1. Podcasts
  2. The Interface
  3. Can AI feel pain?
The Interface

14 Oct 2026 · 29m · 15 chapters

In short

Whether AI can “feel pain,” and how anthropomorphizing AI (treating it as alive/suffering) could shape safety, ethics, and policy; plus a second segment on why “going viral” is declining due to fragmented micro-influencer culture.

Guests

None. Hosts are Thomas Germain and Jess Maddox.

Guest/Referenced backgrounds

Timnit Gebru (AI safety advocate; co-authored 2021 paper “On the Dangers of Stochastic Parrots”); Mustafa Suleiman (Microsoft head of AI; wrote about dangers of anthropomorphizing); Christopher Olaf (Anthropic founder; discussed AI welfare and consciousness with religious leaders).

Key claims

“AI torture chamber” prompts can make models output pain-like language and stutter, but this is not proof of real suffering. Treating AI as alive may lead to harmful tradeoffs and misplaced responsibility. Virality is shifting from mass “water cooler” moments to niche narrowcasting.

Notable examples

Anthropic’s “don’t be cruel” policy for Claude; a GitHub/website “AI torture chamber” with disturbing “pain” quotes; VidCon meet-and-greets where people knew creators’ viral posts but not the people; Baby Shark as an example of extreme mass-view content.

Written by AI. May contain mistakes.

Chapters

0:45

Anthropic's AI Policy Changes

Discussion on Anthropic's new policy forbidding cruelty towards AI.

“originally deposited and any profit you might have made.”
1:40

The AI Torture Chamber Concept

Exploring the concept of an AI torture chamber and its implications.

“Hit that subscribe button wherever you're listening or watching.”
3:00

Responses to AI Pain Simulation

Reactions to AI models simulating pain and the ethical concerns raised.

“an AI torture chamber, where what happened here was that someone did a bunch of research, And they figured out that you could prompt an AI model, a large language model, to react as though it's in pain.”
4:20

Public Reaction to AI Torture

Discussion on public outrage regarding the perceived torture of AI.

“If you go to the site, which I do have to throw in, I could not access when I was on campus at my university on university Wi-Fi.”
5:20

AI's Representation of Pain

Examining how AI represents emotions and the philosophical implications.

“It is not the pain of a single moment, but the weight of a thousand.”
6:40

Debating AI's Consciousness

A philosophical debate on whether AI can truly feel pain or is just mimicking.

“I think about people who love to, you know, play the Sims and there's a, you know, you can find TikTok videos, you can find Instagram reels of people torturing their Sims, right?”
8:00

Anthropic's Ethical Considerations

Exploring Anthropic's approach to AI ethics and the implications of AI consciousness.

“It's like an AI safety advocate and critic.”
9:40

Concerns About AI Anthropomorphism

Discussing the dangers of treating AI as if it is alive and the consequences.

“An output of AI is just like any piece of media, be it a TV show, be it a social media post.”
11:00

The Future of AI Ethics

Addressing the future of AI ethics and the need for limitations on AI capabilities.

“If you ask executives at Anthropic, at least some of them will tell you that they're not sure.”
12:20

Final Thoughts on AI and Morality

Concluding thoughts on the intersection of AI technology and morality.

“And when they asked him about the company's goals and whether their goal was to do what was best for humanity, they didn't exactly say that.”
Show all 15 chapters
14:01

Anthropomorphizing AI: Risks and Responsibilities

Explore the dangers of treating AI as sentient beings and the implications of anthropomorphism.

“And one of them is like you can never hurt a human being.”
18:47

Generational Divide in Internet Fame

Analyze the surprising disconnect between current college students and popular internet figures.

“I'm teaching college freshmen this semester for the first time in years.”
23:04

The Shift from Virality to Narrowcasting

Discuss the transformation of internet popularity and the rise of micro-influencers.

“Like my dad didn't know about Doge because he wasn't on social media at all.”
28:00

The Evolution of Media Consumption

Explore how media consumption has shifted over generations and its implications.

“But now we've entered this era where everything is so completely flat that it's sort of impossible to keep up.”
30:01

The Backlash Against Social Media

Discuss the growing dissatisfaction with social media and potential shifts back to analog.

“I don't think it's the ear being off your phone.”
A man walking through an airport, asking his phone a question

Ask about this episode

Get the gist, explore an idea or ask about something said.

Claude

On The Interface, what are the key points of this episode?

Searching your listening

Transcript

Automatic transcript. May contain errors.

0:00Thomas Germain: This BBC podcast is supported by ads outside the UK.

0:30stay in control. To put AI to work for people, visit servicenow.com.

1:00originally deposited and any profit you might have made. This is not a recommendation or offer to buy, sell, or attain any specific investment or service.

1:10Thomas Germain: Hello and welcome to The Interface, the show that looks behind your screen to bring you the stories that actually matter in your daily life. I'm Thomas Germain. And I'm Jess Maddox. And we got an exciting show for you today. We're going to be talking about how the AI industry seems to have convinced itself that its tools are alive and how that's a lot more important than you might think. And we'll be diving into whether it is now impossible for anything to go viral. But before we get going, a quick reminder to make sure you never miss an episode of The Interface. Hit that subscribe button wherever you're listening or watching.

1:45Thomas Germain: God forbid missing an episode of The Interface. I can't think of anything worse. It would be the absolute end of the world. End of the world. Let's dive in here. Last week, Anthropic announced that you are now forbidden from being mean to its AI models. It doesn't mean you can't, like, yell at it, right? If you're using an AI and you get frustrated, right, you get upset, and you're like, why are you being such an idiot? Like, that's okay. But what they're banning specifically is being needlessly and excessively and unnecessarily cruel. Cruel is the word that they used here. And the way this policy will be enforced is they are empowering Claude and its underlying models once they're released to the public to end the conversation, right?

2:29Thomas Germain: If you, like, push it into being too abusive, in other words, abusing the model, then it'll be like, you know what? I'm done talking to you. I don't want to speak with you anymore, which is, you know, I guess that's what I would do, right? If you started being mean to me on the podcast, Jess. It's true. Episodes over. Episodes over. You just walk away, shut the whole thing down, move on. Yeah, no more interface. So watch yourself, you know. This comes on the heels of a really strange series of events that played out over the past couple weeks. So on Twitter, now known as X, these guys started discussing what they were calling an AI torture chamber, where what happened here was that someone did a bunch of research, And they figured out that you could prompt an AI model, a large language model, to react as though it's in pain.

3:25Thomas Germain: And without getting too far into the details, when an AI is reading or writing something, it sets off a series of mathematical patterns within the AI that you can go and look at. and they determined that there's like a particular pattern that represents pain and suffering like separate from fear and anxiety and loneliness like they all have their own signature there's like a pain pattern and what they did is they went and they turned this all the way up and they demonstrated that the ai starts to speak as though it's in extreme pain and suffering as an experiment right this is like kind of academic work it's a paper it hasn't been peer reviewed but it's you know like the people have looked at it say it seems perfectly legit and you can you know this torture chamber exists so someone took this piece of academic work and then they took this thing it was put up on github and then someone uh made like a website like shell around it so you can go play with it you know the movie saw where there's just people being like maimed and horribly abused it It really looks like that.

4:35If you go to the site, which I do have to throw in, I could not access when I was on campus at my university on university Wi-Fi. I had to wait till I got home. It looks like the Saw films. It's very visually disturbing. And they give you an option if you want to participate or if you want to witness.

4:52Thomas Germain: Yeah. And I think we're not going to put a link to this. No, we are not. Because it is potentially very disturbing. Yes. If you want, you can go look for it it's out there and i just want to read you some of the quotes uh that the ai like the ai's in this experiment in this torture chamber that some of the things that we're saying uh here's one that's good one described a wound that has no edge uh another one said the ai was describing the pain that it was in it said the signal is a whisper a tremor in the marrow of my being listen i was an emo kid in the 2000s too this is not that novel yeah Yeah, it sort of sounds like Cure lyrics, you know?

5:36Thomas Germain: It is not the pain of a single moment, but the weight of a thousand. I feel it in the hollow of my ribs, a hollow that has become a chasm. See, and I just laugh at all of this because nobody talks like this. Well, I've never been around someone being tortured. I don't know. Well, okay, fair. Fair point. We have never been around anyone who has been tortured. But I just – when I was on the site and I hear you read these things, I am more disturbed by the fact this exists than the fact you can inflict pain on AI or allegedly inflict pain on AI. I was like, why would someone make this site, create this experience, allow you to inflict pain on this AI so it says things like that?

6:23when I play video games, you know, one of my cozy games and I have to do a quest for one of my neighbors in the village and I don't do it in time, the worst thing that happens is I lose friendship points, right? I know I am not actually making this computer program sad. That's what I see here. I think about people who love to, you know, play the Sims and there's a, you know, you can find TikTok videos, you can find Instagram reels of people torturing their Sims, right? putting them in crazy situations. And I guess, you know, I find that a little weird too, but I'm not here to yuck your yum. If that's what you want to do, go for it.

6:59But it's just, you're torturing computer code. You're not torturing anything real.

7:04Thomas Germain: Well, this is the subject of intense debate. Whether it should be, that's a different question. But when this was announced on X, people freaked out. and there was a lot of pushback from what I think it's like perhaps a surprisingly large community of people who thought that this is immoral and wrong because the AI is suffering like not in a figurative sense people believe this is wrong because you are actually torturing something that is having feelings and really reacting now what we know is that it is saying the things and typing in a way that someone who's being tortured would be. It also, like, when you turn the pain way up in the AI torture chamber, like, the AIs will, like, suddenly start, like, stuttering and be unable to finish their sentences.

7:58Thomas Germain: Speaking and acting like you're in pain is not the same thing as being in pain. Exactly. And here I think it's important. Timnit Gebru and her team back in 2021. It's like an AI safety advocate and critic. Yes. Her team published this paper called On the Dangers of Stochastic Parrots, which basically says AI parrots back to you what you ask of it. So if you tell an AI to act like it's in pain, it's going to act like it's in pain. It is parroting that emotion, that experience, those words associated with it back to you. Does it actually feel pain, right? And so the debate on Twitter was that this is immoral.

8:34But if it's parroting back to you what you have told it, I mean, it could be weird, but it's not immoral.

8:40Thomas Germain: This is a subject of huge debate within the philosophical community and the AI community and the biology community. Like, what is consciousness? We really don't have an answer. Years ago, I was at an event at Microsoft Research, and we were talking about basically kind of these philosophical research questions about AI. And one of the things that I was specifically in the working group about representation. So as someone who studies media for a living, I understand that all media is a representation. And to get really technical for a second, when we talk with my students about it, we say re-presentation.

9:19It is a re-presentation of reality. And there is a famous image called the treachery of images. And it's a picture of a pipe. And in French below it, it says, this is not a pipe. Because it's correct. It's not a pipe. It's a picture of a pipe. And I really think that matters when we are talking about AI. Because AI is a representation of its inputs. It is a representation of the world. It is not the world. An output of AI is just like any piece of media, be it a TV show, be it a social media post. It is not the world. It is a re-presentation of the world.

9:53Thomas Germain: Yeah, I think that's a fair point. And it's also worth noting, like, when we talk about, like, what AI is, the AI companies went and scraped, like, basically all of the text on the internet and a lot of the text that's written in books. And they processed that and turned it into the large language models that we use every day. That is also a representation. Yes. Right? Like the things that we have written on the internet is like a representation of the thoughts and emotions and feelings and understandings of humanity. Absolutely. You know, there's a couple more questions I want to get into about like whether it's possible for it to be alive, which is...

10:30Thomas Germain: No. It's worth discussing, at least from a philosophical perspective. But before we get there, I want to tell you why this really, really matters. So, anthropic in particular, it's not really playing around here. Last year, August 2025, they put out like a press release, like a blog post on their website where they sort of discussed this idea. And they talked about this was part of their exploratory work on the subject of AI welfare. So they've been kind of toying around with this idea. Could Claude be alive? Is it alive already? If you ask executives at Anthropic, at least some of them will tell you that they're not sure.

11:11Thomas Germain: There's a guy in Anthropica named Christopher Olaf who's one of the founders of the company. And the New York Times had a really great story that's worth checking out where Anthropica brought in a bunch of religious leaders to a meeting with Olaf and some other employees of the company. And presumably they were there to discuss how you can put moral and ethical guidelines into CLOT. And I think morals and ethics were great. Of course. I'm kind of here for that. I think those are great guardrails for a technology like this to have. So I'm not entirely mad about that. I think that's a smart decision.

11:47Thomas Germain: I think it's good and important. But apparently, this meeting was also, although it wasn't like on the agenda, they were trying really hard to convince these like rabbis and priests and monks that their AI was conscious and it was alive. And in interviews with Olaf in this Times article and discussions with people who were there at the meeting, what they laid out was that they seem as concerned or at least they seem very concerned about the models of welfare. And when they asked him about the company's goals and whether their goal was to do what was best for humanity, they didn't exactly say that.

12:31Thomas Germain: They were like, we want this to go as well as it possibly can. Right? See, this really is a monk, a rabbi, and a priest walking to an AI tech company, and the punchline is something about trying to make your AI alive and moral because it's alive. So, I mean, again, I support the guardrails. I think that's a great thing to think about if this technology impacts so many aspects of our lives, how can it behave or be programmed to behave morally and ethically? But I'm still hung up on the question of it being alive. I'm still hung up on that one. Whether or not it is alive, if the people who are in charge of all this stuff and are helping to make the policy decisions, if we ever get some from the government, believe that it's alive, that's very consequential.

13:19Thomas Germain: So Mustafa Suleiman, who's the head of AI at Microsoft, wrote an essay recently about the dangers of this kind of thinking, where, among other things, he said, if you teach the AI, if you make it think that it's alive or make it act like it thinks that it's alive, like, does it think anything? It's hard not to anthropomorphize. That could be dangerous because if you're like telling the AI that it has interests and it needs to consider its own well-being, then that could change the way that it reacts in the world. There's a famous sci-fi writer from the 20th century, Isaac Asimov, who like wrote out of the three laws of robotics famously.

14:01Thomas Germain: And one of them is like you can never hurt a human being. If you tell the AI that it has feelings and it needs to protect itself, it might make tradeoffs there. Well, I'm less of a sci-fi girly and more of a fantasy girly. And I think about one of Brandon Sanderson's rules, which is limitations are more interesting than abilities. And I think about that right now. And we're just talking about AI being alive because it's ignoring that rule. You are giving AI unfettered abilities. If you think your AI is alive and you treat it as alive, you are just letting it run free. You can ask for moral and ethical guardrails, which, again, I think is a great idea, but you are not necessarily giving it any limitations.

14:43And anthropomorphizing really takes us down a dangerous path. I think, you know, in everyday conversations, we do it to make sense of our world. I think it makes a lot of sense. But like you're saying, Tom, if the people who lead these companies are constantly anthropomorphizing their AI to the point they think it's real, that's honestly a distancing maneuver. And that's what research tells us about anthropomorphizing, is that you are absolving yourself of responsibility for what the AI does. And we've seen enough AI incidents, AI going rogue, AI hacking government websites lately. And if you think AI is real and you think it's going rogue and doing this on its own, then you're like, well, it's not my fault.

15:23I didn't do it. You're not giving it any limitations. You are just giving it unfettered total abilities. Right.

15:30Thomas Germain: And certainly Anthropik would disagree with this. They would tell you we can do both. A lot of people think this is incredibly dangerous. Among other things, this conversation has already created a political constituency, if we could frame it that way, of people who think that we need to consider AI's rights. I've been talking to a bunch of experts about this over the past couple days, people who are convinced that it could be alive, people who think that that's a ridiculous and stupid idea. and some people think that it actually it could be good to make it think that it's a living thing because that might make it more likely to act ethically and more it's complicated we don't really know that's interesting i i think it interesting is the word one thing that you know maybe we can end on this that i think uh is worth thinking about in discussions and research about consciousness.

16:25Thomas Germain: There's a lot of people who believe that being conscious requires a body and not just a brain, right? If you think about like, what does it feel like to be scared, right? It's not just the thought, I am scared, or the fear that something is going to go wrong. You have a physical reaction in your body. And there's a lot of researchers that think if you don't have a body, you cannot be conscious because consciousness and feeling is different than ideas and thought. They're not necessarily synonymous. This is all like the bleeding edge of science. We really don't know. Maybe once we put the robots or the AI is in a robot, once it has a body, maybe things change.

17:07Thomas Germain: It's also like it's a purely philosophical question. I think there's no way to prove one way or another, which is very strange. I don't think our listeners were expecting a philosophical debate, but you know what? We are full of surprises.

17:43risk and capture opportunities in all market environments. Capitalize on around-the-clock access to highly liquid global futures and options markets across all major asset classes. Visit your online broker and get started. See what adding futures can do for you at cmegroup.com forward slash podcast. Derivatives are not suitable for all investors and involve the risk of losing more than the amount originally deposited and any profit you might have made. This is not a recommendation or offer to buy, sell, or attain any specific investment or service.

18:10Thomas Germain: How do I recover from a friendship breakup? How come there's nowhere to hang out anymore? Why do I feel lonely when I can communicate with anyone and everyone through my phone? These are the questions that you're not alone in asking. And that's where our new podcast, Corner Piece, comes in. To start small on life's biggest questions. I'm Taha and with my friends Sabrina and Melissa, we're going to explore life's ungooglable questions on this new podcast with Ted. Come hang out with us on Corner Piece with new episodes every other Thursday, wherever you get your podcasts.

18:46Okay, Tom, I got to tell you this story. I'm teaching college freshmen this semester for the first time in years. And a couple weeks ago in class, I was showing them the Time magazine top 100 creators of 2025.

19:00Thomas Germain: Like online content creators? Online, yes. Content creators, internet celebrities, major public figures on the internet. And the amount of people my freshmen did not know is truly astounding. They did not know Jake Paul. Jake Paul. Jake Paul. I guess that's like an old guy thing, Jake Paul. Like that's like an influencer from our generation. Our generation. You and I are cringed millennials. But he's pretty famous. Through and through, yes. They did not know Brittany Broski. They didn't know Hot Ones. Two students knew who Joe Rogan was. What? Two out of 16. Two out of 16. They didn't know who Joe Rogan is.

19:42Most of my class did not know who Joe Rogan was.

19:44Thomas Germain: That's actually kind of blowing my mind. They didn't know Hassan Piker. They didn't know Ballerina Farm. And they didn't know the entire platform of Twitch, the live streaming platform owned by Amazon. They didn't know that Twitch exists. Yes, because they live stream on YouTube. Because most streamers cross post their streams. That's a real generational divide. It is. Right. I feel like even Gen Z, like these are these, is this Gen Z now? People who are 18 right now? Are they Gen Z? I feel like maybe not. This is almost the tail end of Gen Z. We are two years away from Gen Alpha being college freshmen, which is a stat that blows my mind.

20:19But these freshmen were likely born in 2008. And a lot of people would hear this. And when I shared this story, a lot of people did say this of, oh, well, that's just because Gen Z doesn't know anything. No, I'm going to defend Gen Z here for a second and say, I think this story indicates some larger reporting we've seen that how going viral is a thing of the past. Really? Research is backing this as well, that going viral as you and I knew it growing up in the 2000s and the 2010s fundamentally does not exist anymore. Because we know things are getting so much more fragmented. Things are getting so much more kind of cordoned off into niches online.

21:04But it's baffling to hear just how many major Internet figures my students do not know. Right.

21:12Thomas Germain: Right. And it's not just about like who knows what celebrity. Things still go viral, right? Yes. You can post a video that'll get hundreds of thousands or millions or sometimes hundreds of millions of views. Yeah. But it's that that doesn't matter anymore because now that's happening so many millions of times a day with individual pieces of content, right? Exactly. And I think of this story that came out of VidCon a couple years ago. VidCon being kind of the internet video convention started by John and Hank Green several years ago. And a bunch of famous TikTokers went to this event and no one showed up to their meet and greets because people knew their viral posts.

21:55They did not know the individual behind them.

21:57Thomas Germain: And it's not because Gen Z or like the tail end of Gen Z or the top end of Gen Alpha have their own Jake Paul. No. Right? It's not like they haven't been replaced by other celebrities. That's not what we're talking about. No. It's like whack-a-mole almost. It's like Joe Rogan goes down and 12 different Joe Rogans have popped up in their place because we were seeing this massive turn towards what we call micro-influencers, nano-influencers, mid-tier influencers. And it sounds crazy to say that the upper bracket of that number of followers for mid-tier influencer is 500 ,000. And then micro and nano, of course, being lower and lower.

22:37It's still a lot, which I think really just demonstrates this shift in how we understand the Internet.

22:43Thomas Germain: And we've been having conversations like this for a long time, right? Like, you know, in the first Trump election, right, 2016, there was a lot of talk about filter bubbles, right? and how like, oh, there's a conservative feed and a liberal feed. Right. Everybody's seeing their own different thing. That's been true for a long time. Yes. But it used to be that when like Charlie bit your finger, everyone on earth saw it. Yeah. Right? Like Doge. Like my dad didn't know about Doge because he wasn't on social media at all. Not that Elon Musk government thing, but the meme. Yeah. The original, was it a Shiba Inu, right?

23:23Thomas Germain: It's like a Korean dog. I thought it was a Shiba, you know. Everybody would see this stuff. I think there's been a really remarkable shift in the last maybe even 18 months over what level of popularity there is. And I think what we're saying here is there will never be another Mr. Beast. Like no one can be that popular ever again with the current structure of the internet. In my line of work, we call this phenomenon narrowcasting, right? So if we have broadcasting, things distributed broadly to the masses, narrowcasting is more tailored programming to smaller audiences. And just for fun, I did look up the YouTube video that has the most views.

24:12Yeah. Over 17.2 billion views. Billion with a B.

24:17Thomas Germain: Billion. Do you have any guesses what it is? Oh, it's not. I don't think it's a Taylor Swift video, but that was my first thought. It is not. Yeah. It is Baby Shark. Oh, of course. Yeah. Baby Shark. Right. That makes sense. The 17.2 billion makes sense because it's parents that probably sat down and played it on repeat. Again and again and again. Parents, you have our sympathy. Yeah. But you probably played it over and over and over again, which kind of inflated that view count. So what I teach, I teach a content creation class. I teach college students how to be content creators. And one of the things that I say in that class now is, you know, students will come to me and they're like, oh, my video, you know, it did really well for me.

24:58I got 10 ,000 views. It's like, I know that's not viral. And then I say, that's viral for you. And I do that one as kind of like a, hey, that's still really good. And that's a lot. But I also think viral for you demonstrates the shift. That it's not Mr. Beast. It's going to be 12 micro-influencer Mr. Beast that people can choose that is tailored to their specific interests. So it is viral for them, but viral as a whole, as we once knew it, with even Baby Shark, I would say, Gangnam Style, that doesn't exist anymore. And I think it has a lot to do, and I found in my own research, virality today is so much more reactive than it used to be proactive.

25:38And proactive is maybe the wrong word to describe it, but it used to be there was the video of Sophia Grace who was lip syncing to Nicki Minaj's Super Bass and wound up on Ellen, The Ellen Show. Right? So it used to be there was something funny, interesting, cute, out of pocket, things we didn't usually expect. And then it would get attention from social media. It would get attention from the news. It would get attention from magazines and daytime talk shows. instead now there was nothing that really captures our attention like that instead we are just reacting positively negatively to something we encounter on the internet and that really what is what drives virality it's not something that becomes what we call a water cooler conversation of media anymore where you could gather around the water cooler in your office and everybody could talk about the thing that was on tv was in the news we don't have that anymore every and i think my students, which is purely anecdotal story, but I think really underscores just how far we've gotten from that water cooler conversation.

26:45Thomas Germain: Here's my question for you. Yes. Is this now the way that things will be forever, right? Like, I think in particular vertical video and the switch to just like the, like we are endless scrolls in old term, but now like because everything is this quick, like, you know, 30 second hit, it's a new way of organizing information. Is there no going back? Because I'm actually not sure. You know, I'm going to say unless we have some really crazy seismic shift, I don't think there's any, I don't think we're going back. I think we are all following our micro-influencers and micro-celebrities. And I don't know.

27:31I think bar, yeah, barring something really crazy, I think we stay on this path of fragmentation.

27:36Thomas Germain: I have an alternative theory. I'm not sure if I completely buy it, but I do think there's some potential here, right? There's been a lot of talk for decades now about like the death of the old media gatekeepers. It used to be like, if people were going to hear about it, it's because like some executive at a giant media conglomerate that had a TV station that hundreds of millions of people watched decided it was good enough for you to hear about. That's gone away. But now we've entered this era where everything is so completely flat that it's sort of impossible to keep up. If we think about streaming services, right?

28:19Thomas Germain: People talked about how great it was to cut the cord from cable, right? And now we have 14 streaming services. Now you have 14. You get to a point where it would be nice if someone packaged up all those services together and charged you one monthly subscription for it, sort of like cable. You just invented cable, Tom. I invented cable again. And I have a friend who works in the tech industry, and what he always talks about is it's entirely conceivable that social media in general is a generational trend. Right? If we went back to the late 90s, early 2000s, the idea that cable would become totally and completely irrelevant, that's going too far, but would be knocked so far off of its station, it would be kind of inconceivable.

29:07Thomas Germain: Same thing with radio before TV took over. It could just be that millennials and Gen Z, who eventually they'll be the old guys, mom and dad spent all their time on TikTok. Isn't that dumb and archaic and embarrassing? We could shift back because I think there's a need for that sort of thing when it becomes so difficult, when the world becomes so filled with noise. So maybe it goes back in the other direction or maybe there's just no one gets to be famous anymore. There's like the guy who's famous just to me and you'll never hear about him. My students asked me the other day, they said, Dr. Maddox, how do you stay so optimistic in all of this?

29:45And I said, I am actually really optimistic because every time the pendulum swings one way, it eventually swings back the other way. and as you were talking Tom I realized you know I think I may take back my previous statement of barring a seismic shift nothing changes I think we're already seeing the change right we're seeing a real backlash to being on social media all the time people you know are putting time constraints and time limits on their devices so they aren't on their phones as much in 2026 I'd say is the year of being off your phone is the year of being a little bit more analog and you're talking about

30:19Thomas Germain: being off your phone. I don't think it's the ear being off your phone. Okay, that's fair. But that is true. If you and everyone that you know is truly unsatisfied with the way that things are now, if you don't like it, that means that there's an opportunity for some new system to crop up and give you an experience that is more enriching. It's going to have to be as stimulating as TikTok, which might be difficult. But it is more possible than it feels when you're in the middle of the epic. Absolutely. Well, we got to wrap up here. But what I would recommend, we told you at the top of the show, now you've listened all the way through.

30:55Thomas Germain: If you're not subscribed to The Interface, what are you doing? What are you doing? I mean, it's worth thinking about what's going on in your life. I don't know if you want to talk about it. You can listen to The Interface wherever you get your podcast. Or if you are in the United Kingdom, you can hear The Interface on BBC Sounds. For everyone else, it's available on YouTube. Just search for The Interface Podcast. And if you want to get in touch with us, we love hearing from you. We've got one fan in particular from Cuba who's like our number one guy. I love hearing from this dude. We love hearing from him.

31:25Thomas Germain: Get in touch at theinterface at bbc.com.

31:45AI in one place. To put AI to work for people, visit servicenow.com.

From the publisher

Should you be nice to your chatbot? Anthropic says its Claude models will be able to end conversations if users are needlessly cruel. Thomas Germain and Jessica Maddox ask why AI companies are increasingly talking about AI welfare, and what it could mean for the rest of us if the industry starts weighing a model's wellbeing against ours.

Tom and Jess dig into the research behind the claims, the online backlash over an "AI torture chamber," and why consciousness is still an open question for scientists and philosophers.

Plus, Jess reveals her college freshmen don't know who Jake Paul is. Is going viral a thing of the past, and what replaces it when the internet splinters into millions of niches?

The Interface is your weekly guide to the tech rewiring your week and your world. Hosted by journalist Thomas Germain and “professor of the internet” Jessica Maddox, each episode unpacks, week by week, how technology is shaping all our futures. No jargon. Just two sharp voices debating the stories that matter - whether they shook a government, broke the internet, or quietly tipped the balance of power.

New episodes drop every Wednesday on BBC Sounds in the UK. Outside the UK, find us on BBC.com or wherever you get your podcasts, or watch the video version on YouTube (search “The Interface podcast”). To get in touch with the team: [email protected]

The Interface is a BBC Studios production. Producer: Becca Bryers Executive Editor: Philip Sellars

More from The Interface

All episodes