Drew Houston - CEO and Founder of Dropbox발음듣기
대기중
Drew Houston - CEO and Founder of Dropbox
562문장
0%
태국어 번역 0명 참여
출처 : 칸아카데미
Drew Houston - CEO and Founder of Dropbox
So excited to have Drew Houston here.
Very well known figure amongst kind of our team out here.
And for those who are maybe watching this video later, founder of Dropbox.
How many billions of people do you have using?
I know it's more like hundreds of millions only.
300 million. Only 300 million people.
Using it, file sharing, we'll talk more about that.
But, you know, the way I view these things, it's definitely for our team to kind of meet and talk to really interesting people in Silicon Valley and beyond.
But also for kind of just the general Khan Academy user base out there to kind of get exposure to really interesting people in Silicon Valley and beyond.
And with that lens, I always kind of think of myself when I was 12 or 13 years old and I used to watch Johnny Carson.
And I always used to wonder who are these people?
How did they do that?
Like what would I have to major in in order to become that person?
And so I kind of want to have a little bit of that lens.
So I'll start there.
When you were a kid, what did you see yourself doing or becoming, and how did you try to set that up?
So my path started when I was really little.
So my parents had a PC junior.
So very first computer when I was really little.
And my dad - So first I would just play computer games on it but a couple years later my dad got - Showed me how there's this thing on the computer called Basic, and so I was lucky they showed me how to write some of my first lines of code when I was like really, really little.
Like how little?
I was five. That's good. Yeah I was very little.
My son has to get in line. (laughter)
And so as you can imagine, it turns out when you're little like that, you have a lot of free time on your hands.
So I was always like playing on the computer and I really loved playing games.
And so that was actually why I wanted to learn programming is to learn how to make games, how these things worked.
And like sort of my exploration of how the computer worked, I would figure out if there's the files and like when I save a game, like I can go into that file and like give myself 32,767 lives, you know?
So there would be all these little things like that.
If only I could access this file from home and kindergarten.
Moving this floppy drive around, yeah.
So I thought I was going to make computer games.
That was what I - That was what I was interested in in the beginning.
Then as far as a job, I started out babysitting. Really?
So I was very good at like watching.
I was 12, maybe 11 and like I would - The parents would leave, I'd put the kids to bed and then I was very good at watching HBO.
This was a requirement.
I could only babysit for families that had good TV and like Pringles and things like that.
And they didn't view that as a red flag that your focal point was their cable. Exactly.
And then like trying to like - And then sometimes I'd be watching something I probably shouldn't be watching.
There'd be like shooting or bad words and I was very good at listening for when the key went in the lock when the parents got home.
Anyway this has nothing to do with starting a company but then I got a little bit of an upgrade on the job front.
I was beta testing a game, an online game.
I was maybe 14.
And they were taking forever to build the game.
And so I started getting restless and curious and started poking around under the hood of this game, and I found all these security holes.
So I found all these vulnerabilities.
And at first you discover these things.
You're like oh my god, oh my god.
You know I started impersonating the developers and things like that.
And sort of making a little bit of trouble.
But then I'm like guys - I sent them an email and I'm like guys you should fix this, you should fix this, and you shouldn't have done this this way.
You should've done it this way.
And I'm like, you know, really getting into it.
And they're like, well do you want to work here?
And so I ended up working remotely as an engineer.
They're like do you want to fix those bugs?
And I'm like okay.
And then you know, it was a first kind of start up adventure.
First experience of many where my stock options were worth nothing.
Yes, this is 14 years old.
This is 14.
14 years old, that's good.
So it was part time remote.
But it was like a cool thing to talk about in like Biology class.
And you know I had to get my dad to sign all the employment paperwork and things like that.
But that was the path.
I was literally working on a game.
But then, fortunately, as you learn, there's a lot of other things in life beyond like Unreal Tournament which are really interesting. Beyond what?
Unreal Tournament or StarCraft.
Oh Unreal Tournament, yes.
That was a little later on.
PC era was more like the Sierra games like King's Quest 4 or something like that.
For me it was Pong.
I'm a little older.
We're losing him.
Sorry we have a little bit of overlap.
And then you go to MIT.
You know many of us have been there as well.
And I watched your commencement speech.
I watched your commencement speech which was great.
It helped that the guy before you really set a new low in terms of commencement speech quality.
But your speech was a really powerful speech.
And I think it kind of hit this issue of you didn't plan on being Dropbox founder, but you did think you were going to be an entrepreneur.
It seems like while you were in college, you were constantly thinking about starting a business of some kind.
Yes and your speech was great too.
I wasn't fishing for that.
It feels good though.
You know you're sort of - There's no instruction manual for these things.
I'm like what am I supposed - I start like Googling excessively for every commencement speech ever given before.
Anyway so no I did plan to be - I always was excited about doing something that was my own thing.
Whether it was making a computer game, or eventually as I learned more about business and starting a company, I was always it just turned out that I joined start ups as an engineer or like an intern when I was little.
And then just never really left.
And did you keep doing that while you were in college?
I mean you already had that kind of experience.
So were you starting stuff?
Even your freshman year or your sophomore year were you doing stuff?
So the game company unfortunately folded pretty quickly.
But that then gave me - That experience let me get a job at a local company that was another start up.
It was kind of MIT founders.
So it was kind of a relation there.
But yeah in the summers I would take classes and every summer I would go work at - First couple years it was the company that I worked at in high school.
So there were like maybe a half hour away, half hour outside of Boston.
And the one thing that I think is mysterious for everyone, and even sometimes when people ask me I don't have a clear answer because it's still a little bit mysterious is you know, how do you stumble on this thing that just kind of takes off like wildfire?
That just, and you have an interesting story.
You actually started a little bit closer to what we do here.
Kind of your first start up, which I think you started your junior year, was around SAT, it was called Accolade.
Right so my - You never know where these things are going to go because I mean I think anybody who grows up programming, whether it's programming or something else always tinkering with things.
There's always one project after another.
And then the first time I want to start a company, I took my junior year off from MIT.
So I took a year leave, which ended up being this really great experience.
And I promised my parents I would come back and I did but so the SAT at the time.
There was all this change happening, right?
They're merging in the writing section.
It's going from 1600 points to 2400.
Now I know there's all kinds of other stuff happening.
But this was back in the day where to study for the SAT, you would have to show up at some horrible classroom at like eight in the morning on a Saturday to listen to some 17 year old who didn't want to be there either, and was just reading out of a book.
This was the state of the art in SAT prep.
And we're like guys, it's not like the internet's been here for a little while.
It's been here for a long time.
So why are we still?
This is like how my parents study for the SAT.
And so I teamed up with actually one of my former teachers from my high school.
And we started building a new kind of course.
An online course for what was then the new SAT.
And what's your lesson?
What caused you to I guess not do that at some point.
I mean you worked on that for some time.
It was interesting.
So I started working on it because I'm like here's like an opportunity.
Here's something that my co-founder and I knew really well.
I mean I just - I don't know I treated the SAT, like one of the things I really like to do is reverse engineer like how things work.
Take things apart, so whether that's more typical stuff like the kind of tinkering that kids do or like, you know, one of my - Or finding all the security problems in that game.
When I started working on the SAT, or had to study for it, I was like how does this thing work?
And so you can, as you know, there are all these little tricks and little patterns you can take advantage of, and I'm like okay the really hard thing about the verbal section is the vocab.
Because in general, you can teach someone how to do a sentence completion without even really teaching them.
It's the words that are hard.
So when I was 16, I wrote all this software to do kind of flash cards and adapt and remember which ones I was getting right, and which ones I was getting wrong and things like that.
Because I didn't want - I was too lazy to like cut out all the paper flash cards.
And I'm like this is so stupid.
And so I done that, and that code was actually on the shelf for five years.
And then the SAT changed, and I had a friend of mine who was a teacher who had his own SAT course.
And we had this whole schtick going.
I was at MIT.
He had graduated from Harvard.
As a result of all this reverse engineering the SAT, and studying a lot for it, I got a 1600 so we had all like the boxes checked for the parents and we're like we know how this works.
So that, it was really just seeing an opportunity and being like, you know, you don't attach too much of an outcome around it.
Like I didn't have a goal like oh I need to make this much money or whatever.
It was just sort of like another interesting project and probably the best part of it was really just having an excuse to learn about business in like a rigorous way.
And when I mean a rigorous way I mean I didn't know anything about sales or marketing or finance or any of these subjects.
And they all seemed very like remote and mysterious.
And so to combat this, I would just go on Amazon and type in like sales.
And then like find the top three rated or most popular sales books and just read them, and repeat that.
Have you found that valuable?
It was super super useful.
Reading is probably one of the most important.
Reading about business is probably the most important thing that's prepared me for running Dropbox.
What was it?
And you talked a little bit about it in your commencement speech.
But I find this transition is kind of the most interesting one is, you know, what - You were working on this for two years I think.
Through your graduation.
What kind of - You know, it's hard, you know.
You're told to persevere and keep working on something.
Don't give up, but at some point maybe you are kind of putting good time after bad.
So it was super interesting in the beginning because like, you know, and when you first get to work, you know, you get down to such important orders of business of like alright I gotta Photoshop a logo and print out business cards that say founder on them and other very important things.
And then hand them out to people so that they know that I'm a founder now.
This is like the stuff you.
Oh I still do that. (laughter)
But then what you realize is after that kind of.
After you're incorporated and you get your facts back from the Delaware secretary of state.
And these things all seem very official and important.
You're like after, your day number 74 of like I gotta take out the trash.
Gotta erase the whiteboard.
It's like, you're like, actually this is just a lot of work.
And which was good but so I think with my first company, neither of us knew what we were doing at all.
And so we were bootstrapping it.
And was just a lot of kind of manual labor around like I was - We were too cheap to get like employees, so we weren't making any money.
So who's going to write the math questions?
It was like me and I just still remember having to like force myself to like write these math questions and draw these diagrams and be like alright where did the train leave from last time?
Okay this train leaves Sacramento at duh duh duh.
And so I'm like God, at some point I can't do this.
But we have people who do that so it's actually quite fun to do that type of thing.
Seriously I find that, I do.
And so but to, you know.
I feel your pain.
No (stuttering intensifies)
The problem wasn't really the work.
It was just more like we were putting a lot of effort into it and not seeing a lot of return.
So for me, it wasn't - It's kind of lost some of the variety of the work.
And so I just got kind of - It just became harder and harder.
I felt like I had to push myself like so much.
And you know, it just became something I dreaded.
It became something I had to try to trick myself into making progress.
And I start like - It's this like negative spiral, right?
Because you start resenting the work.
You start getting mad at yourself because you're like why am I not more disciplined?
And it's just like which makes you resent it even more.
And think that there's something wrong with yourself.
It's like all this stuff.
All this stuff in your head.
And then what happened was I had a friend of mine who started a company and did things a little bit more properly.
He had a co-founder who he was really good friends with.
They raised money.
They could just set up the company like a good normal company.
And they were just having a blast.
They working every waking hour.
But they were just like having fun every minute is what it looked like and felt like.
And so at first I'm like, maybe I'm really defective.
Look at these guys are having no problem with this.
But then what happened was I actually started - I would find these little side projects and I started working on a poker bot.
Again reverse engineering.
People said this was impossible.
I'm like it is possible.
And so I was really good at the security and kind of reverse engineering part of it.
And could do enough of the AI.
But you know, and it was like this, I was just possessed.
Like I would be like dragging my computers around everywhere.
Like my parents would ask, had me come up to New Hampshire because we have like a little place.
Like this place on a lake you're supposed to get away from technology.
Next thing I'm putting three monitors on the kitchen stove because there's not enough room.
Working on this thing.
And so but it turned out - You never know how these things are going to sort of all line up.
But I was still working on the SAT prep company and my friend Adam who I mentioned moved to San Francisco.
And next thing I know, he's two years younger than me, but next phone call I get from him he's like hey I've raised 5 million dollars.
And like, I was like that's a lot of money.
That number has two commas in it.
Like what the hell?
And this is my little brother in our fraternity.
And I'm like oh my god.
And so but it just turned out that you know in part it was sort of, the set up was there.
Like I really loved algorithms and distributed systems and I study computer science.
And like these are my favorite things.
So sort of fertile ground.
And then one time I lost, I left my thumb drive at home so I couldn't get any work done for my company.
And this sort of non PR, like the director's cut version of sort of the Dropbox origin story.
I was taking a ride on the Chinatown bus.
The Fung Wah from Boston to New York.
And I left my thumb drive behind and I'm just like I hate - You know, what do you do in these situation?
You're not like oh I hate my thumb drive.
I'm like oh I hate myself.
And I'm so, like, stupid.
I'm so disorganized.
I keep doing this, why can't I be better?
And just self flagulation.
And that's like there's nothing else to do for like four and a half hours.
But I had this three gigabyte Linux virtual machine image that I need to keep in sync across all these computers.
Because I worked on a laptop sometimes, sometimes a desktop, sometimes another computer.
And like nothing, there are a million things that claim to solve this problem and none of em actually did.
So it was really personal frustration that lead me to open up the editor and start writing some code that eventually turned into Dropbox.
Although I had no idea that would be the case at the time.
And you know, I think this is an important moment because a lot of times when your read about Dropbox, you'll say oh yeah he had the idea of sharing files across multiple devices and all of that, but whenever I realize like you know that - What's amazing is that the idea is actually a very old idea.
In fact almost anyone who's worked on data storage or anything has for the last 40 years has been trying to do this.
And so to me the power of, or almost the audacity.
This was like a big problem.
A lot of really smart people have kind of tried to do, as you mentioned, maybe even launched companies on it.
Didn't have really complete solutions.
What was kind of going on in your head that convinced you that no look I think the path that I'm about to take on this problem is going to be the one that works.
Well it was just so clear from trying all the other things that like there are 100 things that claim to solve this on paper, but in practice they never really worked.
And you know, at school, at MIT, they've sometimes you have like a campus network or a sort of hardcore Unixy kind of way of solving this problem but like for a Windows computer, or for like a MacBook, you know, you're stuck carrying around a thumb drive, and emailing yourself stuff.
And that was your point.
It was more just amplify the frustration.
I'm like guys this is not a new idea.
We've thought about this since like the '60s.
Like why the hell am I forced to do this?
It really made me upset.
And then there was some conviction.
Like look, you know, the problem with all the online drives is these very specific technical things where it's all a big hack where the way the online drives, the idrives, and xdrives, and for every letter there's been a drive.
For all of them, they all worked on the same basic principle which was like okay we're going to trick your applications and operating system into thinking that instead of like when a program writes something to the disk, instead of writing it to the disk inside the laptop, it actually intercepts it, sends it over to the internet to a server somewhere and gets it back.
So, at first glance you're like this is a great idea.
It will work with everything.
It's not that hard.
But the problem is if the server is 100 milliseconds away instead of five milliseconds away like your hard drive or less, then you get all these inexplicable problems because like millions and millions and millions of lines of code were all written with assumption that this thing is like 5, 10 milliseconds away.
And I'm going into detail on this example, but there's like no way you can ever make that work.
Like it's just like laws of physics.
And so you need a completely different approach and sort of a hybrid where like you have big and cheap and fast storage locally and on a server, but it's like two swimming pools connected by a straw, right?
And so first use the straw well.
And two like make it so that you're not waiting for the swimming pool to make it back and forth.
Use it efficiently, do that in the background.
Syncing, right?
And then it just turned out no one had gotten the technical part right.
But then there's also like a user, there's like a design piece of how do you design the interface?
How do you make this?
What's a metaphor that people can understand?
And then there's kind of, there's like an academic piece to it which is like the algorithms and how do you move files around and things like that and store them.
But then there's like a lot of grungy kind of operating systems work where you know, how do you make it so that you work - The fat 32 file system.
Every other file system when it stores next to the name of the file, it stores when the file was modified.
And it's a time stamp and it's a resolution of one second.
It's actually more detailed than that but I'm just listing - But then anybody just randomly on some random operating system, some random hard drive, you can have a fat 32 file system that stores that modification time with a resolution of two seconds.
And it's like none of the other ones do that but that one does and so it adds all this weird complexity to the code.
Or like Windows XP Service Pack 3, there'll be bugs where like the Swedish version of that, not the Norwegian version, the Swedish version would just cause Dropbox to crash.
And so, you know, it's two things.
One is a fundamentally different approach from what some of the other guys were doing.
And then two is a lot of just that kind of obsession.
Like the same way I was obsessed with making a very well crafted like, you know, SAT question about quadrilaterals and making the diagram look like the actual.
It was like then obsessing over making it really fast, reliable, and fixing that obscure bug with Swedish Windows.
And when did you realize?
You started on this, you see a problem, you start tackling it, and it is an audacious problem to tackle.
When did you say this is real?
This is like I have a solution here.
Like this is already better than what's out there in industry.
So first was just when I made something that worked for myself.
And I'm like I could actually use it.
Like I would save something and it would show up on my other computer, and I'd be like finally.
And then, but it really started to take shape once I made a video, a demo video of Dropbox.
Like a three minute little.
It's still online somewhere, or you can find it.
And I put it on Hacker News, which is this news site for start ups.
And it just got a ton of.
It was like top of Hacker News for like two days.
These were the old days so it was easier to do that.
But that gave me a ton of hope.
I mean I had already decided to quit my job.
So it was complicated.
So I just graduated.
I was working as an engineer at a start up.
And I was moonlighting on my SAT prep thing.
But that was the other thing.
You're asking why did I kind of get frustrated?
It's like we, my co-founder wasn't going to quit his job.
I wasn't really either.
This thing was never really gonna succeed, but it also wasn't going to die.
So like it just got kind of frustrating.
But anyway, so Dropbox, the first real turning point was when I put the video up, and then Arash, my co-founder saw it on Hacker News.
A week later we got introduced through a mutual friend.
I'm sorry I got a co-founder from it, Paul Graham.
There was Arash, my co-founder, and then our first investor, Paul Graham saw the video and emailed me saying because we had just, I had just applied solo to Y combinator with the idea.
So that's when it first...
That was your second interaction with Paul Graham.
I had a complicated series of interactions before. So Paul.
So Y combinator is kind of like applying to college.
Lots of people want to get in.
Not enough spots.
You have this very competitive dynamic.
And so actually the sort of admission system was how I thought about getting into Y combinator.
And so, I had, I had actually been flown out to California to hang out with some other Y combinator founders.
I also wanted to pitch Paul on Dropbox.
So I showed up at Y combinator a little early before one of the dinners, and walked into his office and asked if I could show him Dropbox.
And he was very angry about that.
Visitors were not welcome.
But the problem with that is like imagine, you know, we all remember applying to school.
Imagine having like five minutes with the Dean of Admissions.
And the one thing that they learn is that like you're an asshole.
Right? I don't know if I'm supposed to swear.
No it's cool, we keep it real here. Sorry.
For all the 12 year olds watching, don't say that.
But that first video was like huge in terms of getting started.
Yeah. And then you get get started and I mean.
What year was this?
This was 2007?
And then it's just been kind of, I mean, actually I don't think I've ever seen...
I mean you guys have got to be one of the fastest growing organizations maybe in the history of the Valley.
It's been growing quickly.
People are setting all kinds of records these days, but the company in terms of head count has grown really fast.
And then the user base has grown from 2009 we had a million users, and now we're over 300.
And I mean given to where - Like what are your kind of in hindsight or now that you can look over the last 7 years, I mean what have been...
And I almost view this, and we were talking a little bit about this before we did this is you know, what's kind of your...
What would you have told yourself if it was an 80 person organization or a 90 person organization to kind of ease your stress, let you know that something's coming, or ways to manage it as you grow?
I think there's a lot, you know?
And it's hard to kind of boil it down to like piffy phrases of how this is how you build a company.
It's like how do you play basketball?
It will take a while to explain, right?
But I think the...
I don't really know how to play, I'm not very good at basket ball so. I am.
But I'm not, I'm kidding, I'm horrible.
Dude, like alright.
So what I would say is some of the earliest, most helpful advice that we got was really to focus on the people that we brought into the company.
Like make sure everybody's really, really talented.
That's probably the best, like, hack you can do to make everything else easier, right?
Because put a good person between you and all your problems, right?
And our angel investors and some of our early folks kept drilling that into our heads.
Like make sure the talent bar is really high.
Be very choosey.
And then I think just sort of an attitude of really trying to systematically train to like understand.
Okay like here's what I'm dealing with today, but I'm going to talk to people who are maybe six months ahead, a year ahead, five years ahead, and like what do they do?
Like what would they...
I ask this quite...
I was talking to another CEO, 8,000 person company.
I'm like alright, we're approaching 1,000 people in Dropbox.
What am I going to see between 1,000 and 2,000?
What should I watch out for?
Like what would you put in place now so that as you've done all these things and grown the company, you know, what do you wish you could do today?
So a lot of what I do is think about that.
You have to be able to see around corners and one way to do that is reading.
One way to do that is having mentors and getting advice.
But and then the other thing that gives you some calm is that you know, what you read on Tech Crunch or in the press generally, you know.
When someone tells a story of a company, they're like oh yeah Dropbox.
It started out with not a lot of users and then later they had a lot of users and they did great.
You know, and then they have sort of this like, this narrative arc.
You're like well the real truth is this complicated stuff happened and this is really hard.
They're like yeah yeah no, two MIT kids in a dorm room have an idea.
And so the truth is it's actually very complicated and hard.
But that all kind of gets glossed over.
But when you read like profiles, there's a lot of books that have been written lately, The Everything Stores on Amazon, great book.
Steve Jobs' biography.
Google Apple, Microsoft, Facebook.
All these intel like Oracle.
I've read as many kind of profiles on tech companies and other companies as I could get my hands on of these companies over the last few decades.
And you know, what you learn is that there's nothing, once you sort of look behind the curtain, there's nothing that's that magical.
And so, and that's important for people to understand because you may think I don't know anything about business so I should just go like.
I'm never going to be good at it because I've never been good at it.
And that's just something that I should go get like a business guy or a person to do that stuff because I'll never be good at it.
But it's actually something you can learn.
So I think that mindset of like in a week you're probably limited in terms of what you can do.
But if you think about playing an instrument, or becoming a doctor.
These things all of us can realize yeah it's a lot of work but I can sort of see a path from A to B.
Turns out most things you want to learn have that kind of potential.
Whether it's public speaking or building a company or any of those things.
If you have the attitude that like...
And you give yourself time, you can actually pick these things up.
There's nothing that magical about them.
So and that's one thing about these profiles and the other thing, because you see in the early days they make all these really crazy stupid mistakes.
And like do all these things that are like really, really wrong.
And then that's also therapeutic.
Is like you realize all these companies were like a total mess.
And actually I see Albert in the corner.
Albert was an early Dropboxer.
Albert can tell you what a mess early Dropbox was.
So what you read about is actually not reflective of reality.
It's much messier and kind of less glamorous process.
Yeah and I'm about to go into questions from the team, but my last question is where do you see kind of Drop box in 5, 10, 50 years?
At the end of your life what do you hope Dropbox has done?
50 years. I don't know if anybody's made self-driving boats yet.
Yes that would be an innovation.
But you know, you can't really look that far out.
But we've realized and what we talk a lot about is we have this opportunity where now we have a lot of resources, a lot of really talented people.
We can solve these really big problems.
And what we're really working on is building this home for everybody's most important stuff, right?
Because you think about your house 20 years ago.
Like you'd walk into your house, and be like mail on the table.
You'd go in your living room.
Oh your photos are on your, or in your kitchen your photos are on your refrigerator. Right?
And you'd go down the hall and your briefcase has like your documents in it.
Like all that stuff is now on servers or in your Dropbox.
And so we need this like new kind of home.
So we think all the time about how can we save hundreds of millions of people from these things that are really painful and annoying about technology?
Like all these little paper cuts like forgetting your thumb...
Like the equivalent of these little thumb drive problems that are like everywhere.
And so how do we find the biggest ones and solve them at massive scale?
And then specifically, we started out with this app that you put on your computer or your phone, like it's just this Dropbox folder.
But really what you want is what you're using that folder for is often you're putting photos in there.
You know, you're getting work done.
You're collaborating with other people.
And so we're starting to dig in to some of those use cases a little more with things like we just launched a photo app called Carousel which is like if the gallery on your phone had like every picture you've ever taken.
And if you could text like 100 or 1,000 photos and videos at a time, that's what Carousel lets you do.
For some reason that had not really existed or no one...
Like everybody has this problem.
Just no one solved it.
Yeah, awesome.
So I'll go into some of these questions.
The first one it says this is from Marsha.
You can raise your hand Marsha.
Where's Marsha?
I don't know.
Where is Marsha?
Oh she's remotely.
So what's your typical workday like?
Probably the biggest thing I spend time on is recruiting, and so often there'll be some kind of fire drill at night where it's like okay we're acquiring a company, there's some candidate and we're fighting with Google, or some start up to try to get them.
There's a tug of war.
And like I'm often finding myself down here going big game hunting for people to join Dropbox.
You know so I'll be like in the Google campus.
If you see me at one of these companies, you know what I'm doing.
And so recruiting is the biggest thing.
And then probably the next biggest thing is just meetings in small numbers with people on my team around certain projects.
So every six months we pick a dozen things that are really important that we really want to get done.
And then I'll be spending time with them, being like okay here's new stuff we're going to do in collaboration.
Or here's, you know, answering questions like what is Dropbox?
What should it be in five years?
What will it be in 50 years?
What kind of place do we want to create?
And so it...
And then there's this much longer tale of all kinds of stuff like partners or customers or just other things.
Yeah. It's a hard question to answer succinctly.
No that's pretty good.
Sometimes you show up at the Khan Academy.
I made something that downloads my Google calendar.
I wrote a Python script that downloads my Google calendar into an Excel spreadsheet, like triages my time. Automatically?
Well it's supervised.
You have to triage the stuff or have someone help you triage the stuff.
But then I had these like rolling charts of how I spent the last week, or the last four weeks, or the last 12 weeks so that I can...
Because a lot of people say...
That's a problem, I'll use that.
That's interesting.
You get data analytic on your time.
It's super - Sorry I did that more...
Like one of the most important things you need to think about.
As your job gets more complicated, there's more stuff competing for your time, it's really important to understand where it goes.
There's a book called the Effect of Executive by Peter Drucker.
It's like any book on management will like focus on that.
So there's nothing.
No one's created a tool to see where your time goes.
And it's like everybody, you know, a lot of people have that problem.
So but anyways, I have been at times very rigorous about that.
Very cool. And so I have a question from Dylan.
Dylan's right over there.
So what are some of the most interested unexpected fun technical challenges that you and your team have faced in the process of scaling Dropbox to hundreds of millions of users.
So I think it's just fundamentally the challenge of Dropbox itself.
We're building, we're trying to build like the file system for the internet.
And so how do you connect a billion devices? Right?
And so that comes along with all kinds of technical challenges where people are saving a billion files every day on Dropbox.
And so, you know, that's more than there are tweets on Twitter.
And this is not little 140 characters.
This is like wedding photos.
Nothing, you can't have a bad day.
So even just the core service.
But then as you branch out, there's all these interesting things with you know, how do we make a...
There's far field things like computer vision.
For photos or something.
How do we organize your photos for you?
Search, you know, pretty much every kind of...
We have like client software, we have big data.
We have data science.
We have, you know, a bunch of infrastructure stuff.
Every permutation, every possible platform, if people are really into like hardcore algorithms, we need like that.
If people are really deep into systems, we have like unlimited need for that.
A bunch of storage stuff.
I mean it's really all over the map.
It feels like any permutation of technical challenge, we have it.
At least in software.
Yeah no absolutely.
And I think we have time here for one more.
So this is from Joel.
Where's Joel?
There's Joel right there.
So he'd like to hear about Hack Week.
And this is we've had a little bit of it.
We want to do more of this.
How did it get started?
And what new stuff has come out of it?
Has anything actually been productized?
So Hack Week is basically something we do every year.
You guys have probably heard of a Hackathon or a one day hacking event, but we're like let's take this to 11.
We'll do a whole week.
And because sometimes Dropbox, it would've been hard to prove that in a day, but actually the first prototype had something reasonably working in a week.
And so it's something we do really to sort of get back to the spirit of where we started where, you know, there are no constrains.
A lot of people were telling us that what we were doing...
A lot of investors told me that Dropbox, it's a stupid thing to get into, to have yet another storage company where all the other ones have failed.
But like it just doesn't matter.
Instead of like talking, or debating, actually write code and just like if you do it right, you can change how people think.
And so a ton of things have made it into the product, from having two accounts.
Having a personal and business account in Dropbox, that was a huge project that started out as a Hack Week project.
One of my Hack Week projects was the new the way you do notifications, or sort of like the menu bar in Dropbox is not like the native menu.
You can see what's changed and have a different UI around that.
Things like a bunch of security stuff.
So to factor authentication was something that started there.
A lot of how you can like rewind your files or undo things, that was a Hack Week project.
And then a lot of other stuff that maybe didn't make it one to one, or make it pound for pound into the product, but even just like when we watch people build all these apps on top of Dropbox, like that informs...
Like if someone made like a jukebox kind of music player as a separate app.
And it just sort of like plants a seed that may manifest itself years later in something like oh now instead of having the Dropbox app, you have a whole family of apps like Carousel, mailbox, Dropbox.
So it's really just a time for...a just to work on anything except what you're supposed to be working on.
B, unleash people and remember kind of...
Like lift the constraints of day to day.
And three it's to change how people think about...
You know, if you have an idea, prove it.
And what you learn is people have like a lot of really...
It's a pretty amazing group of people in the company and we have interns take Dropbox and put it in a weather balloon and put an Android phone into space where this phone would be taking pictures of the ride up and down.
And so it was just crazy.
You just have these like pictures uploading into Dropbox of like space.
Well I just want to thank you.
I think for, I really gotta speak for everyone here that you know it's been a real treat to kind of get to know you just now.
So thank you very much.
Great, thanks, awesome.
Thank you. (applause)
번역
0%
But, you know, the way I view these things, it's definitely for our team to kind of meet and talk to really interesting people in Silicon Valley and beyond.발음듣기
대기중
But also for kind of just the general Khan Academy user base out there to kind of get exposure to really interesting people in Silicon Valley and beyond.발음듣기
대기중
And with that lens, I always kind of think of myself when I was 12 or 13 years old and I used to watch Johnny Carson.발음듣기
대기중
When you were a kid, what did you see yourself doing or becoming, and how did you try to set that up?발음듣기
대기중
And my dad - So first I would just play computer games on it but a couple years later my dad got - Showed me how there's this thing on the computer called Basic, and so I was lucky they showed me how to write some of my first lines of code when I was like really, really little.발음듣기
대기중
And so as you can imagine, it turns out when you're little like that, you have a lot of free time on your hands.발음듣기
대기중
And so that was actually why I wanted to learn programming is to learn how to make games, how these things worked.발음듣기
대기중
And like sort of my exploration of how the computer worked, I would figure out if there's the files and like when I save a game, like I can go into that file and like give myself 32,767 lives, you know?발음듣기
대기중
I was 12, maybe 11 and like I would - The parents would leave, I'd put the kids to bed and then I was very good at watching HBO.발음듣기
대기중
And then like trying to like - And then sometimes I'd be watching something I probably shouldn't be watching.발음듣기
대기중
There'd be like shooting or bad words and I was very good at listening for when the key went in the lock when the parents got home.발음듣기
대기중
Anyway this has nothing to do with starting a company but then I got a little bit of an upgrade on the job front.발음듣기
대기중
And so I started getting restless and curious and started poking around under the hood of this game, and I found all these security holes.발음듣기
대기중
But then I'm like guys - I sent them an email and I'm like guys you should fix this, you should fix this, and you shouldn't have done this this way.발음듣기
대기중
But then, fortunately, as you learn, there's a lot of other things in life beyond like Unreal Tournament which are really interesting. Beyond what?발음듣기
대기중
It helped that the guy before you really set a new low in terms of commencement speech quality.발음듣기
대기중
And I think it kind of hit this issue of you didn't plan on being Dropbox founder, but you did think you were going to be an entrepreneur.발음듣기
대기중
It seems like while you were in college, you were constantly thinking about starting a business of some kind.발음듣기
대기중
I'm like what am I supposed - I start like Googling excessively for every commencement speech ever given before.발음듣기
대기중
Anyway so no I did plan to be - I always was excited about doing something that was my own thing.발음듣기
대기중
Whether it was making a computer game, or eventually as I learned more about business and starting a company, I was always it just turned out that I joined start ups as an engineer or like an intern when I was little.발음듣기
대기중
But that then gave me - That experience let me get a job at a local company that was another start up.발음듣기
대기중
But yeah in the summers I would take classes and every summer I would go work at - First couple years it was the company that I worked at in high school.발음듣기
대기중
And the one thing that I think is mysterious for everyone, and even sometimes when people ask me I don't have a clear answer because it's still a little bit mysterious is you know, how do you stumble on this thing that just kind of takes off like wildfire?발음듣기
대기중
Kind of your first start up, which I think you started your junior year, was around SAT, it was called Accolade.발음듣기
대기중
Right so my - You never know where these things are going to go because I mean I think anybody who grows up programming, whether it's programming or something else always tinkering with things.발음듣기
대기중
But this was back in the day where to study for the SAT, you would have to show up at some horrible classroom at like eight in the morning on a Saturday to listen to some 17 year old who didn't want to be there either, and was just reading out of a book.발음듣기
대기중
I mean I just - I don't know I treated the SAT, like one of the things I really like to do is reverse engineer like how things work.발음듣기
대기중
Take things apart, so whether that's more typical stuff like the kind of tinkering that kids do or like, you know, one of my - Or finding all the security problems in that game.발음듣기
대기중
When I started working on the SAT, or had to study for it, I was like how does this thing work?발음듣기
대기중
And so you can, as you know, there are all these little tricks and little patterns you can take advantage of, and I'm like okay the really hard thing about the verbal section is the vocab.발음듣기
대기중
Because in general, you can teach someone how to do a sentence completion without even really teaching them.발음듣기
대기중
So when I was 16, I wrote all this software to do kind of flash cards and adapt and remember which ones I was getting right, and which ones I was getting wrong and things like that.발음듣기
대기중
And then the SAT changed, and I had a friend of mine who was a teacher who had his own SAT course.발음듣기
대기중
As a result of all this reverse engineering the SAT, and studying a lot for it, I got a 1600 so we had all like the boxes checked for the parents and we're like we know how this works.발음듣기
대기중
So that, it was really just seeing an opportunity and being like, you know, you don't attach too much of an outcome around it.발음듣기
대기중
It was just sort of like another interesting project and probably the best part of it was really just having an excuse to learn about business in like a rigorous way.발음듣기
대기중
And when I mean a rigorous way I mean I didn't know anything about sales or marketing or finance or any of these subjects.발음듣기
대기중
And then like find the top three rated or most popular sales books and just read them, and repeat that.발음듣기
대기중
Reading about business is probably the most important thing that's prepared me for running Dropbox.발음듣기
대기중
But I find this transition is kind of the most interesting one is, you know, what - You were working on this for two years I think.발음듣기
대기중
So it was super interesting in the beginning because like, you know, and when you first get to work, you know, you get down to such important orders of business of like alright I gotta Photoshop a logo and print out business cards that say founder on them and other very important things.발음듣기
대기중
And which was good but so I think with my first company, neither of us knew what we were doing at all.발음듣기
대기중
And was just a lot of kind of manual labor around like I was - We were too cheap to get like employees, so we weren't making any money.발음듣기
대기중
It was like me and I just still remember having to like force myself to like write these math questions and draw these diagrams and be like alright where did the train leave from last time?발음듣기
대기중
It was just more like we were putting a lot of effort into it and not seeing a lot of return.발음듣기
대기중
And then what happened was I had a friend of mine who started a company and did things a little bit more properly.발음듣기
대기중
But then what happened was I actually started - I would find these little side projects and I started working on a poker bot.발음듣기
대기중
Like my parents would ask, had me come up to New Hampshire because we have like a little place.발음듣기
대기중
And so but it turned out - You never know how these things are going to sort of all line up.발음듣기
대기중
But I was still working on the SAT prep company and my friend Adam who I mentioned moved to San Francisco.발음듣기
대기중
And next thing I know, he's two years younger than me, but next phone call I get from him he's like hey I've raised 5 million dollars.발음듣기
대기중
And then one time I lost, I left my thumb drive at home so I couldn't get any work done for my company.발음듣기
대기중
And this sort of non PR, like the director's cut version of sort of the Dropbox origin story.발음듣기
대기중
And I left my thumb drive behind and I'm just like I hate - You know, what do you do in these situation?발음듣기
대기중
But I had this three gigabyte Linux virtual machine image that I need to keep in sync across all these computers.발음듣기
대기중
And like nothing, there are a million things that claim to solve this problem and none of em actually did.발음듣기
대기중
So it was really personal frustration that lead me to open up the editor and start writing some code that eventually turned into Dropbox.발음듣기
대기중
And you know, I think this is an important moment because a lot of times when your read about Dropbox, you'll say oh yeah he had the idea of sharing files across multiple devices and all of that, but whenever I realize like you know that - What's amazing is that the idea is actually a very old idea.발음듣기
대기중
In fact almost anyone who's worked on data storage or anything has for the last 40 years has been trying to do this.발음듣기
대기중
A lot of really smart people have kind of tried to do, as you mentioned, maybe even launched companies on it.발음듣기
대기중
What was kind of going on in your head that convinced you that no look I think the path that I'm about to take on this problem is going to be the one that works.발음듣기
대기중
Well it was just so clear from trying all the other things that like there are 100 things that claim to solve this on paper, but in practice they never really worked.발음듣기
대기중
And you know, at school, at MIT, they've sometimes you have like a campus network or a sort of hardcore Unixy kind of way of solving this problem but like for a Windows computer, or for like a MacBook, you know, you're stuck carrying around a thumb drive, and emailing yourself stuff.발음듣기
대기중
Like look, you know, the problem with all the online drives is these very specific technical things where it's all a big hack where the way the online drives, the idrives, and xdrives, and for every letter there's been a drive.발음듣기
대기중
For all of them, they all worked on the same basic principle which was like okay we're going to trick your applications and operating system into thinking that instead of like when a program writes something to the disk, instead of writing it to the disk inside the laptop, it actually intercepts it, sends it over to the internet to a server somewhere and gets it back.발음듣기
대기중
But the problem is if the server is 100 milliseconds away instead of five milliseconds away like your hard drive or less, then you get all these inexplicable problems because like millions and millions and millions of lines of code were all written with assumption that this thing is like 5, 10 milliseconds away.발음듣기
대기중
And I'm going into detail on this example, but there's like no way you can ever make that work.발음듣기
대기중
And so you need a completely different approach and sort of a hybrid where like you have big and cheap and fast storage locally and on a server, but it's like two swimming pools connected by a straw, right?발음듣기
대기중
And two like make it so that you're not waiting for the swimming pool to make it back and forth.발음듣기
대기중
But then there's also like a user, there's like a design piece of how do you design the interface?발음듣기
대기중
And then there's kind of, there's like an academic piece to it which is like the algorithms and how do you move files around and things like that and store them.발음듣기
대기중
But then there's like a lot of grungy kind of operating systems work where you know, how do you make it so that you work - The fat 32 file system.발음듣기
대기중
Every other file system when it stores next to the name of the file, it stores when the file was modified.발음듣기
대기중
It's actually more detailed than that but I'm just listing - But then anybody just randomly on some random operating system, some random hard drive, you can have a fat 32 file system that stores that modification time with a resolution of two seconds.발음듣기
대기중
And it's like none of the other ones do that but that one does and so it adds all this weird complexity to the code.발음듣기
대기중
Or like Windows XP Service Pack 3, there'll be bugs where like the Swedish version of that, not the Norwegian version, the Swedish version would just cause Dropbox to crash.발음듣기
대기중
Like the same way I was obsessed with making a very well crafted like, you know, SAT question about quadrilaterals and making the diagram look like the actual.발음듣기
대기중
It was like then obsessing over making it really fast, reliable, and fixing that obscure bug with Swedish Windows.발음듣기
대기중
You started on this, you see a problem, you start tackling it, and it is an audacious problem to tackle.발음듣기
대기중
Like I would save something and it would show up on my other computer, and I'd be like finally.발음듣기
대기중
But anyway, so Dropbox, the first real turning point was when I put the video up, and then Arash, my co-founder saw it on Hacker News.발음듣기
대기중
There was Arash, my co-founder, and then our first investor, Paul Graham saw the video and emailed me saying because we had just, I had just applied solo to Y combinator with the idea.발음듣기
대기중
And so actually the sort of admission system was how I thought about getting into Y combinator.발음듣기
대기중
And so, I had, I had actually been flown out to California to hang out with some other Y combinator founders.발음듣기
대기중
So I showed up at Y combinator a little early before one of the dinners, and walked into his office and asked if I could show him Dropbox.발음듣기
대기중
I mean you guys have got to be one of the fastest growing organizations maybe in the history of the Valley.발음듣기
대기중
People are setting all kinds of records these days, but the company in terms of head count has grown really fast.발음듣기
대기중
And I mean given to where - Like what are your kind of in hindsight or now that you can look over the last 7 years, I mean what have been...발음듣기
대기중
And I almost view this, and we were talking a little bit about this before we did this is you know, what's kind of your...발음듣기
대기중
What would you have told yourself if it was an 80 person organization or a 90 person organization to kind of ease your stress, let you know that something's coming, or ways to manage it as you grow?발음듣기
대기중
And it's hard to kind of boil it down to like piffy phrases of how this is how you build a company.발음듣기
대기중
So what I would say is some of the earliest, most helpful advice that we got was really to focus on the people that we brought into the company.발음듣기
대기중
And then I think just sort of an attitude of really trying to systematically train to like understand.발음듣기
대기중
Okay like here's what I'm dealing with today, but I'm going to talk to people who are maybe six months ahead, a year ahead, five years ahead, and like what do they do?발음듣기
대기중
Like what would you put in place now so that as you've done all these things and grown the company, you know, what do you wish you could do today?발음듣기
대기중
But and then the other thing that gives you some calm is that you know, what you read on Tech Crunch or in the press generally, you know.발음듣기
대기중
It started out with not a lot of users and then later they had a lot of users and they did great.발음듣기
대기중
But when you read like profiles, there's a lot of books that have been written lately, The Everything Stores on Amazon, great book.발음듣기
대기중
I've read as many kind of profiles on tech companies and other companies as I could get my hands on of these companies over the last few decades.발음듣기
대기중
And you know, what you learn is that there's nothing, once you sort of look behind the curtain, there's nothing that's that magical.발음듣기
대기중
And so, and that's important for people to understand because you may think I don't know anything about business so I should just go like.발음듣기
대기중
And that's just something that I should go get like a business guy or a person to do that stuff because I'll never be good at it.발음듣기
대기중
So I think that mindset of like in a week you're probably limited in terms of what you can do.발음듣기
대기중
These things all of us can realize yeah it's a lot of work but I can sort of see a path from A to B.발음듣기
대기중
So and that's one thing about these profiles and the other thing, because you see in the early days they make all these really crazy stupid mistakes.발음듣기
대기중
Yeah and I'm about to go into questions from the team, but my last question is where do you see kind of Drop box in 5, 10, 50 years?발음듣기
대기중
But we've realized and what we talk a lot about is we have this opportunity where now we have a lot of resources, a lot of really talented people.발음듣기
대기중
And what we're really working on is building this home for everybody's most important stuff, right?발음듣기
대기중
So we think all the time about how can we save hundreds of millions of people from these things that are really painful and annoying about technology?발음듣기
대기중
And then specifically, we started out with this app that you put on your computer or your phone, like it's just this Dropbox folder.발음듣기
대기중
But really what you want is what you're using that folder for is often you're putting photos in there.발음듣기
대기중
And so we're starting to dig in to some of those use cases a little more with things like we just launched a photo app called Carousel which is like if the gallery on your phone had like every picture you've ever taken.발음듣기
대기중
And if you could text like 100 or 1,000 photos and videos at a time, that's what Carousel lets you do.발음듣기
대기중
Probably the biggest thing I spend time on is recruiting, and so often there'll be some kind of fire drill at night where it's like okay we're acquiring a company, there's some candidate and we're fighting with Google, or some start up to try to get them.발음듣기
대기중
And like I'm often finding myself down here going big game hunting for people to join Dropbox.발음듣기
대기중
And then probably the next biggest thing is just meetings in small numbers with people on my team around certain projects.발음듣기
대기중
So every six months we pick a dozen things that are really important that we really want to get done.발음듣기
대기중
And then I'll be spending time with them, being like okay here's new stuff we're going to do in collaboration.발음듣기
대기중
And then there's this much longer tale of all kinds of stuff like partners or customers or just other things.발음듣기
대기중
I wrote a Python script that downloads my Google calendar into an Excel spreadsheet, like triages my time. Automatically?발음듣기
대기중
But then I had these like rolling charts of how I spent the last week, or the last four weeks, or the last 12 weeks so that I can...발음듣기
대기중
As your job gets more complicated, there's more stuff competing for your time, it's really important to understand where it goes.발음듣기
대기중
So what are some of the most interested unexpected fun technical challenges that you and your team have faced in the process of scaling Dropbox to hundreds of millions of users.발음듣기
대기중
And so that comes along with all kinds of technical challenges where people are saving a billion files every day on Dropbox.발음듣기
대기중
But then as you branch out, there's all these interesting things with you know, how do we make a...발음듣기
대기중
Every permutation, every possible platform, if people are really into like hardcore algorithms, we need like that.발음듣기
대기중
You guys have probably heard of a Hackathon or a one day hacking event, but we're like let's take this to 11.발음듣기
대기중
And because sometimes Dropbox, it would've been hard to prove that in a day, but actually the first prototype had something reasonably working in a week.발음듣기
대기중
And so it's something we do really to sort of get back to the spirit of where we started where, you know, there are no constrains.발음듣기
대기중
A lot of investors told me that Dropbox, it's a stupid thing to get into, to have yet another storage company where all the other ones have failed.발음듣기
대기중
Instead of like talking, or debating, actually write code and just like if you do it right, you can change how people think.발음듣기
대기중
Having a personal and business account in Dropbox, that was a huge project that started out as a Hack Week project.발음듣기
대기중
One of my Hack Week projects was the new the way you do notifications, or sort of like the menu bar in Dropbox is not like the native menu.발음듣기
대기중
And then a lot of other stuff that maybe didn't make it one to one, or make it pound for pound into the product, but even just like when we watch people build all these apps on top of Dropbox, like that informs...발음듣기
대기중
And it just sort of like plants a seed that may manifest itself years later in something like oh now instead of having the Dropbox app, you have a whole family of apps like Carousel, mailbox, Dropbox.발음듣기
대기중
So it's really just a time for...a just to work on anything except what you're supposed to be working on.발음듣기
대기중
It's a pretty amazing group of people in the company and we have interns take Dropbox and put it in a weather balloon and put an Android phone into space where this phone would be taking pictures of the ride up and down.발음듣기
대기중
I think for, I really gotta speak for everyone here that you know it's been a real treat to kind of get to know you just now.발음듣기
대기중
칸아카데미 더보기더 보기
-
Great Altar of Zeus and Athena at Pergamon
77문장 0%번역 좋아요3
번역하기 -
The vowel-shift irregular verb
29문장 0%번역 좋아요1
번역하기 -
The Gelatin Silver Process - Photographic Pro...
59문장 0%번역 좋아요6
번역하기 -
Donatello, Mary Magdalene, c. 1455
43문장 0%번역 좋아요2
번역하기
