Back to the Future? (w/ Raoul Pal & Bill Tai)
RAOUL PAL: When you're looking, surveying the landscape, what gets you excited now? We've gone through the VC way, but what on the horizon, or the things that you're investing in, that really starts getting you excited about the future? BILL TAI: I am still incredibly excited about the wave that I would generally call data science. Let me build up to why. Technology, to me, is a series of waves that build upon each other. I got started in that first wave of basically figuring out how to take these big vacuum tubes that got hot and took up a lot of space and burned out a lot to turn them into little dots on a piece of silicon. We had this wave of basically taking complex electronics and turning them into Lego blocks. That was a very big wave that produced companies like TSMC, and Intel and AMD, etc. Once those Lego blocks were available to everybody, you could build things out of them so there was a second wave of building hardware devices, whether they were computers or routers or switches or hubs and things that were the infrastructure that we all use today for moving electrons around. That wave, the second, wave led to the third wave, which was put all that together and then you have internet. You had basically the digitization of the communication system, another wonderful wave to invest in. Then on top of that-- RAOUL PAL: Your first idea of the internet when you first saw it was the digitalization of communication? BILL TAI: Basically, yes. Yeah. Because you had circuit switch networks that were point to point calls. Then once you could chop up the conversations, you could then multiplex them and put many on a line and then you could expand greatly the capacity and you could change the characteristics of what was expressed on the screens. That led to the fourth wave, which was the user interface that went on top of that. If you think about what is Facebook, Google, LinkedIn, anything that you touch on a web page, all it really is is a user interface that controls the ones and zeros that are expressed on the screen into something that's usable to you. That was very high value. Then the next wave, which we are still in and will be a long wave, just like the other ways were 30-year waves is the data science wave where you're trying to figure out what is being done to optimize that engagement. If you think about the winners in the user interface era, the ones I named before. They were the ones that understood how to figure out what was happening behind the front cover, that user interface, and optimize to win. That wave is really massive and just beginning. You see it expressed in everything from blockchain, to physical things like Uber now. If you think about what is Uber, what's Walmart, in the face of like a retail industry that was Sears and JC Penney for many decades, Walmart is a data science cloud where everything is represented by some bits and expressed at the point of sale terminals, so they know what's inside and what's moving around. It's given them like an asset light, low working capital business, everyone and the old guys are out of business. Uber is a data science cloud with cars on the end of it. Airbnb is a data science cloud with rooms on the end of it. June 1st , 2020 - www.realvision.com 4 The Interview: The Electrodollar: Venture Capitalism, Technology, and Silicon Valley We're basically moving today into a world of virtualization of physical assets and efficiency created on top of that cloud, which is what we're doing here. Think of the efficiency right here through Zoom. You and I have been virtualized, we're not 3D but our presence has been totally virtualized in a marketplace of nodes, where you can communicate with anyone anywhere anytime, and have that unit of productivity surface to the top which is why something like Zoom today in the face of no airline travel is worth more than the top seven airlines combined. RAOUL PAL: Yeah, the whole move towards video is one of the things that obviously, I'm in the middle of with Real Vision, but that whole thing, what everybody and every business is now becoming a media company because of the impact of video. The Zoom revolution is part of that whole thing, but video has basically taking over all forms of communication there. BILL TAI: It really has. I think while the capacity was still building in the very early days of internet, it was tough to just get a voice call across that. There would be a lot of dropouts, a lot of congestion and as the speeds increased, and the ability to multiplex increased, many more bits could be put through and you could move from pages to audio to video pretty cleanly. I think what that's done is for the media business in general, like technology has disrupted all these other prior businesses, media, it's in an interesting place today because in my book, the type of media business that we have grown up with for 30, 40, 50 years is an anomaly. If you think about content in general, from the time of the caveman, and let's use music as the main thing, songs were always free. You'd go to the campfire and you'd sing them and you could maybe write down some notes and sell a little sheet music, but it wasn't really a business. It wasn't until vinyl was invented and people could press that on something that had capital intensive characteristics to make that he who control capital control content. All of a sudden, you shifted to a world where the people that were capitalized control the content and could also control distribution. Once you had production and distribution, you could charge for it. We had one or two generations where people were willing to pay $17 for a $3 piece of vinyl and $14 for a 90 cent or nine cents CD, ultimately. We're now going back to what humanity was for 400,000 or 500,000 years, where content is free because production and distribution have been digitized, and it's in everyone's hands now. The business models are rapidly changing. You're seeing that just like it's empowering you to be a big player in this field or companies like Netflix disrupting the old industry. It's remaking industries. RAOUL PAL: How does that work with a subscription model? There is a free element. Obviously, we have a free element as many media companies do, but how does that work with subscription because that's basically saying you're going to have to pay for the curation or some of the work involved? BILL TAI: Yeah, it is, and you do a great job with that. It was always about drawing audience to get people to come to a destination in whatever business model existed and in the '70s, when in the US, it was CBS, NBC and PBS, there were limited choices but now, and we've already gone through that, the choices, the June 1st , 2020 - www.realvision.com 5 The Interview: The Electrodollar: Venture Capitalism, Technology, and Silicon Valley number of choices completely exploded so you really have to curate well and draw audience and build that flywheel of reputation and the ability to draw good content, just like any social circle. RAOUL PAL: Yeah. Where else, outside of the video side, where else are you seeing the developments of this data, and the application of it that you that you're talking about, this big wave? BILL TAI: Well, it's the entire stack, where we're probably 30%, 40% of the way through making available the infrastructure to anyone, and there's a lot of people that still have to learn how to use it, but at the very granular layer, you needed a different architecture to handle the data and there was an explosive wave in a techie nerdy area called Hadoop. Back, if you think about the way data and computation was handled in the '70s and '80s, you might buy a big machine from IBM and load a big instance of Oracle on it, and then get a bunch of disk drives and then store stuff in your company in some back room. When the internet started to explode, the type of data changed a lot, because it wasn't just the things that you'd run on your machine, you started to have things like tweets and JPEGs and user data of when somebody's on or what they clicked on on a page, it was a very different type of data that was not easy to store on Oracle, on an IBM machine, in disk drives. Google and Yahoo changed the game by consolidating that into data centers so instead of having a single big hunk in machines, you had many, many, many little CPUs on cards where a symphony conductor could conduct and distribute the computing. That changed everything about how data is-- how you compute, how you store, how you process, how you read and what information you can get out of it, because there's a lot more ambient data. Google created something called GFS, Google File System. Yahoo created Hadoop. Hadoop became open sourced in the '90s and became the basis for how the modern companies, the ones I mentioned, Facebook, LinkedIn, Twitter, they all use to do the user analysis and the A-B testing and the reading of ambient information to know how to steer their businesses and the ones that we're good at it, won. The ones that never tried were like the JC Penneys and Sears. That wave produced companies like-- Splunk is not specifically Hadoop, but Splunk Treasure Data was a company that I co-founded, where we made Hadoop available as a service on the cloud. Amazon has Redshift. That allowed any engineer sitting in a room anywhere to take all kinds of data stored somewhere through his internet connection on the browser, and then check what's happening. What do you do with it? How do you optimize engagement, so it spread the power of giant data centers, data analysis everywhere. That foundational layer is now available to any tech company, and the ones that really get it win, the ones that don't use it lose. Then now you can build all kinds of businesses on top of it. Whether it's the virtualization of things we talked about, cars, rooms, people. There's all kinds of things now that you can apply data science to, the reading of, the understanding of the movement, where it's headed to optimize so what I described with Walmart where you have a real time cloud, if there's somebody that buys something in Boston, and they're out of inventory, and there's extra inventory in California, they move it so the whole system becomes efficient. June 1st , 2020 - www.realvision.com 6 The Interview: The Electrodollar: Venture Capitalism, Technology, and Silicon Valley In the next decade, I think you're going to see that applied to supply chains. You're going to see a complete remaking because we're in this competitive world now and a remaking possibly of where we source our goods from a political national perspective, where everything is going to need to be identified, where it came from, where it is, where it is on the way. Whenever you get something delivered, you'd love to be able to click on that little thing to see where it is. That type of tracking is going to be pervasive in everything we do, whether it's goods, people services, and we're just really at the start. RAOUL PAL: That may be even forced by regulatory changes as well as either it's deglobalized world or regionalized, however we end up in this world because we're changing right now, we don't know where we're going to end up, but it sounds like people are going to be forced to prove providence, I guess. BILL TAI: Yes, I think so. Then this brings-- RAOUL PAL: Same with food as well. BILL TAI: Yes, definitely. Yeah, there's this movement of course. The sustainability wave is finally also real. Talk about another wave that is increasingly important, who knows whether we're all going to make the earth implode by our activities, but it's on more people's minds today than it ever was before. The safety and quality of your food and your water, everything about the brands you interact with, it matters now. I think this next generation of millennials that, I don't know where they going to get their money, maybe they'll just get it from their parents-- RAOUL PAL: Or basic income maybe. BILL TAI: Yes. Their purchase decisions are not just price performance. They don't buy things just because they're cheaper and a little bit better, they buy things because it means something to them. The brands represent something. If you don't represent something good or if you really represent something bad, it matters. I think the sheen of doing well by doing good matters for your survival going forward. I think that's also quite a big wave. I think back to the provenance point, this whole area of blockchain and cryptocurrency, it's controversial in some ways, because it hasn't become pervasive yet. I think, to me, the blockchain looks like TCP IP for assets. What does that mean? TCP IP is this little acronym you see everywhere, Transmission Control Protocol slash Internet Protocol, and it was and is the fabric around which digital assets, meaning like emails or packets that are sent through the internet are tracked, traced. When I send it into the-- I type an email and hit send to Raoul, the TCP IP stream carries it, transmits it, knows that it got there, and it's just a bunch of bits. I think now in the blockchain world, bits represent physical assets or digital assets of value like a Bitcoin and that fabric underlying that is this blockchain. Some are specific to certain kinds of things like the Bitcoin blockchain is only the Bitcoin blockchain, but there are other blockchains that can be used for other things, those can track those assets. Whether it's a securitization of a real estate project, a building, or it's the trading of a token for another token, or insurance, execution of insurance policies or whatever the contracts may be, those can now become over time, very June 1st , 2020 - www.realvision.com 7 The Interview: The Electrodollar: Venture Capitalism, Technology, and Silicon Valley fast, automatic and guaranteed in a way that the verification was very difficult to do before, very expensive to do. I think those businesses will be remade over time as well.