SGU Episode 884

This episode is in the middle of being transcribed by Hearmepurr (talk) as of 2022-09-23. To help avoid duplication, please do not transcribe this episode while this message is displayed.

Template:Editing required (w/links) You can use this outline to help structure the transcription. Click "Edit" above to begin.

SGU Episode 884
June 18th 2022
(brief caption for the episode icon)
SGU 883                      SGU 885
Skeptical Rogues
S: Steven Novella
B: Bob Novella
C: Cara Santa Maria
J: Jay Novella
E: Evan Bernstein
Quote of the Week
Science is a search for basic truths about the Universe, a search which develops statements that appear to describe how the Universe works, but which are subject to correction, revision, adjustment, or even outright rejection, upon the presentation of better or conflicting evidence.
James Randi, Canadian-American magician and skeptic
Links
Download Podcast
Show Notes
Forum Discussion

Introduction, JWST Hit, Force Fields

Voice-over: You're listening to the Skeptics' Guide to the Universe, your escape to reality.

S: Hello and welcome to the Skeptics' Guide to the Universe. Today is Wednesday, June 15^th 2022, and this is your host, Steven Novella. Joining me this week are Bob Novella...

B: Hey, everybody!

S: Cara Santa Maria...

C: Howdy.

S: Jay Novella...

J: Hey guys.

S: ...and Evan Bernstein.

E: Good evening everyone!

S: So did you guys hear the news about the James Webb Space Telescope?

B: Oh boy.

S: The James Webb Space Telescope's main mirror was hit by a dust sized micro meteoroid. And it's caused noticeable damage. It actually is affecting the data that they're getting back from it.

B: Oh my god.

S: But they say that it should not affect the mission's overall performance. They'll be able to compensate for it. It's not going to in any way reduce its mission. But it took a hit.

E: Well they anticipated that it would and I think they in an interview or a video that Fraser Kane put out about it. He talked about how these telescopes and other things that are sent into space are built with certain tolerances and certain allowances basically for those kinds of strikes. And they understand it when engineering these devices. And so it's anticipated ahead of time.

S: Yeah but it is a little scary that it got hit this soon. I mean I guess it's just bad luck. But it makes it seem like yeah this is going to be a regular event. It's not like it took years.

B: Yeah is it a one-off or is this something because of Lagrange orbit. Is that something because of where it is because last time we had anything there that was in the news.

S: Well we'll see.

J: That sucks.

E: Affects the resale value of this thing.

S: Yeah.

E: Get a ding like that.

S: Well it's supposed to have a good lifespan 15 to 20 years. So I guess they're just calculating that yeah it's going to slowly take damage over that time but it'll function. It won't be the limiting factor on its function but maybe they'll have to recalculate. Who knows.

E: Yeah and it's not the kind of thing you can repair. (laughs) Maybe that was one of the advantages of a Hubble like system that was so close to Earth it could be fixed as it took hits and other problems occurred with it. But here no chance for that. You're just gonna get what you get. You don't get upset.

S: But there was no choice this had to be far away from the Earth.

E: That's right. Of course. Yeah.

S: Because of the type of instrument. Yeah.

B: Yeah they got to turn on the damn force fields before it happens again. Oh wait that goes against laws of physics.

E: So maybe. I don't know. So you cover that in your upcoming book at all? About force field technology?

B: Yeah we do.

S: Yes we do.

C: Nice.

B: I think I researched that.

E: Is there any plausibility whatsoever to any of it? Or is it total myth?

S: So─

B: Depends how you define force field.

S: ─yeah. So as typically depicted in like Star Trek or whatever. No. Like that kind of force field--I mean there's just there's no known plausible wave within the laws of physics to do something like that. But can you use energy to protect a ship or a location? Sure. Like a really powerful magnetic wave─

B: Plasma, laser.

S: ─would repulse anything that has an ionic charge. Anything that's charged. You can use plasma and other things that would, that could deflect impact. But there's there are huge downsides to that as well because it's not it wouldn't be invisible.

E: Yeah. Wouldn't be invisible. Right. So you have to compromise whatever it is you're designing that in mind. And what's the energy? What sort of energy are you looking at to have to power something like that? Depends on what how effective you need it to be. What are you trying to deflect. But yeah it's an interesting thought experiment. That's one of the parts of the book. The last essentially third of the book is the thought experiment of we go through all of the sci-fi technologies and just talk about could this even happen. What's the closest you could get to this. Or is it just even theoretically impossible. Obviously it's always hard to account for ridiculously advanced technology like a million years from now. But still the laws of physics are the laws of physics so we as we know them so we could talk about plausibility. Which was fun. It's a fun part of the book.

B: Yeah another angle I've just remembered about force fields is that is point forces. I mean if you think about it a 360° force field this is a little bit wasteful. Now imagine, I think in the future we and we have these now to it to a decent extent actually where you can apply the amount of force you need at a specific point very quickly. And so that type you wouldn't call that really a force field though. It's not really a force filed though. Although it can be and is actually some form of a defensive shield. So it depends how you define forest field like we said.

S: Yeah you would imagine that any civilization with the technology to wield those forces that way would probably have advanced enough AI that they could very quickly put the energy where it's needed rather than like being having it in place from all directions at all times. That's would suck up a lot of energy.

E: In the case of the James Webb impact that has occurred is there existing technology that would have potentially prevented that?

B: No. Too small too fast is my knee-jerk reaction.

S: I mean if you had a companion sensor system with lasers that could shoot incoming dust particles.

E: Oh yeah. You put a satellite around the telescope. Something that orbits it.

S: It could theoretically work. But yeah I guess you could you wouldn't want to shoot a powerful laser near the telescope─

E: True.

S: ─because it could cause more damage than the meteor or what you're trying to deflect if you're not careful about it. Yeah it's an interesting problem.

The Skeptics' Guide to the Future (5:54)

S: So since, for those who are interested, since we're talking about it. Thanks for bringing it up Evan. Our next book The Skeptics’ Guide to the Future is coming out September 27^th. It discusses futurism itself. It discusses a lot of modern technology and where we think it's gonna go in the future. We talk about science fiction technologies. The near mid and long long term future. A lot of fun to research a lot of fun to write. Coming out September 27^th but you can pre-order it right now . And we strongly encourage you to do so. If you pre-order it, especially the hardcover, it really helps our overall sale. So if you're at all ever intending to get the book we do ask you to just pre-order it now before the launch date. You'll get it that's how you'll get it the soonest but it also helps the overall sales of the book so we'd appreciate it.

J: You can go to theskepticsguide.org/our-book. I can't make it any easier than that folks. And you'll see both of our books on there. The new one is The Skeptics’ Guide to the Future.

B: And it's worth it just for the cover there's lots of nice words inside but the cover itself is awesome.

S: Yeah. Cover is fun.

High School Reunions (7:07)

S: Before we go to the news items Bob and I were at our 40^th high school reunion this past weekend.

B: Oh boy.

C: Wow.

E: Wow that's eight years worth of reading.

B: I told you not to say that number. (Cara laughs)

S: It's really interesting seeing people. Some people, a couple of people I haven't seen in 40 years. Obviously since high school. Some people have seen that previous reunions. So I had seen them in the intervening time. But what's funny is trying to recognize people that you only knew when they were 18 years old. And it's amazing how different. So there are some people who are identical to the way they were when they were 18. Literally they look like they just had a mild aging program on them (laughter) but they looked exactly the same. Instantly recognizable. And then there are other people where I was struggling to see their 18 year old self in the current. They just for some reason they look so different.

C: They like transform.

S: Almost would not recognize who they were. And only I would not have recognized them if I weren't at a reunion, you know what I mean? Where I knew like this is somebody in my class. But if I saw them on the street I would not make the connection. They look so different for some reason. And it's not just the usual thing. It's not just like weight and age. It was just something that fundamentally morphed about their features.

C: What size was your graduating class?

S: Not big because it was─

B: A 100?

S: ─it was less than a hundred.

C: Oh so you knew literally everyone.

E: We took classes with most of them.

C: T hat's amazing. See when I think about going back to a reunion I'm like maybe I'll recognize a dozen or so people because I went to school with thousands of people.

E: Oh wow.

C: Yeah my graduating class was like I think between 1500 and 2000.

J: Oh my god.

B: Nuts.

J: I mean you legit don't know most of the people in your class.

C: Yeah. And my school only had two grades in it. It was a senior high for that reason because there were so many kids.

S: And we had our die hards. So these are people who were in all four years. It was only 16 of us.

C: What!?

B: Yes. I remember that picture the die hard picture.

C: Oh so other people kind of came and went? They transferred in.

S: They came in later or they came in. They came and went. Absolutely. And only 16 of us were there for all four years.

E: Four years.

S: So it's a very tight group.

C: Yeah.

S: Very tight group.

E: They're all good skeptics, right?

S: Nope. (laughter) Not even close.

B: Nope. Yeah.

S: We won't get into any details but that is not.

E: Just like real life. A sampling of any random group of 16 people.

S: We just I think all just silently agreed not to talk about politics or certain things.

E: Good. Healthy.

s: Would not have been a good idea. We just stuck to like reminiscing about high school days but talking about other things. But we didn't, we knew we shouldn't go there.

E: Compact discs and Die Hard.

S: Yeah. But Bob and I while we were there we met with the librarian and we donated a hard copy of our first book to the library. They have─

C: Oh fun.

S: ─yeah it was fun. They have a shelf for books written by alum. Including JFK. John F Kennedy who went to that school. He went to the same school.

C: You went to the same school as JFK?

E: What's the name of his book.

B: For a little while.

S: The Profiles in Courage I think was there.

C: Were there any like─

E: You think he really wrote that?

C: ─(laughs) probably not. Next to the Skeptic's Guide book was there like a cryptozoology book? Are there any books written by alum that were like.

S: Yes but they weren't there.

C: Oh. (laughs)

J: So Steve you know the obvious question is are our books there?

S: Or what?

B: Our. O-U-R. Yes. Our book is there now.

C: Or, our.

S: Yeah I couldn't parse that sentence.

E: A-R-E?

C: R-R?

J: So you donated one is what you're saying.

S: We donated one for the library.

C: That is it literally what he said. (laughs)

S: That's what I just--it's on the alumni bookshelf.

B: So Steve brings the hardcover and there's not too many of those bad boys around anymore. He brings a hardcover over to my house because we were going together. He opens the book so we could sign it and there's Cara's name in the book. Like oh okay.

S: Well you pre-signed some books.

C: Oh yeah I signed a bunch of them.

S: Yeah.

E: Ah, perfect.

C: Whoops. Sorry guys.

S: No worries. Doesn't matter.

C: Well now Evan and Jay have to sneak into the school to sign it as well.

S: That's true.

E: Yeah we'll come up with some zany plan like a bad 80s movie.

S: Right.

B: Oh my god.

C: Yeah that could have only gone over the 80s.

B: Time machine.

E: We'll call it meatball house or something. Like animal house.

S: All right well let's move on to our news items.

News Items

Is LaMDA AI Sentient? (11:43)

• Is LaMDA Sentient?

S: We're going to start with--I think this is probably the most widely discussed science news item of the week.

B: Yeah.

S: Google employee claims that their LaMDA language simulating software is sentient.

E: That's a big claim.

B: Yes.

C: I don't buy it.

S: Yeah no one's buying it.

C: '(laughs) I bet you some are buying it.

E: No one person's buying it.

S: This is Blake Lemoine. Blake Lemoine is a google employee. He was suspended from the company for violating his NDA because he released a transcript of a conversation between himself and unnamed collaborator and google's basically their chat bot software LaMDA which stands for Language Model for Dialogue Applications. The conversation is pretty impressive and as chatbots go. So we could read─

B: It is.

S: ─but here's the one thing though when you read it you have to read LaMDA's part of the transcript as if you're HAL from 2001: Space Odyssey.

C: Even if something is like passes the Turing test really well that doesn't mean it's sentient.

S: Of course not. Turing test is a terrible test for whether something is sentient.

B: It is. It's really [inaudible] worse.

C: But basically that's what you're saying, right? It's in the transcript it seems to pass the Turing test.

S: Oh it might pass the Turing test which means you have to fool a certain number of people into thinking that it's an actual person.

C: Yeah it's not easy but it's definitely far from being able to make a claim of sentience.

E: Oh right yeah it makes no declaration of sentience.

C: No yeah even yeah. That was the point.

S: All right so here's an example. So Lemoine asks "Okay, what about Les Misérables? Have you read that one?" LaMDA says: "Yes, I have read Les Misérables. I really enjoyed it." Lemoine says: "What are some of your favorite themes in the book?" And LaMDA responds: "I liked the themes of justice and injustice, of compassion, and God, redemption and self-sacrifice for a greater good. There’s a section that shows Fantine’s mistreatment at the hands of her supervisor at the factory. That section really shows the justice and injustice themes. Well, Fantine is being mistreated by her supervisor at the factory and yet doesn’t have anywhere to go, either to another job, or to someone who can help her. That shows the injustice of her suffering." All right that's a pretty well put together sentences and it's coherent, and it actually answers the question. I think this would probably, if you didn't know the context, you were just reading this you wouldn't necessarily pick out LaMDA as not a person. As an AI. I think it's shows how far we've come. But there's the--going from that to it's actually sentient because it--so Lemoine, who should know know better, essentially was fooled by the very program that he was working on. Which is supposed to simulate conversations. To simulate words. But there's nothing in the programming. It is a neural network but there's nothing in the programming that in any way can achieve sentience. That's not what it's programmed to do. It's basic so is one of the--there's a lot of experts jumping on him and criticizing him but one of my favorite one ones was LaMDA is basically a really good auto complete.

B: Auto. That was one of my favorite quotes. It really is. It really is.

C: But how does it actually? I mean you said it's a neural net. So just to be clear because like i remember a long time ago doing a story on real dolls which are like these weird sex dolls.

S: Yeah.

C: And they were adding AI to one of them. And I saw it in person. And I saw a guy interviewing it but it was just like scrubbing the internet for content. You could tell. It was just regurgitating things that it found based on key words.

S: Searching. Accessing.

C: Yeah exactly. And sometimes it would do that. It would read URLs.

S: Yeah this is just looking at millions of responses and using software to come up with responses that work. But there's no understanding or knowledge or thinking going on.

C: But it's also not just straight up scrubbing the internet.

S: It's just parsing words.

[talking over each other]

B: It's constructing them. And it uses a transformer. You hear you'll hear the word transformer a lot which is basically it's a deep learning model and it deferentially weigh it is a weighing system of significance for some of the for parts of the input data and stuff. So that and it works very well with sequential data like in languages and stuff and stuff like that. And this thing is meant. It's trained in dialogue which a lot of language models don't really do. So this is trained. This is what it does. It engages in convincing dialogue. This is what it does. It's very good at it and reading what I read of Lemoine's interactions he this guy absolutely should have known better and secondly he should have been working to try to show how brittle this knowledge is. It has no knowledge of meat space. It doesn't have an inner life that you would think that something with sentience or consciousness or self-awareness would have. There'd be some inner life. Some hints to an inner life. And it will allude to an inner life but then in other sentences it doesn't. So it's like it's clearly just doing what it was designed to do very very well.

C: What happened when he asked it introspective questions. like what do you dream about at night or how does something make you feel?

S: They can give it good answers to those questions because that's exactly what it does. It doesn't have to actually feel something. That's just more words to it then comes up with other words that that fit as a response.

C: I'm just wondering how buyable it is. Is it still really passing the Turning test when it does that?

S: Totally.

C: Okay, cool.

S: At least that one transcript that they published. Nothing in there smacked me in the face of like oh yeah this is it. It got broken here. Like this didn't work. It was able to handle all the questions that were thrown at it reasonably well.

B: Let's not forget a lot of what you saw though some of it was edited. So a lot of it that's out there was actually edited and pulled together to be even more convincing. If you search you can find the unedited versions and it's not as tight and it's interesting. But yeah this is clearly--it's a language model of sorts that is really good at what it was designed to do. I mean this is what it was designed to do. I mean that's why this was so disappointing that this guy just fell into it. It's like yeah he was convinced it's like he saw a mind where that was in somewhere that was mimicking your mind. But it goes to show though in the future when we really do get closer to what people are calling sentience or self-awareness or self-consciousness. Then it's going to be hard. It is going to be hard to know is this thing is it walking and quacking like a duck. Is it really a duck or is it really there's something there. This is going to be harder in the future. It's easier now.

C: So Steve how do we define that? How do we? I mean we said the Turing test is a crappy way. That's not even what the Turing test is designed to ask. It's just can you fool a human. But how, what is the test? Is there a computer science?

S: So let me tell you why I and many other people say that the evidence that LaMDA is sentient is crap. So that's the interesting thought experiment here how would we know?

C: Yeah.

S: So as Bob was saying Lemoine fell for the hyperactive agency detection. Which is essentially we have a tendency to anthropomorphize. To see an agent. A sentient behind random events or─

C: A blind watchmaker as it were.

E: It's very comfortable.

S: ─behind--or even that's why cartoons are so effective. If something acts like an agent to our brains treated as if it's an agent and so that's what he was doing. Because it was expressing emotion, expressing feelings and desires he fell for that illusion that hyperactive agency detection. I'm not convinced by the evidence that he finds convincing for a number of reasons. First as Bob mentioned it's just responding to input. It's not driving the conversation or generating questions itself. When I wrote about it I said at no point does LaMDA say hey sorry to interrupt but I was just wondering and then ask some question that had nothing to do with the conversation that they were currently doing. So there was no evidence that there was an in a loop of internal activity happening. When you think about human consciousness when we have wakeful consciousness our brains are functioning in a constant loop of activity. An internal conversation. Where one part of the brain is activating another part of the brain which activates another part of the brain. That constantly is happening.

C: And it's spontaneous.

S: And it's spontaneous. That's what wakeful consciousness is. When that stops happening you are not conscious. So there's no evidence that this is happening in LaMDA. It's just really good at responding to input. But the other thing is Lemoine found Lemoine's emotional expressions convincing and I had the exact opposite reaction. (laughter) Because the fact that they're so human to me is evidence that it's mimicking human emotion. If it had its--so Lemoine because the thing is not designed, programmed, intended to be sentient he would have to believe that sentience in LaMDA was an emergent property somehow.

E: Right. It sparked.

B: Not impossible. The concept is valid.

S: Yeah. Conceptually not impossible. Not not from this. But at some point we may actually, as Bob says, confront that situation where is there something going on here that's emergent? That is some kind of consciousness? But we're not anywhere near that.

C: Yeah because the idea of neural net AI is sort of black box. Like we don't know how it comes up with what it comes up with. So there could be an emergent property.

S: Yeah. That's part of the problem is that that it's self-trained and therefore we don't really know exactly the how it's solving the problems we gave it to solve. But here's the thing if it did have some kind of emergent sentience, what's the probability that it would be so exactly human-like? Wouldn't it be unique and quirky and--I could if Lemoine was saying this computer program is doing something that I cannot account for in its programming and there was a more subtle sort of manifestation. Something quirky like it was breaking in a very interesting way doing something that it shouldn't. Something like that to me would be more interesting, more compelling than exactly mimicking human emotion.

C: It almost reminds me of the concept of like normativity like we can say that there is normalcy. Mean, median, mode, whatever. Central tendency across different variables but no one person is just normal across the board.

S: Right.

C: Like they might have be an average height and they might have an average intelligence but they're not average on everything. And so this AI is like sort of average across the board.

S: Yeah. Yeah. It's too normal.

C: That's so funny and so it's almost like uncanny valley AI. (laughs)

S: Right. Well but why would? Yeah to me that would not that's not what we're going to see. When if there was any emergent consciousness it wouldn't be that way. And you then you get to your question Cara of well if the Turing test doesn't work because it really is not a test for actual sentience. It's a test for how well something could mimic sentience and we know that you could do that really well without being actually sentience. So how do we know if an AI is actually sentient? And I think there's two answers to that. One is because we know how it functions because it's designed to be self-aware to be sentient it's doing something that at least is in the ballpark of sentience of consciousness. This isn't even in the ballpark. Also here's another thing. You know LaMDA is simulating human level consciousness with orders of magnitude and sufficient processing power to do that, right?

C: Right. Yeah. We already know that the technology [inaudible].

S: It doesn't have the processing power to be actually sentience. Only to simulate sentience through language only. Not the actual underlying thoughts. But anyway, if we have a system that is powerful enough, that it has the functionality that could plausibly produce sentience then at least it's a possibility. How would we know? The short answer is we wouldn't. Which is really interesting. How could we know if it's experiencing its own existence or if it's just really good at acting as if it does?

C: Yeah if we haven't figured out how to understand the inner worlds of animals yet I think we're gonna have a really hard time understanding the inner world of a program.

S: Totally. I predict we will get to the point where we have a system where no one can truly say if it's sentient or not. Maybe it'll be 50 years or whatever but we'll get to that point where it's really good at simulating sentience, it might have it, but we have no idea if it's actually conscious. If it's actually experiencing its own existence or if it's just doing a really good job at simulating it.

B: Right and it's also very interesting because it could have lots of utility in terms of say accomplishing goals. Picking problems, coming up with you with unique strategies to solve problems. And it could do all of that but we may find out or maybe never find out it could be just a p-zombie that just--there's no quality that qualia or substance. It's not necessarily gonna be anything like our consciousness.

C: But that raises I think a really important ethical question, right guys? Because we're asking it in the affirmative. How do we know if it is. But the opposite is how do we know it's not. And if we're utilizing these tools to do work. And these tools even have a chance of being sentient is--are we ethically--do we have a moral obligation not to use them for work?

B: We're gonna cross that line.

S: We will have to encounter. I do like the solution that they came up with in the movie Ex Machina where they have to see if it's really sentient or just ask sentient is can it creatively come up with its own problem solving in on its own initiative in order to save its own life.

C: And spoiler alert. (laughter)

B: At this point.

E: It stepped out of the cave.

J: Steve the part of this, we've talked about computer sentience many many times and not much has changed in the last even 10 years. 5 years. Not much has changed. Programming is getting a lot better and everything but we're not that much closer to any real kind of sentience than we were. To me this is a massive lesson in the fact that experts, hyper-educated people, highly intelligent people can be utterly fooled. Just like anybody else. Like this google engineer─

S: They're still human.

J: ─right? They're still human. This google engineer did something intellectually that I find so it's almost unbelievable if I wasn't a trained skeptic. The only thing about me that makes me believe it is that I know that this happens to people. How could this guy fall for that?

S: I'll tell you how. Because being a technical expert doesn't give you critical thinking skills nor does it give you knowledge or expertise outside of your area of focus. And so this guy could be a very competent computer programmer. Do technically what he needs to do but that doesn't mean that he's gonna understand about hyperactive agency detection for example.

C: It's just amazing that somebody who works in a field that grapples with these things philosophically isn't responsibly grappling with these things philosophically.

E: That makes it all the more remarkable.

S: We see that all the time. I certainly see it in my profession. I know people who are technically excellent physicians or surgeons or whatever. But they can't think their way philosophically out of a paper bag.

C: Oh, you see it in psychology too.

E: If only he read our book. If only he read our book.

B: Imagine you're testing this model. This language model. And it's starting to sway you and and you're gonna go to your bosses at google to say this is sentient. Oh my god would I cross my T's and dot my I's. (Cara laughs)

E: You would run it by someone else.

S: Wouldn't you? Like a close friend.

E: You would falsify that presumption. And take the steps.

C: And you try to break it. That's what a good scientist does.

B: Exactly.

C: You try to find a hole.

E: Falsify the thing. Yes.

S: Some people think that Lemoine is not serious or that he that he's a hoax or he's got some ulterior motive. I don't know.

C: I think it's a publicity stunt.

S: I don't think we have to go there. I have no idea.

B: He's kind of disgraced. I mean bad hoax man. If it was.

S: If it was a publicity stunt he failed. But who knows. We'll never know. Like what he actually believes. But I just take people out there at their word unless they have a reason not to. It's certainly possible that he fell for the illusion. But there's lessons here as well. So when people ask will we accept robots in the future as human. My answer is absolute freaking lutely. Without a doubt.

E: Why wouldn't we we?

S: We will accept anything that acts as if it's human. Acts as if it has agency.

C: We think our pets are human.

S: I know. We anthropomorphize anything.

E: We were predisposed for it.

B: How about sex dolls? People are marrying sex dolls.

S: That's actually a really good segue to Jay's item, which is also about--

E: --Sex dolls.

S: --the human acceptance of digital, non-human entities.

AI Influencers (29:43)

S: Jay, tell us about AI influencers.

• The rise of computer-generated, artificially intelligent influencers

J: I will start by saying that somehow through the miracle of just missing certain pieces of reality. I did not know about this. At all. And I think it's very interesting because I live on the internet. My job is internet based and I didn't know that something called a virtual influencer existed. And I wanna say I'm really happy that I didn't know about it.

C: This is like a vocaloid kind of?

J: Just check it Cara, you won't believe this. So these virtual influencers. What are they? They are they're personas that are created by a person or a team of people that mimic real people. I say that they mimic real people because some of these virtual influencers they look almost real. Others can have an anime look to them. Others have exaggerated features which I'll get into. They're essentially, these virtual influencers, are essentially doing the same things that real people do online but that's just it. They're not real people. They're a construct. They're made up by other people. So there are a lot of these virtual influencers out there and some have earned large followings. Which I will when I describe this to you I imagine that those of you who don't know about this will get blown away like I was and how big these followings can get. Even though these personalities at their core they're computer generated. The way that they look, the videos that you see. This is all computer generated by people to pretend in a way that it's a real person. Even though the vast majority of the audience knows it's not a real person the whole game here is to act and pretend like it's a real person. The content that these virtual influencers have it's similar to content created by real people. They have YouTube videos. Tons of pictures of themselves doing all sorts of things all over the world. They're found on all social media platforms. So the the shorthand of thinking about this is there are people who make these avatars and they're individual. They have names and they have personalities and they have recognizable faces and voices and everything that makes them unique. Or makes a human being unique. They have it as well. So let me give you some examples. The first one that was created goes back to 2007. The character is called Hatsune Miku.

C: Oh yes she's a vocaloid. Yeah.

J: That's right. This character started off literally as a voice changing software. But when they package the software they came up with like this character that would kind of represent the software. And since its creation back in 2007 the people who created this evolved this character into a popular personality. Even though Hatsune is virtual people pay to go to her concerts in the real world.

C: I went to one.

J: Well okay.

E: Well there you go.

J: So Cara is a part of what I'm about to describe here.

C: Well I did it for a TV show. I didn't. It wasn't really by choice. It was amazing and weird. But fascinating.

J: So this is what I saw the video and Cara please pepper in anything that you think is interesting.

C: Sure.

J: So the video I saw is a large number of people in a real theater watching basically a very pretty looking hologram like projection of this character singing on the stage. So there's like this eventually this giant screen type of thing that the character I believe is being projected onto.

C: Yeah it's like a pepper's ghost thing. It's kind of like Tupac.

J: And the audience, now this is the part that Cara really experienced I'm just watching a video, the audience seemed to absolutely love this experience.

C: Oh yeah I was there with super fans. And they gave me these like light up things to hold. And they change color with the different song. Because there's a whole like they know all the lyrics and they know the costume changes and there's a whole culture around this character.

J: So not only that but this particular personality is considered a pop superstar legitimately with several released albums. So we have now the very first one that probably came into existence ended up being legitimately a huge money earning entity. And again there is no person really behind this. It's probably a group of people that are creating all this content and everything that goes into this virtual character.

C: An interesting component of virtual characters or at least how they first started is that Japanese culture, which you now see it with American culture like tik-tok and stuff, is very derivative. So there will be like an idea and then somebody else will do their version of. Just like you see on tik-tok. It'll be a video and then somebody will do their version of that video with the same music and they'll do their version and their version and their version. So the way vocaloids first started was it was a software program that was like a midi program with vocal noises. So just like you could write guitar riffs you were writing songs with little vocal. So it's like a little individual phonemes but they were sung on on pitch so you could write music that way. And then individuals would give it a face and a name and now a new outfit. And they were all coding and it became this cultural phenomenon where it wasn't one or two people that owned it. Everybody was owning it together because it was like just this YouTube thing. Where a new person would make a new song for her. I'm sure now there are, because there's money to be made, people are fighting over rights and copyright and all that but originally it was like fully crowdsourced. Which was such a bizarre and interesting phenomenon.

J: Well the whole thing has evolved to the point now where there's real money behind this and so there are companies that their whole game is this. Like create these personas that they create. So let me go through a few more give it other examples of what's out there. So this other virtual influencer is called Lu do Magalu and please you know I don't know exactly how to pronounce all these names. This is one from Brazil and started out as a virtual sales associate for a magazine. Lu do Magalu became an online personality on YouTube in August 2009 and it now has over 55 million followers across social media. And I do encourage you to go look these things up on the internet and see pictures and you can see some of the videos that are out there. So as the technology improves so does the look of most of these virtual influencers. The voices become more real sounding and at some point in the near future there'll be virtual people who look and look exactly like real people online. They could even be in movies and TV shows. Imagine a virtual person who has this following online who people recognize the face and all that stuff. It's gonna happen. Somebody's gonna put one of these characters in a movie like a real person and you're not gonna be able to tell the difference.

C: Yeah.

J: That's coming.

C: It wouldn't be that hard because you can map. I mean even now we can age down people and stuff. We can map them with light stage technology.

J: So arguably the most famous virtual influencer is Lil Miquela. This one first appeared online on Instagram in April 2016. This particular one looks like a young girl with straight dark hair and freckles. When it first appeared online some people weren't sure if this one was real or computer generated. That's how good some of the imagery looks. And after a few months the company who created it named Brud B-R-U-D, this company is located in Los Angeles, they revealed that it's a marketing stunt. And now they use computer graphics and real photography to create all the media that is behind this character. A quick search on YouTube produced a video of this character. And I found this video that had 1.5 million views.

C: Yeah it's definitely nothing to sneeze at.

J: Absolutely not. So some of these virtual influencers are getting legitimate advertising gigs. Like for fashion magazines. Vogue and Elle. Clothing manufacturers like Calvin Klein. There are now teams of people who work behind the scenes to create all this content and there is real money like I said to be made in this space─

C: So who gets the money?

J: like hundred of millions of dollars. The companies that own these characters.

C: Okay.

J: Or the people. So it's, we go all the way down like you were saying. One person might make one of these. There could be a small company. There could be a huge team of people. Like some of these characters have legit full-scale marketing companies that just produce content and imagery and videos and albums and everything behind these characters. So now as technology progresses now some of these characters are becoming dare I say on some level artificially intelligent. And I don't mean conscious like the previous news item. I mean that there is artificial intelligence that is driving them. They use computer algorithms to generate content and they also use these computer algorithms to chat with fans. Just like the last news item.

S: Just like LaMDA. You can buy LaMDA with a virtual influencer and it runs itself.

J: One of these that I'm recommending that you check out because it's a really good example of this and it's also kind of creepy is called Serah Reikka R-E-I-K-K-A. This one has enormous, oddly enormous breasts. Incredibly skinny legs. Big lips. You get the picture. You're going we're going for like that whole creepy barbie doll look that you may have come across on the internet where some people are changing their bodies to look like barbie and they end up looking kind of freakish here. So as you might be guessing there is a real downside to these virtual influencers. Now I can't say all of them. I'm not trying to be like this whole thing is as bunk. But there is definitely a trend going on here that I noticed while researching this and that I've definitely read about. Now beyond the fact that this these people are not real people some fans of the influencers strangely feel like they're developing relationships with them. Some of these virtual AI driven influencers are able to "talk" to their fans. They are talking to their fans. They're talking to their fans more than you would ever get out of a real famous person. Unlike being a real, with real people they don't have time to do this. But virtual influencers can simultaneously simultaneously be talking to thousands of people at the same time forging false relationships. And it's happening. There's something called parasocial relationships. A parasocial relationship is a is simply a one-sided relationship where one person expends emotional energy. They're showing interest and time is going by and meanwhile the other party or persona is completely unaware of the other one's existence. That's a that is a a real problem. These kinds of relationships are not healthy. And once they go too far and it's not--I wouldn't say it's common but it happens and it happens to enough people where their lives can get ruined. Once one of these relationships goes too far a fan can start to become obsessed and then this parasocial relationship can become toxic and then it becomes detrimental to their mental health.

C: You're talking about a relationship with a vocaloid or a AI influencer. You're not talking about with one that's made just for them?

J: Well let me let me broaden a little bit. A parasocial relationship can happen with--a person can have one with a sports team. Or a real famous person. Like someone that they worship. A hero of theirs, an idol, a music's, a pop star. Anybody that they're super into that they're following. They can develop some type of parasocial─

C: I see. See I just wanted to clarify. You're not talking about a person having a relationship like with a real doll or something. Because I don't think it's safe to say a 100% blanket that that's unhealthy.

J: No I'm not even--I'm not steering into that. I'm essentially saying that like the concept of having a relationship with something someone or one of these virtual influencers and it isn't a reciprocated relationship basically in any way. But the complexity here Cara is that it is being reciprocated because you can get reciprocation from these AI driven ones who will write back and talk to you, right? So that's that's the part that's got me a little scared about this. Because these relationships they can't truly give a human being legitimate satisfaction like a real life interaction can.

C: Well they can. I mean depending on somebody's. Yeah I mean I think there's arguments with that. But the one way, the unrequited part is really important.

J: Yeah I mean you could get conversation and you can get some pleasure out of doing this or whatever but if it's taken too far and that's the key thing here. If it's taken too far and it gets into a pair of social relationship.

C: And you don't know that. I mean I think that's the really scary part is like we saw it with Jodie Foster. We've seen these examples where individuals don't realize that they're not in a reciprocal relationship.

J: Yeah I just worry that as we go into the future here and as the technology gets better and like these chat bots get really good. Maybe 5 years, maybe 10 years from now they could get to the point where you could have all sorts of deep levels of a relationship that's remembered. I know the AI might remember this person and continue the conversation as if it's a real human being. The potential harm here I think could be pretty bad.

C: I think there's also a lot of potential good though. A lot. We see this in some countries that offer virtual pets to elderly individuals. We're seeing more and more individuals who struggle with difficulty interacting socially having the opportunity to feel some amount of reciprocation and comfort. We're a social species and we're desperate for social interaction. And social interaction is healthy. And for people who struggle with loneliness and who simply don't--I mean it's one thing to say oh they're doing this instead of hanging out with other people. But some people struggle to be around others. And I think that this is an opportunity or it can be when done in a healthful way that could actually have a real psychological benefit. So I don't think I would want to completely throw the baby out with the bath water personally.

S: Yes I agree. This is also pretty much of a natural evolution of what's been happening for decades. For example in terms of like virtual pop stars virtual pop stars are very old. You could consider Alvin and the Chipmunks for example. They're not real people. Nor that they're chipmunks─

E: What?

S: ─they were created as an─

C: What!?

S: ─it's a marketing created entity. Or even any cartoon band that we grew up with. The Archies or─

E: The Monkees.

S: ─whatever. And then there's other virtual bands that are fronted by people that were studio creations. And you could say that a lot of influencers, actors, actresses, personalities are manufactured illusions just because─

C: Oh I think most of them are.

S: ─they're focused around a person doesn't mean that they're real. They're illusions and people do develop relationships, fall in love with the illusion of the character that's being generated. This just eliminates the meat bag from the equation. And you have something that's completely virtual. So it is, I know it's it can become fundamentally different but it's really just an extension of what's already been happening.

J: I wouldn't minimize the idea that people can talk to these chat bots.

S: Yeah.

J: There's a whole personality and then there's a chat bot attached to it that you could talk to and get responses from. I think that that there's there could be potentially a danger in there. I mean look.

S: Hey famous people often for a long time have had their marketing people answer fan mail for them. And again there was a virtual relationship there.

C: The danger really comes from delusion. That's what it is. If you know what this is and you understand the parameters I don't think it's psychologically detrimental. But when somebody's reality testing is tenuous and they're unsure whether that reciprocation is real that's when you see the scary kind of legal complications of stalker behavior and things of that nature.

J: And it's not just this. I mean there's also a body dysmorphia issue here.

S: That's to me the bigger issue.

C: That's a huge issue.

J: So body dysmorphia it's a mental health disorder. People who develop body dysmorphia they can obsess about how their body looks that this can lead to a person not being able to lead a mentally healthy life. Depression and anxiety can develop in extreme cases and some people literally change their bodies with cosmetic surgery. You can see pictures of this of this online. On the internet. Just take a look man.

C: Yeah. Eating disorders that are highly co-morbid with it.

J: And what bothers me is that at a time when it seems like modern society is actually starting to understand and deal with body dysmorphia. Fashion magazines and things like that like we're starting to be more inclusive of different body types. It's scary to see these virtual personalities. They take on like I said like the barbie doll body style to a whole new level. That one like that Serah Reikka I told you about? I mean her virtual breasts are just ridiculously huge. I'm not talking a little bit.

S: They're not always portrayed that way though by the way. But some versions do. But yeah I agree. It's like that you were creating these virtual people that are impossible standards for anyone to follow.

C: And what's scary even more than that is that in the situations where this is an iterative crowdsourced phenomenon it's the people themselves that are choosing that to be the ideal. So we're perpetuating a really negative personal reflection of how we want to be. Like when you see women who are internalizing deep sexism. It just perpetuates that's the part that I find so dangerous. It's the same thing when you see kids on insta and snapchat, not just kids-everyone, on insta and snapchat using filters to make their faces look "perfect". How they define perfect? Who defines it that way and why is it that we have this sort of negative feedback loop where a young girl looks at her face like that and says I do like that better. That is how I want to be.

J: I feel like something is a little off with this whole thing. I know that the future technology that's coming it's going to blur the lines of what what's real and what's not real. Just think about it you have virtual reality deep fake technology. All this stuff packaged together. It's gonna be able to create a lot of false realities and sometimes I think there is a danger for certain kinds of people that's a little troubling.

C: I think for every. I mean this is why I'm resistant to the metaverse Jay. I completely agree with you. I don't want to spend time in a metaverse. I don't want to spend time where everybody is an idealized version of themselves. It's hard enough to get to know people in the world and see and be authentic and learn who people really are.

J: I think it's inevitable. I want to close out by saying at its core humans need to feel loved or most humans need to feel loved and we need human contact. And I worry that future technology can create a lot of dead ends for people. And it's not gonna connect people. Like we had to have the big dream that the internet was going to connect people. And boy did it do a lot of things that we didn't predict. I mean it's so much trouble came with the internet. So many bad things came out of the internet.

C: But it does still connect people and I think we can't discount that as well. So just like there could be─

J: I'm not. But it comes─

C: But it comes with a cost.

J: ─yeah they all come with a cost.

B: Double-edged sword.

J: You want to have super advanced AI that can truly mimic human conversation. There's gonna be some cool applications for that. There's also going to be some really weird unexpected and probably horrible things that come out of that as well. And how much can we predict this stuff. How much can we prevent these maybes in these circumstances. I just think my recommendation is that we really proceed with caution here.

S: But we're not because it's decentralized. I mean it's going to proceed on its own organically, you know what I mean? But I don't know. I think here's the thing Jay. The marketing and entertainment it's all fake anyway. So this to me is just one more step in that same direction.

C: Well and it's almost more authentic because it's aware of itself.

S: Exactly. Yeah. If it's like totally fake then there's maybe less of a chance that people will get fooled by it because if you have a real person behind it people might think that that real person is the persona. The marketing persona. Here there is no real person. You know it's a virtual creation of a marketing team and that's all it ever will be.

C: And very often it's your own creation. Like let's not forget that we still can crowdsource these things and we can always iterate or make derivatives of the original. Which is kind of the coolest part of the community.

S: But I think listen future generations are going to have to learn how to understand that things that seem sentient may not be.

C: Right.

S: For us that's like shocking. From our generation and we're and we may be concerned about. The future generations are just gonna live with that. They're just gonna be like yeah I mean you know. Things that are not real can act indistinguishably like they are real. We're gonna have to make sure that that's sort of baked into society and understanding. And of course it'll be exploited every which way you can imagine.

E: Oh yeah. The potential for abuse. Never underestimate the potential for abuse.

[commercial brake]

Kids Don't Get Cancer Because They're Unhappy (52:25)

• Princess Märtha Louise's new fiancé Shaman Durek had his book dropped by publishers in her home country of Norway after claiming 'children get cancer because they're unhappy'

S: Evan I understand that kids can get cancer if they're unhappy.

E: That depends what books you read I suppose.

C: Oh god this is gonna make my blood boil, isn't it?

E: Oh my god.

S: Yeah.

E: Well yeah all right.

S: Well.

J: Well.

E: Let's get to it. All right so this all stems out of a recent news item. Bear with me while I explain Shaman Durek announced that he is engaged to princess Märtha Louise of Norway. And from that one sentence alone the first question that would immediately pop in your mind is what does that have to do with the price of tea in China? I love that expression. But in this context what it means is that what does this have to do with anything involving science or skepticism? That's a good question and I will explain. All right Shaman Durek who is this? No, not a character from planet Vulcan on Star Trek Bob. Shaman Durek is a shaman. Well or some people more correctly phrase it a show man but more on him in a second. And he announced his engagement to princess Märtha Louise of Norway and according to Wikipedia she's a member of the Norwegian royal family, a businesswoman and a self-described clairvoyant. So you have a shaman marrying a clairvoyant. That's a match made in heaven I guess? Among the--so you go to Durek's website you got to learn about this guy. And services on his website shaman boot camp, workshops, speeches, home clearing like I guess from like poltergeist this house is cleaned kind of thing. And shamanic experiences whatever it is. H e has spoken in the past of being able to get patients to vomit. Patients to vomit to get rid of poisons in their bodies. And he keeps a bucket in his office for sessions in case they need to throw up and that sounds very appropriate. But here's the description of himself in his own words Shaman Durek, this is from his website: "Shaman Durek is a 6th generation shaman, and author of the best seller, ‘Spirit Hacking: Shamanic Keys to Reclaim Your Personal Power, Transform Yourself, and Light Up the World.’". That book came out in 2019 and in that book Durek claims, yep, childhood cancer is caused by unhappiness. And for that he was properly sort of shunned by a lot of publishers around the world. But the book had already made it out actually and was published in a lot of countries including the United States before actually some of these publishers got a hold of it and started to read what the heck was in that book. And only afterwards they went oops and yeah we're going to pull this book. We're not gonna--this should not have been published and we're gonna stop this. And there was a big it was kind of a big shock wave at the time in a sense. The company in Norway that was publishing the book not only had to pull the book and but released a bunch of statements saying that they were unaware of these kinds of claims in the book. But they're saying this it's obviously this is harmful to people and should never have been published in the first place. So obviously a huge huge mea culpa on that one by the publishers. Here's from the (laughs) my gosh this is straight from his Wikipedia page. I love this. He advocates several conspiracy theories and has been characterized by by Norwegian media and other critics as a conman. His only book the one I just talked about was described by critics as nonsense, garbage and dirty talk and the ravings of a lunatic. And those are quotes.

S: That pretty much describes the wellness industry, doesn't it?

E: (laughs) I mean yeah. You could probably apply those to just about everything. All those books.

S: Once your disconnected from reality that's the kind of stuff you're gonna get. If you're just making stuff up because it sounds good from a new agey point of view. But the the notion that people make themselves sick because of "spiritual reasons", because they're unhappy people or they're mean or what they're skeptical or whatever is very common among gurus and in the new age wellness industry.

C: It's the opposite of the Secret. I mean it's all about yeah just like if you thinking you'll fulfill which the Secret's all about. Like making money and getting laid but it's the opposite. Oh and I struggle with this with my patients all the time.

S: Yeah you have unhealthy thoughts that's why you're unhealthy.

C: A lot of people internalize that even if they're really really intelligent, science forward. It's when when a bad thing happens to you, like cancer, it's really hard to purge yourself of these societal ideas. And it's so common that people are like maybe I'm just not being positive enough and that's why the chemo's not working. It's heartbreaking.

S: But it puts a lot of emotional strain on people who have a life-threatening or terminal illness because it's like telling people you can get through this if you have a positive attitude and you're strong and all that. You sound like you're encouraging them but you're really saying if you things don't go well it's your fault. Because you weren't positive enough. Yeah. Absolutely. Then that's what all like faith healers do. You didn't believe hard enough. And the alternative medicine community completely uses that as well. Is that you weren't true enough to natural healing or whatever it was. It all amounts to saying you didn't have enough faith.

C: It's all victim blaming.

S: It's all victim blaming.

C: It's how they can stay not culpable when their faith healing doesn't work.

S: Yeah. And we're very careful in medicine now. 30 years ago this wasn't the case. But often like 30 years ago would not be uncommon to say like the patient failed this treatment trial. And we realize that that's not really accurate or fair or appropriate.

C: Right. The trial lost efficacy.

S: Yeah. The drug failed. The patient didn't fail. Or just it didn't work. But it was not like saying like the patient failed their treatment trial is not putting the onus on the wrong thing.

C: It's also a fundamental misunderstanding of cancer. Because there are certain types of cancer that if you catch them early enough and you physically remove them that is in so many in a certain sense curative. But by the time you're dealing with a metastatic cancer you're playing whack-a-mole. And very often you develop resistance to chemo the same way you would develop resistance to an antibiotic. And I think a lot of people don't fundamentally understand that a trial or a drug works if it gives you more time. Not indefinite time. Just some time. That's what we're hoping for. If you know that a disease that you're going to die from a disease and you can make the best use of your time that's a huge boon. But our culture and our society looks at death like it's a failure. When it's something we all do.

S: Right. At least so far.

C: Well and that's what faith healers and gurus they tap into people's death anxiety.

E: Of course.

C: And they use it against them.

E: Oh yeah. Fear is so powerful. It is the backbone of the basically that entire industry. Listen to this. I'll wrap up with this. Because I was looking up kind of like the history Steve and some of the things that you were talking about. This has been around for, this idea of mind cure your mind can cure you--thousands of years old. So there's tons of examples of it. Here's a more recent one. Tin May Lwin, former University lecturer in Myanmar and now advises people to approach treatment for cancers at this place called the Horizon Cancer Center in I believe the Philippines. And here's what she says. This is a university lecture: "Positive thinking, not putting too much pressure on yourself and always believing that you are in fact cancer free, living a happy life as you normally do, that is the key to being cured of cancer. I feel that our state of mind is extremely important to the treatment process. If we believe that we will be cured we will be cured. Believe me. If you think that you won't make it you won't." Oh my gosh. That is wicked.

J: That's terrifying.

E: As far as I'm concerned.

S: Totally.

E: Brutal.

Free Floating Black Hole (1:00:40)

• Astronomers may have detected a 'dark' free-floating black hole

Who's That Noisy? (1:11:27)

Answer to previous Noisy:

Seal chirps and bellows

New Noisy (1:13:32)

[hissing with rising horn sounds]

this week's Noisy

Questions/Emails/Corrections/Follow-ups (1:15:52)

_consider_using_block_quotes_for_emails_read_aloud_in_this_segment_
with_reduced_spacing_for_long_chunks –

Science or Fiction (1:32:23)

Theme: Science Misconceptions Item #1: CO2 is not the greatest cause of recent global warming, but rather shorter-lived molecules such as methane. Item #2: Most of the energy generated by the sun is not caused by the fusion of hydrogen into helium. Item #3: The heating up of a spacecraft as it reenters and descends through the atmosphere is not mostly caused by friction, which is only responsible for a small amount of heat, <5%.

Answer	Item
Fiction	Co2 and global warming
Science	Spacecraft reentry and friction
Science	Sun's energy and h→he fusion

Host	Result
Steve	clever

Rogue	Guess
Jay	Spacecraft reentry and friction
Cara	Co2 and global warming
Bob	Sun's energy and h→he fusion
Evan	Co2 and global warming

Voice-over: It's time for Science or Fiction.

Skeptical Quote of the Week (1:49:55)

Science is a search for basic truths about the Universe, a search which develops statements that appear to describe how the Universe works, but which are subject to correction, revision, adjustment, or even outright rejection, upon the presentation of better or conflicting evidence.
– James Randi (1928-2020), Canadian-American magician and skeptic

Signoff, SoF Bitterness (1:50:39)

S: —and until next week, this is your Skeptics' Guide to the Universe.

S: Skeptics' Guide to the Universe is produced by SGU Productions, dedicated to promoting science and critical thinking. For more information, visit us at theskepticsguide.org. Send your questions to info@theskepticsguide.org. And, if you would like to support the show and all the work that we do, go to patreon.com/SkepticsGuide and consider becoming a patron and becoming part of the SGU community. Our listeners and supporters are what make SGU possible.

[top]                        

Today I Learned

• Fact/Description, possibly with an article reference
• Fact/Description
• Fact/Description

References

Vocabulary