we wrote software for personal computers and we believe that we could hire the best engineers

there was a unbelievable amount of software to be written

and we could do it well we could do it on a global basis

and... the original customer base was the hardware manufacturers

and we sold to literally hundreds and hundreds you know... over a hundred companies in japan

and over a hundred companies doing word processors and industrial control type of things

we know in the long run we wanted to sell software directly to users

but we actually didn't get around that till nineteen eighty

when we had uh... our first sort of games and productivity software that

that people would go to a computer store in actually buy the the software package

we actually talked about it in an article

and i think nineteen seventy seven was the first time it appears in print

where we say a computer on it on every desk and every home

and actually the

we said running microsoft software

if we were just talking about the vision we'd leave

that those last three words out

uh... if we were

talking in an internal company discussion

we put those words in

and

it's very hard to recall

how crazy and wild that was you know on every desk and in every home

you know at the time you have

people who are very smart saying

you know why would somebody needed computer or even Ken Olsen

who would run this company digital equipment

who made the computer i grew up with

and you know that we admired

both him and his company immensely

was saying that

this seemed kinda a silly idea

that people would want to have a computer

when

IBM

saw that we had written software for all the personal computers

they came to us sought our advice on the design

but we said you should put the discant

and since they wanted to ship very quickly

another company

uh...

called digital research

had done that work

for the eight bit machines

and they were starting to do a version for this new these new sixteen machines

we commenced by the end of the sixteen bit machine

using this

eighty eighty six eighty d eight processor. Well Digital Research

really hadn't finished the work

and then IBM was getting frustrated because Digital Research

wouldn't sign even a non-disclosure agreement

and then some of us uh... particularly Paul

and uh...

key person named Kozhikode Nishi

uh...

was from japan worked with us

said no no no we should just do that ourselves

and because of a quick timing

we end up liscensing the original code from another company

uh...

and turned that into MS-DOS

and

so then

subsequently MS-DOS competed with

this Digital Research CP/M

uh... after about two or three years and MS-DOS

became far far more popular

uh... then

than CP/M and then eventually we would

take an add

graphics capability on top of MS-DOS

and then integrate the two together

and so today when we talk about Windows

it actually includes

all those MS-DOS things in it. that's the full operating system

although most of you think of the graphics in Windows and stuff there's a lot of

more classic operating system capability that that's built in there

the IBM initial deal is a flat fee deal uh... another flat the deal

it had certain restrictions

that prevented IBM

from selling to other hardware makers

so people did

IBM PC compatible machines

we would get the revenue by doing business directly with those people

and that the deal was very complicated but it was a deal that

Steve Balmer who's a key person of the company by that time

and i thought a lot about

and it was a fairly

junior team from IBM so we tried to make sure they're giving our belief that

personal computers would be hyper popular

that microsoft would get

a lot of that upside so

they felt they got a very good deal, which they did

as the industry expanded

we uh...

for new versions and for different machines, we got that opportunity even

though they did not pay us the royalty

even in the early days if you set a computer on every desk in every home and

you'd say okay how many homes are there on the world how many desk are there on the world

you know can i make twenty bucks for every home, twenty bucks for every desk

if you get these big numbers

but part of the beauty of the

whole thing was

we were very focused on the here and now

should we hire one more person

if our customers

didn't pay us

whould we have enough cash to meet the payroll

we really were very practical about

that next thing and so involved in

the deep engineering

that we didn't get ahead of ourselves we never thought

you know how big we'd be. i remember

when uh... one of the early lists of wealthy people came out

and

uh... one of the Intel founders was there

the guy who ran Wayne

computer actually is still

Wayne is still doing well and we thought hmm... boy, the software business does

well

in fact, microsoft could be

somewhere to that, but it wasn't real focus that

that everyday activity of

just doing great software

drew us in

and some decisions we made, like the quality of the people, the way we were very global

that vision of

uh...

uh... how we thought about software that was very long term

but other than those things you know we just came in to work every day

and

uh...

wrote more code

you know hired

hired more people

it wasn't really until the IBM PC

succeeded and perhaps even into Windows succeeded that

there was a broad awareness that microsoft

was very unique

as a software company that these other companies have been one product

companies

hired

people couldn't do a broad set of things, didn't renew their excellence, didn't do

research

uh... so

and we thought we were

doing something very unique, but it was easily

not until nineteen ninety five or even nineteen ninety-seven that

that there was this wide recognition that we

we were the company that had

had revolutionized software

when i was very young

hadn't been exposed to computers, so i was mostly just reading,

doing math, learning about science

and i wasn't sure what

my career would be

i knew i loved

learning about things, i was an avid reader

but it was when i was twelve years old that i

i first got to use a computer

actually a very

limited machine by today's standards uh... but that

definitely fascinated me when i was first exposed

i was intrigued

uh... by figuring out what it could do and what it couldn't do

and some friends and I spent

lots of time uh... the teachers got intimidated, so we were on our own

trying to figure it out actually we gave

course on computers

uh... to the other students

and it became

you know a fascination where

uh... we

got paid for doing computer work and

talked about forming a accompany

uh... but

there was kind of a magical breakthrough when the computer

became

uh... cheap

and

we could see that

everyone could afford a computer

uh... that was

much later

uh... but it

uh... that's what got us to

really get together and create company for software

yeah, math was the thing that uh... came most natural to me

and

you know you take these

exams some of which were sort of nationwide exams and

uh... i did quite well almost

and that gave me some confidence and i had some

teachers who were very encouraging

uh... they

let me read text books, they encourage me to take

uh... college course on

symbolic math which is actually called

algebra

uh...

so i i felt

pretty confident in my math skills, which is a nice thing because

uh... not only the sciences but economics a lot of things

if you're

comfortable

uh... with math and statistics and

ways of looking at cause-and-effect

uh... that's extremely helpful

computers were immensely

expensive

uh... and cost millions of dollars a machine that

was far less powerful then

then what you have a

a cell phone

today and so that

either you

have a very

important application

or you just share the machine with other people and still you had to pay quite a

bit of money

and so time-sharing is where you connected up and sharing the machine

it's a lot better then

sending your programs in and because you can see

when you make a mistake

uh... pretty quickly

even so because they charge is so much

we'd actually typed the programs

offline on a paper tape

uh... so that we didn't

have any delay for typing

and then when we got onto the computer we'd feed in that tape

uh... so that

there was less less time online

but it gave you a sense of look at what you got right and wrong and you could try

and correct things

uh... we also

because of that time the dominant form of computing was using punch cards

we actually did that quite a bit when we're down

at the university of washington and use some of those

punch card systems

as computers became less expensive so-called mini computers

then more people had access mostly scientists and business people

but also we

managed to find

machines that weren't being used at night. the idea of the machine is something

that an individual would use and that would just sit there idle when they

weren't using them

that only made sense about a decade later

when the work that we and others have done

had gotten the price down so dramatically

that the idea of a computer sitting idle, you know, doesn't feel like some huge waste of

resources

like

it did when they were

so uh...

expensive and rare

i went through several phases of doing more complex programs

where people who were great programmers would look at my work give me feedback on it

and

you get to your

you can be quite a good programmer

and it was kind of a such a

uh... intense activity

between the age of thirteen and seventeen

uh... that