Drew Houston: prediction

Transcript context

…Yeah, right. But then I would talk to people who were actually working on those things, and they were on the same treadmill. I interviewed this director of engineering at SpaceX. I was like, "Oh my God, you're actually going to Mars. This is so cool. How do you guys work together? How are you going to get to Mars?" He's like, "I don't really understand the question." I'm like, "What tools do you use? How do you work together? How are we going to get to Mars?" The answer was basically, "We're going to get to Mars through a lot of emails, and a lot of files." I was like, "Oh, my God." Most of the time, we correctly think about technology as just like force multiplier, like the amplifier of our abilities. But from another perspective, we're only as good as organizations as our tools, or tools can become the limiting factor. It's really a problem when the mechanism through which we were getting work done has gone from being a force multiplier, and amplifier to a suppressor. We are the frogs in the boiling water where this kind happen one day at a time. All the people working on those things, and there are many others, they're all dealing with the same issues of, "I can't focus my tools. I'm fighting with my tools." Somehow the tools went from helping us get the work done to becoming the work, as we layer. It went from 5 of them, to 10 of them, to 500 of them. Yeah. I'm like, "All right. If Einstein were alive today, what would his day be like?" He would wake up. He'd have to delete a bunch of LinkedIn notifications. He'd get down to work, start writing equations, and then someone would Slack him, and interrupt him. Then he'd get back to work. Would we still understand relativity if Einstein were living in that kind of working environment? Probably not. Even just visiting my dad at work when I was a kid, a lot of things were the same. He had a phone, a PC and an office, but he could turn his phone off. He got 5 emails a day, not 500. He would literally come home from work with a briefcase, put it down, and stop thinking about work. Then when he would be in the office, he could close the door, and actually get stuff done. I'm not saying we should go back to the early 90's, but I'm saying it wasn't always like this. The whole frontier of productivity at the time felt like it was just taking us in the wrong direction. Things like Slack, which are wonderful tools, also chop up your day into little fragments, and make you cognitively diabetic. There's a lot of empty carbs in our collaboration, and our collaboration tools. I'm like, "Okay, here is a problem that's not solving itself. Actually, none of our competitors even are framing the problem correctly. What we should be recognizing is that our ultimate non-renewable resource is our time, our attention, mostly our brain power, our creative energy. r competitors even are framing the problem correctly. What we should be recognizing is that our ultimate non-renewable resource is our time, our attention, mostly our brain power, our creative energy. I think as a civilization, we're going to, hopefully in 5 or 10 years be like, "Man, we took this crazy detour where we basically hooked up that brain power to this thing that was just burning off half of it as friction with our tools. " That was dumb. Yet no company was even talking about this, let alone fixing it. Then we came up with a new Dropbox mission. It became Dropbox is designing a more enlightened way of working, because the way we were working is unenlightened, unexamined. Meanwhile, we were patient zero as the company. No one was doing a lot of thinking. Everybody's kind of busy. Everybody's bumping into each other. There were a lot of cultural problems. Like that we were faster. Your company grows, and rises, the easier it is for new people in the company to think they hit a home run instead of starting on third base. A lot of complacency, entitlement, and preoccupation with things other than our customers or other than our products set in. It wasn't all bad. I think there were a lot of good things we had to pay attention to. We were scaling up the business financially. We were getting ready to go public. We had to mature a lot of things, but we lost sight of the place that money is coming from is from our customers, and the thing they really want is a great experience. We've been tackling this in a lot of different ways. First is just things like after COVID hit, suddenly the world is working in a completely different way. We saw this as a big opportunity, not just a crisis. Or once you got past the crisis part of COVID, there was a big opportunity. COVID kind of wrecked everything, but we didn't have to put the floorboards back down in the same places. I'm a big fan of people like Peter Drucker, and lots of people who think about the nature of work, and this idea of we're finally decoupling work from our physical location. It's a big deal. It's probably one of the biggest changes to knowledge work in our lifetimes. Or even since Drucker coined that term knowledge work in 1959. It's like, "Cool. For the first time, we can actually design how we work in ways that our parents couldn't, or the way that most of us just receive how we work from our parents." But like, "Hey, we can actually do things completely differently." We came up with this whole virtual first model where we're 90% remote, but then most importantly, we're trying to rebuild the new product stack for distributed work. I'll talk a little bit about where the new things we're doing, and then I'll come back to how we're rebooting some of the core. But initially, after COVID, we studied a lot of how other remote first companies had worked. We synthesized with a lot of primary research of other companies like GitLab, Automattic, and others who had been doing this for a long time. We collated all that, synthesized a common playbook, open sourced it. If anybody's curious about this, you could find Dropbox. We've open sourced our virtual first toolkit. doing this for a long time. We collated all that, synthesized a common playbook, open sourced it. If anybody's curious about this, you could find Dropbox. We've open sourced our virtual first toolkit. Then since COVID, designing a working model very in line with our new mission with a lot of benefits. It really works. The employee retention, satisfaction, engagement or offer separates all the things we care about have been significantly up since COVID. There's many factors to that, but there's a lot that's working about the model. But really, we're approaching it from a product standpoint. In deciding to turn Dropbox into this lab for distributed work, we're like, "Okay, when you open up your laptop in 2025, 2030, what do you see?" Hopefully, there's some new stuff or some things that are different on there. What are those things, and how do we bring them back? One of the first things you lose when you go distributed is context. We get a lot. You get a lot for free from osmosis. Just, "Oh, hey, how are you doing? How's this?" Or, "Oh, do you have that thing?" But then when you're remote, then a lot of that gets replaced with endless video meetings, and lots of Slack messages, which is super inefficient. Also, one thing that was clear across the board from all the successful remote first companies was you have to be more documented, but then you have a lot more documents. This started turning up very obvious problems, which is like, "Wait, why do I have 1 search box at home, and 10 search boxes at work? Why is it easier to search all of human knowledge than my company's knowledge? Why is this problem getting worse not better every year?" If you think about it, we've got 10 search boxes that you search 10% of our stuff. I'm like, "We should just fix this." Again, even just as a frustrated user, maybe 2018, 2019, I'm having this problem myself. I'm like, "It's really hard to find stuff." Plus there's always cool vector search stuff, and deep learning stuff happening. Even for ChatGPT, I'm going to make a personal search engine that takes us, so I shouldn't have to, right? If I search for strategy, and the title of the document is plan, it should get it right, and the matching should be fuzzy. Then I built something like this. I'm like, "Oh my God, it completely works."…

Stored transcript either side of the excerpt. The highlighted words are the published quote; the surrounding text is unedited source, never generated.