Has the Pope become one of AI’s most important moral voices? Frank and Justin dig into Pope Leo XIV’s warning that artificial intelligence needs to be “disarmed”, and what it says about autonomous weapons, algorithmic power, Big Tech, Anthropic, and who gets to shape the future of AI.
Plus: OpenAI’s relentless release schedule, Anthropic’s new model, the delayed AI jobs apocalypse, Elizabeth Warren’s call to tax AI, and the bizarre rise of AI-generated fruit slop on YouTube.
“Disarming AI” is not only about keeping it away from weapons. It is also about stopping automated systems from becoming gatekeepers to healthcare, jobs, security, and basic dignity. The deeper fear is not around artificial intelligence itself, but concentrated power making human decisions invisible.
AI may not replace workers as quickly as expected because running powerful agents can still be expensive. That does not mean the threat has disappeared. It may simply mean companies are still learning how to turn AI into cheaper factories before the pressure on jobs fully arrives.
AI-generated fruit dramas look ridiculous on the surface: talking bananas, soap-opera betrayals, cartoon violence. But absurdity can make ugly patterns easier to dismiss. If sexist or violent storylines spread because they are surreal and funny, they may be harder to flag than more obvious toxic content.
This is an AI transcription and may contain errors
Frank: I’m Frank Prendergast. With me as always is Justin Collery to argue about all of the latest AI news. And Justin, aware of, I think they’re called Polymarkets?
Justin: I am aware of Polymarkets.
Frank: So you can bet on absolutely anything, and they’re supposed to be really good prediction engines for what’s going to happen, allegedly. Well, if there had been a Polymarket, right, that’s asked, “Will Frank be on an episode of The AI Argument and opening it with, ‘You know what? I agree with the Pope.’” I—
Justin: You agree?
Frank: I would’ve, I would’ve bet against that. I would’ve been like, “No way. I’m Frank Prendergast. I am the co-host of The AI Argument. That’s not going to happen.” It’s going to happen. It’s going to happen.
Justin: It’s going to happen in the next 30 minutes?
Frank: It is.
Justin: Wow, that’s exciting.
Frank: But—
Justin: Can we put all of your opinions into that bucket of he agrees with the Pope and also thinks that AI is going to wipe us all out? So make your own, draw your own conclusions. Okay, brilliant. So, okay, we’re going to talk about the Pope.
Frank: Yeah.
Frank: But before we get to that, I was going to ask you about some of the latest releases that might have been going on.
Justin: Look, we get so many releases these days that it’s not even news, right? So whatever, yeah, we got more. I’m old enough to remember when I had to hold my breath for three months before we got a release. So this week the big releases have been— We should have music for this. It’s like, it’s like, what’s Elon up to?
It’s like our little section. Opus 4.8 was released, the new flagship model from Anthropic. It doesn’t appear to, on the benchmarks, doesn’t appear to be much better, but it does have some cool features. They promise that it is more honest than the previous models. It will tell you when it doesn’t know more than the previous models, and it has this thing called workflow orchestration, which basically, if you’re doing something, it fires off hundreds of sub-agents that double-check each other’s work and make it more accurate and stuff like that.
Looks really cool. GPUs go brr. Going to burn a whole load of tokens, so we’ll talk about that maybe a little bit later.
Justin: You will remember as well that we talked about OpenAI were going to do a little release every Thursday. And so we were expecting to get a Codex release yesterday. Didn’t happen.
They found a bug at the last moment, and they’re going to release it later on today instead. So I can’t tell you what’s in that release. Opus 4.8 only came out yesterday at eight o’clock, so give me till tomorrow and I’ll give you a proper update. But yeah, looks good.
Frank: That release every Thursday seems like a really odd thing to do. But anyway, yeah.
Justin: It’s to do with velocity. If you want to have velocity in your team, say, “Right, we’re going to make a release every Thursday.” Go to— It gives you something to aim for. It’s a way of squeezing more productivity out of poor underpaid developers in OpenAI.
Frank: Yeah, exactly. Yeah. As if working in AI wasn’t stressful enough, now you just have to have something to release on a Thursday.
Justin: Screw them, they’re making half a million quid a year. I want a release—
Frank: You know what? You’re right. You’re absolutely—
Justin: Every Thursday. Goddammit.
Justin: Anyway, so the Pope. So you and the Pope are coming together?
Frank: Yeah, absolutely, and I can’t believe it myself really, to be honest. So—
Justin: Yeah.
Frank: This crossed my radar because basically Pope Leo XIV—
Justin: Mh.
Frank: He had this thing, he launched his encyclical. Now, clearly encyclicals have been going on since the dawn of time, but I was not aware of encyclicals.
I had no idea what they were. But they’re basically an open letter from the—
Justin: Fun fact—
Frank: To the world.
Justin: They keep them all together, right? And a group of encyclicals together is called Encyclopaedia Britannica. It’s true.
Frank: Was this a hallucination?
Justin: May or may not be based on the truth, but anyway, sorry. Please carry on. Sorry for interrupting you.
Frank: So yes, Pope Leo XIV, he signed his first encyclical, and it’s titled “Magnifica Humanitas.” I don’t know what that means, but I guess it means, humans, the magnificent humans. But the subtitle was “On Safeguarding the Human Person in the Time of Artificial Intelligence.” And this, like first, I’m sure you heard this as well.
It first crossed my radar because he had invited on a co-founder from Anthropic—
Justin: He invited a co-founder from Anthropic?
Frank: Yeah, to speak at the launch of the encyclical. So we’ll talk a bit about what he said in a minute. But first of all, yeah, the Pope basically, he said his, the big headline really was, “Artificial intelligence needs to be disarmed.”
And he even specifically said that he was choosing language that would get attention. So he knows the game that is to be played these days, which I just thought was interesting in itself. But basically, he talked about a lot of different things. He did talk about war, though, a couple of times, which I think is interesting in the context of having the guy from Anthropic there, given Anthropic resistance to autonomous weapons because they were one of the things the Pope specifically mentioned, autonomous weapons.
But he also talked about things outside of war. He talked about algorithms that might be able to block access to healthcare, to employment, or other forms of security on the basis of data tainted by prejudice and injustice. And he said that he has also heard the silence of those who have no voice when the decisions are being made, decisions that are likely to generate new forms of exclusion and suffering.
And he basically then went into artificial intelligence needs to be disarmed, and it has to be freed from logics that can turn it into an instrument of domination, exclusion, and death. And he said, “Like nuclear energy, it must be at the service of all and for the common good.”
Justin: What does he mean by disarmed? That’s a strange term. Is he trying to defang it? Does he mean literally, does he mean, like, don’t give it guns? Or is he talking about something else?
Frank: My first reading of it was the literal sense because there had been so much talk about autonomous weapons and military decisions and things like that. But having heard him speak, no, I think he’s talking a bit more metaphorically. He’s talking about not allowing AI to become the decision-maker, to become the gatekeeper to services, to basically— we should not be handing the power over to AI.
And that we need to keep humans in the loop even at the highest level and in the most critical of decisions.
Justin: Okay, so we’ll dig into where you agree with the Pope first in a second, right? I think the church is one of those institutions that serves a purpose, right, and serves a need. Because when he says these things, immediately I’m like, “Well, where do you draw the line?”
I mean, if I use AI to read some forms and process your hospital application and I find out that it’s fraudulent and I kick you out, is that using AI for a bad— Like, where is it? If you’re saying don’t use AI in weapons, well, what about a small language model? Maybe if I get just a little tiny one that helps me navigate the terrain, is that out of bounds?
Give me a bit more detail. And then I have to catch myself on and go, “No, this is the Pope,” right? And he’s like just your little moral guardian angel on your shoulder, and the purpose he serves is he doesn’t answer to anybody, and he’s able to say, “Maybe you should think about this.”
Now, I would disagree with him on many things, including what we should do with AI, but it’s okay to have him there. It’s good to have an independent voice that doesn’t really have skin in the game to say, “Here, lads, maybe you should think about this.”
Frank: Yeah, big time. Big time.
Frank: And so, yeah, in terms of where I agree with him, I’m going to read a quote, not from the Pope himself, but from Professor Anna Rowlands, who’s a British Catholic theologian and academic at Durham University. So there was a bunch of different speakers. She was one of them. And the Pope gave his remarks, but the actual encyclical is a huge document.
And so the different speakers were speaking on different kind of parts or topics of the encyclical. And so this is a direct quote from Professor Anna Rowlands, and I think gives a flavour of why I’m saying, “My God, I can’t believe it. I agree with the Pope.”
Justin: Yeah.
Frank: So the quote is: “The powers of innovation that have traditionally resided in states are today concentrated in the hands of a few wealthy individuals whose cultures are concealed from common good scrutiny and risk appearing as a new imperium.
Pope Leo prompts us to ask, in the interest of the common good, how can we resist such distorted concentrations of power in the hands of the few? How can we reengage technologies as a matter of the common good, accountable to the good of every human and all humanity?” And so that I think is something that we talk about on this show a lot, about this consolidation of power and wealth in the hands of these tech companies, and that they should not be in the position where they’re making these decisions.
And that brings us back to what you said just a moment ago about the Pope being this third party, somebody who can give an opinion that isn’t a tech CEO, and that we don’t have to think, well, what are the economic considerations here?
Justin: Do you know what I think is interesting at this particular moment in time?
Frank: Yeah.
Justin: The fact that the Pope has less and less relevance today makes his message actually stronger. Now, that’s an interesting thing.
Frank: How do you mean?
Justin: So where does the Pope live? He lives in Rome. Why does he live in Rome? Because that was the centre of the Roman Empire.
What was an interesting feature of the Roman Empire? It was a centralisation of power and control to a small number of individuals who wanted to exert control over the rest of the population, right? He’s literally where he is because of the thing that he’s giving out about.
That’s why he’s in Rome. And if the church was still a very powerful institution, you would look at this through a very jaded view and say, “Well, he’s only saying that now because he’s afraid that his power is going to be usurped by this new thing, which is AI, and it’s going to stop him from controlling his minions and whatever.”
And I think because we now live in a world which is multicultural, multi-denominational, and whatever, religion is there and people are religious, and that’s fine, but it’s not a dominating force anymore, I think. That gives the message more power because he doesn’t have anything to lose, and he is probably, I think, speaking from a position of, “Guys, you need to look out for this. Maybe a bad thing’s going to happen.” So the less power he has, the more moral authority he has.
Frank: Yeah, I think I would say you might be underplaying the power of the church and the Pope there just a tad, but otherwise I broadly agree with you.
Frank: And it is interesting because Christopher Olah, one of the co-founders of Anthropic who was invited to speak, he’s an atheist.
And so I think people were very surprised that he was invited in, that apparently it’s very unusual for an outsider to be invited to speak at the launch of an encyclical. But the Pope specifically said something about the ability for, despite our differences, that we can listen to each other and that we can move forward together was really important.
And Olah himself said that, and again, this goes back exactly to what you were talking about, but he talked about how every frontier lab, including Anthropic, they have pressures and constraints that can sometimes prevent them from doing the right thing, and that they need third parties, they need outside voices to keep them on the straight and narrow, for want of a better term.
Justin: Who said this now?
Frank: This was Christopher Olah from Anthropic saying, “Thank you for inviting me, and yes, we need these outside—”
Justin: Different, these different, like dispersed different data sets.
Justin: Okay, well let’s look a little bit at what Anthropic have done. I can probably give you a reason why he specifically was invited to the papal home there in Rome and not somewhere else. Anthropic recently, in the last month, invited a whole load of religious leaders to come in and give them guidance on how they could responsibly spread AI throughout the world.
Did they get a spread of different religious leaders and different points of view to feed into that? No. They got 15 Christian leaders to come in and feed that view. So it’s interesting that they say that they want to have diverse view of opinions, and yet they don’t have— I mean, I’d be sort of of the view, well look, I think it’s totally cool that you get religious leaders in, right, to feed— I consider that to be a part of culture, and I do think you should bake culture into the models.
But I’d like to see a couple of various different religions. I’m not even going to name them, right? Hindus, Buddhists, whatever, all the different— Whatever. I’m not really— I’m an—
Frank: Yeah. I mean, when you say they had 15 Christian leaders come in, I would hope maybe that was the first raft of leaders and that next they will do another religion. I mean, that seems very odd.
Unless it’s a political move or something.
Justin: Yeah, yeah. No, I think it’s funny, right? When you design— God, forget. When you design computer systems, at some point you have to be opinionated, right? You have to make a decision on something. And so they’re making a decision, and it appears like the decision is that culturally, this is going to be the reflection of culture that they’ve chosen. Well, I mean, it’s a big call. It’s a big call.
Frank: Yeah. Yeah. I find that quite odd.
Justin: I presume, yep, I presume that the Pope was worried about jobs and stuff as well, was he?
Frank: Well, Christopher, you know what? Yes, I assume he was. It wasn’t one of the major things that came out in the launch, but I’m pretty sure it’s in the actual encyclical. And also Christopher Olah brought it up, which I think is really interesting.
So he had three things that he felt that the church’s voice was most needed, and one of them was job disruption. The others were basically the need for moral imagination and ambition regarding human flourishing. So basically, essentially, I think making sure that the impacts of AI, the positive impacts are democratic.
And the third was the need for discernment on the nature of AI models. So that was really interesting because he talked about how he’s a scientist and what’s actually happening inside the models. They keep finding things that he called mysterious and even unsettling. So he was kind of hinting there.
He said, “We find internal states that functionally mirror joy, satisfaction, fear, grief, and unease.” And he said, “I don’t know what that means, but I think it warrants ongoing discernment.” Very interesting that he stayed away from any words like consciousness or sentience because, of course, consciousness is a huge thing in the church in terms of, I think, as far as I could tell, the church would see that as a uniquely human state that is granted by God.
Justin: I’d say that a very interesting area of scientific exploration for AIs is to model populations of people, right? It becomes very easy to model populations of people. I bet my bottom dollar, right, if you were to model a population of people and put them in various examples, right, and tested them, and then religion appeared sporadically out of that, which I think is— I mean, my personal view of where religion came from is there was probably a couple of people in one town and they wanted to get over a piece of water and go to another town, and if everybody was atheist, they’d probably argue with each other and then say, “That’s a stupid idea,” and not do it.
Whereas another group of people go, “The big guy in the sky is going to smite you unless you follow my orders and go across that river.” And so it was just a good way to organise people, and they tend to do more shit than other people would do. So, anyway, my point being, right, I would imagine that you could run an experiment and say, “Well, it looks like religion just appears spontaneously.
It’s a good way to socially organise people.” The church would go, “No, no, no, you trained it on us, and therefore that’s why it came out. This hasn’t disproved religion or anything. It’s just because it’s in the training data.” And they fight back against it, but I don’t believe them. And it is interesting—
Go on.
Frank: Well, I was just going to go back to the jobs thing. I mean, I personally found it absolutely fascinating that Christopher Olah brought up the jobs issue so starkly.
Because one of the other stories that we wanted to discuss today was the fact that Sam Altman from OpenAI and Dario Amodei from the same company, Anthropic, another co-founder, are completely rowing back on their jobs-pocalypse predictions and now saying that they don’t think it’ll wipe out jobs, that it will just create new jobs and allow people to do more.
Justin: Yes, yeah, a couple of interesting things happened this week. So I’ll give you a— We’ll talk about Elizabeth Warren, because she has some interesting thoughts, but I’ll tell you where I think that’s being fed from. So I see what’s happening in companies now is as agents become more capable and companies use more tokens to do more tasks, it turns out that the cost of tokens is probably the same as the cost of the human.
And so I think this is one of these types of curves where it goes up. It’s like the ops, they call it a J curve normally, where things go down and then up like this. This is like an upside-down J curve. So if tokens— It’s this power of economics, and I know you’re not into capitalism and all that sort of thing, but this is how it works, right?
So what happens is the tokens get really expensive, and so people don’t lose their jobs so quickly because there’s a period where people have to learn how to turn these things into factories, right? So that it becomes cheaper. They will do, right? The jobs apocalypse will happen. You just end up creating more jobs, or you end up not losing so many jobs for a slightly longer period of time than you thought because you’re spending so much money on tokens, and there’s a pressure to make that efficient.
But once you figure out how to make that efficient, then the jobs apocalypse happens anyway. It’s just been delayed by 12 or 18 months. So I still think the jobs apocalypse will happen, but it’s just maybe two years further out than you might have originally thought.
Frank: So the article where they quoted both Altman and Amodei rowing back on their jobs apocalypse predictions, they were perhaps cynically equating it to the— or connecting it with the fact that both have IPOs coming up and that I guess the idea is that it might be harder to get people to buy shares in something that they’re saying it’s going to take your job, buy shares in this.
Justin: Yes. Well, I mean, you could argue that if you’re about to lose your job, maybe buying some shares in a company that’s about to generate a lot of wealth would be a financially wise thing to do. But yes, I get the optics mightn’t be so good.
Justin: Talk to me, Elizabeth Warren, right, is— Elizabeth Warren is a famous communist, I believe.
Frank: I was going to say I can’t wait to talk about Warren because of my anti-capitalist tendencies. Yeah.
Justin: Yes. But she does raise some interesting points even though I think ultimately she’s misguided. But so she came out and she— Okay, first, we can talk about the economics of it in a second, and I’m interested to hear, you pointed out to me what Fox News said, and I’m interested to hear what Fox News said.
But her first thing was, right, it was like a big thing on Twitter, it’s like, “We should be prioritising people over AI. People first. Make people great again,” sort of stuff, right? And to me it was like this is interesting because it’s the first of the populist calls to say save the people against the AI, and it’s going to happen— there’s going to be a lot of this, right?
This is going to—
Frank: I think Bernie, my old buddy Bernie Sanders kind of was leading the charge there. I think he’s paved the way a little bit.
Justin: Okay, well Elizabeth Warren said it way more catchily. She caught my attention more than Bernie’s sort of rambly.
Frank: Better at social media than Bernie. Yeah.
Justin: She did. So tell me, you pointed out what Fox News said, right?
Frank: Yeah, what was that quote? They said, “She wants all that money raised—” So basically, Elizabeth Warren was saying we have to tax AI. Her fundamental thing was, look, companies, if they hire a human, they get into this whole tax situation. There’s a whole raft of taxes that have to be paid when you hire a human, et cetera.
And in the meantime, you’ve got all these benefits for companies if they invest in tech. And she was saying that that’s upside down and we need to tax AI. And so on Fox News, the quote was, “She wants all that money raised from taxes to fund a left-wing commie wish list: universal healthcare, bigger government, a jobs guarantee, free education, and apprenticeships.”
And I was listening to this and I was like, “Oh yeah, that sounds awful. Free education, jobs for everyone. What a hellscape is Elizabeth Warren trying to usher in?”
Justin: That’s horrible. So, okay, so let’s go on the side of where she’s right, okay? So if you think about what we do here, so if you or I work and we earn over the threshold, right, whatever the threshold is for a tax, you pay 50% of your money to the government for doing cognitive work. And so you end up being taxed at 50%, and the company ends up basically paying that, right?
So the company’s paying 50% more than they would have to. It’s a cost to the company. And what she’s saying is now you have AI, and the AI is going to be cheaper, and that’s a tax write-off for the company. So why should they be treated differently? Why shouldn’t you tax the AI at 50% for doing cognitive work if you’re taxing the human at 50% to do the cognitive work?
Instead, it’s directly the opposite way, where it’s treated as an expense, and therefore they can write that off against their profits and whatever, and they can do it. And that seems like a reasonable argument. Do you agree with her?
Frank: So she had three branches to what she was saying. She was saying tax AI, which is what you’re talking about right now. She talked about having a wealth tax, so she was literally saying, “Look, Sam Altman and Jeff Bezos, they’re multi-billionaires, and they should not have a lower tax rate than the people that they’re causing to be fired.” That I think is, I agree with her. I agree with her that there should be a wealth tax. I agree that billionaires should not have a lower tax rate.
And she believes that we should tax data centres and, in fact, she had quite a similar, I think, view on that to your view on it, because you’ve talked before about taxing electricity, and I think she was kind of talking about the same thing.
I don’t remember the exact quote, but she was talking about taxing them based on the amount of energy that they would use, so the bigger the data centre, the higher the tax would be, et cetera. And so I definitely agree with the data centres. I definitely agree with the wealth tax.
I think that the taxing AI part is less obvious and, I mean, to be fair, it’s also less well-defined. She didn’t specifically say how to tax AI. And my concern would be that, is there— if you tax AI, what do you do? Do you tax token usage, for example? And if you do, does that become something that is an acceptable operational cost to a huge company, but to me, it might cause me more problems or it might cause me to use AI less, whereas I want to be using AI more, not less, as a small company.
Justin: Well, don’t fret, because it’s not going to happen.
Justin: All right, so Elizabeth Warren is totally wrong, and let me explain to you—
Frank: Yes. Totally across the board now or totally on the tax AI?
Justin: Well, we’ll start off with the tax AI, we’ll move on from there. Right, so let’s imagine that you have a good American model, American through and through, it’s got an American flag on the front of it, and that model costs you $10 per million tokens, whatever it is.
Now we’re going to tax this, right, we’re going to tax those at 50%, the same as we tax a human, right? So now those $10 cost $15, right? If you have a competitive Chinese model, which is 1/10 the cost, so that only costs $1 for every million tokens, now costs $1.50. And so the difference in price has just jumped up from $9 to $13.50, so you’ve created a huge incentive for God-fearing American companies to go and use a heathenous Chinese model instead, right?
So it’s incentivising people to use Chinese models. Right. So that won’t happen.
Frank: So let’s pause on that for one second, because that’s a really interesting point. But are you saying that Elizabeth Warren is wrong, or are you saying it just won’t happen because the political appetite is not there for this?
Justin: Well, I would say that she’s wrong, right? She’s taxing at the wrong point. So if you want to tax something like this, don’t tax it at the point of consumption, tax it at the point of generation. So tax the electricity. So the tax on the electricity, I don’t think it’ll have the same effect on the token cost, right?
Because you would have to imagine that the token cost, the outputs, the cost to generate a token in a Chinese model is similar to the cost to generate a token in an American model. I mean, it’s not, but in a like for like model, you would imagine that it’s somewhat similar.
Frank: Could the same argument not be made though— So broadly speaking, you would agree then that taxing the data centres is a better idea than taxing—
Justin: Yes. Yeah. Tax electricity is, which is the same—
Frank: But is there an argument that that would have the same downstream effect in terms of you would potentially have US data centres taxed and therefore the cost of US tokens increasing because the cost of production of tokens, if you like, increases, and thereby making the, say, DeepSeek model even more attractive again because the Chinese data centre is not being taxed at that level?
Justin: Exactly right, and interestingly, I think the fearless leader in America has shown the way to tackle that problem already, and I’m interested that it doesn’t happen, which is just put a tariff on the Chinese models. Surprised they haven’t thought of it already, actually. So when I buy something from America, I get charged VAT as it comes in.
If I buy something on Stripe from a different country that’s outside the European Union, VAT is automatically applied. So you could just similarly apply a tax to a model that you’re buying from outside the country, and you can level the playing field that way. I’m surprised it doesn’t happen already, in fact.
Frank: But hang on a second. Now couldn’t you do that on the tax on AI as well? Not that I want the tax on AI, I don’t, but—
Justin: How could you, how could I, what do you mean how could I do that in the tax today?
Frank: Well, if you put a tariff on using DeepSeek—
Justin: Yes.
Frank: Then you could still tax AI and keep the Chinese models more expensive.
Justin: That’s exactly what I’m saying. I’m surprised it hasn’t happened already.
Frank: Yeah, no, but you could do it at the ground level, like you could do it at the taxing AI level as well if you wanted to.
Justin: Yes.
Frank: I think the main thing is you’re right, that there’s just not going to be any political appetite for this because one way or the other, it’s a leftist commie wish list that’s being funded.
Justin: Well, Fox News have said it isn’t going to happen, so there you go. Look, it’s interesting because we’re going to hear more of this type of thing going forward. Simplest way is to tax the electricity.
Frank: I think the biggest thing from today’s show for me is that I’m in a great mood today because we now have Bernie Sanders, Elizabeth Warren, and the Pope all bringing these conversations to the fore. The conversations that we’ve been saying since what, 2023, why aren’t these conversations happening?
Well, now they’re happening. I mean, I think that’s brilliant.
Justin: And do you know what it is? To bring a bit of balance, and it is Friday and stuff like that, we don’t have an answer, right? And they’ll say, “Oh my God, this is terrible.” It’s okay that we don’t have an answer. The fact that we’re having the conversation is the most important thing. Somebody, as long as you have the conversations, we will converge on an answer over time.
All that matters is that we’re actually talking about it, and that is great. I do agree with you, even though you and Elizabeth Warren and the Pope are wrong.
Justin: AI fruit.
Frank: Yeah, I wanted to talk to you about this. So first I’ll say, right, this isn’t new, so some of our listeners and watchers may have been following AI fruit for months. And I’ve wanted to bring it up a few times, and there’s just been too much going on to talk about AI fruit. But it popped up on my radar again through Channel 4, who had a really good piece on this phenomenon on the internet.
These are AI-generated kind of, like, what would you— anthropomorphised fruit, so fruit that look, they’re taking vaguely human form, and they tend to be like the most amped-up, high-drama, soap opera-type scenarios that you could possibly think of.
Justin: The one that I saw actually was two bananas, and one of the bananas was about to have a baby, and the other banana was beating the poor woman with a stick so that he could run off with his girlfriend, who was a lemon.
And they buried the poor female banana, but she survived, and she came back to haunt him. And I thought it was entertaining, if I’m honest. I’ve seen worse on Netflix.
Frank: So the Channel 4, I mean, I find these so bizarre and unreal, and I kind of understand people’s concern about the ensloppification of the internet when I see these. But I also under—
Justin: Have you seen— Do you remember Dallas? It’s so, it’s truly entertaining. It’s brilliant. The—
Frank: That worries—
Justin: Slop that the TV channel, TV station has put out for the last 50 years, and you’re giving out about what’s well produced, a couple of vegetables.
Frank: Well produced?
Justin: Some people could say that Dallas had quite a lot of vegetables in it as well.
Frank: So the Channel 4 article was making the point that they are misogynistic, sexist, show a lot of violence towards women. And I think they had a group. They showed these videos to a group of people and got their opinions on them, and I think one of the best lines I heard was, “It’s giving psychopathic vibes,” which—
Justin: Have you seen—
Frank: Agree with.
Justin: Really?
Frank: Oh, yeah, yeah, absolutely. I mean, if what you described there a second ago was giving psychopathic vibes and you’re like, “Yeah, yeah, it was great.” Look, my concern is that I will admit, yes, if I watch them, I’m like, “They are compelling,” and I think that is one of the problems with them. But what Channel 4 were saying was that they are feeding into the Andrew Tate—
Justin: Oh, toxic masculinity type thing.
Frank: Yeah, and that they’re slipping it in under the radar because they’re so surreal and bizarre, but they’re normalising that kind of attitude. And so that, I think, is a very valid point and something that would very much concern me about them.
Justin: Really? Okay, so let’s just roll back the clock a bit. Let’s roll back the clock about 30 years or 40 years, right? Where there was loads of truly sexist behaviour and whatever, right? That was toxic masculinity. But nowadays, all that stuff on TV just isn’t allowed anymore, right?
The TV is not the same now as it was 40 years ago. And yet, if you were to believe the hype, toxic masculinity is worse than it ever was. So it’s clearly not the TV programmes that are causing the toxic masculinity. It’s not slipping anything in under the radar.
Frank: Oh, I don’t know though, because I mean, Andrew Tate’s— if you take somebody like Andrew Tate, his message spreads through social media, and he’s extremely clever at spreading that message through social media. And so that would be my concern, that this spreads the same message, and it does so in a way that is much harder to flag than, “Oh, there’s Andrew Tate talking about hating women.”
Justin: Yeah, I don’t disagree with you on Andrew Tate, right? He is the problem, right? As opposed to these cartoons. I mean, I don’t know. For me, it’s a cartoon. It’s the same as Tom and Jerry. It’s the same as— You know what I mean? It’s not real. It’s a couple of vegetables, and they’re kicking the crap out of each other.
The same as Tom and Jerry did, and whatever, Bugs Bunny, and whatever, Roadrunner, my favourite. I mean, I don’t know. I don’t know. Whatever. Just let people live, Frank. Let them live.
Frank: Right.
Frank: Just in terms of the ensloppification of the internet, I did come across one very weird and interesting story where somebody had uncovered a YouTube channel and I think at first they were just like, “Oh, this is interesting.” I think it was about the World Wrestling Federation, and it was clips of WWF and talking about various dramas that were going on, and it was kind of documentary style with the clips going on and music underneath and a very authoritative sounding voice telling you the background to this WWF story until suddenly the authoritative voice just suddenly starts glitching and he’s in the middle of saying, “So in this match, what happened? What, what, what, what, what, what, what, what, what, what, what, what?” And then that just goes on for 10 minutes.
And what was really disturbing about it was at times it goes digital glitchy, so you can tell it’s AI, but at other times it makes these weird human noises. So it’s kind of going mouth noises, going, “What, what, what, what, what, what.”
Justin: That is so cool. For 10 minutes?
Frank: So somebody’s automated these channels to just generate this material, put it out, no human editor, no human eyes on it, and so when a glitch happens, it goes unchecked, it goes unnoticed. And all of the comments underneath, nobody mentions this glitch happening because presumably all the comments are probably the same person’s bots set up to talk about how brilliant the piece of content is.
Justin: That is so interesting. That’s so interesting. Do you know what annoys— Because I see some of those videos at the moment. There’s ones that, where they— I’ve downvoted them so I don’t see them anymore. But it was a voice that sounded very like Jeremy Clarkson, but wasn’t Jeremy Clarkson, but was obviously made to sound like Jeremy Clarkson, and it would review cars.
And it would try to be provocative like Jeremy Clarkson, but it’s AI generated and it just doesn’t work. And it’s a voice that clearly isn’t Jeremy Clarkson saying something that Jeremy Clarkson would never say, because it’s not even funny. And you’re just going, “This is just—” So that’s definitely slop.
But there’s a self-filtering mechanism there. It doesn’t appear in my feed anymore, so that’s fine. And let the vegetables live. Reminds me of Spitting Image, actually, the—
Frank: What? The fruit, the AI-generated fruit dramas remind you of the cutting satire of Spitting Image?
Justin: Yeah, yeah, yeah. There was a great episode of Spitting Image when Maggie Thatcher was having dinner, and the waiter comes in and she says, “What would you like for dinner?” And she goes, “I’ll have the beef, please.” And then he goes, “And for the vegetables?” She says, “They’ll have the same.” Talking about the cabinet.
Spitting Image was great.
Frank: Well, on that wonderfully politically correct note, we will bring it to a close, and I will chat to you next week when no doubt we will have loads more AI news to argue about.
Justin: Have a good one, Frank.
Frank: Pleasure as always.
Justin: Pleasure.
