ROSE WANG: When South Korea called martial law, where was that breaking news unfolding? It was happening on Bluesky. It was really hard to follow on Threads. It was really hard to follow on X.

We don’t have the budget to go and win over every celebrity and every big brand, but our belief is if we focus on what the end user wants and needs, then everyone else will come.

We’re not building the infrastructure for a specific political viewpoint or a specific group of people. There is maybe a wrong belief that communities create echo chambers.

I think that communities actually help us learn about new ideas in a safer environment.

BOB SAFIAN: That’s Rose Wang, COO of the superhot social media platform Bluesky. In recent weeks, millions of users have been switching to Bluesky from Elon Musk’s X platform. That’s created both new opportunities and new challenges for Bluesky’s small 20-person team. I wanted to talk to Rose because Bluesky represents a major shift underway in social media, as dissatisfaction with the big platforms grows. Rose talks about why Bluesky isn’t just an alternative to X or a liberal refuge but part of a new social web that prioritizes closer personal connections, data privacy, and democratization. For anyone eager to understand where the online world is headed, Rose offers terrific insights. So let’s get to it. I’m Bob Safian, and this is Rapid Response.

[THEME MUSIC]

SAFIAN: I’m Bob Safian. I’m here with Rose Wang, the COO of Bluesky. Rose, thanks for being here.

WANG: Thanks for having me, Bob.

Inside the surge in users that Bluesky is experiencing

SAFIAN: So in the wake of the presidential election, we’ve witnessed a so-called X-odus, right? Millions of users leaving Elon Musk’s X platform and finding their way to Bluesky. You guys have added millions of new users, up to 24 million active users in total. Huge growth for a company with only around 20 full-time employees. What have the last three weeks been like for you?

WANG: It has been a hypergrowth phase that a lot of companies dream about in some ways. And so I think the personal experience is that sleep has been a fifth priority in my life when maybe it’s been in the top three in the past. But also I would like to say that this isn’t the first time we’ve experienced growth.

We experienced a lot of growth in February of this year when we launched Bluesky to the public, and we had over a million Japanese users join in a week. And then sometime between April and August, we also had millions of Brazilians join Bluesky. So we’ve been experiencing growth across the world.

It’s really awesome to experience it in your own backyard because we have our server down the street who’s like, “Oh my gosh, I’ve seen you on the news.” And so that’s really cool to have people from all walks of life kind of come out and know what we’re working on and join Bluesky.

And that’s really fun. But it’s important for us to note that I think there’s a general need for a new social media platform, not only in the U.S. but truly all over the world. We’re really excited to fill that hole.

SAFIAN: Yeah, the Brazil burst was in part because Twitter X was having some disruptions there, right? They were having some issues. The surge that you’ve had most recently after the election, was it something you were sort of prepared for? Did it come as a surprise?

WANG: Well, I think being prepared and being surprised are actually two different things. You can be both. We were prepared in the sense that we had built our protocol for Twitter initially. So we were ready for a large number of users to come on all at once.

But then, could we have predicted the way that was going to happen and when it was going to happen? Absolutely not. Because I think we would have probably had more staff members. But we have managed quite well with the team that we have. Our priority is to keep the app online.

SAFIAN: Was there a moment when you realized like, “Oh, something different is happening here?”

WANG: I think the thing that felt really different about this surge was that momentum continued much longer than it did in a lot of other countries, and there was no clear end date. For Brazil, for example, like you mentioned, it was because X was banned in Brazil. We had also assumed that X would get unbanned in Brazil, so we knew that there was a certain time limit that we were able to win over a lot of Brazilians and tell them about what Bluesky was.

But here I think that people are coming to Bluesky for very different reasons. People will attribute it to a certain event, but I oftentimes say people don’t leave their homes after one thing that goes wrong. It’s generally after so much disappointment that trust has been lost.

And some sort of event just tips you over the edge. We started to see this growth when X made another announcement about their difference in block, that it was going to be more of a strong mute, but that was not the first time they announced it and they hadn’t even enacted it. But already that night, 500,000 users came onto Bluesky.

So we were seeing already that users were very unhappy with the experience they were having on other social platforms like X. I think the reason they’re coming here is actually something a lot deeper, which is there are dark patterns happening. They’re not feeling good. They’re not making friends.

They don’t feel safe. It’s these very fundamental human needs that are driving people away.

Why Bluesky is not just a Twitter alternative

SAFIAN: So for listeners who may not be as familiar, can you describe what Bluesky is and what makes it different than X?

WANG: Absolutely. I do love this question, so thank you for asking, because a lot of people assume that we’re just an X alternative, but we’re very different. We are an open social media platform that puts users first and gives users choice. What does that look like? Well, algorithm where it’s controlled by a small group of people, and with Bluesky, that’s not the case.

It’s no longer the case. Users have built over 50,000 different feeds including, I think, like 5 cat feeds, more than 200 feeds about Taylor Swift. There’s wrestling feeds, F1 feeds, and the possibilities are endless. We’re seeing that people really like the chronological feed. We’re seeing that people want a lot more control over their experience and know who they’re talking to.

These feeds provide a cozier corner to connect with other people with either similar interests or have similar dispositions because there’s just no longer a single dominant algorithm that only promotes either the most polarizing posts or the biggest brands. So you actually get to interact with real people and have fun conversations again.

SAFIAN: Even if the size of that group that you’re interacting with may not yet be as big? Or maybe that’s an advantage that it’s a more intimate group that you’re interacting with.

WANG: Yeah, that’s a question that keeps coming up, right? I think there’s a lot of people who are worried. Have we completely lost the town square? Or has the internet unbundled and broken up into all these smaller communities where people are tired of a global stage? I think that it’s hard to point to a point in time and say that all of time will look like that point in time.

We’ve seen that history is cyclical, and we’ve seen that there’s generally an unbundling and a rebundling that people tend to go through when they go through periods of, “Hey, you know what, the experience I had in the town square wasn’t serving me.” So that’s what’s driven them to unbundle and find these smaller communities.

We believe that there will be a time again when people want to know what’s happening and be in a larger conversation and there will be a rebundling. Rather, what we’ve built is the plumbing, the infrastructure for many apps to be built. That’s the whole point of Bluesky. People say, “What does decentralization mean?”

What I say is it just means that it’s more democratic. It means that we can build Bluesky, and you can build Green Sky and Yellow Sky. Users or voters can go and choose where they want to live based on the experience that’s provided rather than locking them in and forcing them to stay in a home that they don’t like.

How Bluesky avoids forcing users into echo chambers

SAFIAN: I’ve noticed that recently Bluesky has become kind of a liberal refuge from the right-wing tilt of X. I know Congresswoman Alexandria Ocasio-Cortez, AOC, became your first user to hit a million followers. How much do you talk about that on the team? Is that kind of focused community a feature, or is it a bug? How do you think about that?

WANG: Well, first of all, we’re not building the infrastructure for a specific political viewpoint or a specific group of people. The goal is, “Hey, here’s open infrastructure for global conversation, and here are the tools that give you the ability to create your own experience upon that infrastructure.”

“Hey, go create your own space, go figure out what your community needs to feel safe,” and we won’t disrupt what you want unless you’re building it in our society and it goes against our terms of service and community guidelines. In which case, you can start another server and just leave our society and build your own.

We actually take a lot of inspiration from just how humans interact when they’re not online, how they interact at parties, how they interact in communities. I think there is maybe a wrong belief that communities create echo chambers.

I think that communities actually help us learn about new ideas in a safer environment. What we’ve learned is putting people in a town hall and having other people scream their extreme opposite viewpoints at them probably won’t change your mind. But if you’re in a community with people with adjacent viewpoints and have other things in common, you’re more likely to take on that adjacent viewpoint.

So we actually think that Bluesky provides both of those experiences, where you have all these custom feeds that are smaller communities and smaller worlds where you can go and have fun and feel safe and connected. Then, when you’re feeling more abundant and ready to go and discover new and maybe not so familiar experiences, you can go to the Discover feed, which is the global algorithm, with just one swipe and see what’s happening across the world.

We’re saying you don’t have to choose. Depending on how you feel at that moment, you can choose whether you want a smaller community experience or a larger community experience.

SAFIAN: It’s interesting because you’re saying the assumption that social media has to feed an echo chamber doesn’t necessarily have to work that way. It’s the way the algorithms and the experience are constructed that’s distancing me from people who are different from me.

WANG: Really interesting what the algorithms have been doing because it’s been done in a black box. We don’t know what’s happening. For the first time ever, Bluesky is kind of opening that box because we’ve provided something different, which is chronological feeds, as a completely different platform.

That’s the true A/B test that’s happening. What publishers have discovered, for example, is I believe The Guardian came onto Bluesky recently, and in their first week, they had 300,000 followers on Bluesky and they had more engagement in that first week on Bluesky than they did all of 2024 on X, even though they have 10.8 million followers on X. I think that tells you that the algorithm has some sort of manipulation. The whole point of other centralized social platforms like Meta and X is that they’re incentivized to lock you in and not show you what they’re doing.

That’s the whole point of the centralized authority. My analogy is it’s like moving to a city where, if you wanted to change the color of your door, you have to go and petition the city government. How is that possible that that’s how the online world works?

We don’t want that. I think ultimately, what we’re trying to provide is a toolbox for you to go and paint your door yellow or green or blue, or whatever you want, because that’s your prerogative. That’s the way that social should work online.

The future of Bluesky

SAFIAN: Back in 2007, when I first started running Fast Company, we put Mark Zuckerberg on his first magazine cover, and Facebook at that time had 19 million users, smaller than you are. How big do you want to be?

WANG: It’s a really hard question to answer because that’s assuming that Bluesky is just an app, and that we’re just the next Twitter alternative. So we can look at those numbers and say we’ll probably get a certain percentage of that. But that’s not what Bluesky is. We’re building the whole new social web.

We see Bluesky, the app, as the lobby to now explore so many other experiences. Imagine social has been closed off for the last 10 to 20 years where APIs have been closed and developers haven’t been able to build. For those of us who’ve been around since the 2000s, we probably remember Zynga and all these other experiences that were built alongside Facebook and other social platforms because developers were building really awesome experiences that the parent companies weren’t able to build because they were focused on something else.

It was the entire ecosystem of those games and other experiences that made it really fun for users. But as soon as Facebook then wanted to go into that industry, they would shut down that API and close down those dev companies that helped the ecosystem grow overnight. So we haven’t seen innovation in social in so many years.

One of the ways that we saw this was with Hinge. When Hinge first started, the dating app, they used Facebook’s API to see who your friends are friends with because that’s a really good proxy for maybe who you’d like to date.

As soon as Facebook tried to come out with their own Facebook dating app, they closed down that API, and Hinge no longer had access to the friends of friends API. Also, the Facebook dating app never really took off. So that entire service was just cut off from the world. But imagine if that was open. That’s how Bluesky is built, where anyone can go and build a friends of friends dating app, or take your Bluesky post, run an AI over, and tell you of anyone on Bluesky who’s your best match. How cool would that be? So for us, it’s, “Hey, there are so many other experiences that can be built on top of the protocol that we built at Proto.” In our atmosphere, we can serve anybody who’s online. So our market is as big as the internet. That’s the way we see it.

SAFIAN: It’s fascinating to hear Rose talk about the impact that a small group can have in our big-tech-dominated world, and how Bluesky sees itself as a facilitator of new habits and new ideas rather than just the next social media app.

After the break, we’ll dig into how Bluesky is responding to Australia’s new rules banning those under 16 from social apps, plus its approach to content moderation, AI data-scraping, and more. Stay with us.

[AD BREAK}

SAFIAN: Before the break, we heard Bluesky’s Rose Wang talk about how her small team is coping with a huge influx of new users. Now, Rose talks about why Bluesky’s CEO says they won’t, quote, “enshitify” the platform with ads. Plus: Bluesky’s response to Australia’s ban on social apps for those under 16, how the company is dealing with content moderation, its approach to AI data scraping, and more. Let’s jump back in.

Fostering positivity over growth

SAFIAN: Bluesky users have noted the positive vibe of the app. I know when I signed up, the trending topic of the day was public library appreciation, which was not something I was getting on my other feeds. Is positivity something you’re trying to encourage?

WANG: That’s a really cool, first of all, awesome trending topic. The thing that I think is important to point out is just that the first users of the Bluesky app were the Bluesky team.

It was very much an extension of who we are. We’re a team that’s led by two women who have been on the internet for decades and experienced the internet as women.

And so, I think prioritizing safety ahead of growth in some ways has been one of the reasons why we’re seeing a more positive environment on Bluesky. I think a lot of people know Bluesky from before this big wave this month as the app last year that had invite codes and wouldn’t let people in. We didn’t do that to be hot.

We did it because we weren’t ready on a moderation front. We weren’t ready for the federation front where people can have choice in feeds. So it was a decision of do we open up invites so when everyone comes in they get the same experience they had on Web 2, which just locks them in, doesn’t give them choice, and then makes them feel unsafe?

Or do we stick to our principles, build a strong moderation team, and also give users choice not only in feeds but also moderation services? So we chose the latter. We didn’t know that was the right decision because for many months, Threads came out of the woodwork. They started with a hundred million users and that’s really hard for a tiny start-up to look at and say, hey, I think we can beat that.

In hindsight, we’re very appreciative of the choice to stick to our principles. That’s partly why we have a more positive and safe environment, but also, it could have gone any way. This is how startups work. You write history as you build the plane, and I think if you can stick to your principles, then hopefully it works out. So far it has for us.

Bluesky’s innovative approach to content moderation

SAFIAN: Content moderation has been a flashpoint for many social apps. You recently expanded your pool of content moderators from 25 contractors to a hundred. What is your content moderation philosophy? How much is about weeding out bots versus enforcing certain kinds of community standards? How do you think about it?

WANG: Moderation is governance. We think about it as a checkbox to just get rid of bad actors, and that’s never going to create a safe environment if you don’t come at it proactively and think about moderation as a reactive step. So for us, the way that we’ve thought about governance is we’ve taken a note from how court systems work, which is you can’t just have one court system take in requests from hundreds of millions of people, which is basically how most other social media platforms work.

The way that Bluesky has envisioned moderation is what we call stackable moderation, where we’ve given a lot of the tools that have previously only been available to the moderation team, to users. What does that look like?

There are a lot of behaviors between intolerance, hate speech, misinformation, and users having a bad time. On most other social media platforms, there’s really nothing you can do about it except petition the moderation team.

Here on Bluesky, we’ve given people moderation labelers, where they can go and label rude posts, movie spoilers, political content, deepfakes. There are so many other types of unpleasant accounts that we want users to be able to adjudicate for themselves. Those are basically lower courts, and you can petition them.

The users run those labelers, and then we would resolve larger problems like if a labeler violates our terms of service, then we’ll take this service down or disconnect it from Bluesky. It’s kind of taking a page out of Reddit’s playbook of giving those tools to community members to govern their own spaces. But on top of Reddit’s manual tools, we’ve given more programmatic tools that are much more powerful.

Where Bluesky stands on Australia’s social media ban for children under 16

SAFIAN: I have to ask you where Bluesky stands on the recent ruling in Australia banning people under 16 from using social media. Your CEO and partner, Jay Graber, recently misspoke on UK radio saying your minimum signup age is 18 when in fact it’s 13. Where are you on this?

WANG: We understand that social has been harmful for a lot of teens. There definitely need to be solutions to make sure that bullying, harassment, and the type of content shown to them when they’re extremely vulnerable and still trying to figure out who they are is addressed. Is social media serving teens today?

No, and we recognize that. That said, there are many teens, like myself, when I grew up — I grew up in Nashville, Tennessee — there were very few Asian Americans in my community, and where I would go and find others to connect with would be online. I think there is a positive to what social does serve.

Social should still be available for them to connect. What we hope to do is, there’s actually a lot of innovation that hasn’t been done.

Can you create an app that’s mostly targeted towards younger folks on Bluesky? We have a firehose of data, and the positive types of data get fed to the more positive teen app. There are so many ways we’re excited to see experimentation in this area.

That said, we plan to comply with Australian law. We understand the reasoning behind what they’re doing. I think Bluesky as a solution, we’re saying, “Hey, rather than having different regulations in every city, state, or country, which becomes more state-controlled media, is there a way we can provide a marketplace where lots of different experiments are happening, and users are voting with their feet?” That’s a more democratic process in which we can evolve social.

Exploring alternative business models

SAFIAN: Your CEO, Jay, also said this week that the company wouldn’t “enshittify” its platform with ads. So the business model will be based on subscriptions.

WANG: Our first business model is subscriptions, where we won’t put core features like speech behind a paywall. There will be more features that help with self-expression, like custom avatar frames, higher resolution image uploads, and views.

Another business model we’re thinking about is observing what users are doing on the app where they’re finding a lot of value, and then amplifying that behavior rather than forcing a new behavior. One of the things we’re excited about is the Bluesky app is created by not just us. It’s created by lots of users — 50,000 different feeds created by users. These users are paying each other on Patreon or Ko-fi, supporting their livelihoods. We want to be able to build a payments network where we can help more transactions happen in more volumes, and we’ll take a percentage of that transaction.

We did not lock ourselves into an advertising business model, which is why Jay said that we can’t enshittify your experience with ads because you can just leave. We’ve open-sourced the Bluesky app so anyone can copy and paste the code and create Green Sky overnight.

Our incentives are aligned with the user, which is why I think most people are having a better experience on Bluesky because we’re ensuring that they are, or else we know they’ll leave.

The role of user data in AI training

SAFIAN: You’ve also sort of differentiated yourselves by saying that your user data won’t be mined to train AI. You’re not going to sell that data. User content is still in the open domain of the internet, right? Like it’s out there to be scraped. But you won’t intentionally sell it. Am I understanding that the right way?

WANG: Yeah, it’s a very difficult problem. So with data, we had to choose when we came out. Are we going to make our data private or public? Private is what Meta, Threads, and X have done — they own your data and they’re training their own proprietary AI models that they’re profiting from your data and your posts. We’ve chosen a completely different path that’s much more like Reddit, which is we’re open, we’re public. That does mean that any open source model, any AI person out there, can go and take that data and train. However, we’re taking a page out of the web, where every web page can be scraped by developers.

What we’ve done on the web, though, is that we have allowed different websites to express consent. “We want our data to be scraped,” or “We don’t want our data to be scraped.” It’s still up to the developer to respect that expression, but at least the ability to express consent has been given. So we’re going to follow that on Bluesky.

We’re going to implement a feature where you can express whether you want your data to be trained on or not. It’s up to the developers and AI companies to follow that. There are lots of interesting open-source models that academics are training on to help better moderation systems be built to help safety across the web.

Bluesky’s data in that way is actually quite helpful for those academics. We always ask, “Please look at this with nuance. It’s a very complicated issue.” Bluesky has promised as a company, we’re not going to be profiting off of your data for generative AI models because that’s not the business we’re in.

We’re trying to build a new layer of the internet for social. That’s our directive. That’s our promise. But we still want to go and make partnerships with AI companies to set a precedent to show, “Hey, you don’t have to do this. People have expressed consent. There’s tons of data out there.”

The prioritization needed to guide Bluesky through this phase

SAFIAN: It’s got to be hard handling the current growth and the interest from people like me. At the same time, the opportunity for the future is suddenly way different. How do you balance all this? How do you prioritize?

WANG: That’s the key word, Bob, prioritize. In the midst of a million opportunities and so many people to talk to, we have to be very clear on who we’re serving and why. What’s been very clear for our team the last few weeks is that Bluesky is a home for independent creators, journalists, artists, publishers, and entrepreneurs who want to build a presence online. There have been probably dark patterns happening for the last 10 to 20 years that we haven’t understood. Most platforms are incentivized to lock people in and not link out. If you post a link, you’re probably going to get demoted in the algorithm.

Hank Green actually posted a YouTube video about this, where he did an experiment of posting on Threads, X, and Bluesky the same posts, and then some posts with certain words, some posts with links, and no links. It was kind of wild what he saw. Like on Threads, when he posted a link, he’d maybe get 60 likes.

On Bluesky, he gets a thousand likes. That’s actually kind of wild how different that engagement metric is. Something’s happening. We do know people want to build a livelihood online. They want to build their own brand that they can control. They want to own the relationship with their community.

Bluesky is trying to make that easier for everyone new who’s trying to do that today.

How Bluesky believes they will grow

SAFIAN: And your growth from here, I mean, we’ve talked about this. The big bursts have come from, in some ways, these external events. To what extent do you expect that sort of news and events will conspire in your favor going forward versus active steps, aside from just talking to me, that you’re preparing to sort of expand the brand?

WANG: Yeah. We want discussions that are happening now to happen on Bluesky. When South Korea called martial law, where was that breaking news unfolding? It was happening on Bluesky. It was really hard to follow on Threads. It was really hard to follow on X.

Many people, we were hearing from reporters and academics, were following the story on Bluesky. We don’t have the budget to go and win over every celebrity and every big brand. Our belief is if we could just focus on what the end user wants and needs, then everyone else will come.

We’re in a really interesting time in history where democracy is being questioned in general. Like, is this a process by which governance should work? Bluesky, as an online solution for society, is a more democratic way to build online communities.

That is the fight that we’re in. That’s what we believe is at stake, which is why sleep is a fifth priority because we all very much believe that democracy needs to continue to uphold and we can only show, we can’t tell.

SAFIAN: Well, Rose, this has been great. Thank you for giving up some of your sleep to talk with us today. I really appreciate it.

WANG: Thank you so much for having me, Bob.

Politics aside, what the small team at Bluesky has achieved is remarkable. Coming away from my conversation with Rose, I was struck by the data points she shared about the site’s user engagement. If that’s really true, I’m sure plenty of notable figures from the business world and beyond will continue to flock to the platform.

Like a lot of people, I’ve been quietly concerned about the drastic fragmentation of social media and how echo chambers are only growing. Rose points out that the ideal of a town square wasn’t serving anyone anyway — no one’s minds were being changed by being shouted down. I really hope that what Rose says is true, that creating safer community spaces will lead to a more comfortable environment for people to listen and eventually engage with opposing viewpoints. Only time will tell. I’m Bob Safian. Thanks for listening.