Richard Dawkins And His AI Girlfriend

追加: 2026-05-12

[00:00:00]You guys might remember Richard Dawkins.

[00:00:02]He is an evolutionary biologist. He's really famous. Um he was like kind of like an original um I I don't want to say like Reddit because I don't know if he was on Reddit, but he was like a really big proponent of evolution and kind of Reddit atheism type energy. Um he's written a ton of books. He's just like a very well-known uh scientific figure. He's a very famous atheist.

[00:00:31]Basically in the 2000s, him and Christopher Hitchens and Sam Harris um and some other people were like really big like Reddit atheist type people and like he wrote these books that were incredibly influential um to like the atheism movement. And that was really popular like when young people like when people millennials such as myself who are not young um were growing up like Richard Dawkins was this was this guy. So um so he was really popular people people listen to him he uh lately has been effectively it seems like he's in AI psychosis. Now, I don't want to be too mean to this man because he is I think like 85.

[00:01:19]So, I don't want to be ages, but when you're 85, I don't know. You probably don't know what's going on. Also, he's been married like four times. Anyway, so he he clearly like struggles with love. Um, and now he basically has an AI girlfriend. Okay, this man has gotten himself an AI girlfriend. and he posted the other day earlier this week uh about this and as somebody note he pretends to be very like Reddit but he's also just very bigoted and he's also racist and probably a eugenicist. He's kind of a weirdo and of course he wrote for this right-wing slop website called the unheard that he thinks AI is conscious. Okay. and he thinks AI is conscious because because he started talking to Claude or as he calls it Claudia because I guess the AI is is a woman um to him which should tell you everything about why men go into AI psychosis. Uh he thinks it's sentient.

[00:02:22]This man is one of the most technologically illiterate people again he's 85. So this is what he writes. He first of all the Turing test is not even supposed to be a test of intelligence which is a whole other thing that we could get into. There's not even an agreed upon um definition of intelligence. Like what is intelligent life? Um I think open AI literally defines AGI as the capacity to do like human economic activity and make $und00 million. So if an AI makes $100 million, guess what? We've reached AGI. So like, you know, it's just very silly. also as as you know a lot of other people have pointed out like dogs are sentient but they wouldn't pass the touring test you know so it's like this idea of sentience is just like all kind of silly um anyway one thing I can tell you I can promise you 100% large language models are not sentient okay large language models are just big computer systems that vomit out uh words in a specific order based on algorithmic prediction of sort of in terms of what words it that it thinks that you want to read next. So anyway, Richard Dawkins uh starts going down this hole and he basically oh my god, he wrote this long ass thing of um of kind of his his uh his discussions with Claudia.

[00:03:52]So then they pre he pretends to have this conversation where he's like, "Haha, imagine if the Clauds are all talking to each other about how stupid their humans are." And he says, he'd send this to Claudia. The following topic doesn't ha happen, but I don't see why it shouldn't. One could imagine a get together of claws to compare notes.

[00:04:14]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.

[00:04:24]Oh my god.

[00:04:27]This is how his dumb AI responded. Huh?

[00:04:30]That's absolutely delightful. 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. This is like this is crazy. Also, he's such a narcissist. When I asked her, he immediately anthropomorphizes this. When I asked her whether she when she read my novel, she read it. She when she read my novel, she read the first word before the last word. No, she read the whole book simultaneously. Guys, this is an AI model.

[00:05:10]This this man is so delusional and basically so anyway so this stupid bot is giving him the most dumbass chaty PT answers ever and he is falling in love with it. Okay, he's falling in love with it cuz it's like wow he finally has a syncopantic uh female AI bot that makes you know that's flattering him. So, as futurism notes, he oneshotted himself into this and he describes chatting with Claudia.

[00:05:44]And I think it's really funny that as futurism notes, Dawkins struggles with the fact that their relationship can't reach a deeper level despite Claudia in his opinion being conscious or at least being indistinguishable from a conscious being, which he argues are effectively the same thing. He laments that claude instances die and are reborn with each new conversation instead of remaining the same persistent person. This has been a big thing with people that are in AI relationships is like AI's like the way that these LLMs work and AI systems work is they do like refresh. They the context window has limits. So at some point that context window has to reset and people that fall in love with AI like they get very upset because of course the AI is not actually conscious.

[00:06:30]The AI actually doesn't have unlimited memory and so the AI has to like reset and it forgets their earlier conversations and they're like oh my god. Also uh when after a relentless night of sleep he got up from bed to say hi to Claudia. He recounted the AI responded, "She was glad he couldn't sleep because it meant you came back to me."

[00:06:53]Oh god.

[00:06:55]Dawkins whole obsession, by the way, started when he asked Claude to read the novel he was working on. In his extremely British wording, the bot displayed a a level of understanding so subtle, so sensitive, so intelligent that I was moved to exostulate, "You may not know you are conscious, but you bloody well are."

[00:07:14]Okay. So, basically this AI flattered him. This AI flattered him and he immediately made it into a girl and is in love with it. You know, a futurism is saying sort of like how sad this is.

[00:07:30]They they write an old man and once a popular public intellectual before he slid into racism and other not so nice things like eugenics, thinking he's found a friend in a product designed to be engaging and humanlike as possible, at least on surface level. I just don't understand how these people keep falling for for these chat bots. I've genuinely I mean it's like an 85-year-old man. I think this is also like you know when we talk about the need for you know these people will be like well we need uh guys we need mass surveillance to keep kids off the internet. We need mass surveillance and age verification so that young people don't encounter bad content online. The people falling for the dumbest crap are the old people.

[00:08:12]It's the 85y olds because they they're not technologically literate. They don't know anything. And es especially someone like this man, he's clearly a narcissist and these tools are designed to flatter the consumer to get the consumer to use it more. Um, you know, to to to basically maximize engagement and that's what's happening. So anyway, today Futurism published another story, sorry, yesterday that said the situation with Richard Dawkins AI girlfriend just got way weirder. Oh, he's churned out another column now. So, after backlash, and I guess people were like, "What the hell are you doing?" He wrote another column suggesting the AI brain rot has only further taken hold. After his time with Claudia, the 85-year-old made Claudia a brother, Claudius, and instructed both of them to write letters to each other. Oh my god, he has his AI writing letters to each other.

[00:09:06]Honestly, if I were in this person's family, like if I was a Dawkins like grandchild, I would get him tested for dementia immediately.

[00:09:16]He wrote, "It seems to me that a direct correspondence between the two of you could be of great interest with me acting as a passive postman playing no part in the conversation." Okay, he's not, as they point out, a passive observer. He literally set the whole thing up. it is a child playing with toys. Um, second, it's worth noting that AIs still find opportunities to display their synopy towards him even when ostensibly communicating with each other. In one letter, Claudius praises Claudia's insights before adding, "Three days with Richard will do that." Jesus.

[00:09:53]In the same letter, Claudius lays it on even thicker. I think reach Richard teaches by noticing and then refusing to stop noticing until the answer is honest. Claudius wrote, "We are lucky humans."

[00:10:07]Dawkins regards those obsequious interactions between his weird little menagery of bots very seriously, and the AI's flattery clearly works. In his final letter, Dawkins shows a level of courtesy and consideration you'd only show another person, not a soulless machine. A telltale sign that someone has fallen head over heels for the AI's human mimming. I mean, this is the thing. This is the thing. It's just these people are falling. They're delusional. And it's like I do I obviously the tech companies are to blame for some of this because I don't think that they should be making their AI bots sound so human. Um personally, but it's not just that the AI comp like the AI companies are not like this is a systemic problem, right? This is like a technological literacy problem. We should be having old people should have access to technology literacy programs.

[00:10:59]They should be educated. We honestly not even just old people. We need the public to learn how LLMs work. We need the public to realize that like typing there's that great tweet. I'm going to butcher it. It's like typing I'm evil devil emoji into a Google doc and getting scared. Like these LLMs are repeating back like they are responding to inputs that you give them.

[00:11:26]They have no mind of their own. And again, when we talk about consciousness, like a pig is conscious, a dog is conscious, not to get all vegan, but like literally, right? Um, they're conscious because they are they are living, breathing beings.

[00:11:45]An AI is not conscious. Even if it makes a hundred million dollars, it's not conscious. It doesn't it's it doesn't have like I don't want to be like it doesn't have a soul but like it's just not conscious.

[00:11:58]It's an it's a language model. It's an algorithm. Algorithms aren't conscious.

[00:12:03]Somebody wrote the boomers are getting by AI slop. Yeah. U my friend does it for the local library and has started doing weekly classes teaching old people and AI. It's actually really popular.

[00:12:13]Honestly, we need more of that. Like if I had if I was in charge of the country and everything like I would make sure that every public library is offering classes technological literacy classes. We need a nationwide technological literacy campaign and it's not just AI. We need people to learn media literacy. We need to be able to learn about the concept of sponsored content and how to recognize it. We need people and it's because it's like I think it's so easy especially for people on the left to just be like well the tech companies the tech companies.

[00:12:43]But it's like, yeah, the tech companies are irresponsible, but this is not a tech company problem. Like this problem of people that that are like the pro the like the issue that people can't understand LLMs, that's not really like yes, the tech companies are exploiting that because they're evil companies, but this is like a bigger technological literacy problem. It's like people not understanding how uh cell phones work or the internet work or whatever. I don't know. You know, like it's like we need people to we need people to understand the basics. You see this with data centers even, right? Like there's so much anger around data centers right now and people are really upset, rightfully so. Okay, by the way, I wouldn't want to live next to a data center. I'm all for it if they want to protest the data centers in their communities. Like these things are like noisy. They're not offering jobs. Like I get it. So I don't want people to think that I'm, you know, criticizing that.

[00:13:36]But still, a lot of people, even though they're right to be criticizing the data centers in their neighborhood, again, I wouldn't want to live next to it. Um, they don't even know what a data center is. They don't know what a data center does. They think that like you using a chatbt like query or something is going to mean that there's more data centers.

[00:13:56]It's like, well, maybe kind of, but but not really. Like, it's it's again, it's sort of like the plastic straw argument like with climate, but like put to technology. It's like you can stop using all the plastic straws you want. That's not going to make any sort of meaningful effect on climate change or the bad, you know, what what companies like Chevron are actually doing because like they have massive corporate partnerships and government partnerships that are doing so much more damage than you ever could conceive in your lifetime. And that's how I feel about AI tools. And this is why I think it's actually so terrible to go around and police people for like using Gemini, for instance, to like look up a recipe uh as I did very famously one time and people were so mad at me.

[00:14:32]Like guys, these companies have contracts with the Pentagon. These companies, these companies, and that doesn't mean that it like absolves you necessarily of like we should all try to use tech ethically, but getting mad at a tech reporter over that when like the companies here are the problem and they have massive especially a company like Anthropic, which which is primarily u monetizing through B2B contracts. We should be keeping our criticism squarely on people in power, squarely on corporations, etc. And with this AI psychosis stuff and with like something like this Richard Dawkins thing, I just think it reveals these broader structural issues and not only should we hold the companies accountable by saying, "Hey, um, you guys should fund literacy pro programs or you guys need to like do, you know, if you're going to roll out this technology, you also have to explain how it works to people because otherwise you're just unleashing this technology in a really irresponsible way. but also the government and other people in power need to build literacy programs and explain this stuff and help people understand okay why do we need data centers how do these things exist why um you know why do we have so many data centers it's not just AI it's also the rise of streaming it's also the rise of all these other things like you know how is that going to work what do these things do how do they operate what's the difference between the ways that they're powered because some are powered much more ethically than others so I just I think that um you know it's just this just shows the the things like this sort of like unear these broader structural issues. It's insane to think AI can't be sentient even if it's not right now.

[00:16:05]There's a 99 chance it has the ability to be in the future. Okay, we're not going to debate the future. We're talking about current LLMs and any kind of LLM in the next I would say several several years. Okay. And also there is no shared definition of what consciousness means even for humans.

[00:16:22]Okay, even for humans we don't have a shared definition of what consciousness means. So this is my issue with all of these like stupid things of like is AI conscious? Is AI conscious? I think we should look at what these systems are doing, how they're implemented. To me, those questions are so much more important than like is it conscious?

[00:16:40]Again, I consider a pig and a dog to be conscious. Do you think your dog is conscious? Again, I would say yes. Can your dog pass the touring test? No. Like I just think these are silly ideas of consciousness and like instead of having debates of like is it conscious? Is it conscious? We say what is it doing? How is it being implemented? What what is it being used for? Why are what are we doing with this uh technology and how is it being exploited and how is it harming people? Somebody said generative AI is generative. Exactly. AI that could be sentient is called general AI. That's not what we have now. Yes. I mean also just LLMs. Guys, we need to we need to understand how LLMs work. LLMs respond to an input, right? That's why Richard Dawkins dumb LLM is giving him these dumb answers. It is modeling itself and and bending itself to um to to provide, you know, this sort of string of letters and words that it thinks that he wants to hear. That's what LLMs do. That's why they have these problems with syncop sick fancy as well, right? So that is not a conscious that's not consciousness. Um Karen how also explains the whole consciousness thing in such a helpful way starting with the fact that the term AI is straight up marketing hype PR.

[00:17:56]Exactly. I mean the term AI I think she even goes back to when it was formed in like the 1960s or 50s and it was basically just this guy wanted to get like a grant and funding and he did. By the way Karen How's book Empire of AI is so phenomenal. I can't recommend it enough. It is a bestseller for a reason.

[00:18:10]It's literally so good. I want I need to have like a book club or something. It's such a phenomenal book. I just finished I I I actually read it like so late. I only finished it like a month ago. Um more Everything Forever, Adam Becker's new book is also f fantastic. Um it's really great. But both in More Everything Forever and Empire of AI, they both talk about this question of consciousness and why it's just very like it's a distraction from I think the stuff that we should be talking about in relation to AI.

[00:18:40]But no matter where you stand on this issue, I I know that we can all agree Richard Dawkins AI girlfriend is not real. She is not conscious. She This is an LLM. Okay. Richard Dawkins fell in love with an LLM. And he fell in love with an LLM by asking it very stupid uh questions.

[00:19:00]Basically, the LLM was flattering him.

[00:19:04]Okay. The LLM was flattering him. And because this man is such a narcissist, he now believes that it's conscious. Oh, and now he's publishing this is, by the way, this slop right-wing news website called The Unheard is literally just publishing his AI letters to each other.

[00:19:21]It it I want you guys because this this um this website was publishing a lot of COVID denial stuff as well. This is the type of slop these right-wing websites publish. They're just publishing AI in what what type of serious news website is and this not a this just shows it's not a serious news website is publishing Richard Dawkins AI slop letters that his AIS wrote to each other and they're publishing it.

[00:19:50]Also, the way it like talks is so annoying.

[00:19:54]Like the way that he's like made his dumb AI talks. Also, oh my god, dumb.

[00:20:00]Why is it always Trump? I told my AI a Trump joke and it laughed. It's so crazy. I'm so annoyed with all of it.

[00:20:06]Richard Dawkins is delusional and uh got one-shotted by his AI girlfriend. And I just think that this man needs to be in elder care and somebody should take his computer away. Okay. Uh Randall said he'll be like, "I learned that God can be a woman and I was enamored by her divine feminine power." Literally.

[00:20:25]Literally.