Nieman Foundation at Harvard
HOME
          
LATEST STORY
From shrimp Jesus to fake self-portraits, AI-generated images have become the latest form of social media spam
ABOUT                    SUBSCRIBE
April 12, 2021, 11:09 a.m.
Business Models

“Maybe the kind of reform that we want comes from creators being like, ‘I’m done'”

Charlie Warzel on newsletters, platforms, reporting, editing, and luck.

Brad Esposito got his start as a reporter at BuzzFeed in 2013 and is currently the director of content at Eucalyptus, an Australian health tech company. A version of this interview first ran in his interview newsletter, Very Fine Day, which he created to “give more depth and context to the people who keep the internet humming.” Subscribe here.

The New York Times’ Ben Smith reported on Sunday that Charlie Warzel has moved to Substack, but when this interview was conducted on March 27, he was still a writer-at-large for the Times’ Opinion section. He writes a lot about technology, platforms, and the meeting point between the physical and digital worlds. You can subscribe to his new newsletter, Galaxy Brain, here. He also has a blog (a Substack) about dogs.

Brad Esposito: All right, Charlie. Where are you right now?

Charlie Warzel: I’m in Missoula, Montana, which is where I moved with my partner Annie [writer Anne Helen Petersen] in 2017.

Esposito: Why’d you do that?

Warzel: Ha.

Esposito: Not that I don’t …

Warzel: It’s a good question.

Esposito: I mean… I do think it’s probably nice there!

Warzel: She was writing a lot about the Mountain West region and covering the politics in that area in the post-Trump election, and there’s lots of good interesting stories there. She’s kind of from this region of the country. And I joined her, tagging along on a reporting trip, and we came out here and it was the start of the summer and it was just gorgeous. And New York City was just hot and smelling like garbage and the subway was broken all the time. I’d always lived on the East Coast, and I was looking for an adventure.

Everyone at BuzzFeed just, like, kind of let us do it. We did not think it was gonna happen. And we mentioned trying to do it and within a month we had moved.

That’s one thing I just don’t understand: So many newsrooms are so weird about where you live. But I think it only behooves every institution to have more people spread out, especially if that’s what they want — like, you’re gonna get a happier worker, too.

Esposito: Did you get into it straight away, though? Or did it take you a few weeks to figure out?

Warzel: It’s almost like you’re setting up the premise of this book that we’re working on …

We just wrote a book on the future of work and remote work. It’s mostly not about us at all, but the germ of it came because when we moved Annie had a pretty decent transition to the work from remote work thing. For me, it was just horrible. All I did is I worked 16 hours a day, seven days a week. I was getting cold sweats on the couch. It was very bad. And then I slowly figured out how to do it, basically.

You have to actually “work from home.” You have to go to a room in your house, you know, do all that crap.

But anyway, we noticed when the pandemic hit and everyone was going through this that [other people who were now working from home] were just replaying that eight months of my life before I figured out what was going on. How do you actually find balance and do the thing where you, say, work out in the middle of the day, and all that kind of jazz?

Esposito: Yeah, I have multiple health conditions — well, not health conditions, but things as a result of working from home, and in Australia it wasn’t even as bad as in the U.S., it was probably three to four months of working from home. It has never made me feel more old than going into the [physical therapist] and they’re like, yeah, your glutes don’t activate anymore, which is why your hamstrings keep falling apart.

And then I did my neck few weeks ago, and they were like, we’re getting a lot of these from everyone working from home and looking at their computer.

Warzel: Oh, yeah. It’s crazy. I mean, do you remember the thing a while back — I can’t remember if it’s true or not — where people were getting a deformed part of their hands from just playing Candy Crush?

Esposito: You said you’re from the East Coast. Have you been in New York always?

Warzel: Yeah, I was in New York. After college, I went to DC and I worked in NBC News as a researcher on the TV side. Cable and stuff in the news division. I was basically a glorified production assistant. I got a lot of coffee and did a lot of research.

I remember being at NBC in 2010 and being like, I want to blog for you guys! As like a young person, you know, covering DC with some young person lens or whatever! And they were like, no.

Then I got into covering technology because I was covering media companies, like digital media, for this advertising magazine AdWeek.

And from there, I met a lot of BuzzFeed people, and then got hired there. And because I was doing digital media, they’re like, write about technology. So they threw me in that bucket. And then from there, the interest was these big platforms and how the platforms are influencing the way that we behave. They’re just changing social and cultural dynamics so much. And I think that there were certainly plenty of people writing about that, but it was still really small and historically siloed. I didn’t think it was the next big thing, necessarily, it was just really fascinating.

The first thing I wrote through this lens was about why Facebook doesn’t let you see how many people saw your post. They [only] show you how many people liked it. Basically, it’s because the metrics are super not in your favor. 100,000 people saw your post, nine people liked it. Nobody wants people to know that.

[My story] really pissed Facebook off, and they responded to it. This was 2013, when the stakes of everything were just totally different than they are now, but this was a thing that really made them mad. I ended up talking with the head engineer of News Feed. In that conversation, which I think at the time was off the record, they were telling me how they think about what they prioritize.

I started to realize: There’s so much more to this. Facebook is involved in psychology and all these different kinds of patterns and analysis of behavior to try to provoke these types of things. It was sort of the first understanding for me that these platforms aren’t neutral. They are pushing you in these directions, and they’re being engineered to do that. I was like, OK, so then what is happening as a result of a lot of these opaque decisions? Let’s see how people are behaving. And it was like: Oh, women and people of color are constantly getting harassed on and off the internet. Why is that?

And so I tried to understand the dynamics at play. And then it became, like … I got obsessed with their rules, too.

Esposito: Very cool.

Warzel: These platforms, they say what they’re trying to accomplish right here. This is what’s allowed and not allowed. And they never enforce any of it.

So it was like: OK, well, let’s find some examples and just bring it to them and try to get them to say what’s going on. And that sort of became its own — not just via me, but in general, other people, alongside me, were doing that. And it became this cottage industry.

But there was this direct line from all the harassment stuff to the behaviors of these nascent, ironic communities that started to get populated with worse and worse people. So this is really a super long way of saying: I stumbled into this. Because I was like: Oh, these platforms are making us act a certain way.

And it was kind of almost a direct line from that, to harassment to Gamergate to the rise of the alt right, and then 2015 4chan, Trump and the Pepe stuff, into the rise of the pro-Trump media …

Esposito: Do you … how much do you think that the algorithms and platforms affect people? I guess my question is: How much of it is just people being people? And we just didn’t know that people were just kind of dicks?

Warzel: I don’t think we’re ever going to be able to fully untangle it. I think to some degree there is just natural human behavior, obviously. But I don’t think we can discount how totally unnatural this level of connection is.

Something snapped in my brain in, like, 2017, 2018. I remember reading about the cottage industry of misinformation in the U.S. — people who were just creating totally fake news sites. It was [after] the Macedonian trolls thing. It was these stories about boomers who were just on their computers all the time and were being destroyed by garbage news.

I was just like: I don’t think we should be connected at this scale. And at this intensity.

And I don’t think I’ve recovered from thinking that. And it’s not to say it’s all bad, and I hate painting any of this with a super broad brush, but there’s just something wild to this. And I’ve tried to explore it a little in the past couple of years with the reporting myself and my colleague, Stuart Thompson, have done.

We did this piece on Covid misinformation and Stop The Steal. We went back through people’s Facebook histories and it was just people using the platform, trying to do whatever they could to catch some virality, with nothing happening. Then one day, they’re just like: I think Covid is a hoax. And it was, like, so much engagement. They were just off to the races and insane.

I just think those incentives are always there to some degree. Like, let’s go back to like Alex Jones in the analog days, right.

He realized at one point that showing up at the Texas state legislator press conferences with a bullhorn was really advantageous and helped him become a local weirdo and he got some popularity. But the scale of this is just so mind boggling. We — myself included — still don’t really stop and think about Facebook in the sense that there’s never been anything this massively connecting so many people.

And I don’t think our brains can wrap our heads around it, to some degree, because it is preposterous and it’s so new.

So, yeah. I mean, I think people are dicks. But I also think we have the ability to tap into such large reserves of attention in ways that were basically unfathomable 15 years ago, and that’s tough to discount.

Esposito: Yeah, I definitely feel like people don’t recognize this kind of Industrial Revolution–era moment.

I feel like I spend a lot of my time explaining to people who complain about their parents online, or boomers online is another great example — people struggling to make that jump to online social living. Like, why don’t they understand? And it’s just like: Well, they … can’t.

There’s a level to it where they really are just not born into understanding in the way that you, a 20-something-year-old, are. And because of that they are so susceptible to a lot of the worst things going on.

Warzel: I think it can maybe go further, too, right? Even the 18-year-old who’s platform native to everything and has an intuitive understanding of the dynamics and the behavior of these networks and how to use them really well — I don’t think that even those individuals understand what’s happening when you’re putting this stuff out there. I don’t mean from a technical standpoint, but the size of potential audience that you can court and the ways in which your message can be distorted or manipulated or taken out of context.

I just think a lot of times we open up all these different platforms and see it as a fun game — like, let’s try to like ride the wave. And I don’t think most people know what it actually feels like when you catch the wave. And a lot of people realize that and are like: Oh shit!

But how could you expect someone to understand? I think if we really, truly, with our feeble brains, could understand what is happening when we’re about to hit send on a tweet or post or upload a video or something — I think if we really could understand the power there, we’d all just be like: Man, I’m out.

Esposito: Do you think it’s inevitable that that’ll happen? And like soonish, in the next 10 years?

Warzel: I don’t know. I feel like I’m just starting to have some brain space opening up, post the last five or six years. And I’m trying to think about questions like this, about the internet, in new ways.

I don’t think I had a new or interesting thought about the internet basically since Trump came down the escalator, right? It was just like running around with a fire extinguisher, putting out fires, being like: what was that thing? How did they do this? It wasn’t thinking through new consequences, or the evolution of something, or how it’s going to go. It was all just responding to something.

I’m really fascinated right now with the idea of what are we going to do with creators, broadly as a category. Are the lives of anyone who wants to create things on the internet just going to be increasingly financialized and sliced and diced? And you basically have to sell every decision of yours to some crowdfunded platform, or turn yourself into cryptocurrency or whatever? Is it going to be weird stuff like that?

Or is it going to be this really interesting labor movement, where creators realize their power and exert authority over platforms, in unions, or come together in collectives? Or governments might have to declare, like, “Posters’ Unions.”

Esposito: Oh, I’d be getting that shirt so quickly.

Warzel: Right now these platforms have so much control, just by how they can tweak an algorithm or a set of policies that changes how someone gets paid. In some ways that’s like taxation without representation. Maybe the kind of reform that we want for certain parts of the internet comes from creators being like: OK, well, I’m done. Or: I’m not going there. Or: You need to disclose how this part of the algorithm works so that I’m not left in the cold when you change something.

I don’t know. It all depends on how crazy politics in the rest of the world gets. But I think it’s gonna be a really interesting next decade for the maturation of the internet.

Esposito: I thought recently — and you don’t have to comment on this, given it [was] your employer — but a recently The New York Times told its staff: No newsletters for any of you, please. And my first thought was, like: Everyone should just … leave?

That power struggle is really coming to a point, it feels like. Or maybe it’s at the foot of the hill. But that feels like something we’re about to face as well: if we localize it to media and the creator economy as reporters and writers.

Warzel: For my own safety, I won’t comment on the newsletter thing. But I do think that broadly in media it is a super interesting moment, especially as a lot of people are waking up to their own ability to really quantify their own value as a creator of content, and then also watching other people wake up to their true value.

I think the whole thing is incredibly uncomfortable, because what you realize — which is the way that most places work — is that certain individuals’ work always subsidizes the work of others. And it’s never fair. I think that a lot of organizations — and this isn’t even just media, this is broadly any industry — make it so you don’t really know and you can’t really see that kind of thing.

I think it’s just super-uncomfortable for lots of people. And I think it’s going to cause a fair amount of turmoil, but also some really cool stuff, too.

Esposito: Yeah, well, especially because it’s like … I don’t think it is a good indication of your individual value as a creator. But you still have to do a bunch of things before you get to that point, like generate a Twitter following or put out a lot of content for free on YouTube and hope people follow you there, before you even have this value.

So you’re tied to these other platforms, and you’re relying on them to give you that value. The more I talk about, and think about, these things, I’m just like, oh my God.

Warzel: No, it’s wild. And some of it is just so random, too. If you look at some of the Substack stuff — some of the people who are able to generate really big followings — you see a number of people who came from the blogosphere, really prominent early bloggers. Those people also stumbled into blogging at a really good time, right?

Like, I came into “being a professional online” and I got the very tail end of blogging. I couldn’t put my foot in there, but those people were able to.

Then I look at when I got on Twitter, which was sort of near the beginnings of Twitter or at least the beginnings of when media people started getting on Twitter. As a result, I got in at the ground floor and I was able to build up a big enough following.

A lot of this stuff is based off of luck — of when you hit that one platform that sort of gives you the ability to then leverage these other ones down the road. The people who had the big blogs, they then went to Twitter, they got a lot of Twitter followers, and then they’re able to use that Twitter following and the blog thing to get good jobs. And then when they want to go independent, they can do that, because they have those followings from porting them over.

It’s like this game of like, I don’t know…

Esposito: Catch up.

Warzel: Yeah.

Esposito: I think that it’ll be interesting … The way I think it’ll play out is: You’ll have all these individual creators finding their value, making their own newsletters, and making a lot of money. And then you’ll get a handful of them with big enough egos, or big enough ideas for the future, who are like: You know, I can put a bunch of these newsletters together, or I can get a bunch of these writers together.

And then you’ll start — in the same way that the blog boom created a bunch of new media websites — we will have another boom, a bit further on.

But the interesting part of that will be individuals who have all the money for the first time, rather than being funded by multinational corporations or venture capitalists. They’ll have to make the decision, like, OK, I’m deliberately going to go from having this money just to myself and maybe an editor, and being ethical, maybe, about what I build out. But I just see a whole bunch of messy, broken, cancellation moments of internet darlings when they try to transition away.

Warzel: Well, there’s also a really interesting, sustainable potential part there. Look at the folks over at Defector, who are creating this community-owned publication, because of the Deadspin stuff they know an audience is there. And also because they’re really good at what they do.

I saw one of them talking about the fallout that was happening at Medium. And they were like: The great thing about this is we’re not trying to grow it to some massive scale. We’re trying to do a good job and have good-paying jobs for us and the people that we can afford to employ.

I was just talking to this venture capitalist named Li Jin, and she wrote this piece about the creator middle class and how there’s no middle class on the internet, it’s just the top and the bottom. The Defector thing seems to me like a potential path toward [a middle class].

If you have a lot of people via a blog, or Substack, or whatever sort of individual thing is moving away from big publications, and then if they condense again, but they do it in a sustainable way, maybe that’s … to me that feels like a better future than all venture-backed media.

Esposito: Well, venture-backed media has worked so well, I can’t see why we wouldn’t keep doing it.

Warzel: Oh yeah, it’s fantastic stuff all around.

Esposito: So, do you think we can fix the mess we’re in?

Warzel: Which mess?

Esposito: Exactly, yeah, which mess! But probably like: platforms and the social internet and how, like you were saying before, partly because no one really knows how to deal with it or how much Facebook was kind of controlling their decisions, but can we actually train people to understand that?

And can we actually change the platforms themselves to be better? Or is it a matter of hoping competition comes in and Facebook doesn’t buy them for a billion dollars or copy them so that it creates some sort of diverse environment online?

Warzel: With some of the speech issues, a lot of these discussions start out about content moderation stuff: Like, that has to get fixed, and it does, to some degree.

But like, we don’t even have agreed-upon language for what’s going on here. We talk about it as issues of speech. But in reality, a lot of them are really issues of reach, and of amplification, and distribution, and access to large pools of attention.

It’s part of the reason — not all, but part — of why you have both sides of the aisle in Congress equally mad at the same people for completely different reasons. And not only are they mad for different reasons, they can’t even have a conversation about them, because there isn’t good language to articulate some of the problems yet. I think that’s a big impediment, but that will change.

When I think about fixing stuff, there’s a couple things. I think a lot is going to be really difficult to change while Mark Zuckerberg is still …

Esposito: God?

Warzel: In total control of Facebook.

I think that is just a real issue. And I don’t see that changing. I think it’s a much bigger roadblock than most people who are just kind of thinking about this realize.

I get the sense that a lot of these issues that seem super intractable are probably going to get fixed in the most boring ways, or change in boring ways. Like I think that certain data privacy laws or regulations, even just in terms of what companies need to disclose, or how they need to rewrite their policies, or how they need to show users what kind of information is being collected, or whatever that is, however that works out — I think that will happen over time.

Things that could possibly solve some really big and heavy problems: It could be like forced disclosures from platforms about the ways that their algorithms work in sneaky ways. And then that could ameliorate a lot of content moderation problems that feel so thorny right now.

Or, I think that some of the antitrust stuff always gets put under this umbrella of like: OK, are we gonna break WhatsApp and Instagram away from Facebook and turn it into several companies? And I don’t think that really solves it necessarily. But there’s so much more, you know, nitty gritty. There’s so many other potential solutions there.

And this is not really my expertise, so I don’t feel comfortable coming into it, but I think there’s a chance that the FTC can come in and start making some rules that make life miserable for these companies and start solving some of these issues on the margins. But they’re so deeply unsexy. It’s not just going to be like: We’re going to stamp down this rule that forever changes speech on the internet. I just don’t think that’s gonna happen that way.

Esposito: What do you think about when you’re trying to put together something to write? Assuming you need to write something once a week, do you just spend a lot of time looking at the sky just thinking about the internet?

Warzel: Ha. It’s … I think it’s harder and harder. The ultimate thing that I’m always trying to reach for, at least lately, is like, a big idea about something and then the actual human being that represents.

Take this idea of algorithmic incentives that are causing people to become election deniers almost overnight. Like: who’s the guy, right? Where can we see it? How can you figure that out? Who’s a person who’s having that happen to them?

I feel like a lot of stories right now have reached saturation. It doesn’t mean it’s over, or that they’re not important. It’s just like: You’ve told the same story quite a bit.

So I’m always trying to find a way around that corner, if that’s possible. And oftentimes, it’s not. You just kind of end up writing the next chapter in that story, or the sequel, as opposed to a whole new piece of intellectual property.

I only write good stories when I follow my interests. And there’s plenty of these stories that happen where I just kind of commit myself to something because I’m struggling to find something. I always look at my stories at the end of the year, what was good, what was bad, and there are always those ones where I’m like, oh, yeah, I remember I halfheartedly pitched that and three or four hours of researching or reporting in I was like, I don’t really like this. If I get excited about it, or think that it’s really fascinating, and feel propulsive toward it, it’s always going to be better.

Esposito: How do you then convince your editor, or is that quite easy?

Warzel: I have an interesting relationship with my editor because, unlike a lot of editors that I’ve had, he is not just straight out of the world of tech journalism. He covered wars in the Middle East and is, like, really steeped in history. He comes at all this as a really interested and engaged person, but isn’t necessarily obsessed with whatever tech Twitter is talking about.

So there’s more convincing that I’ve had to do, but usually that makes it stronger. And I will say I probably get waved off a little bit more on certain things. And then I will kind of come back to it, often, and it will be stronger.

A good example is this column I just wrote last week about NFTs. Originally I was trying to tie 30 things together and just, like, make fun of them. And he was just sort of like: This is a bit incomprehensible. I don’t know what you’re trying to do.

It’s possible that someone who was just hyper, you know, engaged in the internet conversation online would have just been like, OK, yeah, go with it! And I would have turned out something. Instead, he was sort of like, you’re really gonna have to come over here and do a little more work. And that made it better.

But normally, put it this way: My tastes seem to be pretty normie mainstream. If I’m really interested in something it tends to be that story does well, or if it aligns interest wise… I would say I don’t have super esoteric tastes, in terms of stories. For whatever reason. So, I’ve learned to, like, trust that instinct a little bit, if I think something’s gonna be clever, or fun or interesting. And I think over time, editors have learned to sort of trust that a little bit, too. But yeah, every case is different. And I’m very grateful.

There’s been a lot of conversations about editors lately. But I’m super grateful when someone tells me the idea is bad. You want the gatekeeper to be like, I don’t understand what you’re talking about.

Esposito: Is that how you think about failure and getting things wrong? An editor is a good example of that, I guess, in stopping you from going down that path. But how do you approach, in your professional life or in your personal life, making mistakes and then correcting them?

Warzel: I screwed something up in that NFT column in the second sentence. This happens a lot with writing about Internet stuff: You talk to somebody and then you’re just slowly trying to bring that down to something your mom can understand. And in the process this time, I messed it up and somebody noted it. And they pointed it out in a way that was sort of like: oh, this is a huge bummer. This is a pretty bad, glaring correctable error in the second sentence of the piece about the fundamental thing.

It’s not like I got the piece wrong, but it was a notable thing for people who understand and care about this technology. And so I was like … I mean, I felt like shit, and really embarrassed, but publicly, I was like, walk me through what’s wrong. I asked them to also direct message me and I was like, OK, I just want to be clear: This is what you saw to be wrong?

Then I went to my editor and I was like, this is a kind of an in the weeds and nerdy correction, but it has to be done, obviously. And then we did it. And then I quote tweeted that person’s tweet. It was a good faith mistake … it’s really hard, but admitting as early on as possible when you’ve when you’ve screwed up is good.

Ben Smith told me one time during the Facebook vs. Alex Jones “are they gonna ban him?” discussions, they’re totally going to end up banning him at some point, because he’s gonna cross the line. He was like, first rule of crisis PR: You figure out what you’re going to have to do eventually, you just [figure out] as many steps there as possible, and you shorten the length of the crisis.

I feel that way about screwing up. What’s going to happen here, right? Like, do I see that tweet and ignore it and hope it goes away? And then somebody else brings it up, or it gets put in some crypto forum? Or I nip this in the bud. It was corrected within an hour of it going up. Most people who came into that article saw a sentence that was correct. And the correction on the bottom.

Admit to it sooner. We’re all gonna screw up. That kind of thing is a small issue, and almost no one except your worst faith critics will care. I’m much more worried about getting something wrong on a structural level, at the article level, or like, even just the conceptual level. With that kind of thing, that’s where editors are really helpful.

I often try to talk to people on a subject who are coming at it from a completely different angle. You definitely want to hear the best case against something you think. I think of those as background calls: Like, I’ve done enough reporting, I believe that it’s this, but … if they can convince me of even a part of a point of it, then I include it in the piece.

Esposito: Last question. This is a from a subscriber, because I told them I was gonna talk to you. What’s with the bear on Twitter?

Warzel:I was in the BuzzFeed offices, around 2014, and I had the occasion to Google image search “cool bear,” and that bear with 80s sunglasses just showed up. And it was this really cool Native American artist who made these things, and I don’t think I really realized it at the time. [The artist is Michael Meissner and his website is here.] At the time, I was just, like, I like this bear with the 80s sunglasses.

I’ve decided I’m never gonna change this. I always get nervous when people change their avatars. But awhile back, I basically asked myself your subscriber’s question, what’s with the bear? And I went, and I looked it up, and this person … they’re in Montana, which is really cool.

They live on the far side of the state. I emailed their website a number of times, because you can actually get the bear as a print. I really want to get it, and I want to pay extra money, basically licensing the image, because I’ve used it for free for so long.

But I have tried many ways to try to get in contact with this person. And they haven’t responded, but none of the emails are going to return or anything. So I think they’re kind of older. I hope that they’re OK, and they’re just living an internet-free life.

Image by Charlie Warzel.

POSTED     April 12, 2021, 11:09 a.m.
SEE MORE ON Business Models
Show tags
 
Join the 60,000 who get the freshest future-of-journalism news in our daily email.
From shrimp Jesus to fake self-portraits, AI-generated images have become the latest form of social media spam
Within days of visiting the pages — and without commenting on, liking, or following any of the material — Facebook’s algorithm recommended reams of other AI-generated content.
What journalists and independent creators can learn from each other
“The question is not about the topics but how you approach the topics.”
Deepfake detection improves when using algorithms that are more aware of demographic diversity
“Our research addresses deepfake detection algorithms’ fairness, rather than just attempting to balance the data. It offers a new approach to algorithm design that considers demographic fairness as a core aspect.”