 Thank you. Hi everybody. It's great to be here. I want to talk to you about design now, and I'd like to do that with a bit of a history lesson. So for the past two decades, I have been practicing designer, building products and services, largely for the web and mobile and things like that, but for the last three years, I have been a partner at a VC firm in San Francisco called True Ventures. I still very much identify as a designer, and it can lead to some sort of awkward conversations when people ask me what I do. If I get into an Uber, typically they ask me things like, you know, so what do you do when I get into the car? And I say, well, I'm a designer. And they say, well, what kind of design do you do? And I say, I guess, you know, I've worked with startups, and they say, so what sort of things do you design for them? And I'm like, well, I'm more on the kind of the investing side of things now, and they're like, oh, like a VC? And I'm like, yeah. And they often say, great, I got to show you this app I'm working on. And I respond with this. So now when people ask me, so what do you do? I say, well, I'm in finance, and that's pretty much the end of the conversation. So my transition in my career from designer to investor, I think, is kind of reflective of how the technology industry is changing, that the skills we need to be successful in the products that we make, as well as the businesses that we create, are changing. And I want to do that with a sort of history lesson that goes from three separate eras of how digital technology has shifted over time, over the history of technology. Because the skills that we need are still very much centered on technical competence, but increasingly an understanding of humans and human behavior is more important than an understanding of technology. And that while kind of being the smartest is still going to be the way that we win, that emotional intelligence is going to become even more important than just raw intelligence for making the products. So like I said, a history lesson, let me start with where my career started back in the early 90s. I remember going to a party just after I got out of university and seeing this magazine on the table at the party. It was the first issue ever of Wired Magazine. And I was shocked that there were other people in the world that believed that technology was going to become incredibly culturally relevant and not just about all the gadgets. So I picked up the magazine. I put it cover to cover, very antisocial at this party, and I was like, this seems like the future. You have to understand that at the time in the early 90s, there were about 2,000 websites out there. So it was conceivable that you could have seen everyone. And if you wanted to see what new websites there were, you would go to the What's New page for the whole internet, and that every day about 12 new sites were launching and they would list them here and you'd get a sense of this. So it was very, very early on. I wanted to go work at Wired Magazine. So I sent them an email with a URL that pointed to my resume as HTML on the web, and as it happens back then, I was the first person to have ever done that. I got their attention and I got the job. I was fortunate then to work on the first commercial ad-supported website, which was HotWired, which was called HotWired and looked like this. Now, to give you some sort of historical context or technical constraint that we were under, if you were to take kind of a modern 5K iMac and translate this over to those pixels, it would look something like this. This made quite a splash in 1994 when we launched this because at the time in technology, design was not taken very seriously. It was considered something you do at the end with fonts and colors and layout and things like that. The popular sites at the time were Yahoo, which looked like this, Amazon, which didn't even have a picture of a book on it, and this was where technology was. This is because in the first era of technology, the human interaction was not considered at all. The first era of technology is something I like to call computation. It was strictly big machines that were designed to do math better than humans. In fact, the earliest machines were machines of war. They were used for fighting Nazis, so, unfortunately, I think history might be repeating itself a little bit, but they were designed to do things like crack enigma codes and plot out missile trajectories. Then they moved into business where they were used for things like inventory management and accounting systems. These were systems that were designed for expert users. These were often systems that were used by the people who designed them, and the interfaces that they used looked something like this. They were direct representations of the data structures that were inherent in the systems, inputs for the algorithms, and that was it. The users who would use them were trained to be able to use them. They were experts themselves. Very little was considered about the human interaction between them, let alone things like trust or delight or a sense of accomplishment. Even up to 10 years ago, people thought of using computer, using technology as work. You would go sit at a desk, you would do the task, and then you would get on with your life. That all changed at that moment when I was starting to work at HotWired and some of the first early websites came around. That heralded the second age of technology, that of connection. We linked all the machines together, and as it turns out, that's what popularized digital technology with the entire population because we were able to take the experiences that we have every day and share them with the people we care about. That completely exploded when we moved those computers off of the desktop and into our pockets and purses and fundamentally changed how we started thinking about the interactions we were having with those devices. Now that is because of the unbelievable scale of what we were doing. When I started in the industry, there was about 19 million people using the internet. Today, about half of the world's population, 3.8 billion people are using it every day. So that means that people are no longer experts that are trained in the systems that they're using. They are the full range of humans with all of their cultural understandings, their amazing human abilities and disabilities, and levels of expertise all over the place. Okay, poor user experience affects us even more, affects more people even more often. We would use our computers 10 years ago, two or three times a day to do some work and then go on with our lives. But today, roughly 80 times a day, the average person pulls out their phone, has an interaction with it. That interaction is either successful or not. And when they're not, it causes a set of emotional responses in people. Whereas when it was two or three times a day, people might say, oh, I'm not a computer person. I'm not really into technology. But now you need it for every single thing you do in your life. That causes people to feel angry, frustrated, or even ashamed. These are feelings people don't want to have, and therefore they reject digital products that make them feel that way. That has brought design and user experience research into the forefront of how it's a way to have successful products that make people feel empowered. When you make people feel powerful, it changes everything when it comes to designing for technology. We do this by designing for humans, by bringing people in and understanding how they succeed and fail with our products, or us going out into the world and understanding the problems that people have, what they're experiencing, and how we can provide solutions for them rather than the other way around. That's important today for the interactions we have, but even more important as technology goes into the third phase that I like to call comprehension. This is a video of the best human go player in the world losing to a machine in a way that he cannot imagine happening. He's utterly perplexed by the solution that the machine is used to defeat him. We are teaching our machines to learn, and we are doing that in a way that is inherently rife with the cultural biases that are around us in the world. When a couple of weeks ago, our iPhones started responding with this weird autocomplete, this alien message that we didn't understand, not only did we not understand it, but we didn't understand why it did that. It was looking at billions and billions of data points and coming up with the best recommendation based on the patterns that it saw. It turned into a sort of jokey kind of meme on social media, but it has significant implications when applied to systems that people use every day that help define sort of whether they think they should be using those products or not. Google Translate, for example, when it goes from Turkish to English, does this. Turkish has no gender for its pronouns. All the pronouns are just, oh, but English does have gender for pronouns. So Google looks at the billions of web pages that are out there, looks at the patterns that are there, and makes recommendations that cement the cultural biases that are rife in the language that we use today, as opposed to the world that we're trying to create in the future. This makes people feel left out. It's even more important when it's a personal rejection saying you are not allowed. So this passport application system which uses computer vision to help automate the submission of portraits that people put on their passports has clearly been trained only by a subset of the people who would possibly be using the system, and probably trained by a subset of people that look like the people who are designing the system. This happens in the lab all the time where researchers aren't out in the world understanding how people are reacting, but instead simply doing their research in isolation. As you can see by a group that was completely surprised when people rejected the idea that they should be using computer vision systems to determine a sexual orientation based on people's pictures on the internet. It's never been easier and cheaper to start a company. Many of you are probably in startups and realize that now, but it's never been more expensive to grow them. So why on earth would we design systems that fundamentally exclude portions of a potential audience as we get started? So how do we switch this around? There's a great quote in a Wall Street Journal from Ben Weizmann recently saying that all business is a bet on human behavior. So now as an investor and formerly as an entrepreneur, very interested in how to reduce the risk on that bet. We call that the product market fit, which is we take the things that we have built and hope that there is an overlap with the things that people need and that that overlap in the middle is a market that we can create that provides value. When we hear pitches all the time, there's traditionally a lot of engineering hubris that goes into it that people say, I can build this thing right, but very rarely do we hear people say, I can build the right thing. And that's what we're listening for now, is are you out in the world? Are you engaging with people and listening in a way that is respectful of what they need so that you can go back and create solutions rather than taking the solutions and pushing them out on people. Humans before machines. The reason that we are in this kind of mess that we are is that so much business is based today on the practices of companies that look like this 100 or 150 years ago. This is when human resources were applied to physical resources to produce products that would go out into the world. And the most important person in this picture is this guy right here. You probably can't see it where you are, but he's sitting there with a stopwatch and a clipboard. And he's measuring the productivity of his easily replaceable human components in the system that he's got, which leads, of course, to reactions like this poor guy who is like, my God, if I have to make one more of these things, just kill me now, kill me now. But you can imagine that the focus on unrelenting productivity of the people was the thing that tried to drive the value and success of these products. And that is no longer the case anymore. Productivity obviously is important, but creativity is where the next wave of value is going to come from. And the design systems, the design methodologies, and the human-centered research that we have done in the past is the way that we're going to get there. Google has a good example of trying to understand this. A few years ago, they commissioned a study of all the projects that they had ever done to see what were the qualities of the ones that succeeded versus the ones that failed. They called this Project Aristotle. Google is a computer science-driven organization, so obviously they took a quantitative approach to this. They looked at the grades of the people on the teams, the size of the teams, their IQ, the methods that they used. They looked at all of these things and found no patterns in what caused success. It wasn't until they took a more liberal arts approach and found that the one thing that all of the successful products or the projects had in common was a sense of psychological safety. This is one of the most computer science-driven technical companies in the world, but the only time products succeed were when the individual members of the team felt that they had a sense of confidence that the rest of the team would not embarrass, reject, or punish them when they spoke up. What that meant was the teams where there was safety led to much, much higher levels of creativity, meaning that the teams could evaluate many, many, many more opportunities for solutions and apply them to problems that people had in the world to be successful. There are many ways that you can do this. It is the language that we use in meetings, and it's a way that we conduct our product reviews and code reviews. It is the interactions between people that's modeled from the top down as inclusive and open as opposed to opinion-driven and aggressive. There are a couple more ways that we look for companies to succeed at this. I've been talking about this human-centered process. That means actually taking people, whether they use your products or not, and bringing them into your organization to watch them use the products and either succeed or fail. It is unbelievably powerful. It leads to better product instincts when everybody on the team, and I mean people in sales, engineers, literally everybody in the team, spends time watching users because then we get united with a sense of empathy for the people that we witness struggling to use the things that we are creating, and it changes your motivation for why you want to do it. It's incredibly powerful. And finally, the teams themselves, we need to increase the diversity of the teams that we have, and I mean on economic, gender, cultural, age, all the dimensions of diversity for the teams themselves. When we have people who are not like us on our own teams, it forces us into a form of communication that is more robust, more open. It also increases the amount of experiences that can go into making decisions. Now, I don't mean having users in your company and having diverse teams just as a matter of doing better design. I mean as a way of developing high, emotionally intelligent companies that are gonna succeed in this next era of computing. That means creativity through safety, human-centered design for decision-making, diverse teams for problem solving. That is the roadmap. That is the kind of thing that reduces risk in the world of investing. And I encourage you to be able to do that with your own teams, to practice this way. Because frankly, the future depends on it. And thank you very much for your time. Thank you.