Jack Dorsey Latest Interview On Bitcoin

Thank you so much.

Well, it was a long time in the making. When I was a kid, I was always fascinated by how cities worked — just how these things that were so large to me at the time actually functioned. And I got into maps and I would obsess over them and stare at them and try to understand what was happening in any particular neighborhood or area. And that led me to programming because my dad brought home a computer one time and I wanted to draw a map on the screen so that I could start playing with it. And I taught myself how to do that, I figured it out, and little by little I worked my way to a field called dispatch — dispatching taxi cabs and ambulances and police cars. And it was a way to put information on the map, information about the city. And I eventually ended up working at the largest firm in New York and had this very rich map of everything happening in New York City, but it was missing one thing, which were the people and what they were doing, where they were. And that's really the question that started Twitter — if I could see and visualize what the people were doing and what they were thinking or where they were going, it would help me understand the city a little bit more. And it took probably about 10, 15 years for the right time and the right technology to really make the idea something worth pursuing. But in 2006 it did, and I was working at a podcasting company that kind of lost its way because it suddenly became competitive with Apple when Apple decided to get into podcasting. And we had this great team, we started a hack week, I brought up the idea for this thing that you could update from your phone and other people could follow it and see what you're doing and what you're thinking and where you're going. And I was given another programmer and two weeks, and we built it. And we fell in love with it, and little by little we took more and more resources of the other company, and we eventually spun out and took venture capital.

I didn't really have a vision. I wanted to play with it. I mean, ultimately I wanted to understand what was happening in the city — that was the original question — and I thought this was a way to answer the question. But I just wanted to use it. I wanted to be able to see what my friends were doing. I wanted to share what I was doing. SMS was just becoming big in the US — you all in Europe had it for 10 years prior, but the US, because of our different protocols for carriers, Verizon and AT&T, GSM versus CDMA, we didn't have text messages that actually worked together, so we didn't have text for quite some time. But when we did, that was the technology — like, oh, I could have this cheap little Nokia device and I could send a message and it would fan out to all the people that cared about what I was saying or what I was doing. That was super cool. And we just launched it and we invited our friends and they invited their friends and so on and so forth, and then it continued to roll. And we decided as people were using it that there's something that resonates here, there's something interesting, and now it's up to us to really pull the thread on what that is. And what's interesting about Twitter is it wasn't really us who determined where it went — it was the people using it. We saw them try to converse with each other and address one another. And I took three hours over breakfast and I wrote some code — anytime someone tried to address someone else, it would be linked and it would be filtered into a replies tab, is what I called it. And that started conversation on Twitter, this public conversation. Whereas I originally was thinking, oh, this is just a way to update my friends of what I'm doing, see what they're doing, but people wanted to talk. And they invented the syntax, the at symbol, to talk, and then they invented another syntax, the hash, for grouping — topic grouping tweets by topics. And then they invented another thing called the retweet, which is rebroadcasting someone they're following to people following them. So little by little, as we observed what people were doing with it, we made it easier, it spread, and it made the service much bigger because more and more people got value out of it. And there's just an interesting way to co-create with the people that are using your product, and I learned a lot from it.

I think people don't want to just consume — they want to help. Participation means creation, and they want to leave some mark. They may not intend to do that, but when you use something like Twitter, you develop this feeling for it and you want it to last, you want it to succeed, and you become very emotional about any changes that are made that you think go against the way it should be or the way it's used. And we saw so many of those when we tried to do a thing that we thought was important — the network just threw up on it and said it was wrong. And as we observed what they were trying to do, we had much better insight into what Twitter wanted to be. And as we got more new audiences and audience types, that density of information would grow even further.

When Obama was elected. 100 days after he took office, he got up in front of the country and did the State of the Union, 100 days after. And we had a number of senators and Congress people on Twitter already, and I was watching the speech, and the camera panned around and there were people on their phones. And in that moment I got a tweet — my pocket buzzed, I pulled out my phone, I got a tweet from my senator from Missouri. I'm from Missouri, and she was tweeting about Barack Obama giving his first address to the nation. I was like, whoa! I'm watching this thing unfold in front of me and I'm feeling it physically in my pocket. And it just really shortened the distance between something that can be very abstract and feel far away — which is how our government operates and works — to now being in my hand, and having this sense of like, wow, this tangible humanity. Someone typed this out, it buzzed my phone, and I'm reading it, and I saw it, you know, they're thousands of miles away.

Oh, I felt excited. I got back to work right away because I was like, oh man, this feels great. I mean, that's what you want in terms of building any product — you want to see people use it in a way that you can see that there's value and you can feel it. It feels electric and you feel tingly and you're like, oh, I'm going to keep going, I'm going to keep going.

These things aren't important by themselves. We make them important and we choose to by giving them attention, give them importance. And it spoke to a gap that I think existed before Twitter, which was everything was entirely scripted, it was behind a media wall. The information that you got was through another point of view, typically a corporate point of view, corporate media point of view, paid for by advertisers and seen on television or radio or whatnot. Podcasts were the first, in my experience, to really break that wall, but they weren't talking about newsworthy stuff at that point — they were just doing interviews and whatnot. And they matured eventually, but Twitter kind of pierced through it very quickly. And it pierced through that wall of abstraction, media abstraction. And I think for elections, again, our political process, our government can be extremely abstract. And when you have something that you know is 99 times out of 10 coming from the person's thumbs directly, which is connected directly to their brain, it's authentic, it's unfiltered, and it's open to your interpretation instead of being filtered by someone else's interpretation. So it was less about Twitter providing that — it was more like that was the need and that was the desire, and anything at that time that could fill that would have taken off. I think we had a lot of luck, we had a great product because it was simple, it was straightforward, it was easy to use, and it was accessible. But it also was the right time — the frustration with the media wall was at a peak, at a high, and there were cracks in it, and people wanted more of that. They wanted more authenticity, they wanted more directness, they wanted more realness.

80% of the organic, non-reply content — like, there are replies too, and that's not often counted in that number. So me just sending out a tweet, yes, you're right. But I think the real power of Twitter is the conversation underneath it. And it's not necessarily about, you know, what is the news of the day direct from the source — it's like, what are people talking about? What do people think about the news of the day? That's the power, and that is extremely large. And then there's the — yeah, like, Twitter has a bunch of influential people on it and people want to listen to them and such, that these tweets were being broadcast on television and we were unable to count the impressions of CNN or Fox putting a tweet up, but we knew the reach was much larger than our analytics and our measurements accounted for. So that was pretty interesting, pretty special.

I mean, there's a lot of things I would do differently in retrospect. I think we became — I wasn't there at the time running the company, but we raced to monetization way too quickly and we force-fitted a monetization path that created a bunch of issues for us, which was advertising. Google had an advertising business at scale, Facebook had one that was growing. Twitter's goal at that point was to grow and go public. Being public meant you had revenue, so we copied the model of the day, model of the internet, which was advertising. And that just set up a bunch of issues in terms of growth, in terms of competition, in terms of how we were valued. It didn't really express Twitter's true value, that model. And I wish we took a little bit more time to figure out a better model, or that we had multiple at once. The second thing is, in the early days Twitter felt like a protocol. It felt like HTTP, it felt like SMTP. Yet we owned every single part of the stack, and I think that was a mistake, because I don't think a company can necessarily control this without having some sort of choke point that makes it open to manipulation in some way. So if you have an open protocol for social media — which I think Twitter could have been, instead of just a company, it could have been both — you have a lot more freedom and you have a lot more agility to do the things that the platform actually needs to do and the protocol needs to do. But if you wrap all that up in one company, you create a lot of issues for yourself, including heads of government and advertisers wanting to dictate content policy, or get the numbers of journalists who are critiquing them, or shut you down because you won't take content down in their country. If it was a protocol instead that was owned by the public and you're building a business on top of that, you couldn't do those things. This is where I learned a lot from Bitcoin.

Oh yeah, yeah, definitely.

There's tons — there's subscription, there's commerce. Because the internet didn't have, was not born with a native currency protocol, advertising was the way to subsidize free services. And it was taken as a given that these internet businesses had to be advertising businesses in the early days. Netflix showed a different path, obviously others — eBay showed a different path. But for an information tool like we were, advertising was it. It was the way to get the biggest network because you could offer it for free and subsidize it. But the reality is, the people that are using it are truly the product, and your customer is the advertiser, and your customer is dictating where the product should go.

Experiment as much as possible. I mean, like, keep pulling the thread on what people want to use it for. I'm a big believer in — I think it's less about free speech, I think it's more about do we get to choose how the algorithms are programming us, because the algorithms are definitively programming us. And it's very, very hard to tell what an algorithm is going to do and predict what it's going to do, and that's where the danger comes in. But if I can choose what algorithm filters my tweets and what I get to see, and I have instead of like an app store, an algorithm store, where I get to put these on my timeline, then I have a lot more agency in the content and the experience. And I think it ends up creating a much healthier relationship with the technology. And I'm not just speaking about Twitter — this is for YouTube and for financial technologies. We are being programmed. These algorithms know our preferences better than us. That's only going to increase. How do we increase agency at the individual level? It's by choice. It's by being able to turn off the algorithm. It's by being able to choose different algorithms. It's by being able to make my own algorithm.

Well, I mean, a simple one — like, if I were more left or right leaning, I could choose a more right or left leaning algorithm. If I hated politics and I only wanted to focus on science and arts, I could choose an algorithm that filtered everything out. But those things — I could go deeper into that. I could only do physics, I could only do quantum physics. And you might say, well, you can do that through taxonomies and categories, but there's just too much information at this point for those to really be useful. You have to get into the algorithmic aspects of it and really effectively create these agents that go off and fetch the information that you're truly interested in and deliver it back to you and take feedback on whether it was useful to you or not, if you learned something from it or not, or you interacted with it or not. Right now, most of these internet companies and most of these internet services, the algorithm is driven towards one thing, which is maximizing impressions to maximize advertising revenue. That is not maximizing my experience, my learning of the world, or aligned with my goals as an individual.

Oh, it's a great thing. But I want to be able to choose that. And you can build a great business around that. You can sell the algorithms, you can subscribe to them, you can put more directed topical advertising on these algorithms because they actually are programmed in a way that you're actually interested in the outcomes. They're not just random and they're not trying to guess — they know. So I think it lends itself to a much better business, but you have to have companies and individuals who are willing to let go of the secret sauce, which is the algorithm, and democratize it a little bit more.

I think we're seeing the extent of that right now. I think we're at the edge of one horizon. I think the next horizon for these tools is probably around agent systems and software that I task to do things for me and they work with other agents on the internet to complete those tasks. So I'm more interested in that. I think the generative AI stuff, especially the large language models, have the ability to obviously create a bunch of information, and it becomes important — like, what is real, what is reality, what is truth, what is fictional. And these tools are going to get so good that you won't be able to tell, and we'll truly feel like we're in a simulation because it looks like and we perceive that anything literally can happen or could happen or did happen because these tools are so realistic and feel so authentic. So I think the only way around that really is to just have more information and better filters, which leads to agents, and that they'd be directed by me, by the individual, and do more of that work. So I'm less interested right now around LLMs and generative AI because I do think we've hit some horizon on it, and more interested by how you take this intelligence and make it mobile and make it something that people can send off and it can actually do work for them.

I mean, I'm not a journalist, but I think having more information, more perspective, and getting better tools to get more information and better information and distill it only helps your craft. I think it does become more and more challenging because you again have to sift through what's real and what's fake. But people have been lying since the dawn of time and the dawn of language, and journalists have to figure out what's the lie and what's truth, and they need different perspectives. So I think the only thing that changes here is velocity — nothing else. Velocity and breadth.

I think the best content moderation is an algorithm, and I should be able to determine my own algorithm and what I see and what I don't see. I think there should be a corpus of content that exists that only the author of it can delete, and they have the sole right to do so. But I have the right as an individual to filter out what I don't want to see or filter up what I do want to see. But that's all dependent upon whether I get to choose the algorithm. If a corporation is choosing it for me based on maximizing advertising revenue, then it's not truly serving me.

Yeah. About three years before I left Twitter, I realized that, like what we were talking about earlier, a company cannot be all parts of the stack for social media. It'd be better if we were built on top of a protocol that was shared between multiple companies. And what's an example here? HTTP — the protocol that runs the web, the protocol that we're using to talk right now and record this podcast. It's a common understanding and agreement that this is how we transmit data in a very particular way. And then you have browsers on top of it — I'm using Safari, you might be using Firefox or Chrome. These are built by different companies. These companies can build business models on top of that, but they're using the same protocol and they don't control the protocol. They can't influence the direction of it as a single entity — it has to be done through this consensus model. So if we have that for the web, if we have that for email, if we have it for file transfer, for video, why not have it for social media? Why not build a protocol that any company, whether it be Facebook or Google or X or open source tools, can use around this concept of social media? And that was the idea, and we called it Bluesky. We took about a year to find a leader for it. We found a leader in this woman named Jay, and she built a team and actually built it and built the protocol and built a client. And it's working. There's another individual from Brazil — I don't know who this person is, it's a pseudonym — he created this other protocol called Nostr, N-O-S-T-R, had a different take on it, which was he kind of looked at every single one of Twitter's failings as he saw it and built his own protocol. And what was interesting about that protocol, I found it about a year and a half ago, was that not only did it address what I perceived on every one of our problems, but it was completely open in terms of the development model. Bluesky was effectively this corporation building the API and building the protocol. Nostr was just a bunch of individuals building the protocol and building the clients on top of it. And I resonate a lot more with that because I think at the protocol level it has to be done not through a corporation but through a collection of interested and obsessed individuals. And that's what Nostr was. So eventually I stepped off the Bluesky board and really focused a lot of my efforts, which amounted to funding this ecosystem around Nostr. Nostr is still very small because it's a lot more technical at the moment, but I have a lot of hope for it because of the way it's built. It follows the way that Linux was built, it follows the way that Unix was built, it follows the way that the C programming language was built, versus these other cathedral-type models which tried to achieve perfection but actually no one ended up using because they were built by these companies and these companies had these particular agendas and there was no real individual drive behind them. So Bluesky is a very different model than Nostr. I think the Nostr model lasts longer and wins, but it's good to have multiple experiences. And there's others — there's Mastodon and there's Matrix and what Meta is doing with Threads. All these things are kind of competing with each other, but one of them will win.

I think people are running away from X rather than running to something on Bluesky. And that's not a great way to build a product, unfortunately. We want people that are running to us for a particular thing that they couldn't do before. Now, Bluesky has built something that I wanted it to build and I'm excited about, which is this algorithm store — being able to choose your own algorithms. So that is a reason why people will run to it eventually, and I think why people run to these services eventually, because they get more agency, more control. But it's not something that people care about right now. What they care about right now is not being in X for whatever personal reason. And again, that sort of behavior kind of comes in waves. You have to want to stay there for a reason. You can stay there because you have moved your entire community over, which is rare. But when we have these different factions of social services — if X represents the right and Bluesky represents the left and they can't actually talk with each other — there's going to be crossover naturally because people are going to keep their accounts because they want to hear what the other side is saying. It's just natural. Only way to solve this: one protocol, different views, different algorithms on top. This can be the right, this can be the left, different business models, this one works, this one doesn't work as well. But we need a common protocol, and that's the whole idea.

I think it's better technology. I think it's a much healthier developer ecosystem. I don't think it's just social media like Bluesky is trying to be — it's more of this ecosystem of apps with a common protocol. So people have built Twitter-like things, they've built spreadsheet-like, Google Sheets-like things, they've built GitHub-like things. And it's really amazing because what it does is it allows you to really prove authenticity, because it's a public-private key pair system. It's not an email and password. I have a public key and I have a secret key, and I maintain the secret key. And every message I send on Nostr, I'm signing with my secret key, so you know it's actually coming from me and my device, which is very unlike a Twitter or a Facebook where you don't actually know — it could be intercepted, it could be the company doing it on the person's behalf in the worst case. But because I control my key, you know anything I write or anything I post is coming from me directly.