Hank Green: What No One Tells You About the Algorithm

Ten years I've been trying to figure out how to do science on weather but hair prevents chafing and I still haven't gotten it done. [music] Because when you had chemo, you lost your butt hair. It's true, yeah, yeah, it was great. Way preferable to the alternative, I have to say. Did it, did you have any less chafing? Well I wouldn't know, I wasn't jogging. You know what I'm saying? Did it use science, jazz, with our favorite nerds? Yeah. Hi, I'm Wendy Zuckerman and this is Science Verses. Today on the show, Hank Crane. Hank is one of the OG science YouTubers. He has been doing it for almost 20 years, racking up millions of followers, bazillions of views as he talks about the weird and wonderful world of science. And Hank has had this remarkable front row seat into the internet as it's morphed from this curious, exciting new thing to the cesspool of AI slop that it is today. And by making viral content for all this time, nearly two decades, he has been able to see how the algorithm has changed and what it takes to make a viral video these days. So in our chat, we talk about what people tend to misunderstand about the algorithm and we're going to talk about big stuff, like where AI is heading and will a computer ever suffer. Plus, a few years ago, Hank was diagnosed with Hodgkin lymphoma and became his own science experiment, learning many things about his body and cancer, including why chemo can make your peak cloudy. And it is a surprisingly inspirational science story. So my chat with Hank Green is coming up just after the break. Today, we are talking to the one and only Hank Green. Hello. Welcome to the show. As far as I know, it's just me. So before you were a YouTuber, you worked at a fungicide company. I did, yeah, correct. Was it fungicide? No. Oh, no. No, no. It was a really weird job because I was the only person in the lab. Oh. So there's a lot of like managing supplies. Oh gosh. But then in addition to that, there was nobody else in the building. Like it was just me and fungicide. And liquid nitrogen, four armadillos out the back door, a bunch of wasps, and a lot of produce. So you and the armadillos not having the best time of your life? No, yeah. Pretty, pretty bored. It was weird because I'd been in college for four years and I'd been socializing all of the time. And then it was just like back to my parents' house, long commute, by myself and building, yeah. So then in 2007, you see a little baby website called YouTube. There's so many years went by, by the way. I went to grad school and a bunch of other stuff happened. But yes, I saw this little website called YouTube. The site itself was pretty crappy in 2007. Yeah. And in fact, like our biggest inspiration, Zay Frank, he hosted everything on his own. He just like uploaded quick time files to his server and paid the bills. And he got people like crowdfunded money so that he could stream on his own website. And that is how the biggest video blog project that had ever existed happened in 2006. It's crazy, it's crazy, it's crazy. So what made you think I want to be on this? It was John, mostly. Your brother. My brother John was obsessed with a couple of these video blog projects. And I am three years younger than him and it took me 40 years of my life before I was willing to accept that there might be something that he thought was cool. That wasn't extremely cool. You know, like just like that, like younger brother vibe where if John liked a band, that's because that band was the best band. And if he liked online video, that's because the online video was the best thing. Which was amazing for me because I'd never questioned it. I was like, I am in all the way. And I was in all the way and remain in all the way. So I thought we could watch your very first video. This is science first together. This isn't YouTuber versus. We'll get to that. We'll get to the Wii in science. Okay. But first, trip down memory lane. Okay, there it is, there it is. Okay. Hey John. I guess you've heard by now. A lot of power off. There's still some glitches to work out. Hello John. By now you have received my message that we will no longer be communicating through any textual means. No more instant messaging, no more emailing. Only video blogging. I've learned a lot. Now that I know the background, it really does have little brother energy. When you do think about kind of where YouTube and the broader internet is today, and you watch that kind of beautiful dinky video, what does it make you think of? It really makes me want to re-edit it. And then there's also the piece where I feel quite strongly that we never could have made it if we weren't starting at the very beginning. And that the quality of a person's first video now has to be so much higher. It has to be much more written. It has to be much more charismatic. There's nothing charismatic about that for a child. But it's just gotten, there's so much value in being a YouTuber that was not there then. And that's great, but at the same time, it means that you don't get to have it happen unless you're really hungry for it, which we weren't that hungry for it. We were just messing around. The very first viral video was "Acio Deathly Hallows." Correct. Whoa, Axio Deathly Hall is in Sendio. Book sales and barbers, it'll be like Phoenix Tears on a broken nose. So when you think of what makes a video go viral back then versus now? Yeah, so the big difference now is that it used to be that virality was like a contagion, which is why it was called that. I saw this and so I'm going to share it with five people. And then those five people are going to share it with five people. And it is a cultural contagion. And that doesn't happen anymore. Yeah. That is no longer how virality happens, virality happens entirely through recommendation algorithms. And so that's the big difference. And so that does to some extent mean that you are playing a different game. Everyone has already seen 20,000 videos this week. They're going to see 20,000 more by the end of the week. And the recommendation algorithms are sort of trying to find the thing that is holding on to people's attention the most, getting the most clicks. There's like a viral YouTube video that's been slow burning for the last six years of a high school band playing a Weezer song. And they start playing in the moment the first thing is about to think the first line, "Stand collapses." And everybody at the auditorium is just like perfectly silent. Okay, but then that's the beautiful reality. So that's just like a regular viral video. But like most of what is happening is people telling a story that is trying to capture and hold your attention. And also on YouTube trying to put it inside a side of some kind of frame that will you know be like, "Oh, I kind of actually have like, I feel like I need this information." And a lot of times that information comes in the form of a threat. So if I like made a video that was called like, "Here's why you can't trust the AI companies." People would watch that because it's like, "Oh, a threat." Like I want to know, like why can't I? Hank has some thoughts about why I can't trust the AI companies. And it's all very salient. It's very of the moment. And so that's like a title that I would have if I wanted to get a couple million views. And then I'd interview an AI safety researcher and it would be extra long, which is also something that YouTube now rewards. Have you ever used one of these tricks and then just felt real bad about yourself? Yeah, yeah, yeah. So there's this thing in SciShow that we call, or that I call, a headline. So it's a headline that is a lie until you watch the video. And you have to be careful with these. So sometimes this is okay. Like sometimes it's fine. Like we did a video recently that was called like, "Everyone on Earth has the same commute." And the video was about the fact that like, everywhere, no matter where you look in the Himalayas or in South America or in Texas, there's like an average commute time. And it's like 47 minutes. It's very strange. But like if you took that that headline seriously, you'd be like, "Well, I guess everybody like the person who lives five minutes away from work and a person who lives 45 minutes away from work have the same commute time." I'm not worried about that one, because you don't think that. You're like, "You are confronted with. " The stakes alone. - Yes. - But there are also times when I have written a headline where it's like, if somebody just saw that headline and they didn't watch the video, and that's most people who read the headline, most people aren't gonna click. They're gonna walk away with an incorrect idea in their head, and we like, you know, like sometimes I'm like, but I don't get more of viewers. - What's an example? - I can't, there was like, it was something about a B, Wendy, I'm sorry, it was something about a B, but I can't remember what it was. (laughing) - I'll give you another example, which is recently John made a video where he went and like toured some facilities where they're testing people for tuberculosis and trying, like doing a bunch of work to help people know when they get to tuberculosis, make sure they know what kind they have, treat them, 'cause the treatment's different depending on the strain of TB you have, and he put this video out and it was getting like no views, which of course, you know, that's just not like a super sexy topic for the average American. - TB is hard, yes. - Yeah, and so I changed the title to Elon Musk canceled this, we're doing it anyway. Which I feel a little bad about. - Did Elon Musk say anything about tuberculosis? Was it a big deal? - Oh no, I mean, he canceled all the USAID. - Oh, okay, okay, right. - Yeah. - And it was connected to USAID. Okay, so it was true, it was true. - It was true, yeah, it was true. And like it took a lot of work to get the money together to fill the gaps to make it happen. - Okay. - But like it's a little bit like this is a story of humans doing an amazing thing and I hate that I have to frame it inside of, like you hate this guy and so watch, like hate watch this content. - Yes, yes, did it work though? - Oh yeah, yeah, yeah, yeah. - Okay, you know, yes. - Yeah, it was one of our biggest videos of the month. - Okay, well, all right. So you talked about how you are worried that these platforms, social media, YouTube, really optimized for making money. - Yeah, I mean, I would even not, I would say that they're optimized for keeping you on the site even. It's just like they know that if they get more of you, then they will eventually get more money. And that's like, that's the sort of get that they optimize for more than anything is like having you not close the app. - What is something that people misunderstand about these algorithms? - Oh, that they have, that like you are in control of them, I think is a big one. Like people, there's this phrase, when you see a TikTok or a reel that's just like really, like exactly, like something very weird that nonetheless you are extremely into. - I built this feed brick by brick. - Yeah, exactly, yeah. (laughing) - Tell me how you feel about, explain this concept first. - So the concept is like, you give the algorithm your choices. You say like, I'm gonna watch this one and I'm gonna not watch this one, I'm gonna heart this one and I'm gonna comment on this one. And I'm gonna scroll past this one super fit, fast. And that's how that is, how the recommendation algorithms, like every input they have is a human decision, but they are not really conscious human decisions anymore, whereas they used to be, it used to be like, did you click on this video? Did you make that choice? Did you like this video? Did you comment on this video? And did you dislike this video? So you have those like decision points, whereas these modern swipeable feeds are like by far the most important data point is whether you keep watching the video and for how long. And so these imperceptible, I mean, the most obvious example is just you see boobs. You see boobs and you stop for a little longer than you want it. And you were like, I was here to learn about cool science. - Boops. - Yeah, yeah, and then it's like, wait, maybe I should show this person more boobs. - Yeah, here they are, bigger and better, let's go. - Which look, I mean boobs fantastic. So I'm not that worried about boobs in particular, but I am worried about it. So like salience is the word that I use for this. So like anything that is easy to pay attention to, boobs, great example of a thing that are, a thing, two things, they're easy to pay attention to. And then, but they're so, but like, what are the easiest things to pay attention to was a really interesting sort of set of thoughts to go down and then what makes something easy to pay attention to and am I in control of that? And I'm not really. - What are some examples you're worried about, yeah? - Crime committed by people who look different from me. - Yes. - Is very easy to pay attention to. And very easy to sort of like, once your feed starts being that that's all you get and then you're like, wait a second, I think that these people who look different from me are committing an awful lot of crime. And it's like, well, are they or is it just that that is something that people watch because it's scary. And so it's very easy to pay attention to. Another thing would be stories about how I'm better than other people, which we all want. And that is not like a, it's not just a racism thing. It's also just a political thing. Like if you show me content about how like, people who vote differently than me are dumb, I'm like, well, I did watch that. And so there's like this, this like combination of, like outrage is really, like this is cute, is one very salient thing. Outrage is another very, like these test very highly, sexual stuff also test very highly. And the reason, like I think about a lot about outrage, like what is it actually? And it tends to have a component of superiority tied along with it. So like I am angry about this. The world shouldn't be this way. If I were in charge, none of this would be happening. Like it's like, there's like a morality to outrage. - Yes. - Yes, and you do use outrage in some of your videos. And what I would say to good purpose, I mean, there was a graph, some garbage graph about climate change. - As you can see, they're not occurring more than in recent years. You don't know if they're occurring more than recent years because there's all kinds of things that mess with this. - The disease is decreasing. What would you expect to have happen if the earth was getting warmer? - You were very angry about that. And I was too. And it didn't pull me in. Your anger and outrage. And the fact that you wouldn't go to bed. You just saw that graph. - It was. - Did a lot of, I let it do a lot of graph research at 11.30 at night. Yeah, it was not the best choice I've ever made. But. - So I want to talk about just staying with the algorithm. In the early days of social media about 2010, do you remember when this idea of filter bubbles by Elecarrisa sort of entered the fray? And it was, Johnny explained it. - Sure, I mean, the filter that, yeah, the basic idea of the filter bubble is very simple, which is like if you are existing inside of a curated environment, then you see a different world than the people around you. And you never interact with the world of people who have different worldviews than you, which is, I feel like the version one of filter bubbles was a good first step. And then it has turned out to be all so much more complicated than that. For example, I see lots of right wing perspectives. I just see the absolute craziest ones. - That's what your algorithm is giving you. - It's selects for left wing people, finding the craziest right wing perspectives and then showing that to left wing people, you know. So it's like here and that like that's like, you know, never would we have expected that this was one of the ways that it would function. - So this is fascinating because to prepare for this interview, I definitely feel like, I think we all feel like we live in a filter bubble. It is the deep story we've got this echo chamber. It's used to explain why there's this deep polarization when two people, political issue happens in the world. People have these very different views of it and there's this story that it's because we're seeing different images of what is happening. That's how divisive the internet is right now because of these personalized algorithms. But I was looking into the science of it and researchers cannot clearly find evidence for filter bubbles. - Isn't this crazy to me? Let me give you some quotes. So one report out of Oxford Uni said the research quote suggests echo chambers are much less widespread than is commonly assumed. Most people have relatively diverse media diets. Another academic in a chapter said that filter bubbles were a quote, unfounded moral panic. And basically the idea being that, as your case in point, is that it's much more complicated. Sometimes we are fed information from the other side. But then we often dismiss it out of hand. And so this researcher was saying, we need to, we are a part, we are complicit in our polarized views. - Yeah, of course you sure. - Yeah. And so, and then the other point was they were saying we're actively seeking out information that further policy. arises out of use, what do you make of this? - I mean, like the thing that really turned me from being like, ah, the version of filter bubbles that I heard in the TED talk 15 years ago is like not entirely like gonna be the version of reality. Is when somebody said to me never before have our information diets been more diverse than they are right now. - Interesting. And I think that the, by far the best frame for all of this is what holds your attention. And it's really hard to, like, it doesn't seem right, but like let me pitch you this. So we have this understanding now that we are all kind of suffering due to the hyper-palitability of food. So we got too good at making food too delicious. And so people ate more of it because it's so good. And if you have a french fry and then a Coke, you can have more french fry and then pay are very well. And that doesn't seem right 'cause Coke is also food. But so we have this like hyper-palitability problem in food. - Yes. - We also have a hyper-palitability or like a hypersalience problem in media where I and all of my colleague creators are in collaboration with like the most sophisticated software of all time to create stuff that is just a little bit more salient than the person next to me is creating. And I try to use my powers for good. But I don't know if I always do, but like not everybody's trying to use their powers for good because they're more ambitious than me. They're younger than me. They didn't get lucky as I did. And they have to compete. And it's harder to compete when they're them than when they're me. And so like imagine like trying to understand what a sort of hyper-salient information object is like is really interesting to me. So for example, you would want something that's not just threatening to you, it's maybe threatening to your children. Like that'd be hyper-salient. You'd also want something where it's like, there's like a big nefarious like thing that is more powerful than you that you can't control. So that's like another level of threat. And then like my favorite hyper-salient object to identify is what if this thing that you've been told is good your whole life is actually bad. And so that's like a different scale or like like flavor of threat. - Yes. - And what if I told you a thing where you could tell a story, all three of those stories at the same time about an object and that object is vaccines? - I was gonna say, or it could be fluoride in the water. - Yeah, yeah, yeah. So many things, or like NPR, you know, like USAID, you know, all of these things that are like FEMA, you know, there's conspiracy about the federal emergency management agency, the thing that like saves people when things go wrong, that's like all they do is like negotiate bottled water contracts and case something bad happens. And it's like, I don't know, that sounds pretty scary to me. - Oh goodness. - So this is like the form of the thing that worries me is not really so much that we're polarized. It's the thing that we all have started to agree on, which is that like institutions are bad, experts are bad, like completely losing faith in all of society is really bad for society, it also bad for systems. And I feel like that's just sort of a lot of what gets traction in a world where everything has to compete with everything else to be as attention grabbing as possible. - Yes. And while we are on the topic of slop on the internet, let's go into AI. What is worrying you about AI right now? - Can I tell you a weird thought I had that's not worrying? - Please. - Do you know about brand new laws? - No. - It's the, the, the, the, the, the, the, the, the, the, the, the creation to, like, refuting BS requires an order of magnitude more effort than creating BS. - Oh God, I live. I live in Bradley. He's a little bit more do, right? - Yeah, but I can't believe this is your entire life. - Yes. I swim in this law. Oh my gosh. Okay. Yes. - I, so, so there's, because I would like to make the case that perhaps creating BS isn't just like an order of magnitude easier than refuting it. I think that it's, it's free, and I think it's always been free, and I think that it's never going to be easier because it's as easy as it can be. Like, there's nothing easier than just saying some BS. - Yes. - So it's, you can't actually make that easier. You can't make, say. - You don't need to research BS. You can just say, okay. - Now, at the same time, I think that they're like, I've seen some true BS artists who will do a lot of research so that they can say clever sounding things to support their BS, and oftentimes they are absolutely certain of, like they, they like eat the sh*t they're full of. Like they, they believe it. - It is the new brand and super dangerous brand of misinformation out there right now. I, I think, I mean, this, I, when you use these complicated sounding words, and, and you really, really sound smart, and at the base of what you're saying is just feces, you know, it's nothing, it's crazy. - I mean, that's what pseudoscience really is. Like, that's when we talk about pseudoscience. We mean things that are, like, putting on the cloaks of science and saying a bunch of big words, and they're like, "Metabolism, inflammation." Then you're like, "Yes, I guess, I don't know." - "Midocentriol." Yeah. - Yeah. - You know, what? - Yeah. - Yeah. - Yes. So, wait, where do we get to the goods? I thought you was talking something good. - I have a, I have a, I have a, like, this sense that, like, you can't, you can't make it easier to create BS. - Yes. - So, so actually AI doesn't make creating BS easier. - Right. The bar is locked. - Because it's already been at the bottom the whole time. - But you can make more of it. You can pump out more. - You can make more, but like, more isn't necessarily, like, I think that we have like a, like, the, there's already an infinite amount of information also. Like, we've already hit. - We're at capacity. - Compacity for information. - In more craft. - And I do hope, maybe, that there will be tools that will make the other side of the brand delineas law equation easier. So, if your job as a researcher becomes easier, because of tools with which you can research more quickly, you can find information, you can digest papers, you can, you can, like, if that part becomes easier, then like, that is, and I'm not saying, because we're going to talk about other things that I am very dependent. But, but, but like, that's like what, what I'll say, that's a version of the story you can tell where it doesn't actually make it worse. And so, like, one of the things I feel about any big transition like this, having lived through a few of them is, boy, do I not predict what's going to happen? Because I'm always wrong. It's like, so, so weird. It's so strange. - I'm going to put a prediction out there, though, that I do hear that AI produces a bunch of BS, then maybe AI produces a bunch of facts, and they're just the AI just talking to each other while the world burns. - Yeah, a little bit of that. What I will say is I think that absolutely AI will make scamming way easier. - Yes. - So, I think scamming is different than, like, information be asked, but direct person to person scams and grifts, I think. It's so, it's so scary right now. - Yes. - Like, I don't know. I feel like, you know, I could like, watch someone's YouTube videos, clone their voice, call their mom with their voice, and be like, "Hey, mom, can you send me 20 bucks? What's your bank account number?" - 100%. You got a safe word? You got a safe word? - Yeah, everybody needs a safe word. - Everyone needs a safe word. Okay, so when we zoom out and think about where AI might be headed, I have heard that you have a video that you have not released with the question of, "Will a computer ever suffer?" - Yeah. - But you haven't put it out. What are you wrestling with here? Do you think a computer will ever suffer? I think a computer can suffer. But I don't, like, I think the kind of interesting thing I want to say here is maybe a computer won't ever suffer because we won't need them to, they won't need themselves to, there will be no reason to build suffering. And there is like two sides to this, like maybe there will be no reason to build suffering inside of a non-biological thing. Why would you create it? And then there's the other side which is it isn't an inevitability. If you have it prefers one thing to another thing, which they already do. Then if you tie experience to that, then the feeling of not having the thing that it wants, or the feeling of failing, or the feeling of, you know, not being able to achieve whatever series of goals it has will feel like suffering. - So. - So. - On the topic of suffering. Shall we talk about your cancer diagnosis? Sorry. Take away this is so wrong. - That's great. That's a great one. That's a great one. I just feel bad that I did. I just want to say before we leave AI behind, I am legitimately worried about three weird guys having this much power. Okay, that we can move on. and they've got bunkers or structures. - A lot of them seem to have structures, which is a little worrying. - Yes, yes. (laughs) So, in 2023, you had a-- - Wait, can I do a segue? - Absolutely, please. - But their structures will not protect them, from cancer, segue, segue, thank you, thank you. After the break, how Hank peed out his cancer. Literally, coming up. (upbeat music) Welcome back today. On the show, I'm chatting with veteran science communicator, YouTuber, Hank Green. Let's jump back in. - So in 2023, you found a lump in your lymph nodes and did some tests. And the doctor told you, we know you have cancer, we just don't know what kind. - Yeah, in fact, it was better news than that a little bit. She said that they say that you have lymphoma, but they're not sure what kind yet. - Okay. - Which is good, because if you have cancer in your lymph nodes and it's not lymphoma, then that means it's probably a metastasized from somewhere else. And that would be worse news. So, that was a little bit of good news, and she didn't tell me about that. - So you ended up having to do chemo. You were ultimately diagnosed with Hodgkin lymphoma, which your doctor told you. - Yeah, well, my doctor did not say that. I knew this. In fact, they did not need to say anything to me. My doctor called me and was like, we just got it back. I just heard it's Hodgkins. And I was like, thank you so much. Thank you so much. And I just hung up and I told Catherine. And I knew that that was the best outcome. - Yes. - By the time I got staged, like a 5% chance that I would not be alive in five years. - What possibly, what went through your mind? When this was unfolding? - Oh, so many different things. - There's weird practicalities of like, I got to make sure my passwords are all I got in front of her. In fact, I had had a weird experience before I found out that I, so I went to see my doctor just to get anti-anxiety pills. And I was like, this is a crazy thing to go to my doctor and be like, I need you to help me with my anxiety. And then I was like, second thing to help me with my anxiety. Can you tell me that I'll be alive in three months? Because otherwise, I got some stuff to do. And she was like, he was your prescription for your anti-anxiety medicine. I cannot. And so I was like, okay, well, I guess I'm gonna do this stuff. So there was stuff to do during that period of time because if it was like metastasized lung cancer or something, I really might not have lived for three months. So that was terrifying. And then when I knew, though, the landscape was like, okay, I've got years at least. And that looked very much like, okay, so I have to prepare for treatment, I have to understand treatment. I have to, I'm worrying about like, okay, my son's not gonna be grown up for a decade. Like, I thought maybe I would be more worried about work. I wasn't worried about work at all. I wasn't worried about how everybody was gonna handle life without me at work. I was worried about my family. I was worried a little bit about all the information that's tied up inside of my head and feeling bad for not having mentored more people. But I didn't feel, another thing that I was surprised is I didn't feel worried about missing the future. I was worried about not seeing my son grow up. I wasn't worried about missing a mission to Mars or the next scientific advancement. Which is like, I love that stuff. So that's like so much what I enjoy about the world is like, finding some new thing that we've discovered. But I wasn't, I'm kind of freed to find out that I don't, like when I die, maybe won't actually care that I'm gonna miss out on the next set of discoveries, even though that is really a source of joy in my life. That's very interesting to me. - Yeah. And you did discover many things about your body and chemo. So one of, I think the most fascinating is, tell me what happened when you're peak of cloudy. - Sure, yeah, I thought I had a UTI 'cause it burned when I peed and my pee was super cloudy. And I went in, I called, and so when you have, when you're in chemo, they take everything very seriously because you basically have no white blood cells and so little things can become big things really fast. And so I called and they were like, come in and do it. You're in test and I came in and I did. And luckily, by chance, my doctor, I was like free at that moment. So they did it and he was like, hey, come on in and we chatted and I was like, he was like, you're all good, your pee looks fine. And I was like, well, why is my pee cloudy? And he's, well, first I said, why is it burned when I pee? And he was like, that's the chemo. It, you like pee out the chemo and it irritates you, your re-thra. You pee out the chemo. There's a lot of pieces in there. Yeah. Yeah, I mean, yeah, your kidneys filter it out and you know, it doesn't all the way broken down by the time it gets out. So it's, it does some damage on the way up. And, but the reason it's cloudy is you are looking at literal pieces of, little bits of cancer. Yeah, and so I was like, why is it cloudy? And he was like, that's the little, that's like the parts of the cells. It's like fats and proteins from the cells that have broken down in your body and you are, your kidney filters them out. And Michael, a lot of that stuff actually gets recycled by the body, but when you have like a big release of it, it overwhelms those systems and you actually can see the physical effects of the cells exiting you. Did it change the way you looked at your pee? I mean, I find this very inspiration. Everything changed the way I looked at everything. I know it was very good. It was amazing to, to be like seeing the physical results. And a lot of times you don't, you don't really know that you're getting that. And so you're like waiting for the next scan. I was very lucky not just to have the pee, but also had a lymph node that I could feel. Most of my lymph nodes were underneath my pectoral muscle so I couldn't like palpate them. I couldn't, couldn't feel what they were up to, but I had one in my armpit that I could feel and I could feel it changing shape and texture. Wow. And now here's a weird word. But yeah, it was softer, I should say. It was, it was, you could feel improvement. And how are you now? I'm good. I've been in remission for almost three years and I remain much more medically anxious than I was. I have had some, you know, almost everybody has a little bit of like some level of disability that they carry out of cancer treatment. It turns out, which I also was the thing that I never thought about. I have some pain that I didn't used to have. The anxiety is definitely a thing. Yeah. I got, I got shingles this year so bad, which is apparently a thing that can happen to people who've done chemo. And it was, it was, it just, like I've had shingles before. And this was like, I don't know, like 30 times worse, it was so bad. It still hurts. And it's been, like, two months. I'm so sorry. Yeah, it's crazy. Thank you to everybody who's ever done a clinical trial, by the way, to refine these treatments so that they are now so much more tolerable. And gosh, I'm going to try another segway. Hank, I'm really sorry. Just while we're on crappy, crappy things, let's zoom out, let's zoom out. Yeah. What is it like in the US, I guess? I mean, it's so crazy right now. It feels pretty bad. It feels like something I never would have expected. It's heavy, for sure. And maybe I'm just beating this drum because this is my anchor. But I think that when you have a communications revolution, like existing institutions take a pretty hard hit every time. Mm. Like right now, I feel like I kind of wish the internet hadn't happened. But I don't think that that is what I feel will be the long-term opinion of it. I think that the long-term people-- Well, I'm right now. You're right. I don't have any more ability to feel as if things, as if it has made the world better like I used to. And I'm like an optimistic, hopeful person. But this is a mess. And it is a mess that is created by the fracturing of communication. You think this is all due to the fracturing of communication as you put it. There's social media spaces, the breakdown of traditional media. Is that what you're referring to specifically? I think that those are two things that are also emergent of the thing. But the thing is many-to-many communication being free. I don't want to participate. And I write that wave. And I don't think that on its own does it. I think that that combined with social media algorithms, social content platform algorithms, doing this recommendation and promotion, I think that we do not, to any extent, understand how much we live in a world constructed by AI. But it is not large language models. It is recommendation algorithms. Yeah. How did you feel that you are in the middle of this? I mean, you are one of the biggest YouTubers in the world. It's not true. Thank you. One of the biggest YouTubers in Wendy's documents were old for sure. I don't know about the others. I mean, you are in this space, and yet, I mean, are you thinking of just pulling the plug and, you know, becoming a god? No, like me, leave, no, never, no, yeah, no, like the, that's not how I think about it at all. I think that it is vitally important for thoughtful people to be engaging with it in a way where you will try and understand it and understand how it's affecting society and, like, that's how we get through. It's like an ex, you're breaking out, you're in that middle breaking up period, I'm still obsessed with you and you maybe you could be good. Yeah, I don't, I don't actually like, that's not, it's, I can fix you, yeah, no, it doesn't sound like it's okay, yeah, yeah, all right, we're going to end with our oddball questions. Quick questions to cap us off, are you ready? I'm ready. Okay, what is a conspiracy theory that you secretly have a bit of a soft spot for? Um, that caterpillars and butterflies used to be different species and merged. Ah, ah, ah, I just tell me we didn't land on the moon. I mean, it's definitely not true and there's also all kinds of other bugs that have larval, there's fish who have larval stages, larval stages are totally a thing that happens that are normal, but, like, I love the idea, like, I agree, you look at a caterpillar and you look at a butterfly and you're like, there's no way, that's no way, that's our different things, they go into the concave, yeah, yeah, it's like, you're not going on in that, no, there was a worm and there was a butterfly and then they somehow merged like the mitochondria and the cell finish this sentence. Now that I know blank, I'll never look at blank the same way again. Hmm, now that I know how cancer works, I will never look at life the same way again. And not mine, just for clarity, the existence of multicellular organisms. Tell me more. Do you know what cancer is? No one knows, but can you give me the Wendy's Agreement answer to that question? What is cancer? Sure, I would say cancer is when the DNA mutates in your cells in a particular way that creates a cell that will replicate faster than other cells in your body. Yeah. So there is, there is at the root of that a mechanism at work that nobody talks about, I think just because it's terrifying, which is that natural selection is acting on these cells. And so any cell that can replicate faster than other cells, there will be more of that cell, any trait that allows it to replicate faster, there will be more of that trait. And and thus, if you understand how natural selection works, if you're like, wait a second, then multicellularity shouldn't work. Because if you just be having these cancer types cells, the moment that a multicellular organism starts existing, some cell in that multicellular organism is going to be like, well, if I, like, it's not going to think this, if it acquires a trait that allows it to multiply faster than the other cells in the organism, it will just do that, that, that because of anything, just because if the, if it's able to make more of itself, there will be more of it. Yes. And there are so many cells in our body, surely one is going to go rogue. Yeah. And indeed, it does happen in about 40% of people, by the end of their life. But what this means is that like the existence of multicellularity is from the first little weird jelly thing in the etiacaran has been all based on controlling cancer. It's been controlling cancer the whole time. That's what multicellularity is. It is cells working together to try and pass their genes onto the next colony of cell rather than the next individual cell. And in order to do that, shouting at each other and working constantly to stop each other from becoming a cancer. It's fucking wild. Yes. That is, I mean, that, it does kind of make me think of when I was in college, I had this theory, I worked at a bakery that everything is bred. And now that you've got this answer, you're like, everything is cancer, everything is hiding cancer. Yeah, yeah, for sure. Yeah. You look in like tumor suppressor genes are among the oldest genes in, in, it's crazy. Yeah. It is crazy. It is a crazy thought now that you think about it. Yeah. We just, that is the, the bait, the almost the basis of multicellular life, just fucking cancer. Yeah, just like bread. What's your favorite title to a scientific paper that you read recently? I was rereading the scientific paper that Thomas Hodgkin wrote about Hodgkin's lymphoma when he discovered it, which was titled on some morbid appearances of the absorbent glands and spleen, which we used to have better titles to papers. No. I love that. Wow. Yes. All right. Can I tell you, Hank Greene, my theory as to why when you fought and light a match, this mail goes away. Okay. Hit me. I have, now I have a guess, but. Okay. You want to, have you thought about, like, do have you used this strategy when you do a fart? I have not. I've heard it. Are you a farty person? Oh, yeah, we're just, we're very chill about the fart smells in our house. Yeah. You're not, that means you're not a really funny person. You don't know. You don't know what it's like, mate. If you're, we're, we're all fine with bonds, but there comes a level you need to use the match. Okay. Why do you think it works? I have, I, so it is not the match lighting, but the match going out would be my guess, that when the match goes out, it creates a smell that overwhelms the fart smell. Okay. The mask, the typical masking theory. Yes. It's the masking theory, yeah. It's been, it's been proposed. This is what I think the internet thinks, but I think it goes deeper. Okay. It goes, it goes catapillabot of like deeper. And now I finally have someone to tell. We should have a podcast together. This is ridiculous. I don't know why we don't. This is, I'm so excited. Do you know how many people I have tried to tell this theory to and they could not give a ****. And, and I'm, and I'm like, I want to spend every, I was spent an hour with you a week forever. Is that why I respond? Okay. All right. Here it is. All right. Here it is. Here it is. Here it is. Okay. So when you light a match, red phosphorus is a beginner and the ignition point of red phosphorus is 260 degrees Celsius, which is 500 Fahrenheit. The reason a fart is smelly is because of hydrogen sulfide in the fart, in pot, in pot. That's often the smelliest part. Often the smelliest part famously rotten egg smell, guess what the ignition point of hydrogen sulfide. I think you're burning hydrogen sulfide in the egg. I think it's exactly 260 degrees Celsius. I think through the magic of chemistry, the smell, the hydrogen sulfide is splitting apart and you are, you are creating whatever. I think it becomes water H2O in the air. And some kind of sulfur, I think the smell disappears. I think it's more than masking. Hmm. I don't know, Wendy, because I feel like you'd have to get all of the air to 262 degrees Celsius, which would definitely be bad because it's not like it's not going to spread around. It's not, the fart smell isn't just at the, at the strike point on the matchbox. The good point, at grades, got it again. We need to do some actual science, yeah, actual science. Exactly. I've been, I've been ten years, ten years, I've been trying to figure out how to do science on whether but hair prevents chafing and I still haven't gotten it done. Because when you had, cameo, you lost your butt hair. It's true. Yeah, yeah. It was great. It was incredible to the alternative, I have to say. Well, I wouldn't know, I wasn't jogging, you know, it was, it was in bed, watching Seinfeld and eating Chick-Fil-A. It was not like a much opportunity for chafing when I went up and down stairs once a day. Thank you so much, Hank. Thanks so much for your time. Thank you. Thank you, Fan. It is so great to be here, keep doing the good things you do. That was my chat with the science communicator, Hank Green. Now, soon after I interviewed Hank, some people got a bit miffed because of how he's used AI to help make some of his content and you might have heard about some of that controversy. Hank has in the meantime released a new AI use policy, which you can find on his YouTube channel. We can also find Hank at SciShow, and his comedy special is called Pissing Out Cancer. It came out in 2024. We'd love to know what you thought of this episode. You can find us all over social media, or on Instagram and TikTok. We're also on YouTube now, so come and find us at Science vs Podcast, if you're already watching on YouTube. Hello, please like and subscribe. And if you like this episode, then why don't you share it with one of your friends. This episode was produced by me, Wendy Zookerman, and a Ketie Foster Keys, with help from Merrill Horn, Rose Rimmer, and Michelle Dange, editing by Black Terrell. Mix out. and sound design by Bobby Lorde, music written by Peter Leonard, Bobby Lorde, so wily, Boomi Hidaka, and Emma Munga. An extra thanks to the Zuckerman family and Joseph L. Wilson. I'm Wendy Zuckerman, back to you next time.