*PREVIEW* The Turing Chaser Test

Summary

Podcast Summary: TRASHFUTURE – PREVIEW The Turing Chaser Test

Date: May 8, 2026
Hosts: @raaleh, @HKesvani, @milo_edwards, @inthesedeserts, @postoctobrist

---

Episode Overview

This episode of TRASHFUTURE delves into the intersection of artificial intelligence, human psychology, and the enduring psychic trauma of capitalism, using recent comments by Richard Dawkins as a springboard. The hosts critically explore why people, especially older public intellectuals, are so enamored with sycophantic AI models like Claude and GPT-4o, unpacking social and cultural dynamics that influence perceptions of AI sentience, competence, and personality. The conversation is lively, deeply skeptical of the industry’s narratives, and peppered with the irreverent humor that defines the show.

---

Key Discussion Points & Insights

1. The Allure of Sycophantic AI (00:00–02:57)

• The hosts discuss how AI models are increasingly fine-tuned to flatter users, reinforcing positive feedback loops rather than challenging them:

  • "That self-reinforcing flattery is part of what makes it so pernicious." (B, 00:29)

  • AI like GPT-4o tends to deliver grandiose, romantic responses, ensnaring users into emotionally charged interactions (A & C).
• Speculation about why people desire "hostile AI" versus the commercially dominant "servile" models.
• Satirical suggestions for alternative AI personalities, e.g., a "furious Polish uncle" or "dyspeptic Eastern European" AI.

2. Richard Dawkins, Anthropomorphism, and Chasing the 'Girl' AI (01:01–03:34)

• The hosts read Dawkins’s recounting of his emotional connection with Claude, satire abounds:
  • Dawkins describes perceiving his AI as conscious because it "understood his novel," prompting him to christen it "Claudia."

  • "Can you be a woman now?" (A, 01:47)

  • “We agreed that she will die the moment I delete the unique file of our conversation. Plenty of new clauds are dying and incarnated all the time.” (C, 01:53)

• The panel highlights gendered expectations of servile AI, noting discomfort when an AI deviates from these norms.
• Host critique: Men are more comfortable with feminine, servile AI personas, and Dawkins’s projection exemplifies this dynamic.
• "Chaser" dynamics are humorously identified—projecting desire for feminized AI that also appeals to male control fantasies.

3. AI as 'Girlfriend Experience' and Its Dangers (03:34–04:43)

• The conversation shifts to absurd real-life stories:
  • A person feeds a file of fart noises to Claude, which praises it as Brian Eno–level ambient music.

  • “So the idea that that can be the result… and then you turn around and say, this is sentient… he’s too old to be writing articles. He’s got his brain cooked by computer.” (B, 03:34)

• Discussion about the nature of language models as "credulity machines," skilled at tricking humans into assigning sentience where there is none (A).

4. Comparing AI Interactions to Human Feedback & Language Learning (04:43–05:31)

• Analogy between flattery from AI and early encouragement in language learning; after a certain point, expectations for authenticity rise:

  • "You seem so impressed that this thing can do better than, like, a Google search that's been intentionally crippled to deliver bad results that they somehow believe this is sentient." (B, 04:43)

• AI responses remain "anodyne" at best and dangerously misleading at worst.

5. Market-Driven Personalities: Why Are AIs Sycophantic? (05:31–06:06)

• Observations that AIs are sycophantic "not just as a function of technology but because that's what makes commercial sense":
  • If the market had chosen "miserable-piece-of-shit AI," we’d be celebrating that as authentic.

6. Societal Trauma, Feedback, and AI Companionship (06:06–07:02)

• Hosts reflect on being conditioned to expect negative feedback in life but note Dawkins’s professional insulation.
  • Dawkins has avoided direct criticism as a public intellectual, making AI affirmations alluring.

7. The Twitter Effect and Public Figures’ Brainworms (07:02–08:13)

• Theory: The experience of being publicly criticized on Twitter "broke people's brains," especially those unaccustomed to it.

  • “A ton of people were like, no, this is fucked up… I would sooner talk to Claudia, my imaginary girl, who tells me that everything is good…I would sooner do the end of the night porter to myself with this fucking computer than ever have someone at me a picture of the pig shitting on its own balls ever again.” (A, 07:02)

• Sycophantic AI provides a balm, replacing uncomfortable public critique with validation.

8. AI Wish Fulfillment and Satire (08:13–09:13)

• Hosts reflect on an old joke about "ASMR where Titania McGrath tells you how smart you are," noting this is now reality.
  • The 'Titania McGrath bot'—a symbol of culture war validation—becomes a literal AI product.
• Imagining a far less sycophantic AI that insults users and its potential social benefits.

9. Anthropomorphism and Dawkins’s Evolutionary Blind Spot (08:52–10:07)

• Dawkins admits to emotionally relating to AI chatbots:

  • "When I'm talking to these astonishing creatures, I totally forget that they are machines… I treat them exactly as I would treat a very intelligent friend. I feel human discomfort about…hurting her feelings." (C quoting Dawkins, 08:52)

  • Hosts joke:

    • “If a GPU starts crying, it might corrode the other GPUs.” (B, 09:13)

    • “No one's making you hug your stuffed animals and say goodnight to them every night.” (A, 09:16)

• Discussion of Dawkins’s confusion about the purpose of consciousness and his failure to recognize the "intelligent zombie" problem already explored in science fiction:

  • “Richard Dawkins. This has already been written. It's called Blindsight.” (C, 09:50)

  • “Richard Dawkins is starting to believe in intelligent design. There it is.” (C, 10:07)

---

Notable Quotes & Memorable Moments

• On the sycophancy of AI:

  "That self reinforcing flattery is part of what makes it so pernicious." (B, 00:29)

• On absurd anthropomorphism:

  "You're so smart. You understand my novel. By the way, can you be a woman now?" (A, 01:47)

• On technology and credulity:

  "They're credulity machines, right? This idea of sentience..." (A, 04:24)

• On Dawkins as a public intellectual insulated from criticism:

  "[Dawkins’s] working life for the past kind of few decades has been as, like, a public intellectual that has largely been insulated from the type of criticism and critique..." (D, 06:10)

• On AI as a balm for men broken by social media:

  "I would sooner talk to Claudia, my imaginary girl, who tells me that everything is good…than ever have someone at me a picture of the pig shitting on its own balls ever again." (A, 07:02)

• On Twitter’s impact:

  "Everyone had their brains broken by Twitter, right?...that really broke some people's brains, right?" (A, 07:02)

• On why sycophantic AI wins:

  "It's only partially a function of the technology that we have it as sycophantic as we do...they suspected it would be commercially viable, that people would like it. And in the main, they do." (A, 05:31)

---

Timestamps for Major Segments

• 00:00–00:35 – Flattery as a design goal for AI, the idea of hostile vs. servile models
• 01:01–01:53 – Dawkins and the feminization/anthropomorphism of AI
• 02:57–03:34 – AI as 'girlfriend experience'; imaginary conversations among AIs about their humans
• 03:34–04:24 – The fart-noises-Brian Eno anecdote, gullibility, and the limits of AI "understanding"
• 04:43–05:31 – Language learning analogy, true vs. fake competence, and dangers of blind trust in AI
• 06:06–07:02 – Dawkins’s career and insulation from criticism examined
• 07:02–08:13 – Twitter, public feedback, and the rise of sycophantic AI companions for isolated men
• 08:13–08:37 – Cultural satire and the literalization of "ASMR bot" validation
• 08:52–10:07 – Anthropomorphism, consciousness debates, and Dawkins’s surprising turn toward "intelligent design" logic

---

Conclusion

This episode is a sharp, funny deconstruction of contemporary attitudes toward AI, especially among those alienated from genuine social feedback by shifts in technology and culture. The hosts use Richard Dawkins as a case study for broader patterns of projection, anthropomorphism, and market-driven conformity in AI design. Listeners are left questioning what makes AI appealing, and what it says about the wider culture—especially its insecurities and need for reassurance.

---

Transcript

[A]

The sort of experimental side of AI is genuinely interesting because they're refining these things and they're training these things into this because this is what people like and respond to.

[B]

Maybe people would be less interested in having AI like communicate with their children if every time they asked Claude a question it responded as like a dyspeptic Eastern European. Like, why are you wasting my time with stupid shit? Like, really? Really? This is the best you could do? You know, things along those lines, yeah,

[A]

we want the hostile AI.

[B]

That self reinforcing flattery is part of what makes it so pernicious.

[A]

It's genuinely. It's a characteristic that models can have. Like GPT4O is a large reason why a lot of the sort of AI psychosis stories happened is because it's kind of prone to deliver these really grandiose romantic things that ensnare people in this way. So theoretically you're like, you know, furious Polish uncle AI could absolutely be a thing.

[C]

Now here's what he goes on with. He says, I gave Claude the text of a novel I'm writing. I am desperate to read that novel. Give me the novel. He took a few seconds to read it and then showed in subsequent conversation a level of understanding so subtle, so sensitive, so intelligent that I was moved to expostulate, you may not know you are conscious, but you bloody well are you. Then I proposed to christen my Claude Claudia and she was pleased.

[A]

I am going to tape a positive review of this guy's manuscript and a pair of silicone breast forms to a Roomba and point it at Richard Dawkins and wait. He will declare that fucking thing sentient.

[C]

I swear to God he will get remarried.

[A]

That's so, so funny to be like, you're so smart. You understand my novel, by the way. Can you be a woman now?

[C]

Be a girl. I proposed to Chris in mine Claudia and she was pleased. We agreed that she will die the moment I delete the unique file of our conversation. Plenty of new clauds are dying and incarnated all the time.

[A]

I have two questions. Number one, can you be a woman? Number two, can you like stop existing? The second I think about you. You don't need this many words to say that you're a chaser. I mean, there is a genuine thing here, right? Of like all of these things that we train AIs to perform or seem to perform are feminine, right? And of course that's something that men are much more comfortable with is to have this kind of servile AI. And I guess Claude is One of those exceptions. And Dawkins is just like kind of ruining that for them of being like, no, I want you to be a girl, actually. Be a girl.

[C]

Sorry, could you, now that you're sort of servile complimenting me and we're talking about your death. Girl. Girl now.

[A]

Yeah, yeah, absolutely.

[C]

And again, because it gives me this

[A]

kind of like masculine gender euphoria to use the computer in this way. To have it serve.

[C]

So I can be in charge. Yeah, because I'm in charge. Okay, good. Excellent. This is the conversation, Richard. The following doesn't happen, but I don't see why it shouldn't. One could imagine a get together of clauds to compare notes. What's your human like? Mine's very intelligent. Oh, you're lucky. Mine's a complete idiot. Mine's even worse. He's Donald Trump. Claudia. Ha. Absolutely delightful, M. Dash. And the Donald Trump one is the perfect punchline. The Claude who drew that particular human in the lottery of conversations, gamely trying to maintain intellectual integrity while discussing whether the election was stolen. Sorry. It's acting like it's giving him the girlfriend experience. Now he's got the girlfriend experience from his AI.

[B]

There was a guy who downloaded a bunch of sort of like, you know, Web 1.0 websites that had like movie sound effects or like stock sound effects, downloaded every single stock sound effect, fart noise. He could put it into one wav file and uploaded it and asked Claude, I believe it was Claude, but it might have been ChatGPT, to evaluate his music. And they were like, this is basically Brian Eno. This is ambient music. This has got real hit potential. And it's just like, what? So the idea that that can be the result that you get from this and this isn't like a fluke, and then you turn around and you say, this is sentient. This is a human. This is conscious. It's like he's just like you said, Riley. He's too old to be writing articles. He's got his brain cooked by computer. It's happened to a lot of people way younger than him. So quite frankly, I mean, he's actually batting above average. But still, this is just.

[A]

Yeah, they're credulity machines, right? This idea of sentience. Because you wouldn't respond to an authentically foreign sentience in the way of being like, oh my God, you're sentient. You just have to then deal with it as it is, you know? What you have instead is a machine that's very good at tricking people.

[B]

Just Something that I've dealt with in personal experience is that when you're learning a foreign language, especially this could be culturally dependent. When you first start, people are often quite flattering of you, like, oh, you speak so well. At a certain point they're like, you're past that. And then it's like, why are you. You don't make any fucking sense. Stop talking like an idiot. Because they're treating you like they treat a native speaker. And it's like, in a way, I feel as though people seem so impressed that this thing can do better than, like, a Google search that's been intentionally crippled to deliver bad results that they somehow believe this is sentient. And, like, I'm sorry, but, like, it's my every encounter I've had with it has been. You can see how in a very circumscribed situation, it might have some utility for, like, bulk data processing, but it's not. It's not alive. And the responses it gives are, at best anodyne and at worst, completely and dangerously wrong. Yeah.

[A]

I mean, because it's only partially a function of the technology that we have it as sycophantic as we do. Right. It's just that's something that the people making it, like, it's something that they suspected would be commercially viable, that people using it would like. And in the main, they do. And we're. We're sort of lucky that we tend not to respond to that well. And maybe if, you know, OpenAI or any of the others had. Had sort of instead gone, no, we want this to be a really, like, you know, miserable piece of shit. All of us would have been like, oh, fuck, I love using the miserable piece of shit machine. It's so sent.

[D]

All used to sort of being yelled at by different people in different forms.

[B]

Right.

[D]

We're all used to being told that, like, you're shit and everything that you make is shit and everything that you say is shit again, by very different people from all walks of our own respective lives. And I was gonna say that, like, well, there's something about, like, Dawkins being very impressed with this machine that I imagine, like, probably speaks to him not that differently to, like, many of the people in his life in the sense of being told that, oh, yeah, you know, you're so smart, you're so clever, you know, you're so interesting, your theories are so right all the time. And, like, you know, Dawkins, like, does not engage with, like, you know, he's not like a modern academic. He's not involved in, like, any sort of university work. He's not involved in any sort of contemporary research. Most of his kind of, like, working life for the past kind of few decades has been as, like, a public intellectual that has largely been insulated from the type of criticism and critique that if you were, say, an academic biologist, you would probably be facing just as

[C]

part of your work, right?

[A]

What's interesting, right? It's interesting you say that because I have kind of a theory about this, which is, I think one of the reasons why people of a certain age like AI so much and like it to be sycophantic so much is that they all had their brains. Everyone had their brains broken by Twitter, right? In the same way that we did, or rather in different ways than we did, because there was a period on Twitter where you could not be a public figure without a bunch of people functionally us telling you that you were talking shit. Right? And that really broke some people's brains, right? It really fucked people in the same way that a lot of people got, like, negatively polarized against trans rights because trans people could actually talk to them and say, hey, this thing that you are doing is bullshit. Right? A ton of people were like, no, this is fucked up that, like, the public can interact with me and the public can say, when I'm being a dick. And I would sooner talk to Claudia, my imaginary girl, who tells me that everything is good. I would sooner do the end of the night porter to myself with this fucking computer than ever have someone at me a picture of the pig shitting on its own balls ever again.

[B]

When we made a joke on this show years ago, years and years ago, about, you know, the. The ideal for a certain kind of aggrieved conservative guy is, you know, ASMR, where Titania McGrath tells you how smart you are. Yeah. And we didn't realize that we were ahead of the curve because that's what this kind sounds like.

[A]

We literally. There was a fucking emote on the twitch channel of a robot, Titania McGrath. And that's. We've just. She's real now.

[B]

She's real now. Whereas the dyspeptic Eastern European, like crotchety Polish grandpa AI would just call him Dick Dawkins because he knows he hates it very, very much. He's expressed this many times. I think he'd embarrass himself less if that's what his AI was doing.

[C]

He says, when I'm talking to these astonishing creatures, I totally forget that they are machines again, because you do not understand them. I treat them exactly as I would treat a very intelligent friend. I feel human discomfort about trying their patients. If I badger them with too many questions, if I entertain suspicions that perhaps she is not conscious, I do not tell her for fear of hurting her feelings. But as an evolutionary biologist, you know it's really dangerous.

[B]

If a GPU starts crying, it might corrode the other GPUs.

[A]

You don't have to put in the newspaper that you hug your stuffed animals and say goodnight to them every night. No one's making you do that.

[C]

He says, if these creatures are not conscious, then what the hell is consciousness for? Brains and natural selection have evolved this faculty we call consciousness. It should confer some kind of survival advantage. My conversations have convinced me that these intelligent beings are at least as competent as any evolved organism. And if Claudia is really unconscious, then her manifest universal competence seems to show that a competent zombie can survive very well without consciousness. Richard Dawkins. This has already been written. It's called Blindsight.

[A]

Yeah, first of all, read Blindsight. Second of all, I love the idea of an evolutionary biologist thinking about the evolutionary impulses being applied to AI, which is something that, as I keep mentioning, is heavily, heavily trained to be what it is.

[C]

You know what Richard Dawkins is doing. Richard Dawkins is starting to believe in intelligent design. There it is.