
The YouTuber does a “reverse OpenAI.”
Loading summary
A
Good morning, John. Remember when vlogs were just vlogs? When there were just people talking to the camera, Almost like a diary entry, but you were just talking sometimes just what was going on in their life. It's like a video, almost like a diary entry, like a video blog.
B
If you know the name Hank Green, that is probably the context in which you know him talking on the Internet as one half of vlogbrothers with his brother John, or as someone explaining science on YouTube, or as a guy who's written some novels and hosted Crash Course and scishow and started vidcon. There is a lot more. The list, honestly, goes on and on. But I wanted to call him up to talk about his company complexly and AI and our very weird information environment. And we did this interview while I had a cold and Hank was sick with shingles. Hank Green, thanks for coming on the show.
A
Hey,
B
you're such a good sport. You have shingles and you're soldiering through this?
A
I put on my nicest fun shirt just to convince my. I need to be distracted. Look, I could tell you all about the progression of the shingles virus now, though, and why it evolved the way that it did. It's very old. We've had the virus that causes chickenpox since as long as there have been humans.
B
Wow.
A
And it is in no way connected to chickens. They did not do this to us.
B
No. I don't know why it has that name. Varicella. We'll just call it Varicella.
A
Wow. Look at you knowing stuff.
B
I got kids. Oh, kid.
A
Yeah. Hi.
B
I had this whole intro written out for you. I was gonna start the show, and I was gonna have you introduce yourself, yada, and then you dropped a video that blew up my plan, and I want to start with that instead.
A
Okay. You may have seen this article or this news story about this tech guy who saved his dog's life by engineering a cancer vaccine specifically for the dog. And now the dog's tumors are shrinking. This is very cool. It is also very misunderstood.
B
In this video, Hank goes on to explain how this dog owner and tech guy used AI the science of how this likely worked, and why it's sort of applicable to humans and sort of not. We are more complex than dogs. Our cancers are vastly more complex. So the thing about this that I also found fascinating, and I think your video did a really nice job of explaining this, which was like, there's one narrative that's like, dude, cures dog's cancer using A.I. hey, we should all do that, and it just missed every single slice of nuance and science. And the fact that there are MNRA vaccines for pancreatic cancer, and there's.
A
And there's all this, like, all of this is like resting upon a huge amount of oncology that's only like 10, 20 years old.
B
Right.
A
You know, the fact that we can do sequencing this fast and this cheap, very, very new. Like, I was alive when the first. When the Human Genome Project was like a twinkle in Al Gore's eye or whatever.
B
So what do you think this episode says about how people get and receive information on the Internet in 2026?
A
I think it says the same thing that everything else on the Internet says, which is that we think that we live in the attention age. You know, but, like, what that actually means is, like, one layer below that is when you're creating both when you're creating the content and when the content is being selected for, like, whether or not you see it, if that's both the people creating that content, which includes me, or the algorithms deciding what to recommend, it all comes down to a very fairly simple idea of salience, which is the word that we have for how easy is it to pay attention to something? And so, like, if a dash cam video comes on your screen and it holds your attention for 15 seconds because you know that there's going to be a car crash soon, because otherwise, why would it be showing up in your screen? And, you know, who doesn't like watching a car crash? I guess you're not necessarily going to feel proud about it later, but it's going to hold your attention. And so that's going to be the thing that comes up on your feed. And so in the same way, like, there are just things that are easier and harder to pay attention to. And one thing that makes something really easy to pay attention to is when it's like confirming all of your priors, it's saying, like, yeah, everything's messed up. It doesn't have to be this way. We could be healthy if it weren't for the nefarious forces in our way. Whether those forces are bureaucratic piling on itself or deep, dark, nefarious big pharma companies hiding the cure for cancer. It's all really. It comes down to two things. One is the visceral, which it can be like, cute. Like cute is also visceral. Or it could be like car crash videos. And the other thing is stories. And, like, what stories do we believe? What stories do we immediately fall into and want to know what's gonna happen next, which is like this is why true crime is so powerful, because it's both visceral and story and you get to be a part of that story in a weird way. Or you self insert into the story, you imagine how you would handle the situation, etceter. And also it like tells you, tells you a story about how scary the world is. And of course fear is, is the ultimate salience engine when you feel that you may, may be under threat, but so is hope, you know, And I think that this also tapped into that.
B
Today on the show we are going to get complex about YouTube, AI education and the Trump administration with Hank Green. I'm Lizzie o' Leary and you're listening to what next tbd, a show about technology, power and how the future will be determined. Stick around. So let's talk about the salience engine and storytelling platform on which you primarily exist, which is YouTube.
A
Exist is a strong word, but yeah, I get what you mean.
B
Fair enough. On the one hand you can find anything on YouTube, you can fix your sink, you can learn about the Roman Empire, but also there's almost no oversight from an editorial standpoint on the platform. And pretty much anyone can say anything right or wrong. And I wonder when you think about your work, your company after all, which we will talk about is complexly right. You don't want to squish down to the shortest version.
A
No. Yeah.
B
How do you think about it?
A
I get criticized for this a fair amount that I like. You're like, you don't have to make this so complicated, Hank. And I'm like, there's lots of other people who will give you the simple story. If you want that, go get it. But like, it's not, it never is. And like, maybe the simple story is important if you want to or if you're like doing politics and you don't want to complexify, you want to make a story that people will understand so that they will do the thing at the ballot box that is better for the world from your perspective. But that's not what I'm. That's not what I'm here for. And I remember thinking that YouTube and social media were going to be such clear goods and they were going to bring people together and they were going to help us understand each other better. And it was all going to be about if I was going to be able to understand the suffering of someone 20,000 miles away that might not be on earth as, as easily as someone, you know, my next door neighbor, then like, that would be a Good thing, right? But the. That's not what happened. Like, it's really deeply super not what happened. And I think it's important to sort of like, look at that. But the bigger thing is I, it, like, it was really sort of like challenging to my identity where I was like, I'm here trying to do this thing and I've contributed a lot of my attention and efforts to this thing succeed and big and bringing people and people follow me to new platforms. And so I feel that conflict inside of me. But I feel like what I've kind of come to is this isn't like smoking or non smoking. The Internet isn't like, hey, you're going to come here and you're going to get poison injected into you or you're going to go out into the world and not. It's like food where like, information can be good and food can be like you need information and you need food, and yet some food is not good. And in fact, you might imagine like a particular kind of food that's like, very easy to, not like very hard to ignore and is always there in your mind. And it's like I kind of imagine the sort of current state of Internet information as if, as if, like, it is. There's just like a lot of Doritos, except that in order to eat a Dorito, all you had to do was look at it and you never got full. And also there's a man in your pocket who's always got more Doritos and he's just like. And he's sort of like, hey, you want a Dorito? Look at, look right here and you'll be eating them. I feel a little bit like that.
B
I feel that must keep you up at night.
A
We are junk fooded. You know what? I. It, it doesn't really. No, the pain keeps me up at night.
B
Okay, fair. When you don't have shinks.
A
No, it doesn't. It doesn't keep me up at night because I don't feel like I'm making Doritos. Yeah, I feel like I'm. I don't know that I'm making like, you know, like the best superfoods ever. But like, sometimes I am. Sometimes I really do like to make a superfood. I make a thing where I'm like, if you watch this video from start to finish, it's gonna be 20 minutes and you're going to have a much better understanding of how your world works. And that's just good. That's just good work. Now a lot of Times I'm making stuff that's not quite like that. But I think that usually what I'm making is pro social. And I watch a lot of people in my business make stuff that I think is specifically antisocial. I think that it is bad for the people who watch it. It makes them less successful humans. And. And I understand why they do it. And I don't think that they're evil. I just think that they are following their incentives. But I don't do that. I really don't think I do. I think sometimes I fall into it, especially on a Twitter like platform. I fall into being a jerk. I fall into the tropes of the platform and I regret it later. That absolutely happens.
B
I mean, who doesn't?
A
Yeah.
B
I want to talk about what Complexly has just done. It's what I would call a reverse OpenAI.
A
Yeah, we did the reverse OpenAI.
B
You did the reverse OpenAI. You went from for profit into a nonprofit. Why?
A
Oh, so many different reasons. John and I started Complexly with the goal. This is an educational media company, and the goal was to help people understand their world better, to lower barriers to understanding. And there's a lot of different barriers to understanding. And part of it is like, you know, one way to lower a barrier is just like a really good explanation, like a really compelling video that holds your attention and teaches you things in a way that you hadn't heard before. But there's also like a bunch of other, like, cost structures to education. And there's lots of educational media companies. Most of them are based on by, like, selling the content and putting it behind paywalls of various kinds. And we really didn't want to do that. And we kept feeling the business pull to do things for the. Not the people who are our actual customers, but to do things for the business case. So, like, don't do.
B
Can you give me an example of that?
A
Yes. Anybody who had, like any normal educational media company that had Crash Course, which is a series of videos that are used in pretty much every high school in America, lots of colleges, definitely every school district in America, we've got that in. What would you do with that? Every person I ever said that to in the education industry said the same thing. Like some version of create a bunch of additional products around it and put that behind a paywall and then use that as like the beginning of the paywall and then slowly start to only release content inside of the paywall. And that's the way to make a billion dollars, no doubt. And like, maybe that would be the. That money would then be the fuel to, like, take on the industry in some way. But there's no doubt in my mind that we would make more money while having less impact in that world. Like, we would do less good while making more money. And that. That just seemed so obvious to me. And we kept not making choices like that. We kept saying no to advertisers who we didn't think were good partners to work with. We kept saying no to certain sources of funding that we felt would lead us into the wrong direction. And also, weirdly, we kept not doing things because we thought it might be too close to that stuff, even if we thought it were good. For example, we haven't done this, but here's an example. If we wanted to create a bunch of worksheets that went along with the crash course, but we couldn't justify the cost, we still felt like we couldn't do that. We still felt like we couldn't sell that product because it would be starting down the path of the dark side. And eventually it wouldn't be me and John in Jarge. We'd have a new CEO, maybe we'd be financially distressed for some reason. Suddenly we'd have to sell the company off. And that would lead us. And that would be like we'd already have been set off down this road. And it just seemed like we were so clearly making a bunch of choices specifically not to let profit be the motivation that it started to seem very silly that we weren't just formalizing this.
B
I don't know if the timing is accidental or if it just feels like some mantle being handed to you, but this is happening right as the Corporation for Public Broadcasting is, you know, kind of under assault, they've closed their doors.
A
Yeah.
B
Do you. Do you see what you're doing as filling that gap?
A
No, no, I don't think there is any filling that gap. I think that the. The way it. That'd be a little like saying if you had like a, you know, if you had an npr, is PBS filling that gap? You know, like, it's a. There are different media, there are different distribution systems. There's a real loss there. But also that gap doesn't have to be open. Like, you know, the gap left by the Corporation for Public Broadcasting obviously is open, but the content that. That funded can be funded and is. People are working very hard to get that content continued to be funded in other ways to continue using those distribution systems to continue to reach the people who need it. And so I think that yeah. And also we are, I just want to emphasize, way smaller than that. And I've had this question from a different angle, which is that of course the biggest harm done by the sort of Doge sensibility was USAID by orders of many orders of magnitude. And John and I have also helped raise $30 million to fund a hospital in Sierra Leone. And that might seem to some people, because it might be just as attention grabbing in some ways, where it's like, oh, this couple of brothers doing this great thing. And that seems like a big story. And I'm not diminishing the work that we did or especially the work that Partners in Health has done to help make that hospital a reality. And it recently opened and babies are getting born there. Probably, I mean, literally there's probably gonna be being born there. Right now. It's like 15 babies a day are being born there. So the, that I just like, I need people to understand the scale of the difference here. It's just like so much was lost and like, there isn't a way to fill that gap.
B
No, I mean, I'm a public media veteran. There's not. It's gone.
A
The wonderful thing about government is, like, we're pooling our resources so that a very small percentage of what we are contributing can do amazing goods. And then just to like have the richest man in the world unplug that, it's like the greatest moral failing I could possibly imagine.
B
After the break, here comes AI. So let's talk a little bit about the future. Right? So like, whatever happens in the next five, 10 years is probably not going to unwind that damage, Right. I don't know if that's fixable, but let's say that our children or their children can or are trying. How do we get them to have the kind of critical thinking skills that are necessary to think through those problems in an age where literally my 5 year old was like, mommy, this thing says I can talk to an AI and it will be my friend. And I went like running into the other room and slammed the gyoto closed off the table, actually. Yeah. How do we do that part? Because you are in an environment that is now suddenly overrun with a tremendous amount of AI I worry about the atrophy of those critical thinking skills.
A
Yeah, it's super scary. I think that we. One of the crazy things about AI as it currently stands, especially in terms of it being an engine of media. So when you're talking to a chatbot, it is creating media. It's only for you. So it's kind of hard to call it media because it's not mass broadcasted, but this is content that is being consumed. And if in aggregate, it is a kind of media. Now the crazy thing about this is that we are looking at that as if it is this big new thing, which obviously it is. But at the same time it feels like we completely ignored the fact that AI took over our brains 10 years ago. Like, what gets recommended to us? These are content recommendation algorithms. They're not like AI in terms like they're not an LLM, but they're AI. They're machine learning that are figuring out how to make you stay on a website. And that's really their only goal, is to get you to stay on the website and then theoretically also come back to the website. So hopefully also make sure you have somewhat of a positive experience. Otherwise maybe you stay, but then you never come back because actually you didn't want to just watch dashcam videos for two hours. And so like we're there now and that feels like deeply unexamined, even though, like we're in the world that that created now. And it's not a good one. It's way it is worse it is. And I know people, I will not even say the name of a platform because I'm getting too close here. But like, I know people in this world who are like, oh no, is this what I have wrought? And they were working really hard the whole time to try and not have it go this way, but the incentives are just too strong. If you're trying to get people to keep watching something, you are going to incentivize a lot of outrage bait. You're going to incentivize a lot of feelings of superiority. You're gonna incentivize simple stories. It's just the way that it's gonna be. And you're gonna incentivize a story where like, the new stuff is the pure authentic stuff and the old stuff is actually bad and evil. And like one of the strongest, most salient stories is like this powerful thing that you thought was good is actually evil, which is why we don't like vaccines anymore.
B
Right?
A
And so it's almost unsurprising that we would have this giant media revolution and then also a bunch of trust and expertise would fall apart. Oh no, when did that ever happen? Except other media revolution that's ever happened. Not that I expected it, but in hindsight it makes some sense. But you're right that the critical thinking part of it is really important.
B
How do you Talk to your kid about this.
A
I mean, I try to have. I mean, so he's nine, so he's old enough to start to really be looking into this stuff. I try to. And it feels like, so, like, I don't know. This feels almost like tech, bro. But, like, I try to talk about incentives. Like, why did that creator make those choices? It's almost like I'm bringing him into the family business, which I don't want to do, by the way. I hope that he becomes a civil engineer or something. But. There's a reason why everything is the way that it is. And that goes double or triple or 10x for these very popular YouTube videos that you are watching. Like, they are structured this way for a reason. You're paying attention to it for a reason. And you should understand, like, what your brain is doing and what they are doing to. To your brain. And I think that, like, in terms of, like, how a thumbnail is structured, like, Like, I. I know all of this stuff, and so I. I can think and talk about it. But another thing. And this is great. So kids. I think that kids invented this term brain rot.
B
Oh, I was gonna ask you about Brainrot. So here we are.
A
Yeah. And it's really powerful to be able to talk about brainrot with my son. Because when we first started talking about Brainrot, we. When he was talking about brain rot, it was like the traditional brain rot of, like, AI sloppy ballerina cappuccino. Like, just. Just like, nothing. Like, just like, oh, my God, I can. I'm just staring at this. And there's nothing, like, wrong with it. You know, it's like Doritos, where it's fine to eat Doritos. Don't only eat Doritos, but it's fine to eat Doritos. But one time I called something else. He was watching Brainrot, and he was like, that's not Brainrot. But. But it was like, it's basically like techno music while marbles fall down, you know? Okay, where to me, he's nine. He's not learning anything from techno music. While marbles get distributed among different. And they're not even physical marbles. They're like 3D generated marbles. And so I called this brain rot. And he was like, no. And then I had a discussion where it was like, no, a brain rot's anything where, like, you're watching it and this is fine, but you're watching it because it just keeps your attention without giving you anything. And so when we watch a Veritasium video that's not Brainrot. Like, you get that that's different from marbles, right? And he was like, well, yeah. And I was like. So, like, to me, Brainrot is ballerina cappuccino, but it's also techno music while marbles fall down. Like, you're not getting anything out of this. It's just holding your attention. That's what brain rot is. It is something designed only to hold your attention while doing nothing else, which is actually better than some things I've seen on YouTube, which is, like, things that are actively doing destructive, antisocial. But it is brain rot. And he got that almost immediately. Huh. Which was great. So now I can say, like.
B
Because there's a critical thinking skill.
A
Yeah. Like, I sit down with him and I'm like, okay, just let my brain rot. And then we'll go to watch, you know, a SciShow or, you know, Kurt's Cassav video or something.
B
So how do we think about these platforms then? Particularly now in a moment where they are not the platforms that they were of 2010, 2016. The owners of these platforms, very close to the administration, in the case of TikTok, basically giving them a kickback. How do you navigate that tension?
A
I think the traditional. The traditional answer to that question is competition. So if people start to.
B
That assumes a working antitrust system, but. Go ahead.
A
Well, it assumes a working antitrust system, but more than that, it also assumes, like, not an oligarchy. So you could imagine beyond antitrust, that monopolies are being enforced by the government, which happens. It's a thing that happens. It happens. It is almost an inevitable consequence of the centralization of power under an autocrat. So it basically never doesn't happen. And it's very bad. It's very bad for the economy. It's very bad for everyone. And of course, media is a place where it happens usually even faster because they would like to have control over those things, so they enforce the monopoly. And also they can use that as a way to enforce speech. So this is a. This is something that we should, I don't know, maybe look to our forefathers to see if they were concerned about that at all, and if maybe they passed some laws or maybe we might call it an amendment, specifically saying that we shouldn't do that and we shouldn't, like. I don't, like, go out and sue a company for an unrelated reason because they did something else that we don't like. That should not just be illegal. It should be imagined as the corruption of the system, that it is, but we are where we are. These things are all leaning on each other in a way that is scary. But I don't think that that should say then thus we should not do any competition. The network effects are very strong. I don't create off of YouTube very much because I create where I'm gonna make money. I create where I'm gonna make impact. I create where I'm gonna get views. To jump over to Vimeo would be very silly. And I benefit a lot from the recommendation algorithm. I reach people who I wouldn't otherwise reach, but I'm interested in other models which we've seen with smaller scale things. Dropout TV is an interesting example where it's just like, this is a place where you go to watch a specific kind of comedy content. You've got Nebula, which is like, this is where you go when you want to watch a specific kind of educational content. You've got CuriosityStream. I know I'm more inside of the educational world than other places, but RuPaul's Drag Race has their own over the top streaming service called WOW Presents, where you could just get Drag Race content. So there's something there that's interesting to me. I don't think that it's going to look like what it looks like right now, but you could maybe see if the costs of developing software get a lot cheaper, which they seem to be maybe be doing. You could maybe see a world where things like that pop up that are more sort of demographically specific, like more for certain kinds of folks who want certain kinds of things, or just for people who, like, you know, hell, like, are a little sick of the rest of it.
B
Well, is that something you hear from your audience?
A
I hear it from young people.
B
That's interesting.
A
And like, we see that in the sort of time spent online. It started to go down for the first time ever. And it started to go down among young people before, for clarity, from a very high peak. But it started to go down among Gen Z, before Millennials and Gen X. But Boomers are actually still going up, but in part because their market saturation was lower to start with. But that has happened. I think that the Real Hope competition is a part of it. But, you know, I remember when these companies first started to say, like, you know, we're not competing with Facebook and Instagram and Netflix, we're competing with like, every moment of time people have, we're competing with like, dinner parties. And I'm just like, I'm out. So that is the saddest sentence. But like we saw happened, you know, like we do. We are much less social than we used to be. And not that that trend also, for clarity, isn't just about social media. Predates social media. I remember. Do you remember bowling alone?
B
Of course.
A
When that happened?
B
Yes. Yes.
A
Yeah. But I do think that the real hope lies in realizing that and be like, what if I went bowling? Which I do love, though I hurt my shoulder last time I did it in.
B
All right, that is it for our show today. If you want to hear the full version of my interview with Hank Green, you can find it on our plus feed. We talk about cancer, mortality, masculinity, big stuff. Just head on over to slate.com what next to sign up. And if you want to see the video version of this interview, you can find it on Slate's YouTube channel. Hank Green is a YouTuber, educator and founder of Complex. What Next TBD is produced by Evan Campbell and Patrick Fort. Our show is edited by Paige Osborne, who is the senior supervising producer for what Next and what Next tbd. Mia Lobel is the executive producer here at Slate. TBD is part of the larger what Next family, and we will be back on Sunday with another episode. I'm Lizzie o'. Leary. Thanks for listening.
Host: Lizzie O’Leary (B)
Guest: Hank Green (A), YouTuber, educator, and founder of Complexly
Date: March 20, 2026
In this candid and insightful conversation, Lizzie O’Leary sits down with Hank Green to dissect the contemporary information environment, the impact of algorithms and AI on media, and the evolution of educational content online. Green reflects on new challenges for critical thinking in a tech-saturated culture—touching on YouTube’s power, personal responsibility as a creator, the nonprofit transition of his company Complexly, and the cultural consequences of both viral content and the collapse of old public media institutions. The episode weaves through personal anecdotes, sweeping societal changes, and Hank’s philosophy on fostering genuine understanding in the digital age.
[02:00-06:04]
[06:04–11:25]
[11:25–15:00]
[15:00–17:50]
[17:50–24:57]
[25:07–30:22]
[28:39–30:22]
[03:37] Hank Green:
“If a dash cam video comes on your screen and it holds your attention for 15 seconds… who doesn’t like watching a car crash? I guess you’re not necessarily going to feel proud about it later, but it’s going to hold your attention.”
[07:16] Hank Green:
“There’s lots of other people who will give you the simple story. If you want that, go get it. But like, it’s not, it never is.”
[09:00] Hank Green:
“It is. There’s just like a lot of Doritos, except that in order to eat a Dorito, all you had to do was look at it and you never got full. And also there’s a man in your pocket who’s always got more Doritos and he’s just like… hey, you want a Dorito? Look right here and you’ll be eating them. I feel a little bit like that.”
[13:47] Hank Green:
“We would do less good while making more money. And that just seemed so obvious to me.”
[17:24] Hank Green:
“The wonderful thing about government is, like, we’re pooling our resources so that a very small percentage of what we are contributing can do amazing goods. And then just to like have the richest man in the world unplug that, it’s like the greatest moral failing I could possibly imagine.”
[19:09] Hank Green:
“AI took over our brains 10 years ago… These are content recommendation algorithms… figuring out how to make you stay on a website.”
[24:43] Hank Green:
“It is something designed only to hold your attention while doing nothing else, which is actually better than some things I’ve seen on YouTube, which is, like, things that are actively doing destructive, antisocial [things].”
[30:02] Hank Green:
“I do think that the real hope lies in realizing that and be like, what if I went bowling? Which I do love, though I hurt my shoulder last time I did it.”
For more in-depth discussion—including further reflections on cancer, mortality, and masculinity—check out the full episode on Slate’s What Next TBD Plus feed or watch the video interview on Slate’s YouTube channel.