Martine Paris is a freelance tech reporter who covers AI, consumer tech, crypto and blockchain for Fast Company, VentureBeat, CoinDesk, The Block, Modern Consensus, Pocket Gamer and other media outlets. She is a frequent speaker at leading tech conferences and serves on the CES Digital Money Forum programming committee. Follow her on Twitter @contentnow.
Picture a world where there’s a composite “mini-me” of you running around the web, interacting with your favorite sites, logging activity into your digital brain, and compiling a comprehensive personal data store of all of your interests, likes and dislikes, whether you’re naughty or nice, when your Uber driver dinged your rating because you asked him to turn right, which Instagram stories you’re watching, what you ordered from Postmates last night, etc. Every thought, feeling, preference you ever had forever inscribed on the immutable ledger of life.
In this utopia, you, of course, get the keys to you, with the ability to keep all of your data private or monetize it across an ecosystem of digital delights. Here there are intelligent agents that you can permission to use your data to act on your behalf across the web. Perhaps to find you clothes that match your selfies, in your exact size, shipped to your exact location in time for that event you need to attend. Would you do it? Would you trust a bot you hardly know to represent you online and not do something nefarious with your most intimate data, like copy it and sell it to the highest bidder?
Sounds a bit Orwellian but it’s the actual investment thesis of Internet pioneer, Brad Burnham, and his partners at Placeholder VC. Burnham founded Union Square Ventures with Fred Wilson in 2003 after the dot-com crash. The two became legend when their first fund returned 14x and became one of the best performing funds of all time.
Known for their Midas touch, they were early investors in Zynga, Tumblr, Twitter, Etsy, DuckDuckGo, Coinbase, and today are a founding member of the Libra Association. With their feet firmly planted in the future, it was no surprise when they hired Joel Monegro in 2013 for their blockchain practice. In 2017, when Monegro left USV to form Placeholder VC with Chris Burniske, Burnham followed and helped them raise their first $150 million fund.
When the three of them spoke at SF Blockchain Week last week, Burnham shared his “aha” moment when he realized that data could live outside its application.
“When we first met Tumblr founder David Karp, he had four employees and 20 million users and it was a hard for me to wrap my head around it because my experience had been that it took a lot more resources to build a network of that size. When we started to dissect it, it became clear that the difference was that they were able to create a global communications network without paying for any communications because individually, we were funding access to this common infrastructure so they could build this application on top of it. As that network started accumulating a lot of data, their S3 bill went right through the roof. All of a sudden they had to raise a ton of money just to support the infrastructure they were operating in to continue to provide the service. In doing that they eventually had to figure out what the business model was and ultimately came to the model we’re all so familiar with which is essentially surveillance capitalism: turning users into product to sell advertising. Similar to how Tumblr separated communication from the application, we’re now separating data from the application.”
He was asked what it was like to launch USV during the aftermath of the dot-com crash. Burnham responded, “Real investment opportunities happen when things are out of favor not when they’re in favor. When Fred and I went out in 2003 to raise our fund, it took us 80 meetings. People fell asleep on us, people walked out on us, people were incredibly rude. The challenge was that we were advocating an investment thesis that said we were going to invest in the applications layer on the web where people had just lost money. After we raised the fund and went into the market to start investing, most VCs were retreating to chips, routers and hard technology, we had an open field.”
On advice he’d give to crypto founders during this crypto winter, Burnham commented, “I'm a believer in decentralized emergent bottom up startup innovation and I think we’re in an era of consolidation of the data monopolies on the web. If an open public communications medium was the key that unlocked innovation in the media industry when the internet came out, I think open public data stores will unlock another Cambiran explosion of innovation. If you can separate the data from the application, you can unlock the market power associated with the exclusive control over that data set. I think it's an enormous, profound, transforming opportunity. “
Monegro explained how the data monopolies don’t have visibility into the full data set of users because they’re proprietary. For example, Google has your search data but not your Amazon purchase history. Because they don’t share, they’re unable to capture the full value of information. Seeing an opportunity, Burnham discussed his desire for an insight generation engine that could bring him all the information he desired instead of having to search for it.
“My dream is to capture all of the data I create as I interact with services before it gets embedded in a server somewhere. If I had a comprehensive user data store that compiled information from all of my interactions across the web and I permissioned that out to service providers, I can imagine a world emerging where there were AI engines that did very highly specialized things, working against that data set and providing services to me, working on my behalf. One of those services would be a news reader that had the benefit of knowing everything I already consumed and could do natural language processing and sentiment analysis to figure out that this is the one article I haven’t read yet that I should pay attention to. I think there are a set of personalization services coming our way in a world where we have control of our data set that we haven’t even begun to think about. A product feed is another one, it completely changes our concept of advertising if I have a permissioned data set that is producing for me a set of recommendations that I’m interacting with on my terms.”
He talked about re-imagining a property right in data to help the market work correctly and added, “I think the most interesting thing we're going to see over the next five years is the convergence of functionality within protocols. I'd like to think we end up with a set of protocols that basically models human behavior in a way that is efficient and captures a lot of value, and then a large number of much smaller applications competing on a level playing field. That would make for a much more exciting, innovative world.”
I then had a chance to talk with Burnham about his relationship with Placeholder and how they were planning to gain trust in the data agents to drive mass adoption. What follows is an edited transcript of our interview:
Paris: What is the relationship between you and USV and Placeholder?
Burnham: I have the incredibly good fortune at this time in my career of being able to work on both platforms as a venture partner at Union Square Ventures and a venture partner at Placeholder.
Paris: What’s the investment thesis of Placeholder?
Burnham: It's investing in both the applications layer and the protocol layer of the Internet, and it's a bet on crypto being a new infrastructure in the same way that TCP/IP is infrastructure that we can build lots of services on. It's obviously a provocative thesis, there's a certain amount of timing and risk about when and how this becomes mainstream but we believe it will become mainstream and will have a profound impact on the way we do everything. USV is also investing in this space but is somewhat constrained by not wanting to become a registered investment advisor. Therefore, it has a limitation in terms of the number of investments that can make. Placeholder has bit the bullet and said we're going to have to deal with this so we're all in on crypto.
Paris: What size checks are you writing?
Burnham: We will write anything from a $250,000 check to $10 million. We tend to take the larger positions in the more liquid tradable currencies and so we have a mix of investing in well-established networks that are liquid and tradable. We are also investing in brand new startups that have a year or two to go before they even get to maintain that. We'll do that full range, we're very interested in both applications that sit right on top of these protocols that create value. We're also interested in the protocols and invest in tokens as well.
Paris: Can you tell me about the projects in your portfolio that are helping you create your decentralized utopia?
Burnham: One very interesting protocol were working with is 3box. It’s creating personal data stores on a distributed network and I think they're a key piece of infrastructure for any crypto application that wants to separate the data from the application and not take on the responsibility or the cost of managing that data. This is for the crypto application that wants the user to manage their own data but wants to provide services to that user based on that data.
Paris: Your utopia is based on people owning their own data, but how practical is it for me to manage my own data? There’s so much of me.
Burnham: Oh, you can’t. We don’t have infinite memory. There's no way any of us could, but you should be able to choose who you delegate that responsibility to. You should never put yourself in a position where you're solely reliable on a single vendor. If an architecture emerges where people aggregate personal data stores on a system like 3box, you can take your data with you anytime you want and have complete control over it. It would be stored on a distributed network so even if the data store goes down, you wouldn’t lose all of your data. You control the keys, you can give the keys to a user agent, and you can take them away. You can permission them to have access to your health data or just your location data. But I think we will see a class of applications emerge that represent us to the web that can make choices on our behalf, so if you don't want the cognitive load of thinking should I give this data or that data, the agent can do that and you won’t have to train it.
Paris: To have a personal data store, does everything I do need to be recorded on chain or can data be stored off chain?
Burnham: Key management maybe tied the chain in some way but the data store doesn't need to be tied to the chain. It's your data store.
Paris: How will all of this work with the new privacy laws?
Burnham: I think most of the consumer regulation around privacy is actually backfiring and entrenching the existing incumbents because they're the ones with the resources to manage those relationships. I believe the half life of most data is short. I want a market-based mechanism that creates an incentive for the agent not to do the wrong thing. And that incentive is I have the keys, I can shut you off. If you’re a bad actor and move my data to your side, I can cut you off from future data and now I'm accumulating data in my own data store that you don't have access to it.
Paris: If the value of the data stolen from you exceeds the value of the future relationship to that agent, then that market incentive may not work.
Burnham: There could be something contractual that if you get breached, the agent would owe thousands of dollars per record, thus a bankruptcy situation. Why would the agent take that risk?
Paris: Right now, I trust Facebook more than 3box because I don’t know who 3box is. How do you overcome that friction?
Burnham: I agree there’s friction but I believe you’ll see the use case to delegate control of your data keys to an agent whose principles you know and already trust. I’m arguing for an architectural model that is inherently trustworthy. It’s not just market incentives, it’s also the distributed nature of the data. Bitcoin is the only network that’s never failed because it’s built on 400,000 servers around the world. You have multiple copies of the same data set in multiple places, the only difference is that there is an agent managing those keys on your behalf.
I think the smartest thing in the world for the data monopolies to do is to become data agents themselves as the competitive marketplace emerges ...but I’m not offering Facebook any strategy tips.
© 2023 The Block Crypto, Inc. All Rights Reserved. This article is provided for informational purposes only. It is not offered or intended to be used as legal, tax, investment, financial, or other advice.