Welcome back to another episode of Terms of Service. I'm Claire Duffy. Today we're going to be talking with longtime tech investor and entrepreneur Kevin Rose. He's here to talk about something that, if you were an internet user in the early aughts, you'll probably be excited to hear about. Kevin is relaunching a platform he founded in 2004 called Digg. Digg was similar to Reddit, a platform where users could share links, discuss, and upvote (aka dig) or downvote content. The homepage was basically a list of the most popular links from across the internet, and for a while, Digg had the power to send so much traffic to publishers featured there that their sites would crash. Eventually, users turned on Digg after an unpopular redesign, Reddit overtook it in popularity, and it basically fizzled out. Now, Kevin is relaunching the platform with an unlikely partner: Reddit co-founder and former rival Alexis Ohanian. With this revamped version of Digg, Kevin and Alexis have a goal that feels especially ambitious in this moment where people are so divided and AI-generated content is flooding the internet, and that is making social media human again. So what exactly does that mean? How could it change all of our experiences online? And why now? That's all in today's conversation with Kevin. Stick around.
We are here today with longtime tech entrepreneur and investor Kevin Rose. He is relaunching the platform that he founded in 2004 called Digg. We are going to get into why you're doing it, but it all has to do with hopefully making social media and the internet a bit more human, right?
Yeah, 100%. I mean, this is one of those things where I don't think Alexis or myself would have dared to enter into the arena again, just given the chaos that is social media. But it was a very unique time in the world where AI is coming online in a way that we're pretty optimistic can be quite helpful in solving a lot of these bigger problems, as long as we're actually doing it right and really making sure that we have a chance to go back to the bare metal, to first principles, and build a network from the ground up. What would you do today that's differently? And the answer is quite a lot.
So before we really get into the plan, you launched Digg more than 20 years ago. You're making me feel old now. In a few words, how do you see how the internet has changed in those past 20 years?
Well, I mean, we had no idea what we were doing. So in 2004, when the dig button was the first button to vote on things, and this was like pre-Facebook likes and everything else. I remember when Zuckerberg came to my office and he wouldn't sit in a chair. He was like a kid and he would just sit on the ground, which was so odd. But whatever. And so I'm like sitting in a chair, he sits on the ground. He was asking about likes and not likes but digs and voting and how we thought about that. And like at the time, we had gone and we had filed a patent for this idea that someone would say 'I like this piece of content,' but more importantly, that would feed back into an algorithm that would then say 'I understand this person's interests and I'm learning more and more about this person.' And our naive thought was that we will put better content that the user actually really loves and enjoys right in front of them based on the stuff that they've dug in the past. We had no idea that this stuff was going to be weaponized the way that it has been, and that social media will turn into basically, well, some of these networks they just hand the microphone to the loudest mouth, and it can be pretty damaging. So you know, Alexis and I both oddly have girls that were just within a few months of each other too. We both have two, and we were like, we need to build a better world, a more human world, especially given how much of these agents and AI and bots are just going to be taking over the entire internet and eroding all trust if we're not careful.
I have to know too, where did Digg, the name, the idea of that's what you call a like, where did that come from? As a product of California surf culture myself, I love it.
Yeah, well, there was no term like that hadn't come out yet. And so I thought of dig as like, there's all this great content underneath the surface, like under the ground, like hidden, that needs to be discovered. And so the idea I had was like, okay, dig will just be a way to unearth that. And the more votes something receives, the more exposure it will receive. And so it was just the idea of finding the gems in the rough and kind of bringing those to the surface that turned into that. So it was very much like a Reddit homepage. And so Reddit came on the scene a few months later. We competed with Reddit and we beat them for about six years, so much so that they sold the content ass and we thought we had won. And then they kind of respawned it back up. By that time, we were like old and we were pretty mature, and our investors were pushing for us to go more mainstream and less of the weird news, which is what we were so known for. And then Reddit opened up the taxonomy, basically allowing for any community to spin up. So you could say 'I want a community about knitting' or whatever it may be, your weird little interest or niche interest that you have. And turns out the long tail of those niche interests in aggregate was quite a big community, and it overtook Digg. And then Digg kind of fell away and became an old relic at that point.
So you announced in March that you were teaming up with Reddit co-founder Alexis to relaunch Digg. Obviously, he has not been with Reddit for a number of years. Why now, and what are you hoping to do together that you weren't able to do before?
Yeah, a lot of things. There's probably, I would say, a good 20 things that are uniquely ours that we're going to do over the first couple of years. A lot of the early stuff will be kind of like table stakes, let's get this thing up in the air and flying. But then very fast follow with a lot of the new novel stuff that I think will force Reddit's hand to make some pretty dramatic changes. There's a lot of things that you can read about it online, but Alexis's departure was very public and that he was pushing for change and he ended up quitting and leaving the board, and they actually removed him from the S1 filing, which was like a really dick thing to do. And there was a bunch of stuff there that just didn't land well for me reading from the outside. And then, you know, I'll let him speak to that on his own terms. But like, as you can imagine, he's fully divested and separate from Reddit. So when we started thinking about these things, it just, you know, it's very apparent that Reddit is really powered by about, you know, call it 50 to 75,000 non-paid moderators that do so much heavy lifting. And they are one, their tasks are very laborious and boring, and they're banning people and they're just dealing with a lot of archaic tools on that side. And two, they have no ownership over these things. So like, you know, when Wall Street Bets got big, the first thing Reddit did was trademark it and then file a lawsuit against the guy that wanted to write a book about it. And it's like, those people created this community. Like if anything, with my Digg hat on, if a community gets to a certain size, I want to pay for them to trademark their own stuff. Like I want to thank them and say, 'You're awesome, you did amazing work here, your effort should be rewarded. Let's help you run with this brand and monetize it and own it in a unique way that you have some say in the future of this thing.' And so for us, it's like we see a much more aligned kind of a Patreon and Substack world where people should actually have more ownership and control and say over these communities that they're building. And so if we can get there, you know, I don't think this is a winner take all, so we're not trying to destroy Reddit. But if we can do some things on that front that really force them to go in and really put community first, it'll be a win for the entire internet.
And how did the partnership come together? Did you call Alexis?
So Alexis and I had been at first enemies for many, many years. We really didn't like each other, even though we didn't know why, we just like competitor. And then it was one of those things where I saw him start posting about his girls and the love for his girls, and I was just like, this is a good dude. And so I reached out to him and we were chatting, and then he was like, 'I'm raising another fund, do you want to invest?' And so I invested in his fund 776 and we did some deals together on the investment front. And then it was one of these things where we did a podcast together and we're like, 'Okay, let's break bread.' And we ended up just jamming so well on that podcast, we were like, 'Oh, we should be friends.' And it was just like, it turns out the reason why we built these things in the beginning was because we have a lot of overlapping interests. And so it kind of kicked off that way. And then when this opportunity to buy Digg back came yet again, because it's come on my plate a few times, just what we can do to help these moderators, what we can do to change the entire game, is so unique right now with the application of AI that it made sense to go and fund this thing and go for it.
Right now if you go to the Digg website, it's like 'stay tuned' basically. Can you give us a teaser on like when do you think you're going to be able to relaunch, and will it look at all like the old Digg?
Yeah, I mean, there'll be definitely, I mean we at our peak we had 38 million people a month that were coming to Digg. So in today's standards that's tiny, but it's a decent size site. We want a little bit of nostalgic kind of the action going on there, so you'll feel the old Digg vibes kind of shine through for the hardcore. And that's like, we've had a couple hundred thousand people sign up to be a part of the early beta. So those are the people that were there and they were like, 'I remember Digg, bring back those days, a simpler internet.' And so that will be there, but then we'll try to pepper in new advances and push things forward in ways that make sense very quickly thereafter. And so we've got 25,000 people that are part of this Groundbreakers community, which is like the early access community that we capped at 25,000. And we're going to move all them over and get them kicking the tires, squashing bugs, and giving feedback right away.
Got it. That was a question for me, like how do you, once this is up and running, how do you attract users away from all of the other platforms that are occupying people's time right now?
Well, I think everyone wants to build the X killer, Twitter killer, whatever. There's been so many clones of that type of interface, but no one has really gone after what I think Reddit is quite good at, or was good at. And I say 'was' not as a slam, but we're entering into a world where agents and bots are taking over a lot of these conversations without people knowing it. We need to make sure that human verification is a very important piece of the future of this. Interesting. So trust is gone, it's dead. And so we need to start first and be very deliberate and very intentional about who we onboard, really making sure that they're human to the best of our abilities, and making sure that's a gradient of trust that we have that people need to maintain over time. Because in our opinion, when humans come together around topics and have conversation, they want connection. And we think connection is a big part of this, real human connection. If they're having conversations about these things, trust needs to be at the front and center. So if I say, 'Hey, I've had a Whoop band and I've owned it for 5 years,' I should both be able to prove that inside of Digg, kind of like the verified purchases that you see on Amazon. And you can do that in an anonymous way. So we very much see a world where people are conversing about things, having conversations, they're debating different topics. And yes, AI can be there to fact check, yes, AI can be there to extend the information and always be flagged as such, or find me other relevant articles that I may be interested in reading, like ask questions of the corpus of data. If the new Switch comes out and I want to know what type of GPU it has from Nintendo, I shouldn't have to go and look that up. I can ask AI in two seconds directly on Digg. Those are no-brainer features, right? But I think that the humans are always going to be the most important piece of this network. And if we can do that, then we will have the most trusted platform out there. Because we believe, and we've seen this, LLM prices are falling. The cheapest LLMs now can act like a human and convince you and hang out in your DMs for 6 months. They don't breathe air, so they can hang out in your DMs for as long as they want and just have conversations with you and slowly convince you over time, months or years, to eventually buy something, take action on something, change your opinion on something. And I don't want to be a doomsday scenario type person here, like this whole dead internet theory that is all going to be run by bots and AI. I think that's a little wild, but I think it is a real risk that we need to pay attention to right now. And so that's top of mind for us.
It makes sense that you would want to know, am I in a space on the internet where I'm interacting with people or AI? Obviously there will be both. How do you do that human verification piece? Because obviously lots of platforms, I mean Elon when he bought X said that dealing with the bots was a big motivation. We've not seen all that much action on that front, at least publicly. How do you do that, and will you also be trying to preserve the ability for people to interact anonymously even if they've been verified on the back end?
Great questions. Yeah, and I think you're hitting two very important parts. One, there is no silver bullet for humanness on the other side, because you're going to have agents that are acting on your behalf, and in some cases that'll be okay, like 'go find this information on Digg for me, you're an authorized agent to go and do this.' But there are a whole slew of technologies that kind of overlap each other that you have to apply and look at this kind of gradient of who we know or what we believe to be true about this particular person. Then there's other things where, like, you can do as simple as a $2 fee on a Stripe charge, where Stripe's seen 90% plus of the internet's credit cards. They give you something back called the Stripe Radar score, and the Radar score says, 'Okay, does this person...' They look at over 200 different pieces of data out there, and they're the most important thing to Stripe is to prevent fraud from happening, right? So they know what disposable cards look like, they know if they've seen somebody that has an email address that has had an active Adobe cloud subscription for the last 3 years. Like they can tell, are these real people behind here? And then you pair that with unique device identification along with where does this person, how does this person normally show up. So the models behind the scenes on our end, they take all these things into account. They say, so let's say for example, you have an account, you do some of these things, you're like, 'Yeah, I'll give Digg a couple dollars because I want to get that next level of verification. I'm really verified on Digg, I'm trusted.' And then all of a sudden out of nowhere your email address changes and your country changes and you're posting completely different topics than you've ever posted before. Like we're going to know that and we'll be able to flag that and really run that down through a series of checks to make sure either you just moved or something really weird changed about your personality, or you know. So there's a bunch of models that you have to have running 24/7, 365 on the back end to check this stuff. So another great example is, as you're about to post a comment, in 200 milliseconds or less from Google right now with some of their ML tools, I can do complete sentiment analysis of that comment, soup to nuts. I can tell you, is it hate speech? What topics is it mentioning? Is it mentioning products? Is it trying to convince someone? I get all of these different parameters back, and I can even prevent that comment from being posted. I can tell you, 'You're not showing up correctly for this community in the way that you're about to post some really outrageous attacking another user or something like that.' I can give them a timeout. There's all of these pieces of tech that exist that need to be implemented right now that no one's just really taken the time to put in there that should be happening right now. Like it should catch this stuff as it's coming in the front door. And so that's on us to build out.
