

25:23

are getting leads to another

25:25

opportunity at another timeshare computing company

25:27

based down in Portland. They

25:29

ask the kids to write a real

25:32

piece of software to write a payroll

25:34

billing program for all

25:36

their clients that are using the

25:38

timeshare system. And Bill

25:41

now, who's the de facto leader of this

25:43

group, he negotiates a deal with the help

25:45

of his dad, Bill Gates, senior, prolific corporate

25:48

attorney in Seattle, that rather than

25:50

just being paid hourly for their time, they're

25:52

going to get a royalty on the revenue

25:54

that their client makes on the software. I

25:57

can't believe it. These kids are teenagers. They're figuring out the whole

25:59

software business. this model here, they end

26:01

up making at least $10,000 from

26:04

these royalties, which the average

26:06

household income in the US at the time

26:08

was below $10,000. These kids are rolling in

26:11

money. Wow.

26:14

So the next year, Paul graduates from

26:16

Lakeside and goes off to college at

26:18

Washington State. But he and

26:20

Bill decide to team up on a new

26:23

venture that they're going to do together called

26:26

Traf-O-Data. They've identified

26:28

a market opportunity, and that opportunity

26:30

is reducing traffic. Yes. So

26:33

the business plan is that municipalities

26:35

count cars that go through intersections, use

26:37

that to make decisions about how they're

26:39

going to do city planning. Bill

26:41

and Paul are like, we can take this

26:44

new thing that's coming out of

26:46

Intel, a microprocessor, which is promising

26:48

to be a full computer on

26:50

one chip, and we can use that.

26:52

We can build a machine that is

26:54

going to be a computer and it'll process

26:56

and analyze that data, and then we can

26:58

sell it to governments. Like, great, big market.

27:00

And listeners, are you sensing what's happening here?

27:03

Mainframe, mini computer, microprocessor. We

27:05

have to keep using smaller

27:07

and smaller words to represent

27:09

the fact that the computer

27:11

is getting smaller and smaller

27:13

here. It wasn't until we started

27:16

doing research for this episode that I finally

27:18

realized, oh, microcomputers, which

27:20

is the original term for the personal computer

27:22

for the PC, it was called microcomputers before

27:24

PC caught on. They're called microcomputers because they're

27:26

based on the microprocessor. Yes, absolutely. It's not

27:29

just that micro is smaller than mini. And

27:31

it is funny that it kind of stopped

27:33

there. The computers that are sitting on all

27:35

of our desks are microcomputers. Right.

27:39

Yeah. So while they're waiting for

27:41

the AD-08, this new first microprocessor

27:43

from Intel to come out, or at

27:45

least for them to get access to it, they

27:47

want to get a head start on programming

27:50

their traffic data machine and programming

27:52

this microprocessor. So Paul's like,

27:54

I got this. I can find a way to make

27:56

this happen. He takes the

27:58

PDP-10 at Washington. Washington State, and

28:01

he writes a whole emulator

28:04

program to mimic the

28:06

instruction set for the 80-08

28:09

from the manual, and they get

28:11

a full emulator up and running, and they can code

28:14

even without the microprocessor actually

28:16

being there and having access. It's just like

28:18

NVIDIA, you know, like when Jensen was like,

28:20

no, we've got to build an emulator and

28:22

simulate this, and then we're going to ship

28:24

it site unseen. They're doing the same thing.

28:26

It's funny, in many ways, at this point

28:28

in history, getting a manual was actually much

28:30

more valuable than getting the processor itself, because

28:33

the processor would arrive, and unless there was

28:35

documentation, you would have no idea how to

28:37

interact with it to take advantage of its

28:39

power. But if you had a manual, well,

28:42

sure, you couldn't actually test the stuff you

28:44

wrote for it on the hardware, but

28:46

if you wrote an emulator on a

28:48

bigger, more powerful computer that could sort

28:50

of mimic the computer that you're actually targeting,

28:53

you could go years before actually

28:55

ever running the software on

28:57

the target device and just work off

28:59

of what the manual says, as long

29:01

as the manual is correct and matches

29:03

how it actually works. Yes, and

29:06

this is going to become very important

29:08

to Microsoft in just a second.

29:11

So traffic data is not a huge

29:13

success. I think I read a few places, they

29:15

make about $20,000 in revenue from it. So

29:18

like, again, great money for high school and

29:20

college kids, but, you know, not world-changing

29:22

stuff here. Not what Bill aspires to

29:24

for the company he's going to start.

29:27

But Bill and Paul are getting

29:29

experience with the microprocessor. Bill

29:32

actually has the idea for Microsoft when

29:34

they're working with it. He's like,

29:37

oh, this is a computer,

29:39

why don't I go off and write

29:41

an interpreter for BASIC here, and we

29:43

can sell the BASIC interpreter for the

29:45

microprocessor and build a big business. The

29:48

AD-08 just wasn't powerful enough

29:50

yet to do that. But

29:52

spoiler alert, that totally becomes

29:55

Microsoft. The seeds of Microsoft

29:57

are selling language interpreters for...

30:00

new processors, new hardware, new computers

30:02

that enable you to write familiar

30:04

programming languages on that new hardware.

30:07

Yeah. And Bill and Paul

30:09

are not the only ones having this insight here

30:11

too. Another Seattle guy

30:14

named Gary Kildall, who they had intersected

30:16

with, who they knew from CQ in

30:18

the University of Washington, he kind

30:21

of had the same idea here. We're going to

30:23

bring up Gary and his company Digital Research a

30:26

little bit later. Yeah, he cared a little bit

30:28

less about programming languages and a little bit more

30:30

about operating systems. So that's how they diverge for

30:32

the few years here. Yep. But

30:35

Bill and Paul, they absolutely

30:38

see the vision for what this

30:40

can grow into and become. Bill

30:42

has a great quote. Paul

30:44

and I had talked about the microprocessor. And

30:47

it was really his insight that

30:49

because of semiconductor improvements, things would

30:52

just keep getting better. I said

30:54

to him, oh, exponential phenomena are

30:56

pretty rare, pretty dramatic. Are

30:59

you serious about this? Because this

31:01

means, in effect, that we can think of

31:03

computing as free. It was

31:05

a gross exaggeration, but it was probably the easiest

31:07

way to understand what it means to cut costs

31:09

like that. And Paul was quite

31:11

convinced of it. Yes. Is

31:13

this in the Smithsonian interview? Yes, so good.

31:15

It's so good. This quote is incredible because

31:18

basically Paul Allen brings up Moore's law to

31:20

Bill Gates. They don't use that language there.

31:23

And in 1971, that is what's happening. And

31:26

for Paul, this is just an observation of, hey,

31:28

there's an exponential thing happening here. Seems

31:30

like it's going to keep happening. It's been happening. And

31:33

Bill's shaking. And he's like, what? Exponential

31:36

phenomena don't just happen. That's

31:38

incredibly, incredibly rare and immediately

31:40

gets Bill's wheels turning on. What does this mean

31:42

for the world? If that's actually

31:45

true, we need to act and

31:47

do something profoundly different than anyone's ever done

31:49

before, because this enables new things that no

31:51

one ever thought could be possible. This

31:54

moment is the genesis of the vision for

31:56

Microsoft. Even though Bill doesn't say the words

31:58

in this quote. division of a computer

32:01

on every desk in every home.

32:04

That's the famous part and the part

32:06

that got left off later when the

32:08

DOJ started sniffing around was running Microsoft

32:10

software. But that is the

32:12

vision here and it is crazy at

32:14

the time like a computer on every

32:17

desk and in every home.

32:19

Bill sees that this is what this exponential

32:21

phenomenon what Moore's law means that that is

32:24

going to happen. You know, we're still in

32:26

the era of teletypes. Nobody

32:28

else sees this. Yeah. And it's also

32:30

the reason why Microsoft is going to

32:32

form into such a different type of

32:34

company that's ever come before it. Why

32:36

they can break all the rules, why

32:38

they can sell just software, even though

32:40

that's never been a thing before, why

32:43

their business model can be so much different

32:45

than everyone else's business model. I mean, in

32:47

prepping for this episode, we got to talk

32:49

with Pete Higgins, who ran Excel and was

32:52

an executive overseeing office for a long time

32:54

in the early days. And he had this

32:56

great quote to us, which was, compute on

32:58

every desk was wackadoo stuff. People laughed at

33:00

it. It was absolutely wild. People thought, I

33:02

don't know, maybe one in 10 people in

33:04

their finance group or something will have one

33:07

at some point. This is the profoundness of

33:09

an exponential decrease in price or increase in

33:11

power of computing is it's going to become

33:13

universal. So all that said, even

33:17

Bill and Paul know the 80-08. It's

33:20

not there yet. It's not powerful enough to

33:22

really be a general purpose computer on a

33:24

chip, but they know it's coming. So

33:27

in the fall of 1973, Bill goes

33:29

off to college at Harvard famously.

33:31

It's funny at Harvard, you know, kind

33:33

of like Jim Simons that we talked about

33:36

at MIT on the RENTEC episode, Bill thinks

33:38

he's going to be like a world class

33:40

mathematician and set the world on fire. It's

33:43

literally a quote from Paul Allen. So Bill was

33:45

the number one math student in the state of

33:47

Washington. And he gets there and he does this

33:49

theoretical math class, Math 55, and

33:51

gets a B. And Paul says, when it

33:54

came to higher mathematics, he might have been

33:56

one in 100,000 or better, but

33:58

there were people who were one in a million. one

34:00

in ten million and some of them wound up

34:02

at Harvard. Bill would never be the smartest guy

34:04

in the room and I think that hurt his

34:06

motivation. He eventually switched his major to applied math.

34:10

Yes. So while Bill's at Harvard, he's

34:12

also doing a bunch of the typical college kid

34:15

stuff. He's playing poker, he's cutting classes, he's making

34:17

friends. And one of the friends he

34:19

makes there is a kid down the hall from him named

34:21

Steve. Steve, bomber.

34:24

And bomber, even everybody knows bomber. He's

34:26

kind of everything that Bill is not.

34:28

He's super social. He's super outgoing. He's

34:30

in a final club, which is like

34:33

a big thing in the social scene

34:35

at Harvard. He's gregarious. Anyone who's ever

34:37

met Steve or seen a video of

34:39

Steve, you are well aware that this

34:41

man has presence. But the thing

34:43

that people don't know about him is he is

34:46

so unbelievably analytical.

34:50

Steve is the guy that outscored

34:52

Bill Gates on the Putnam exam.

34:54

Which is the annual math competition

34:56

for college students. Yeah, Steve

34:58

isn't a programmer, but he is every bit the

35:01

mathematician that Bill Gates is. And that is one

35:03

of these things where I think when people try

35:05

to set it up as well, you know, you've

35:07

got the brilliant programmer

35:09

genius and the marketing guy. It's

35:11

just like, those are the

35:13

roles they took. But I think when you're

35:15

getting a sense of who the original crew

35:18

was at Microsoft, they were all brainiacs and

35:20

they were all wildly analytical. So

35:23

then, in the spring of 1974, Bill's

35:25

freshman year, Electronics Magazine

35:28

publishes big news about

35:30

a new Intel chip, the next generation, the

35:32

next turn of the crank on Moore's law,

35:35

the 8080. And in

35:38

Bill's words here, all at once

35:40

we were looking at the heart of a

35:42

real computer and the price was

35:44

under $200. We

35:46

attacked the manual. I

35:49

told Paul, Dec can't sell any

35:51

more PDP-8s now. It seemed obvious

35:53

to us that if a tiny

35:55

chip could get so much more

35:57

powerful, the end of big unwieldy

35:59

machines. was coming. And

36:02

this is really where Bill Gates commits to computers

36:04

to be his life's work. I think what's

36:07

often lost in the story is, Bill,

36:09

even though he was good at computers

36:11

and spent tons of time programming computers,

36:13

he never fancied himself a computer guy

36:15

until this moment in history, he went

36:17

to Harvard because he felt like, hey,

36:20

if I ever want to be a lawyer or something

36:22

else, like they've got a lot of great programs there.

36:24

And this was the moment where I think it really

36:26

clicked for him that I'm just in

36:28

the middle of the right place at the right time

36:30

with the right skill set. And this is my way

36:32

of having the most impact on the world. So

36:35

they think, okay, what's clearly going to happen here

36:38

is all the big computer companies, you know, IBM,

36:40

deck, the big Japanese computing

36:42

companies, they see this,

36:44

they're going to get into this business, they're

36:46

going to make machines, and they're

36:49

going to make micro computers, surely, they

36:51

will jump on this opportunity, right? If

36:54

it didn't destroy their existing business model,

36:56

sure they would. Exactly.

36:59

So Bill and Paul kind of are

37:01

sitting around waiting through 1974 and 1975 and like,

37:06

hey, when are the 8080 computers going to

37:08

come out? Where are they?

37:10

It's just crickets. Yep. Paul

37:12

is so convinced that the revolution

37:15

is coming that he actually drops

37:17

out of Washington State, moves

37:19

to Boston to be close to Bill so that they

37:21

can be ready when it happens. And

37:24

that summer, they both get summer jobs at

37:26

Honeywell as programmers. Paul stays

37:28

on into the next school year when Bill

37:30

goes back to school, he's just like waiting,

37:32

waiting, waiting. And then in

37:35

December 1974, Paul

37:38

is walking across Harvard Square,

37:40

and he sees

37:43

in a newsstand the January

37:45

issue of Popular Electronics on

37:48

whose cover is the all-tier 8800, the world's first real,

37:56

honest to God, commercially available

37:58

for sale. microcomputer.

38:02

And the legend has it that Paul grabs

38:04

the magazine, runs over

38:06

to Bill's dorm, throws open the door,

38:09

throws the magazine on the desk and is

38:11

like, it's here. And Bill just says,

38:13

oh my God, it's happening without

38:16

us. We need to get on

38:18

this right now. What's so funny,

38:20

he thinks they're already behind because clearly they're not

38:22

history would show I think Bill Gates even says

38:24

we might have actually started a year or two

38:26

too early. The market actually hadn't materialized yet. And

38:29

the funniest thing is the starting gun went off and

38:32

Bill and Paul ran and everyone else

38:34

is still standing around. And

38:36

this is it. This is the moment

38:38

the revolution is here. Microsoft is about to

38:40

be founded. But first, this

38:42

is a great time to thank our presenting

38:44

partner this season, JP Morgan payments. Yes,

38:47

just like Bill's vision of Microsoft to

38:49

put a computer on every desk and

38:51

in every home, JP Morgan has a

38:53

vision for making payments real time 24

38:56

seven everywhere in an every currency.

38:59

So JP Morgan has experienced a

39:01

massive digital transformation and is now

39:03

much more than just a global

39:06

bank. They invest $15 billion a

39:08

year into technology and R&D. So

39:11

with JP Morgan payments, you are getting their

39:13

history of service stability and scale, but

39:15

also innovation and new technology. And

39:18

scale is really the key word here. Every

39:20

episode we talk about Hamilton, Helmer

39:23

seven powers and boy, does JP

39:25

Morgan have scale. What's

39:27

unique about their position in payments is

39:29

that they can make massive infrastructure investments

39:31

that benefit their customers in the near

39:34

term. In addition to the long term

39:36

technology bets like blockchain or quantum computing,

39:38

since they move $10 trillion a day in

39:40

160 countries, these investments create a

39:44

lot of shared value across their large

39:47

customer base. Yeah, not to mention the

39:49

second of the seven powers network economies.

39:51

Yep. If you're a JP Morgan payments

39:53

customer, whether you're a startup or a fortune

39:56

500, you can access their payments partner network.

39:58

This is a set of third party integrations

40:00

across virtually every industry and payment use

40:02

case. It makes it easy to discover

40:04

what specific set of tools are best

40:06

for your business. It's basically

40:09

a B2B marketplace that connects their

40:11

clients with the broader payments and

40:13

technology ecosystem, such as treasury management

40:15

systems, ERPs, point of sale hardware

40:18

solutions, payment gateways, and so

40:20

much more. And given JP Morgan scale,

40:22

you can imagine the value that comes

40:24

with unlocking access to their industry leading

40:26

partners and technology. Let's take

40:28

Visa, for example, which listeners may remember.

40:31

We talked a lot on that episode

40:33

about the power of a network. Well,

40:35

Visa works with JP Morgan to provide

40:38

global payment processing services. And now to

40:40

the partner network, you can easily add

40:42

Visa as an acceptance method through a

40:45

streamlined implementation with consolidated reporting. Other

40:47

companies in the partner network include

40:49

Oracle, American Express, MasterCard, Accenture, and

40:52

many more, including Friends of

40:54

the Show, Modern Treasury. Across

40:56

any industry, the most successful companies

40:58

we've talked about are the ones

41:00

who think and invest strategically through

41:02

a long term lens. Like Hermes,

41:04

durability and reliability are paramount for

41:07

building hundred year companies that can

41:09

reinvent and continue to grow. And

41:11

that is the case for JP

41:13

Morgan payments. Their history of trusted

41:15

solutions and investments in technology bring

41:17

clients and new FinTech companies together

41:19

to drive innovation and growth for

41:21

businesses of all sizes. So

41:23

before we jump back into Bill and

41:25

Microsoft who essentially wrote the Modern Playbook

41:27

on building platforms, be sure to check

41:30

out JP morgan.com/acquired or click on the

41:32

link in the show notes. So

41:35

David, they're like, it's happening without us.

41:37

What do they do? Well,

41:39

they do the natural thing

41:41

that to super excited, ambitious,

41:44

high octane college kids would do.

41:47

They call up the main

41:49

phone number of Altair's manufacturer,

41:51

a company called MITs MITS

41:54

and ask for the president, a

41:57

man named Ed Roberts and

41:59

Bill and Paul They get him on the phone and

42:01

they say we have a basic

42:04

interpreter Ready to go

42:07

ready to ship for the 8080 Intel

42:09

chip and we want to provide it

42:12

for you for your machine Which

42:14

of course they don't they don't have a single lot of

42:16

code written. They don't have anything but it's a market test

42:19

They want to know like what's the response if this were

42:21

true? Exactly, and

42:23

he's a bit of a character himself. He

42:25

says okay Well guys a lot of other people

42:27

are calling me and saying the same thing what

42:30

I'm telling them and I'm gonna tell you too is

42:33

that anybody who can come here to my

42:35

office in Albuquerque, New Mexico and demonstrate

42:38

a working version of basic

42:40

on My Altair

42:43

will get a contract with us to distribute

42:45

it when they go on sale And

42:47

so Bill and Paul they say okay

42:50

great. We'll see you soon Hang

42:53

up and by see you soon.

42:55

They mean let's go get to work And

42:58

this is a big deal for myths too if

43:01

this works because right now they've just announced a

43:03

machine For which you

43:05

can't really do anything on it the hardware

43:07

is powerful But they're not gonna have a

43:09

lot of customers unless there's stuff you can

43:12

do on the machine and a basic interpreter

43:14

running on it You know, it's

43:16

quite valuable to then make the claim you can

43:18

program basic on our computer So they're very excited

43:20

about this even though they're kind of playing coy

43:23

I was gonna talk about this in a minute, but let's

43:25

talk about what the Altair is What

43:28

did they just announce in the magazine here? This

43:31

is the Altair is the first mass

43:33

market commercially available personal computer again

43:36

No screen or anything. Yeah does

43:38

not have a screen does

43:40

not have a keyboard doesn't have a display of any

43:42

kind what it does have is

43:45

it has a set of 16 lights on

43:47

the front of the machine like Christmas lights and

43:51

16 switches and you can flip

43:53

the switches to flip bits and then the

43:55

machine will respond by lighting up different Patterns

43:57

of lights. It doesn't come with any thought

44:00

There's nothing. That's all it is. 16 lights,

44:02

16 switches. So

44:04

in order to use it, you got to hook

44:06

up your own teletype. You got to get the manual. You

44:09

got to hope that the manual is right. And

44:11

you got to code to the machine instructions, like

44:13

literally the assembly language for the chip

44:16

inside, for the Intel 8080. Now,

44:19

back to the TRAFO data days and

44:22

Paul writing the emulator for

44:24

the 8008 at Washington State.

44:27

Paul's like, we got this. I'll just write

44:29

an emulator on the Harvard PDP-10

44:31

for the 8080 instruction set. So

44:34

he does the same thing again. They get the

44:36

manual and they have an emulator and they write

44:38

it against an emulator. Yep. Bill

44:41

writes the basic interpreter. And

44:43

in a couple weeks, they've

44:45

got it working. And that's

44:47

like, okay, come on out to Albuquerque.

44:50

So Bill and Paul, remember Bill still

44:53

looks like he's 12 at this point in time.

44:55

They decide that just Paul should go. And does

44:57

Paul have like a rocking beard at this point

44:59

yet? Paul is super 70s. He

45:01

is like into it. And as we'll

45:03

see, he's going to fit right in

45:05

at Albuquerque and MIT. So

45:07

Paul gets on a plane, flies

45:10

from Boston to Albuquerque. And,

45:12

you know, in a total like epic

45:14

legend moment, they didn't have

45:17

a bootloader written for the basic interpreter.

45:19

So they had the basic. It was

45:21

all written. They'd done it on the

45:23

emulator. And Paul's flying out with the

45:25

tape, the computer tape with the code

45:27

of the basic interpreter on it. But

45:30

he's like, oh, shoot, we

45:32

can't just feed that right into the

45:34

machine. There's got to be a bootloader

45:36

to load up this thing. So he

45:39

writes the bootloader on the plane. On

45:41

paper by hand, he is hand coding

45:44

octal, not even assembly language instructions.

45:46

Like he's hand coding in

45:49

pure octal the instructions

45:51

to load their basic

45:53

interpreter program into memory.

45:56

Yep. So he lands in

45:58

Albuquerque. drives out to MITS,

46:01

they load the bootloader onto the

46:04

prototype Altair there that loads up

46:06

the basic interpreter, and

46:09

it fails. It doesn't work. Paul's

46:12

like, oh, let's try it again. Let's try it again. So

46:14

they try it again. This is how early computing is.

46:16

It works the second time. Who knows what the bug

46:18

was the first time? It didn't change anything. It just

46:20

didn't work the first time, and it worked

46:22

the second time. So it loads

46:24

up. Paul writes in the

46:27

instructions, print 2 plus 2. It

46:29

spits out 4. And by spits out, I mean

46:31

the lights light up and say 4. And

46:34

both he and Ed, the jaws are

46:37

on the floor. Paul's like, oh my god,

46:39

the basic works. And Ed's like, oh my god,

46:41

the Altair works. Neither of them believed this was

46:43

actually going to work. And

46:45

Ed actually has more eggs in this

46:47

basket than he sort of let on,

46:50

because when Bill and

46:52

Paul call and say, hey, can you

46:54

give us the teletype instructions, he

46:57

reveals they're actually the only ones who called about

46:59

that. So everyone else who said they were writing

47:01

a basic never got far enough to ask, how

47:03

do we actually interact with your computer?

47:06

Yes. So now

47:09

is the time to say a few words

47:11

about Ed and Mitt. Like, what is this

47:13

company? Bill and Paul originally

47:15

thought that it was going to be the

47:18

IBM's, the DEC's, the Japanese companies who are

47:20

going to make the first microcomputers.

47:24

Mitt's is about as far away from

47:26

IBM as you can

47:28

possibly imagine. Mitt's basically

47:31

was a model rocket company.

47:34

Which Albuquerque is a great place to do that. Ed,

47:37

Robert's the founder. He had

47:39

been in the Air Force and stationed in

47:41

Albuquerque. And that's how he got involved in

47:44

model rocketry. And the reason

47:46

that they're introducing the Altair and

47:48

they made this big splash in

47:50

the magazine was this is a

47:52

last ditch gambit to try and

47:54

save the company. So you got

47:56

like a bankrupt model rocketry company.

47:58

And their little gambit worked and got

48:00

a couple of college kids to pounce. Totally.

48:03

And like, why did it work? They

48:05

had two things going for them that

48:08

really Ed, I think probably

48:10

personally made happen. One,

48:13

they got this splashy popular electronics

48:15

magazine cover. That was through a

48:17

relationship that Ed had. And two,

48:21

the sticker price was $397, which

48:25

is about $2,300 in today, $20,

48:28

$24. Yes, that's a lot of money, but

48:32

the next cheapest computer that anybody could

48:34

buy at this point in time was

48:36

like a deck, you know, like $120,000

48:38

mini computer. So

48:42

the idea that somebody could buy

48:45

a computer for $400,

48:48

I don't care who's selling that thing. Like, I

48:50

want that. So what did they get some kind

48:52

of sweet deal from Intel? Yes. So this

48:55

is all Ed's doing. The list

48:57

price from Intel for the 8080 chip

48:59

was $360. So like, I

49:03

think this is part of what was deterring

49:06

the market of how would anybody sell a

49:08

kit that was affordably priced when

49:10

so much of the cost of

49:12

goods would go to Intel with the processor

49:14

deal. He managed to negotiate a

49:17

volume deal with Intel to get 8080 chips

49:19

at $75

49:21

a pop. Whoa. And that was

49:23

the key unlock. That's like a

49:26

5x price reduction. Totally. Huh. I

49:28

wasn't able to find how that negotiation went

49:31

down or why like Ed Roberts in Albuquerque,

49:33

New Mexico got this sweetheart deal from Intel.

49:35

I mean, either the list price is

49:37

like wildly wrong or they were cutting

49:40

deals all over the place or one

49:42

thing it could have been is just that I'm

49:44

totally speculating, but chips are

49:46

the ultimate high fixed

49:48

cost investment, low marginal costs

49:50

next to software. You

49:53

could imagine maybe Intel had already put all the

49:55

money into the fixed cost of spinning up the

49:57

fabs and was

49:59

expecting. a certain amount of market demand

50:01

and they weren't seeing it. They were like, crap, we got

50:03

to recoup our investment. I don't know, lower the price, let

50:05

them just maybe we'll make it up in volume. I

50:08

like that. I have no idea. But that's

50:10

a totally viable, I think, train of

50:12

thought here. Either

50:14

way, he gets the sweetheart deal. And

50:17

not only does it make computer history

50:20

and enable and create Microsoft, it saves

50:22

the company of myths. So they were

50:24

on the edge of bankruptcy after

50:26

the popular electronics article comes out,

50:29

they get 4000 preorders

50:31

in the first month or two, which

50:33

is one and a half million dollars

50:36

in revenue cash

50:38

paid upfront. That is

50:40

pure cash hitting the bank account.

50:42

That'll save a company. And also

50:44

just proves, hey, 4000 people just

50:46

paid cash sight unseen for

50:49

16 lights and switches. There's a

50:51

lot of demand for a home computer here.

50:53

And at decent margins, too, I mean, if

50:55

they're getting the processor for 72

50:57

bucks and they're selling it for what did you say? Yeah.

51:01

So I mean, everything else in there is much cheaper

51:03

than the processor. So I don't know, depending

51:05

on how much they have to give to the sales channel they're

51:08

selling through, if it's retail or

51:10

distributors. I think they're selling direct. I

51:12

think people are just sending money orders.

51:14

Yeah. So it's decent margin business. Unlike

51:17

what the PC business would become over

51:19

time, they managed to have nice margins.

51:22

So Paul and Ed, they kind of hit

51:24

it off. Paul decides to move

51:26

out to Albuquerque to be close to

51:28

the action here. And he actually joins

51:30

MITs as their vice

51:32

president of software. Now,

51:34

he's vice president of a software department of one.

51:36

He is the software department. Yeah,

51:40

software department of one. But, you know, he's got

51:42

his buddy Bill Gates, who is not employed. But

51:44

Bill's definitely working on software for the altar as

51:46

well. Yes. So Bill

51:49

stays at Harvard, but keeps

51:51

cranking on enhancing the basic

51:53

interpreter and adding more functions

51:55

and functionality to the version

51:57

of basic that they had just written for the altar.

52:00

Altair and then once the

52:02

school year is over, he comes out to

52:04

Albuquerque too for the summer. Now, the Altair

52:06

is getting ready to ship with the basic,

52:08

the Microsoft basic included in it. Bill

52:11

and Paul need to set up a company, but

52:13

Paul is an employee of MITs at

52:15

this point. So what did they do?

52:17

They set up a partnership. So

52:20

the founding of Microsoft,

52:22

at this point, Micro-Soft

52:24

is a two-person partnership between Bill

52:27

and Paul. As we record

52:29

this, that was 49 years and one day ago. So

52:32

we are sitting here on April 5th recording that was April

52:34

4th, 1975. It

52:38

is very funny to look back at

52:40

some of the original signatures when Bill

52:42

writes on letters. It's Bill Gates, the

52:45

general partner of Micro-Soft, which is great.

52:47

I think it's actually a Paul Allen

52:49

name where he wants to put together

52:51

microcomputer and software and Bill's like, that's

52:54

perfect, we're immediately just going to run

52:56

with it. As it was

52:58

a partnership, originally they were going to call it something

53:00

like Allen and Gates. And then

53:02

they ultimately are like, no, Microsoft is

53:05

perfect. Microsoft has become

53:07

like Kleenex. But like, no, no,

53:09

it's like Microsoft means microprocessor software.

53:13

I will say like it's a nice

53:15

clarifying North Star because it

53:17

really draws the line in the sand

53:19

and says, we're in the software business.

53:21

And Gates makes this really clear to

53:23

Paul Allen, who is often tempted to

53:25

do hardware stuff. And Bill is very

53:27

hardcore about saying, no, what

53:29

we're uniquely good at in the world is

53:31

software. And we should stick to that. I

53:34

also suspect Bill is starting to realize there's

53:37

an amazing business model here, if we can

53:39

pull it off, where we don't have to

53:41

make the hardware and we can charge for

53:43

every copy of the software sold. But

53:45

that insight, I would say, has not

53:47

yet fully materialized. Yeah. OK. Well, let's

53:49

talk about business model here in one

53:51

second. First, though, on the partnership, again,

53:54

we've been saying all along that Bill is clearly

53:56

the leader here. They set up the partnership initially

53:58

at 60. 40 ownership, Bill

54:01

is 60%, Paul is 40%. Later

54:03

it gets changed to Bill is 64%, and Paul is 36%.

54:07

So, yep. Bill is the leader here.

54:10

Yeah. And Bill's case that he makes on

54:12

that to Paul is, hey, you took a job, and

54:14

you were doing this on the side, I was all

54:16

in. And Paul's an agreeable guy, and

54:18

36 is still a nice percent. And so he says,

54:20

Sure. You know, in the long run here, everybody

54:23

gets. It's all sort of a rounding error.

54:25

But to

54:28

that point, back to the business model. So

54:30

once the partnership is set up, they

54:33

sign an exclusive

54:35

licensing arrangement with MITS.

54:38

This is super important.

54:40

This is a big lesson that young

54:43

Bill and Paul are going to learn here. So

54:45

MITS gets exclusive

54:47

license to the basic

54:50

interpreter to the basic as they

54:52

call it for the 8080.

54:55

And MITS is the one

54:57

that can then decide whether to

54:59

sub license the basic out to other

55:02

companies or not. Huh?

55:04

Essentially, this is a distribution

55:06

deal with MITS

55:08

where MITS becomes the exclusive

55:10

seller and distributor of Microsoft's

55:14

basic. Microsoft doesn't

55:16

have any direct sales

55:18

control here. That's going to

55:20

become a big, big, big issue. And

55:23

the terms of the deal are Microsoft

55:25

is to get $30 for each copy

55:27

of basic that MITS sells, plus

55:29

50% of the revenue that comes

55:32

from the sub licensing deals that MITS may

55:34

or may not do with other companies who

55:36

want to use the basic. Which why would

55:38

they ever do any sub licensing deals? Like,

55:40

why would you give it to your competitors?

55:43

Well, that is a really good question, Ben. So

55:46

let's just round that part of the

55:48

revenue to zero. Yeah, this is a

55:50

big, big, big diverging of interest between

55:52

Microsoft and MITS. And

55:55

the kicker on this contract is

55:57

that the total amount of lifetime

55:59

revenue that Microsoft can make from

56:02

the basic. Oh, it's capped? From MITS is

56:04

capped at $180,000. Ha.

56:08

So, and in MITS really

56:10

have the upper hand in this deal.

56:12

Or phrased another way, it's we will

56:15

give you $180,000 for

56:18

you to hand over exclusive rights to all that

56:20

cool basic stuff you just wrote to us. But

56:23

if we sell fewer than X machines, we're actually gonna pay

56:25

it out to you on a prorated basis at $30 a

56:27

pop, rather than giving you

56:29

the full $180,000. Yeah, that

56:31

is another way to frame it. So,

56:34

definitely a great deal for MITS. On

56:37

the other hand, what are Bill and Paul

56:39

gonna do here? Right, kind of a great

56:42

deal for them too, given the position they're

56:44

in. Right, MITS is the industry. Yeah. Now,

56:47

there is one very,

56:49

very important clause in the contract

56:51

though, protecting Microsoft's

56:54

interests. And that

56:56

clause is that MITS must

56:58

use its best efforts to

57:01

license, promote, and

57:04

commercialize the basic, broadly

57:07

in the marketplace. And that any

57:09

failure to do so by MITS would

57:11

be grounds for termination of the

57:14

contract by Microsoft. Hmm,

57:17

thank God Bill's dad is a lawyer. Indeed,

57:20

indeed. So,

57:22

the Altair comes out for sale later in

57:25

1975. Microsoft

57:27

does $16,000 in revenue that year from

57:31

there, $30 a pop, the

57:33

basics that are getting sold with the

57:35

Altair. Which is

57:37

great, especially the first year that they're starting. And

57:40

the next year in 1976, everybody's so excited about

57:44

this new market, the vision that's happening,

57:46

the demand, the Altair, the sales, that

57:49

Paul Allen resigns from MITS

57:51

to join Microsoft full-time. Bill

57:54

drops out of Harvard officially. He moves

57:56

to Albuquerque. They're all in on this.

57:59

But... For the year in 1976, Microsoft's

58:01

revenue is still only $22,000. So

58:05

it was $16,000 the year before, $22,000 in 1976. Not

58:10

a high growth company. This is less

58:12

than they were making in high school. What is

58:14

going on here? One, it's sort of like

58:16

MITS is the one at the controls of sales, not

58:18

Microsoft. But two, MITS

58:21

is selling 1,000 computers a month.

58:23

This is taking off. This is creating a

58:25

new industry. Despite MITS

58:27

selling thousands of computers a month, only a

58:29

few hundred copies of BASIC are selling per

58:32

month. What's going on? People

58:34

are pirating the software. This is

58:36

the discovery of software piracy. And

58:39

this is a pretty interesting time

58:41

to pause and say, well, are

58:43

they pirating software? Because this is 1975.

58:48

So piracy implies that you are

58:50

running afoul of some particular legal

58:53

protection for the good. And

58:56

you might say, well, with today's legal

58:58

frameworks in hindsight, you would say, of course, if they're

59:00

copying the software and not paying the money for it,

59:03

it's piracy. That was actually

59:05

not established yet. And this

59:07

is the craziest thing. So Bill basically

59:09

has an opinion that it's piracy. And

59:11

he writes letters to the computer community.

59:13

He writes an open letter to hobbyists.

59:15

Yes. And tries to basically

59:17

guilt trip people. He tries to use

59:20

that as a recruiting method and say, if you're so

59:22

excited about pirating our software, maybe you should just come

59:24

work with us. And nothing would make me happier than

59:26

making the best software in the world. And please join

59:28

us on this mission. But ultimately, the

59:30

legal standing that he has to say, hey,

59:32

what you're doing is illegal is not fully

59:34

established. And so it would actually take a

59:37

couple of years for the courts

59:39

to look at software and say, what

59:42

about this is protectable? And

59:45

if you think about it, it is a little bit weird.

59:47

So you've got source code that looks kind

59:49

of like English, you know, basic, it's letters

59:51

and numbers. It gets

59:54

translated to machine code. That

59:56

machine code ends up running

59:58

and it's basically a lecture. It's

1:00:01

voltages that are flipped up and down.

1:00:03

And so what about that are

1:00:05

we trying to protect? Ultimately, the way it

1:00:08

gets litigated through some case law from court

1:00:10

cases is that the

1:00:12

source code is a

1:00:14

copyrightable creative work

1:00:16

that is expressed through some

1:00:18

sort of tangible medium. That's

1:00:21

the important thing about copyright law. It's a

1:00:23

creative work expressed through a tangible medium. So

1:00:25

a book, the creative work is the words

1:00:27

and the tangible medium is printed

1:00:30

on paper. And so with software, it

1:00:32

actually took until 1980, Congress

1:00:36

changed the law. We'll put a link

1:00:38

in the show notes to the literal

1:00:41

congressional change that happened. And

1:00:44

it is in Title 17, Copyrights

1:00:46

Chapter 1, Subject Matter and Scope

1:00:48

of Copyright. And in

1:00:50

1980, they include a defined term,

1:00:52

which is a quote unquote computer

1:00:54

program is a set of statements

1:00:56

or instructions to be used directly

1:00:58

or indirectly in a computer in

1:01:00

order to bring about a certain

1:01:02

result. And once you have that

1:01:04

passed by Congress codified into law,

1:01:06

you now have the standing legal

1:01:08

framework that all the whole computer

1:01:11

industry used going forward, in particular

1:01:13

the software industry, a computer program

1:01:15

is copyrightable work. Wow,

1:01:18

I didn't know all that's awesome. It's totally

1:01:20

crazy how recent that is. But when you

1:01:22

think about it, why would that have any

1:01:24

legal, you know, software is such an abstract

1:01:26

idea before the whole business model of computers

1:01:29

was good luck, just replicating an IBM PC

1:01:31

and everything that comes with it. You don't

1:01:33

need any legal standing. But if you're going

1:01:35

to pursue this software only business model, what's

1:01:38

the protection around your abstract product? That's

1:01:40

exactly what I was going to say here. This is

1:01:42

the other element of what's going on. This

1:01:45

is the first time software has

1:01:47

ever been sold. Right. Other than

1:01:49

the IBM accounting machinations to protect

1:01:52

themselves from antitrust, which was

1:01:54

just accounting, nobody had ever sold

1:01:56

software before. This is the first

1:01:58

time. Right. Certainly. to

1:02:00

build like a legitimate business around it. The other

1:02:02

thing that's useful to know is when

1:02:04

you're selling an IBM PC, you're literally selling

1:02:07

a PC to a customer. The same way

1:02:09

that when I'm selling you this glass from

1:02:11

Crate and Barrel, I am selling

1:02:13

you the glass and the glass is now

1:02:15

yours. I've transferred property to you. Software

1:02:18

is not that. So the whole

1:02:20

world of software is built on

1:02:22

a license agreement. So the source

1:02:24

code that computer program, the actual

1:02:26

right of that is retained by

1:02:29

the creator and you license the

1:02:31

copyright to your customer to be

1:02:33

able to use that on their

1:02:35

machine. So there's this dual idea

1:02:37

that computer software is copyrightable

1:02:39

and you can grant a license

1:02:41

under certain conditions for customers to

1:02:43

use it. That is the

1:02:45

legal framework for which the next 50

1:02:47

years of technology at

1:02:50

large would operate under. For the moment

1:02:52

though, they got a piracy problem. Right.

1:02:55

The law isn't going to change until 1980. Microsoft

1:02:57

would be dead if they didn't figure out

1:02:59

a solution to this before 1980. So

1:03:03

this is when Bill ultimately kind

1:03:05

of realizes, shoot, we did the

1:03:08

wrong business deal with MITS here.

1:03:11

MITS has to sell and customers

1:03:13

have to make the decision to

1:03:15

buy our basic. It

1:03:19

is a key critical part of the

1:03:21

value of the computer. It's like the

1:03:23

whole thing. It makes that machine useful.

1:03:26

Right. Without it, it is not

1:03:28

useful. It's totally setting up the

1:03:30

wrong incentives and value equation that customers

1:03:33

should be buying this

1:03:35

themselves. It should be included by

1:03:37

the hardware OEM in the

1:03:39

machine that they are selling and in the

1:03:41

total purchase price. And then if that happens,

1:03:44

we no longer have a piracy problem because

1:03:46

we're just getting paid as part of the

1:03:49

purchase of the machine. Right. It

1:03:51

should be a royalty. Right. The problem is that

1:03:53

is not the deal that they had with MITS. Right. Or

1:03:56

framed differently. Instead of saying, hey, consumer,

1:03:58

do you want to buy something? else

1:04:00

too and make a new purchase decision,

1:04:02

they should be saying, hey computer manufacturer,

1:04:05

we make your thing actually useful, so pay

1:04:07

us for it. Yes. So

1:04:10

during 75 and 76, MITS

1:04:14

pretty much had this

1:04:16

new microcomputer market all to

1:04:18

themselves. There were a couple

1:04:20

other competitors who sprung up, but nothing

1:04:23

made. The Altair and MITS

1:04:26

was the microcomputer company. All

1:04:29

of that changes though in a big way in 1977 when what

1:04:31

Bite magazine calls the 1977 Trinity

1:04:37

hits the market. And that is

1:04:40

three machines, the Radio

1:04:42

Shack TRS-80, Tandy slash Radio

1:04:44

Shack TRS-80, the Commodore

1:04:47

Personal Electronic Transactor or

1:04:49

the acronym PET, and

1:04:52

the Apple II. All

1:04:56

three of which machines were like

1:04:58

the Altair, low cost, mass

1:05:01

market, unlike the Altair,

1:05:04

they were not kits. They were fully

1:05:07

assembled, fully functional, right out of the

1:05:09

box. And they

1:05:11

each had their own major distribution

1:05:13

advantages. And in Bill's words, these

1:05:15

three machines, the 1977 Trinity ignite volume in the

1:05:17

market. Bill

1:05:22

loves these really dramatic verbs. We attacked

1:05:24

the manual. They ignite the volume in

1:05:26

the market. Oh, and the press just

1:05:28

latches onto them. Whenever Bill has a

1:05:30

leaked memo or something where he talks

1:05:32

about all this war terminology, those

1:05:34

all become headlines. So great. Now,

1:05:37

earlier during 1976, Microsoft

1:05:39

had started getting approached by a

1:05:42

few of the bigger computing companies

1:05:44

like NCR, National Cash Register, GE,

1:05:47

Control Data Systems. I

1:05:50

want to license Microsoft Your

1:05:52

Basic for the 8080 microprocessor here

1:05:54

so that we can experiment with these things.

1:05:58

And each of these deals... would

1:06:00

have been revenue to Microsoft of like $100,000-ish.

1:06:04

But Ed and Mitts, they

1:06:06

keep dragging their feet on negotiating these.

1:06:08

They've got the exclusive license. Everything's got

1:06:11

to go through Mitts. And

1:06:13

most of them, they're turning down, because Ben, like

1:06:15

you said, they don't want anybody to come in

1:06:17

and compete with them. Yep. There's

1:06:19

the misaligned incentive. There's the misaligned

1:06:21

incentive, and there is the clause

1:06:24

that Bill and Microsoft, and I presume

1:06:26

Bill Sr., put in the original agreement,

1:06:29

is Mitts using its best

1:06:32

effort to commercialize

1:06:34

the basic and gain

1:06:36

adoption in the market. And you can

1:06:38

make a pretty strong argument that they're not. So

1:06:41

Ed, though, unbeknownst to Bill and Paul, he has

1:06:43

another reason that he's dragging his feet on these

1:06:45

deals, which is that he's about to sell

1:06:47

the company. So in May of 1977, Mitts gets acquired by

1:06:52

the tape drive manufacturer, Pertech,

1:06:54

for $6.5 million, and Ed

1:06:56

kind of rides off into the sunset. And

1:06:59

Pertech, they know about this

1:07:01

dispute with Microsoft, and they

1:07:03

sort of come in and they figure out, who is

1:07:06

this Bill Gates? He's a 21-year-old kid,

1:07:08

a college dropout, like we're a big

1:07:10

company. We can deal with this. And

1:07:13

Robert says an amazing quote later. He

1:07:15

says, Pertech kept telling me they could deal with

1:07:17

this kid. It was a little like

1:07:20

Roosevelt telling Churchill that he could deal with Stalin. Oh,

1:07:23

my God. Oh,

1:07:26

boy. And I also don't think they realize that

1:07:28

this kid's dad is one of the best corporate

1:07:30

attorneys in the country. Yeah.

1:07:32

Well, Bill Gates was just constantly underestimated,

1:07:34

which kind of worked to his advantage

1:07:37

in those early days. Yes,

1:07:39

totally. That's the thing

1:07:41

about Microsoft. People kind of forget

1:07:43

how insanely young Bill was. He

1:07:46

was just 20. And to put

1:07:48

that in context, he's only

1:07:50

seven years older than Jensen

1:07:53

Huang, but they

1:07:55

feel an entire generation apart since

1:07:58

Microsoft was started almost. 20

1:08:00

years before Nvidia. When you

1:08:03

start bending your mind around like, oh,

1:08:05

Bill Gates is still pretty young, considering

1:08:08

what an institution Microsoft has become in

1:08:10

the world. Right. I mean, Microsoft

1:08:12

next year is going to have

1:08:14

its 50th anniversary. And I

1:08:17

believe that also be the same year that Bill

1:08:19

turned 70. That's wild.

1:08:21

Yeah. So that

1:08:23

fall in 1977, the

1:08:26

dispute between MIT slash per

1:08:28

tech and Microsoft goes to arbitration

1:08:30

in Albuquerque. And

1:08:32

the interim months while this arbitration is happening

1:08:35

are the only moment in Microsoft's

1:08:37

history where like cash gets tight.

1:08:40

They're sort of running out of money because they

1:08:42

can't really make any sales

1:08:44

here. Right. They don't control their

1:08:46

destiny. People aren't paying MIT for

1:08:48

the basic, their quote unquote, pirating

1:08:50

the software. They can't do

1:08:52

deals with all the other computer companies

1:08:54

that want to come license directly. And

1:08:57

so things get a little tight. Microsoft

1:08:59

ends up winning the arbitration, I believe

1:09:01

in maybe like November 1977,

1:09:05

meaning they are now totally free to

1:09:07

go license basic to anybody who wants

1:09:09

to buy it on any

1:09:11

terms that they want. So they

1:09:14

turn around, they immediately license it

1:09:16

to the Trinity, you know, Apple,

1:09:18

Commodore, Radio Shack and Tandy. They

1:09:20

license it to all the big

1:09:22

companies, the GEs, the NCRs who

1:09:24

want to experiment with micro computers.

1:09:27

There's a really funny story with

1:09:29

Apple that apparently

1:09:31

was had more or

1:09:33

less written like 95% of their own

1:09:37

basic, their own basic, but it

1:09:39

didn't have floating point numbers. It

1:09:41

only had integer numbers and jobs

1:09:43

is like totally writing was

1:09:45

like the basic is really important. Like can

1:09:47

you just finish it? Can you do floating point and

1:09:50

was just doesn't do it. So

1:09:53

jobs has to go license Microsoft's

1:09:55

basic. Amazing.

1:10:00

The first feel that happens within the company

1:10:02

is a me. There's so many deals done

1:10:04

both directions, commercial deals, equity deals, legal disputes

1:10:06

in both directions, and this is the very

1:10:09

first time that they do something together. Job.

1:10:12

So. Bill. And

1:10:14

Paul Microsoft did you all. These deals may do

1:10:16

them all as cash up front. Fixed

1:10:19

Costs: All you companies you're going pay us

1:10:21

and then you include the basic in

1:10:23

the machines that you're selling and were going

1:10:25

to get all the money upfront. Super.

1:10:28

Presently though, Bill.

1:10:30

Does not. Value. Maximize

1:10:32

on these deal. So like the

1:10:34

Apple deal, Is thirty one

1:10:37

thousand dollars for eight years.

1:10:39

Of. Whoa Access for Apple for

1:10:41

the Apple to Heart and Microsoft

1:10:44

Basic. They're. Not price gouging

1:10:46

here because Bill sees he's like.

1:10:49

The play here is we want

1:10:51

to make it a no brainer

1:10:53

for everybody everybody who's selling and

1:10:55

microcomputer to have Microsoft Basic on

1:10:57

it because we want to set

1:10:59

the standard if we are the

1:11:01

standard. Programming. Environment

1:11:03

That anybody who's using these computers and

1:11:05

can anybody who is using these computers

1:11:07

is programming them. There. Used

1:11:10

to the Microsoft version of basic. Were.

1:11:12

Going to have so much power that you

1:11:14

become a self fulfilling prophecy. All of our

1:11:16

competitors will just. Wither. Away nobody a

1:11:19

want them because it's not going to be

1:11:21

compatible with the language everybody knows and them

1:11:23

once people start trading and and ultimately. Developing.

1:11:26

And selling software that they've written.

1:11:29

It's. Only gonna run on our basic

1:11:31

interpreter. not anybody else is sasson eating

1:11:33

actually didn't know that started this early.

1:11:36

And one cracks in there. You don't know

1:11:38

for a fact it's only gonna run on

1:11:40

Microsoft Basic interpreter, but you do know for

1:11:42

a fact that it will run on Microsoft

1:11:44

Basic Interpreter. And so it's It's cheap enough.

1:11:47

Why would you take the chance on a

1:11:49

clone that might have one or two things

1:11:51

wrong with. It. Just. Basically his

1:11:53

vision is I want to remove

1:11:55

any oxygen from. Any. Argument

1:11:57

anyone could have about not using.

1:12:00

Microsoft Basic on a microcomputer who.

1:12:02

He thinks about this concept as

1:12:04

a positive spiral that he really

1:12:06

in his mind is the reason

1:12:09

for Microsoft success. He says success

1:12:11

reinforces success in a growing my

1:12:13

gift. One. Way of doing something

1:12:15

gets a slight advantage over it's competitors.

1:12:17

This is most likely to happen with

1:12:19

high technology products. They can be made

1:12:21

in great volume for a very little

1:12:24

increase costs and as he get that

1:12:26

slight advantage in on compound and this

1:12:28

is what he's playing for here and

1:12:30

it's interesting In the earliest days. What?

1:12:32

Was stopping someone else from writing a

1:12:34

basic interpreter and licensing get to Apple

1:12:37

or Radioshack? Nothing. There were other smart

1:12:39

people out there and so was just

1:12:41

a. Very good business decision to

1:12:43

say. We're. Close That door. We

1:12:45

just gotta make this a no brainer for people

1:12:47

to buy from us because if we're value maximizing

1:12:49

and it's starting to feel expensive, they're going to

1:12:51

turn elsewhere in till we get a lead. Race

1:12:54

the Apple Store is the perfect example. you

1:12:56

know the was thing is cute and makes

1:12:58

for good story and like all he didn't

1:13:01

finish the basic but when make a sauce

1:13:03

sold them the basic which was already getting

1:13:05

established as the standard for thirty one thousand

1:13:07

dollars serve my not no brainer. Apple could

1:13:10

have gone out and hired another programmer to

1:13:12

finish the basic. The daily quip. Boot.

1:13:14

Do that are I could just get the standard one for thirty

1:13:16

one thousand dollars on a do that. Just. Because

1:13:18

it's important to establish there's tree asr everything.

1:13:21

If you're running a start up right now,

1:13:23

you might think to yourself, oh great, I'll

1:13:25

just run that exact strategies. The important thing

1:13:27

here is a most of the work was

1:13:29

already done for the original basic. Be.

1:13:32

Bill was doing it himself Bill and

1:13:34

Paul So the importance of technical cofounders

1:13:36

and their overhead was crazy low and

1:13:38

so they could do these deals where

1:13:40

they don't make very much money because

1:13:42

they were I think at the end

1:13:44

of seventy seventy or five employees. So.

1:13:46

Their overhead was just so unbelievably low

1:13:49

that they couldn't take a really long

1:13:51

lens. So. Yes, even though it

1:13:53

takes until the very end of Nineteen seventy

1:13:55

Seven when the Mets dispute gets resolved and

1:13:57

Microsoft can actually. Make money, In.

1:14:00

May and Nineteen seventy seven with three

1:14:02

hundred and eighty one thousand dollars in

1:14:05

revenue despite zero for the first like

1:14:07

eleven months as. They

1:14:10

just rolling in task. This is when

1:14:12

Bill famously goes out and buys agree

1:14:14

and person nine eleven and is motoring

1:14:16

around. Albuquerque Journal sense of speeding ticket

1:14:18

said. you know hilarious stuff. This is

1:14:20

when he got his mugshot right? Yeah,

1:14:22

the good. So this is a classic.

1:14:24

Bill Gates or not, They? yeah this

1:14:26

is just. Amazing because he got

1:14:28

i think three speeding tickets in one day,

1:14:31

two of which were some the same police

1:14:33

officer. Yeah. And there's funny stories

1:14:35

of the Albuquerque Police thought this: You know

1:14:37

how his kids driving a Porsche nine eleven?

1:14:39

He must be like a drug dealer or

1:14:41

something, because even when he was five twenty

1:14:43

twenty one, he looked seventeen. right? Is

1:14:45

like do you even have a driver's

1:14:47

license Wilde Ah, But back to what

1:14:50

you're saying. It's a really, really important

1:14:52

point. To make

1:14:54

this dynamic works he needs

1:14:56

to be able. To.

1:14:59

Afford the investment in the

1:15:02

fixed costs for the software,

1:15:04

for the technology. To. Make

1:15:06

it that little bit superior. Like

1:15:08

Bills talking about the slight advantage

1:15:10

over the competitors, and at this

1:15:12

moment in time the industry is

1:15:14

completely brandy like the software industry

1:15:17

is brand new. So. The

1:15:19

amount that that six costs seats,

1:15:22

the cost is quite low really.

1:15:24

It's just. Bill. In Paul's

1:15:26

time and dedication to this

1:15:28

industry that nobody else is

1:15:31

making that investments. You. Can't

1:15:33

run this playbook today because. In

1:15:35

any market, even a brand new market,

1:15:37

even a speculative Martha, the minimum viable

1:15:39

fixed cost is. Billions. Of

1:15:42

dollars, right? Yeah, it's interesting.

1:15:45

I. also think this moment galvanize

1:15:47

something important which is bill and

1:15:49

paul could sort of look around

1:15:51

and see there is gonna be

1:15:53

so much value created by microcomputers

1:15:55

and by software they really found

1:15:58

religion around saw is magic. The

1:16:00

things that people can create now

1:16:02

that we've done this basic interpreter

1:16:05

and these machines are cheap and plentiful,

1:16:08

the magic will take care of itself as

1:16:10

long as we ensure this industry can just

1:16:12

exist and do its thing. And so they

1:16:14

flipped from this mode of we need to

1:16:16

bite and scratch and claw and make sure

1:16:18

that we win in deals to, huh, how

1:16:22

can we enable software as a thing

1:16:24

to thrive? And I'm sure we can

1:16:26

position ourselves well to capture some or

1:16:28

a lot of that. And

1:16:30

I think they became almost stewards of

1:16:33

the software industry and evangelists from this

1:16:35

point forward. Yes. They

1:16:37

also do another really prescient thing the next

1:16:39

year in 1978, which is they

1:16:43

go global. Yes. Nobody

1:16:45

else is going global yet. And the way that this happens

1:16:47

is so fun. Bill gets a

1:16:49

call one day. Also, how crazy is it you've

1:16:51

got five people, you're operating out of Albuquerque, you

1:16:53

just finally expanded from having one customer and you're

1:16:56

like, you know what we should do this year?

1:16:58

Let's open in Japan and become an international

1:17:00

company. Yes. Now the

1:17:02

way this happens is one day, Bill

1:17:05

gets a call from a guy

1:17:07

named Kazuhiro Nishi or Kay

1:17:09

Nishi, who's a computer enthusiast

1:17:12

in Japan, has gotten

1:17:14

a hold of Microsoft's

1:17:16

basic totally shares the same vision

1:17:18

as Bill and Paul. He doesn't have a

1:17:20

Paul, you know, nor is he technical himself.

1:17:23

He's like, I'm going to bring you

1:17:25

guys to Japan. I'm going to bring you to all

1:17:27

the big computer companies. They agree

1:17:29

that Kay will become Microsoft's

1:17:31

exclusive distribution partner in Japan.

1:17:34

And by the next year in 1979, half

1:17:38

of Microsoft's revenue is coming

1:17:40

from Japan, which is

1:17:42

wild. It's unbelievable. And it stayed

1:17:44

at that very high run rate

1:17:46

of international being a huge chunk,

1:17:48

you know, close to half always,

1:17:51

basically forever. This is

1:17:53

a huge cornerstone of Microsoft's success that

1:17:55

they were an international company from year

1:17:57

three of their existence. Yes.

1:18:00

Totally. So, revenue in 1977, that

1:18:02

last month of revenue was almost $400,000. 1978

1:18:07

revenue is $1.3 million. They

1:18:10

have 13 employees at this point. 1979

1:18:13

revenue is $2.4 million. And

1:18:15

at the end of that year, they're like, all

1:18:17

right, we got to get out of Albuquerque.

1:18:20

And they've got 25 employees, I believe,

1:18:22

at that point? Yeah, something like that. And

1:18:25

this is when they moved to Seattle. And

1:18:27

it's interesting to hear Bill talk

1:18:29

about this. He actually really liked

1:18:32

Albuquerque. And specifically, there

1:18:34

weren't any distractions there. No distractions,

1:18:36

weather was great. Yeah. Everybody

1:18:39

was happy there. But the big problem was

1:18:41

recruiting. Yeah. He was like, you know, if we're

1:18:43

going to build this into the opportunity that

1:18:45

I see, the vision that I see and

1:18:48

that Paul shares with me, there's no way

1:18:50

we're going to do that in Albuquerque. We

1:18:53

got to move to a hub. Right. And

1:18:55

so he's got three reasons for why Seattle in

1:18:57

particular. By the way, it worked. Every single person

1:18:59

except for his secretary did make the move. So

1:19:02

one, he grew up in Seattle. He's like, I just

1:19:04

want to go home. And then

1:19:06

he justifies it in two other ways, which I found

1:19:08

pretty fascinating. This is from an interview in the early

1:19:10

90s that he did. He said

1:19:13

it basically came down to Seattle or

1:19:15

Silicon Valley. And in

1:19:17

Silicon Valley, it's hard to keep secrets because there's

1:19:19

a rumor mill. And in Seattle, we can be

1:19:21

a little bit more removed. And we can announce

1:19:23

things when we want to announce them. And

1:19:26

two, in Silicon Valley, people switch around

1:19:28

companies. I don't want that. I want people

1:19:30

to just work at Microsoft. There

1:19:32

was a disadvantage to not being able to recruit from

1:19:34

your competitors. But for a

1:19:36

while, they were really the only game in

1:19:38

town in Seattle. Right. But not

1:19:40

really, though, because pretty quickly, Microsoft is such an

1:19:42

important part of the industry. They recruit from Silicon

1:19:44

Valley, too. We're going to talk about some of

1:19:47

the people who come up. But

1:19:49

you're totally right. People stay at

1:19:51

Microsoft. They don't leave. This

1:19:53

continues right through to this day. And

1:19:55

the other thing that I think is really important

1:19:57

to say that makes it work for Seattle.

1:20:00

in a way that I don't know that this could

1:20:02

have worked in too many other places in the country

1:20:05

is the University of Washington. The

1:20:07

computer science department there was really

1:20:09

good. There were great people, Steve

1:20:11

Russell had come out there, there

1:20:13

was real talent and they were

1:20:15

churning out graduates out of

1:20:18

the U-Dub that would go on

1:20:20

to populate Microsoft for decades to come. Oh,

1:20:22

and then Bill of course reinvested in

1:20:25

that flywheel, donating tons of money to

1:20:27

the university. I mean, there's buildings, there's

1:20:29

whole new schools. Yes, absolutely,

1:20:31

they played right into it. And so today,

1:20:33

I mean, it's always a top 10, if

1:20:36

not top five computer science program in the

1:20:38

country, but unlike other top computer

1:20:41

science programs, it's a state school, so it

1:20:43

just has huge volumes. I think more students

1:20:46

come out of the University of Washington and go

1:20:48

to big tech than any other program

1:20:50

in the country. That has stayed this

1:20:52

amazing advantage. Yeah, I think the only

1:20:54

thing even close to it

1:20:56

is Berkeley in the Bay Area with a lot

1:20:59

of the same dynamics, but there's Stanford there too.

1:21:01

So it's kind of like a dual university system

1:21:03

in the Bay Area, but yeah, cannot overstate

1:21:06

how important the University of Washington was to

1:21:08

this decision and then the ultimate success of

1:21:10

coming to Seattle. Yep. So

1:21:13

this brings us to 1980, in

1:21:16

the beginning of the year when they move to

1:21:18

Seattle, just in time for,

1:21:21

I think you can make a

1:21:23

very strong argument, the single

1:21:26

most important deal ever

1:21:28

done in the history of

1:21:31

technology. Absolutely. The Microsoft

1:21:34

IBM, I mean, it's crazy

1:21:36

to even say it now, the Microsoft

1:21:38

IBM PC partnership.

1:21:42

It's crazy. You have this absolute

1:21:45

behemoth partnering with someone

1:21:47

that's not really relevant.

1:21:49

And if you're standing here today, it

1:21:52

sounds like I'm talking about Microsoft partnering

1:21:55

with IBM, but at the time it was IBM

1:21:57

partnering with Microsoft. computer

1:22:00

company that mattered in the entire

1:22:02

world got themselves into a particular

1:22:04

situation where they came to Microsoft

1:22:06

looking for help. It's the

1:22:09

craziest set of events that made

1:22:11

this possible and I can't wait to dive into

1:22:13

it. Oh me too. But before we

1:22:15

do this is the perfect time to

1:22:17

talk about another one of our favorite enterprise

1:22:19

software platforms, ServiceNow, who is one of our

1:22:22

big partners here in season 14. Yes, ServiceNow

1:22:25

is the AI platform for your

1:22:27

enterprise, helping automate processes, improve service

1:22:29

delivery, and increase efficiency. Over 85%

1:22:31

of the Fortune 500 runs on

1:22:33

them and they have quickly joined

1:22:35

the Microsofts and the NVIDIAs as

1:22:37

one of the most important enterprise

1:22:39

technology vendors in the world. And

1:22:42

just like those companies, ServiceNow has

1:22:44

AI baked in everywhere in their

1:22:46

platform and they're also a major

1:22:48

partner of both Microsoft and NVIDIA.

1:22:50

I was at NVIDIA's GTC last

1:22:52

week which by the way

1:22:54

was insane. I felt like a VIP

1:22:57

at a Taylor Swift concert. It was

1:22:59

crazy. And Jensen brought up ServiceNow

1:23:01

and their partnership many times throughout the

1:23:03

keynote. So why is

1:23:05

ServiceNow so important both to NVIDIA

1:23:07

and to Microsoft? Well, AI in

1:23:09

the real world is only as

1:23:12

good as the bedrock platform infrastructure

1:23:14

it runs on and for enterprises

1:23:16

around the world, ServiceNow manages that

1:23:18

infrastructure. So all the AI applications

1:23:21

for enterprises, whether that's supercharging developers

1:23:23

with generative AI, empowering and streamlining

1:23:25

customer service, allowing HR to deliver

1:23:27

better employee experiences, ServiceNow has already

1:23:29

done the hard technical work to

1:23:32

make rolling them out not just

1:23:34

possible but easy and fast. Yeah,

1:23:37

employees can get answers quickly

1:23:39

with smarter self-service like changing

1:23:42

401k contributions directly through AI-powered

1:23:44

chat or developers can

1:23:46

build apps faster with AI-powered

1:23:48

code generation and service agents

1:23:50

can use AI to notify you that someone's

1:23:53

product needs replacing before they even chat with

1:23:55

you about it. With the ServiceNow platform your

1:23:57

business is a hundred percent ready for AI.

1:24:00

simply turn it on and it all works. They've

1:24:02

built AI directly into their platform. So

1:24:04

all the integration work to prepare for

1:24:06

it that otherwise would have taken you

1:24:08

years is already done. So

1:24:10

if you wanna learn more about the ServiceNow platform

1:24:13

and how it can turbo charge the time to

1:24:15

deploy AI for your business, go on over to

1:24:17

servicenow.com/acquired. And when you get in touch, just tell

1:24:19

them that Ben and David sent you. So

1:24:22

the IBM PC. Why

1:24:26

is IBM getting into

1:24:28

the personal computer here in 1980?

1:24:31

A fun quote we heard in our research

1:24:34

was that IBM was the sun, the moon,

1:24:36

the stars of the computing industry. And that

1:24:38

meant the hardcore enterprise

1:24:40

mainframe computing industry. Yes.

1:24:43

David, are you gonna attribute that quote? Are you

1:24:45

gonna leave listeners just hanging? All right, I think

1:24:48

it's time. We talked to probably 10 early

1:24:51

Microsoft people in research

1:24:53

and preparation. And one of

1:24:55

those folks was Steve Bomer himself. And

1:24:57

he used those words in describing IBM.

1:24:59

Yes, and it's hard to imagine a

1:25:02

better person to get their perspective on

1:25:04

what IBM meant to the world at

1:25:07

this point in time. Because 1980 was

1:25:10

also the year that Steve joined

1:25:13

Microsoft. So literally at the same

1:25:15

time in 1980, you've

1:25:18

got the management team coming together

1:25:20

with Steve and Bill and Paul

1:25:22

Allen. And you've got the IBM

1:25:24

thing going on. And you've got

1:25:27

them moving to Seattle. And we

1:25:29

haven't even talked about Charles Simone yet, but

1:25:31

this was the year he joined. This is

1:25:33

the year. Yes. I mean, every

1:25:35

year for Microsoft until this point is the year, but like 1980 is

1:25:37

big. Okay,

1:25:40

so why is IBM the sun, the

1:25:42

moon, and the stars of computing? Why

1:25:44

are they getting into this PC? It's

1:25:46

way cheaper than anything else they sell.

1:25:48

It seems to be like a totally

1:25:50

different business strategy, a different customer set.

1:25:52

What's going on? So all

1:25:55

the early micro computers we were just

1:25:57

talking about, you know, it's all ter-

1:26:00

Thappel II, the TRS-80,

1:26:03

these are all 8-bit machines. They're

1:26:06

running the Intel 8080 processor or

1:26:09

competitor making a similar

1:26:12

8-bit processor. The problem

1:26:14

with an 8-bit processor

1:26:17

is that the maximum data

1:26:20

size for a given instruction cycle

1:26:22

in the processor, this is called

1:26:24

a data word in computer

1:26:26

science terminology, is 256, 2

1:26:28

to the 8th. You can't represent any number greater than

1:26:33

256 in any given

1:26:35

CPU clock cycle in an 8-bit

1:26:38

machine. Right, it's effectively a bandwidth

1:26:40

limitation where if you're in a

1:26:42

single clock cycle trying to do

1:26:45

some particular instruction, it's a

1:26:47

very, very small amount of data that you

1:26:49

can move through the arithmetic logic unit or

1:26:51

that you can move through the processor in

1:26:53

that clock cycle. Totally, you can think of

1:26:55

it like an hourglass or something. There's like

1:26:57

all the data sitting there in memory at

1:26:59

the top of the hourglass and

1:27:02

then there's this small little funnel that it

1:27:04

goes through, that's the processor and then it

1:27:06

comes out. That's a good analogy. It's going

1:27:08

to take forever. Yeah, into the application of

1:27:11

software that the user sees. You're just not

1:27:13

really going to process it that fast. Very

1:27:15

primitive machines. Yes.

1:27:18

And so for a company like IBM,

1:27:20

they eclipsed the 8-bit computing

1:27:22

cycle a long, long, long

1:27:24

time ago. Mainframes, even mini

1:27:26

computers with DAC, all these

1:27:28

machines are at least 16-bit,

1:27:30

if not 32-bit computing

1:27:33

machines. So 8-bit is not

1:27:35

interesting. Which is why they only cost 375 bucks

1:27:37

or whatever for an Altair. Right. In late 1979,

1:27:39

Intel announces that they're coming

1:27:45

out with the 8086

1:27:48

processor, which is a 16-bit microprocessor.

1:27:50

With 16-bits, you

1:27:57

can really start to do some

1:27:59

damage here. in terms of the applications

1:28:01

that you could put on this thing to eat

1:28:03

into business software

1:28:05

use cases. In

1:28:07

16 bits, you can represent numbers up to

1:28:09

65,536. So

1:28:13

that's two to the 16th. You

1:28:15

can do interesting things passing 16

1:28:18

bits around at once. There's some really, really

1:28:20

fun aspects to this. If you

1:28:22

look at pictures of these processors,

1:28:25

what did the 8080, the 8-bit

1:28:27

processor look like? And

1:28:29

then you look at what did the 8086, the

1:28:31

16-bit processor look like? You can see this in

1:28:34

the 808 processor that's only 8 bits.

1:28:37

You see only nine pins coming off

1:28:39

of the little chip. There's

1:28:41

the eight pins for the data bits, and then I

1:28:43

think there's one more control pin. If

1:28:45

you look at the 8086 processor,

1:28:48

it's a much longer rectangle

1:28:50

with 16, 17, maybe 20 pins

1:28:54

coming off of it. You see this physically represented

1:28:56

in the chip. The further we get

1:28:58

in the computing world, the more abstract stuff

1:29:00

becomes. So it's always fun to go back

1:29:02

in history when these concepts were

1:29:04

so grounded in our physical reality that's

1:29:07

sort of easily observable since everything was

1:29:09

so much bigger too. Yeah.

1:29:12

Back to 1980. The 8086 has

1:29:15

been announced. 16-bit microprocessor

1:29:17

is coming. IBM

1:29:19

has already lived through

1:29:22

missing a computing expansion era

1:29:25

once with DEC and the mini

1:29:27

computing computer. They just

1:29:29

let DEC take that market. Of

1:29:31

course, that didn't really hurt IBM, but man,

1:29:34

it would have been nice to also have that

1:29:36

market too. And the thing that

1:29:38

they're observing about the microcomputer market is it's

1:29:41

exploding. People in our

1:29:43

industry know about DEC. People in the broader world

1:29:45

never knew about DEC. But

1:29:48

I think it's a very different

1:29:50

rate of adoption and rate of

1:29:52

demand with microcomputers where IBM

1:29:54

started to kind of look at

1:29:56

and go, oh, this might

1:29:59

be like a real market, like a really

1:30:01

big computer market for people.

1:30:04

They're finally observing the same thing that Bill

1:30:06

and Paul did, you know, all the way

1:30:08

back in the traffic data days of this

1:30:10

is an exponential cycle. Like

1:30:13

Moore's law, this is exponential, and

1:30:15

exponential gets real big very

1:30:17

quickly once you get a few years in. Yep.

1:30:20

And the mini computer cycle never was that. So

1:30:23

it's hard to remember today, but just to underscore

1:30:25

this again, in 1980, IBM was the most

1:30:28

valuable company in the entire world.

1:30:31

The highest market cap company, bigger

1:30:33

than all the oil companies in

1:30:35

1980. The

1:30:38

sun, the moon, and the stars. Yes.

1:30:41

Do you know what their market cap was? In

1:30:43

1980? $150

1:30:45

billion? You are

1:30:47

almost an order of magnitude off. $34

1:30:49

billion. That

1:30:52

was the most valuable company. It is wild

1:30:54

what a different world we live in today.

1:30:56

Yeah, even inflation adjusted. It's interesting that the

1:30:58

rate of growth of the most valuable companies

1:31:00

in the world in terms of

1:31:02

market cap has far outpaced inflation. So

1:31:05

that company, the most valuable in the world, they're

1:31:07

going to tiny little Microsoft that's just moved to

1:31:09

Seattle for this partnership. What's

1:31:11

going on here? So when

1:31:13

IBM decided, this is just

1:31:16

so amazing, like they deserve

1:31:18

so much credit here. They

1:31:21

got the Clay Christensen, disrupt

1:31:23

yourself, disruptive technology thing intuitively

1:31:25

decades before Clay writes any

1:31:28

of this stuff. The

1:31:30

way that they decide to compete is they're like, all

1:31:33

the things that Clay wrote about are

1:31:36

working against us here. What we need

1:31:38

to do is we need to create

1:31:40

essentially a skunkworks division, just like our

1:31:42

Lockheed episode. We need to do something

1:31:45

outside the company, completely removed from the

1:31:47

politics, sure, but like the business

1:31:49

incentives not to disrupt ourselves and

1:31:52

create a new division. They call it the

1:31:54

entry level systems division. Actually, it may have

1:31:56

existed before, but they repurpose it. This is

1:31:58

in Boca Raton. In Boca Raton. Florida. Very

1:32:00

nice place. We were just there a

1:32:02

couple months ago, but not a technology

1:32:04

hotbed in the world. And they

1:32:07

create a secret project called Project

1:32:09

Chess. Secret from the rest

1:32:11

of the company, the whole world. And

1:32:13

the goal is to develop the IBM

1:32:16

microcomputer or the personal

1:32:19

computer as people are starting to

1:32:21

refer to microcomputers. Yeah. And this

1:32:23

is wild. They're going to do

1:32:25

it in secret with a small

1:32:28

team with no other IBM resources.

1:32:30

And so that means this small

1:32:33

team, the only way they can do it is

1:32:35

to use all off the

1:32:37

shelf components from technology providers,

1:32:40

basically play on the same level

1:32:42

playing field as all the other

1:32:44

microcomputer manufacturers out there. And

1:32:47

oh, yeah, one more thing. IBM

1:32:49

leadership tells this team in Boca, they

1:32:52

have to ship the PC

1:32:54

to customers within one year.

1:32:56

It's a crazy constraint. Is

1:32:58

a total crazy constraint. A couple quotes

1:33:00

on this Don Estridge, who was one

1:33:02

of the leaders of Project Chess, he

1:33:04

would later say that the company realized

1:33:06

that if you're going to compete against

1:33:08

people who started in a garage, obviously

1:33:10

a reference to Apple here, you

1:33:13

have to start in a garage yourself. And

1:33:15

then Lou Gerstner, who later would take over IBM,

1:33:18

would describe this whole Boca

1:33:20

project as the way

1:33:22

you get an elephant to tap dance. And

1:33:25

the question is, are they

1:33:27

playing from behind and thus have to

1:33:29

adopt a flawed strategy? Or

1:33:31

is this strategy of assembling with all off

1:33:34

the shelf components actually a good strategy if

1:33:36

it works? Hmm. Well, let's

1:33:38

tell the story and then come back to it.

1:33:40

Great. Okay. So what do they do? The

1:33:43

hardware aspect of this is trivial,

1:33:46

basically. Hell, Ed Roberts could put

1:33:48

together the hardware to sell a

1:33:50

microcomputer and do this deal with

1:33:52

Intel. I think IBM can do

1:33:55

a deal with Intel. Not

1:33:57

necessarily trickier, but

1:33:59

the The more important part is

1:34:01

the software. And thanks to

1:34:03

Bill's genius strategy about be the volume

1:34:05

player, don't optimize on per unit price,

1:34:07

set the standards out there. They were

1:34:10

like the world's leading provider of programming

1:34:12

language interpreters, right? A hundred percent. There

1:34:14

is one game in town and one

1:34:16

game in town only, and that is

1:34:19

Microsoft in Bellevue, Washington at this

1:34:21

point in time. Now interpreters are

1:34:23

notably different than operating systems, but

1:34:26

Microsoft definitely had sort of raised the

1:34:28

flag and everyone could see if I

1:34:30

want to go buy software for my computers,

1:34:32

broadly, they're an interesting group to talk to.

1:34:34

Oh, yeah, obviously operating systems are going to

1:34:36

become really, really big here in just a

1:34:39

sec. But again, the

1:34:41

8-bit generation operating systems weren't that important because

1:34:43

people were writing their own software. The

1:34:46

standardized software packaged application software

1:34:48

doesn't happen until the 16-bit

1:34:51

era and the IBM PC.

1:34:53

So that's why the basic, the interpreter is

1:34:55

so important. So what happens? IBM

1:34:58

calls up Bill Gates and by August 1980, the two

1:35:00

companies are in serious talks to partner

1:35:06

and work together on the IBM PC. And

1:35:08

we referenced Steve Bomer a minute ago. The

1:35:11

timing is just crazy. Steve had just joined the company

1:35:13

in June, the first time he had

1:35:15

ever met a company. Which

1:35:18

is crazy. Steve had just joined the

1:35:20

company in June of 1980. He's

1:35:24

employee number 30. The Microsoft

1:35:26

team of 30, which the whole

1:35:28

company of Microsoft pivots to like

1:35:30

work on the IBM PC partnership

1:35:32

is bigger than the project

1:35:35

chess team in Boca. It's

1:35:38

amazing. Wow. So okay,

1:35:40

Bill has just convinced Steve to drop out of

1:35:42

business school at Stanford and come help

1:35:45

him and Paul run the company. So

1:35:47

Microsoft at this point in time is still a partnership.

1:35:50

Steve is the first person besides

1:35:53

Bill and Paul to get equity in

1:35:55

the company when he joins and

1:35:57

it's eight and a half percent and it's a handshake.

1:36:00

at this point. Yes, Bill really, really

1:36:02

wanted to bring Steve on, knew

1:36:04

him from the Harvard days, knew

1:36:06

what an asset he could be. He is

1:36:09

the end of Bill Yang. Yes. And so I

1:36:11

mean, frankly, eight and a half percent, it's a

1:36:13

big grant. Who's out there

1:36:15

running a 30 person company, and you're giving

1:36:17

away eight and a half percent slugs? It

1:36:20

just doesn't happen. Those are founder shares. And this

1:36:22

is really a sort of reflection that the way

1:36:24

that Bill thought about Steve was as a founder.

1:36:27

In fact, it created some tension with Paul Allen,

1:36:29

where Bill asked Paul if they could go to

1:36:31

5%. Paul said, Sure. And then Bill actually offered

1:36:33

him eight and a half percent. And Paul got

1:36:35

upset and Bill said, I'll eat the three and

1:36:37

a half percent, it can come out of my

1:36:39

share because I want him that bad. But

1:36:42

he's the perfect person for this point

1:36:44

in time. Bill was like the only

1:36:46

salesperson doing these OEM deals. Now

1:36:48

they're dealing with IBM, they're entering the enterprise

1:36:50

world, this needs to be a real business.

1:36:53

Yes. So back to the IBM

1:36:55

negotiations. Obviously, Microsoft is

1:36:58

very interested. IBM is not

1:37:00

just very interested in working with Microsoft, they

1:37:02

have to work with Microsoft. They're the only

1:37:04

game in town. Yes. And specifically, they asked

1:37:06

Microsoft for programming languages. They're like, we're making

1:37:09

this great PC, we're gonna need a basic,

1:37:11

we think you guys are working on a

1:37:13

cobalt, we'd like some cobalt. Yeah, Fortran, like

1:37:15

give us the whole thing. And

1:37:17

there's some debate on whether it was the

1:37:19

Microsoft side or the IBM side that really

1:37:21

saw the vision of, hey, the 16 bit

1:37:23

generation is going to enable real business

1:37:26

software use cases on the personal computer.

1:37:28

But it doesn't matter.

1:37:30

That's the plan here. And that is absolutely

1:37:32

what happens. So these initial

1:37:35

discussions, right, are for the

1:37:37

programming language, Microsoft doesn't make an operating

1:37:39

system at this point, because in the

1:37:41

eight bit generation, the operating system, I

1:37:44

think was kind of like a glorified bootloader

1:37:47

to just get into the programming environment so

1:37:49

that you could either write

1:37:51

or load up the basic programs that

1:37:53

Microsoft was going to interpret and then

1:37:55

run. And in the

1:37:57

eight bit generation, most if not

1:37:59

all. all hardware

1:38:01

providers of microcomputers

1:38:04

just wrote their own operating system. So yeah, this wasn't

1:38:06

a big deal. Now, there

1:38:08

was one off the

1:38:10

shelf operating system out there from

1:38:13

a company called Digital Research,

1:38:15

which was run by Gary Kildall, who

1:38:18

I think as we talked about earlier, Bill

1:38:20

and Paul had actually intersected with back in Seattle.

1:38:23

Yep, I think they were reasonably friendly. Yeah,

1:38:25

I think they were quite friendly because they

1:38:27

partnered. You know, you needed the operating system

1:38:29

to get into the programming environment. It wasn't

1:38:31

that big a deal. And so, yeah, whenever

1:38:34

anybody needed one off the shelf, Bill

1:38:36

and Microsoft would just refer people over

1:38:38

to digital to get it. Yep, CPM,