The Skeptics Guide #1064 - Nov 29 2025

You're listening to the Skeptic's Guide to the Universe, your escape to reality.

Hello, and welcome to the Skeptics Guide to the Universe. Today is Tuesday, November 25th, 2025, and this is your host, Stephen Novella.
Joining me this week are Bob Novella.

Hey, everybody, Jay Novella. Hey, guys, Evan Bernstein.
Good evening, everyone. And we have a guest, Andrea Jones.
Roy, Andrea, welcome back. Thanks for having me back.

Always love hanging hanging out with you guys so how's it going oh you know

everything in the United States and the world is totally fine so

nothing interesting to report yeah no but this is fine this is fine yeah I'm hanging in there despite I'm making an apple pumpkin bread as we speak so

once a year

I don't cook but once a year I'm seized with the desire to do so and today was a magical day for that so do you happen to have a good recipe no I just found something online I was gonna just make a pumpkin bread but I had two apples that were going to go bad, and so I was like, we're having apple pumpkin bread.

Nice. Yeah.
If it's any good, I'll give you the recipe. And if it's not, I'll keep it to myself.

You had me at bread, so I'm intrigued. At least take a picture of it and send it to me.
I will. All right.
Excellent. How are you all doing? Good.
So, Andrew, we are four episodes in.

We recorded four episodes of our political reality podcast. Wow.
Yes. Yes.
Four episodes and some intro material. Yes, that's right.
Yeah.

And what's funny is, like, you know, because we couldn't record the last two weeks because Steve was in Dubai.

And it's really good that nothing, like, impactful or interesting or noteworthy has happened in the United States over the last two weeks. I'm glad we're not behind the eight ball on anything.

Very slow talking. Very slow news cycle, yeah.

The administration had the courtesy of not doing anything out of it. Oh, my God.

Are you guys going to release those four? Are they just like hopelessly out of date at this point? It's a good question, Bob. No, the podcast is largely evergreen.
Yeah. It's not news.

It's not like, yeah, it's, it's, although sometimes we reflect what's going on in the news, but the topics are more, yeah, they're more evergreen. Yeah, the hope is that they're evergreen.

I suppose if we undergo a massive political change, things will stop being evergreen. But if the world continues, then it's evergreen, evergreen enough.

But it is something that we, you know, we've all been talking about is like to what extent do we need to tether to the news?

And I think to do anything political that's current, it all but has to be a daily show, I think. Definitely, yeah.
Because things just change too fast. The point of this show is, you know,

it could be in the context of things that are happening on a weekly basis, you know, because there is so many different things happening that it would bring up a lot of different topics.

Like, for example, with the Epstein votes going through the House and the Senate, you know, that would be a great episode to discuss just how, you know, a bill is created and how, you know, what's its pathway to being approved or denied.

Didn't we see that on Schoolhouse Rock already? I was going to say, only if you let me hear the song, Jay. Cool dated, you know.
But you get the idea.

The point is, like, we can't be sitting on top of the news as it comes out because that would mean that we're YouTubers that are recording six hours a day, and that's not the point here.

Yeah, I mean, and some of the topics, you know, we're choosing them with an eye to what's relevant. So one of the, I don't know if we're allowed to reveal what we've done.
Of course we can, sure.

Yeah. So one of the episodes is about voting systems and, you know, where does ranked choice voting fit in in the other types of systems out there? What are the pros and cons of them?

And ranked choice voting is something that has been in the news because New York City uses it locally. It's been used in different elections around the U.S.
It certainly

has legs overseas as well. And so it's something that's both in the news, but also not like the most cutting-edge news ever in the world.

But I am reading that Democrats are thinking about using ranked choice voting for more elections in the future in the U.S. So it's

yeah. Well, Steve and I disagree about ranked choice of voting, but it would be better than what we have, I can tell you.

Yeah, not necessarily the best possible option, but a lot better than what we have. That's right.
Okay. That's right.
I'd love to see it at the presidential level, at national elections. Ooh, geez.

Yeah. I mean, that's the thing.
It's history.

I've only participated in it in very local elections, and apart from the mayoral race in New York City, there were a lot of races where I didn't know who any of the candidates were, nevermind, knew how to rank them.

And so for something like president or senate or governor, it's a little more fun, I think, because you know more about the candidate.

You can really, you know, think through third-party candidates and other things like that. So, you didn't agonize over your fifth choice for comptroller? I did because I'm a dedicated citizen, but

one might not.

I did actually sit in the booth forever because the service was terrible, and I was Googling all the comptroller choices.

I couldn't tell you the name of a single, I couldn't even tell you who I ranked first, but I'm sure they're doing doing a great job. Nope, they haven't even been sworn in.

Yeah, I'm not sure what's going on on that item.

Well, then they are doing a good job. Yeah, I just can't believe they spell controller wrong all the time.
It's embarrassing.

All right. Well, let's get to some science news.

Bob, give us an update on the Hellion Fusion. Is it Helion or Helion? Probably Helion.

Yeah, probably. I'm guessing.
Thank you, Steve. This is your Quickie with Bob.

Back in episode 932, I talked about the company Helion Energy, which was working with Microsoft to commercialize Fusion Energy. Now, their design was interesting.
It was a hybrid technology.

Jay, say it. It's a hybrid.
They call it magnetoinertial fusion, which is a hybrid of magnetic and inertial confinement, which we've talked about on the show.

You know, many times.

It basically borrows the ideas of magnetic confinement like Eiders, or is it Eater? It's Eater, right?

Eater's Takamak, and inertial compression, like the laser implosions at Lawrence Livermore, that actually

had some notoriety.

God, that was a bunch of years ago, now, a few years ago, right, where they

hit ignition. But in their process, which is interesting, electromagnetic fields smash deuterium and helium-3 plasma together to fuse some of it, right?

So you got some fusion going on, and then the motion of the expanding fusion plasma pushes on the surrounding magnetic fields.

And they're taking that energy that's inherent in the magnetic field when it's being pushed.

They take that and convert it directly into electricity. And it's this direct conversion into electricity that's one of the more interesting attributes of their technology.

This cuts out the inherent inefficiencies that come along with using heat to make steam and then to run turbines and then using that to generate electricity.

So it cuts out a couple of the middlemen in the middle there and just goes kind of directly into electricity, which is really interesting. So all of this ties into their general strategy.

They want energy production to be so efficient that they don't need to chase ignition like every other, everyone else is doing.

And ignition is the point where you've got basically a self-sustaining fusion reaction going on. And also there's other little attributes as well, but that's the main thrust of ignition.

So they think they don't need to really chase that.

So back then, they were, I talked about, hey, they're working on, they have their sixth generation fusion generator and it reached 100 million degrees Celsius and rah, rah, rah, wasn't that awesome?

And they were talking about that they will create their seventh-generation fusion generator called Polaris, and it should be ready by 2024.

And they said with this first prototype, they wanted to demonstrate electricity produced directly from fusion, potentially even net electric.

Okay, so that's what they were, that's what was happening a few years ago in 2023. So now we're seeing Helion back in the news, and they have, in fact, created their Polaris fusion generator.

And it's a doozy.

They have a monster capacitor farm delivering 100 gigawatts into the machine. Now, think about that.
100 gigawatts, that's like grid-scale power, right, in one building.

Of course, it's only for microseconds, but it's still quite impressive. Their CEO says it runs at 100 million degrees.
I thought that's what Trento was doing.

So that didn't really increase at all, it seems, if these numbers are right, but that's fine. 10 times the sun's core temperature.
That seems like enough right there, isn't it?

And they also mentioned that their coaxial cables can carry these pulses. Those coaxial cables are 720 miles long if you put them all together, laid them end to end.

So that's a lot of cables, a lot of coax.

Now, this is one thing that was kind of discouraging. The details of their progress are only kind of minimally and cautiously available, right?

Because they, and they, they, that's because, ostensibly, because they had Chinese competitors steal bits of their intellectual property in the past.

So that, that theft, of course, is totally believable, right? We read about that in the news all the time.

But it does make it hard for the scientific community to properly assess their chances of success. And it also seems like, oh, really? You can't give us a lot of information? That's just,

it's definitely a red flag. But I understand if it truly is because of

Chinese competitors stealing their intellectual property, then I would understand why they would be so reticent. Let's see.

Also in the news, they're still aiming for this idea of minimal minimal fusion, which actually wasn't very, I don't think they clearly discussed that a few years ago because that really wasn't in my notes when I looked at them from a few years ago.

They were just pushing for fusion. It didn't seem like they

didn't hear the term or the phrase chasing ignition, that they're not chasing ignition.

So, if they didn't make it clear a few years ago, they're definitely making it clear now that this is that they're going to rely on these efficiencies.

So, here's a quote from Helion, CEO and co-founder David Kirtley. He said, We can recover electricity at high efficiency.
We require a lot less fusion. Fusion is the hard part.

My goal, ironically, is to do the minimum amount of fusion that we can deliver a product to the customer and generate electricity. So very interesting.

Of course, the question remains, you know,

is that going to be enough?

Is a little fusion enough? So Helion has also broken ground and started work on their Orion plant.

They mentioned that a few years ago, that they were going to do that. So they're on track for that.
The Orion plant has been worked on. And this is going to be their first commercial plan.

The technology, the idea is that the technology that they develop in Polaris, which has been running for most of 2025, I believe, that technology will be ported into Orion.

At Orion, it's going to be, the plan is to have a 50-megawatt class plant online. They predict by 2028, so that's when they...

a few years ago they were saying 2028 as well so it's kind of um i don't know what to make of that it's it's i'm encouraged that they that they're not saying well it's going to be 2030 or 2032 now,

which is what we would expect, right, Steve, to expect these initial dates that are a few years old to be pushed back. So they're still saying 2028.

And when it's done, it's going to be under a power purchase agreement with Microsoft so they can feed their hungry data centers. But Helion's competitors are, of course, skeptical

that their competitors are going to hit the 2028 goal. I got one quote from Ben Levitt.
He's head of RD at Zaps Energy.

He said, Ben said, I don't see a commercial application in the next few years happening. There's a lot of complicated science and engineering still to be discovered and be applied.

He says he doesn't see it happening in the next few years. Does that mean it could potentially happen in five years? Or is he thinking more 10 years or more? I don't know.

I couldn't find any other quotes around that. Other people are worrying that if helium screws up badly, it could embarrass and taint the whole industry.
So yeah, that's a concern as well.

I mean, that's something that

we seem to be making such cool progress in the past decade. I'd hate to have, you know, what was that called, the AI winter where, you know, resources dry up.

Bob, let me ask you a couple questions really quick. So I'm reading that they have not achieved net energy production.
Yeah,

I didn't see, like I said, I didn't see too much information about what the status is. And because they're saying that because they're trying to,

like, that's not information, that's not technical information that could be stolen. Just saying we have achieved net energy, or this is how close we are to net energy.

That's sort of the bottom line here. That's like a big thing.

It is. And I am discussing.
I'm suspicious that they're not saying it. Yeah,

that's definitely

a red flag right there. And they've been running this thing, apparently, from what I can tell, Polaris, they've been running this thing all year, like every day, they say every day.

So here's the other thing. So the other question I have: so this is producing electricity directly, right? It's not going to heat steam and turn a turbine like every other other reactor.

It's just the fusion will create a

magnetic field that can induce current directly in

their coils.

Right. As a plasma is expanding, because the plasma is going to be energized by

the fusion that's happening inside, right?

So the fusion is going to be happening within the plasma, and that's going to expand the plasma field, and the magnetic field around it is going to kind of try to hold it in.

It's going to be expanding. So the energy that the plasma is putting into the magnetic field to expand it is what they tap into and convert directly into electricity.
So that's one of their...

Does that work? Is that just in theory or has that actually happened? They have a proof of concept there? Yeah, I think they proved that with

Generation 6, I think.

But how efficient it is

is another question.

I'm pretty sure that that's actually happening. And that's one of the key interesting aspects of their technology.

But

I don't know how well it's happening. You know what I mean?

So, yeah. But like I said, I'm fascinated by their approach of not chasing ignition.
But will it be good enough?

Is that going to be enough if not having ignition?

Just because your whole entire process is super efficient, so you're kind of like making the maximum use of what little fusion is actually happening.

I will say I'm optimistically skeptical about this.

We've been burned too many times.

Some of this looks really interesting, but I'd like to have some other people, some other real scientists looking at this because we had some, I looked at it, I remember a few years ago, some scientists were skeptical.

So check out that episode for some of their quotes. So yeah, so they were skeptical a few years ago.

I haven't come across too much of that right now, but it's still kind of, you know, it's still kind of early in terms of this

Polaris. I mean, it's been running for a while, but I haven't read too much about it.
But if this works, this would be a hell of a coup.

And they could actually say, you know, our fusion is on your power bill. There will actually be power bills going out that might list fusion as one of the sources, which would be really cool.

All right. Thanks, Bob.
Jay, we have another sort of cutting-edge science news item. You're going to tell us about using CRISPR to make GMO wheat.
Yeah,

this is really cool. So, wheat draws almost everything it needs from, of course, the soil, right? Its roots take up minerals like,

what would it be, Steve? Do you have any idea of what wheat would need from the soil? Nitrogen, phosphorus, calcium. That's pretty good.
Anybody? Yeah, potassium, zinc, iron.

Right? Yeah. So, so the soil microbes and the fungi, are you guys like fungi or fungi? Fungi.
I don't know, man. I go back and forth.
Yeah, it depends on the time of day. I don't like that.

What about evening, Devin? I think I might like that. I think that's a good thing.
No, definitely fungi. Yeah, that's right.
Yeah, I want them more of a fungi.

So you have the soil microbes and you have the fungi, and they help unlock these nutrients that the wheat's roots either physically can't reach on their own or can't use in the form that's present

in the soil, right? And water acts as this delivery system that moves those nutrients into the roots. As

the water is moving around in the soil, or if someone is watering,

that helps distribute these nutrients around the roots. Now, the atmosphere contributes carbon dioxide for photosynthesis, right?

But there is one critical nutrient we cannot access directly from the air, even though it surrounds the plant. Now, I will ask you again, do you guys know what nutrient that is?

Talking about nitrogen. Steve, you're so smart.
I'm so proud of you.

Yeah, it is. It's nitrogen.
So nitrogen is the main ingredient wheat uses to build the proteins

in the healthy leaves.

Now, protein,

as you may know, is present in wheat, and gluten is a protein, but gluten doesn't improve the flavor. It only improves the chew.
And there's other reasons why bakers want it. But that's a side note.

We can't use atmospheric nitrogen because it's locked up as N2. And this is why modern agriculture has to rely heavily on nitrogen fertilizers, right? And there's a ton of money in that.

The problem is that wheat is not an efficient user of it, meaning that, you know, wheat,

your typical wheat field takes up about 30 to 50 percent of the nitrogen that is fertilized to it, right? Or that's applied to it. And the rest goes into.

Well, the problem is, Bob, that

there's downstream effects, literally, like runoff into waterways. Oxygen runoff, okay, yeah, yeah.

It's a terrible groundwater. It really is a big problem.
Do other plants have a more efficient use of the nitrogen? I did not check all 3,292.

No,

I'm sure some do, Ev.

But, you know, wheat being a massive crop, like this is like a worldwide, unbelievably necessary crop. It's used for so much.
Yeah, it's just one of our most important crops.

And in fact, it was one of the crops that let humans not be,

you know, not have to be nomadic and let them stay and

grow wheat.

And then, you know, there's a, yeah, without a doubt. There's a couple of other things that happen.

It converts into nitrous oxide, which is a greenhouse gas. And there is a

process called volatilization, which means that the nitrogen goes into the air as ammonia. So there's lots of different things, lots of chemistry happening here.

And of course, a real monetary downside here is that farmers lose money because they're buying a lot more nitrogen than that is actually being applied to the thing that they're trying to fix.

And like I said, waterways will get polluted, atmosphere picks up more nitrous oxide, just a lot of nasty things happening with all that fertilizer.

The good news here is that, like Steve said, these researchers at UC Davis found a way to push wheat into working with soil bacteria that can convert atmospheric nitrogen into a form that the plant can actually use.

So, this is not a full replacement for fertilizer, but it changes how the plant interacts with the microbial world around its roots. And this is really interesting.

So, let me get into some of these details here.

Their work is built around a pretty simple idea. Instead of trying to convert the wheat into something, it's not like a bean that grows nitrogen-fixing nodules, right? You know,

sure, they could borrow some programming from a a bean and try to push it into the DNA of the wheat and all that, but that's a big deal and it's really difficult and

it's expensive, time-consuming, and there's no guarantee that's going to work. So they decided to make a small change in the wheat that influences the soil microbes.

And they decided that if they could pull this off, that bacteria would do most of the work here. And that's what the study actually tested.

The researchers started by first testing thousands of natural plant chemicals to see which ones could influence the bacteria that live in the soil.

And this was, of course, like a very time-consuming process, but they scrolled through all of these different chemicals. Only a few of the chemicals actually panned out, and one in particular called

an easy word to pronounce for people like me. This one stood out as the best.
And epigenin is a natural compound that many plants make, And it belongs to a family of chemicals called flavones, right?

And flavones are substances plants use for communication. It's kind of like their pheromones.
And they also use it for defense and dealing with stress.

In the soil, these chemicals can act, they can act like messages that guide or influence nearby microbes.

When wheat roots release apigenin into the soil, certain nitrogen-fixing bacteria respond very strongly to it.

So the bacteria that responds, it actually draws them towards the roots and encourages this bacteria to form a protective biofilm. And this is crucial to this whole thing.

The biofilm gives the bacteria the low oxygen conditions that they need to convert nitrogen that's in the air into a form that the plant can use.

So the team used CRISPR gene editing, and what they did was they increased the amount of apigenin that wheat produces in its roots. And it's, you know, it actually worked.

It was brilliant how they figured this out. The roots will then release extra epigenet into the surrounding soil.
It's apigenin,

epigen, and they know what I mean.

Under these nitrogen-limited conditions, though, the edited wheat performed actually better than regular

quote-unquote normal wheat. It had higher nitrogen content in its leaves and the roots.
It had stronger photosynthetic activity and better grain yield, which is fantastic. Yeah, the soil, right?

I mean, I was really surprised to read about that, that it didn't just equal and solve a problem. It was better yields, I mean, all across the board with everything we needed.

And to clarify, that's under what they called limited nitrogen fertilization conditions. So that's with less fertilizer than you would normally give.

So the soil around the engineered plants, it also showed increased nitrogen conversion activity when they measured it using standard isotopes and biochemical tests.

The pattern actually held across all their experiments and more apigenum in the root zone.

Apigenin. There's more of it in the root zone.
It allowed more bacterial nitrogen fixation, which leads to healthier plants.

When, you know, let's say fertilization is scarce or the farmer can't afford it or whatever the problem is, the plants can survive,

more likely to survive with a lot less fertilizer. And there is, of course, a practical angle here.
So the wheat plant that can supplement part of its nitrogen requirements through this

co-op bacteria, this is a really big deal. It doesn't eliminate the need for fertilizer, but it really does reduce how much it needs.

And of course, the downstream effect, like I said, all of that decreases when you use less fertilizer. So the CRISPR wheat stayed productive.

When they cut the nitrogen to half or even to 30% of the normal application, it continued to function perfectly fine.

And under those low fertilizer conditions, it still outperforms conventional wheat by a wide margin. Of course, you know, we have to be cautious here.
You know, these are controlled experiments.

They're useful, but

trying this in the real world.

It's a lot more complicated. Yeah.
Yeah. These field trials, there's a lot more variables in the soil.
There's unpredictable weather.

You know, there's complex microbial communities that are all working with each other and maybe even sometimes competing with each other. You have soil ecosystem shift, you know, not just from

global warming, but like just as you go, you know, throughout a region,

the ecosystems are different from here to there. It could be even 50 miles away.
You have a totally different scenario going on.

And of course, these bacteria might not behave consistently from season to season. But, you know, again, this is the beginning.
Like, they found something. It works.

It's a definite thing. The mechanisms work.

It's one of those things that it seems likely to pan out. I know that they're going to have to make modifications and it's going to take some time, but this is a real win here.

So, just to clarify a couple of things that you said, but I just wanted to emphasize them a little bit. So, there are crops that fix their own nitrogen.
They all do it through these bacteria, right?

The plants aren't doing it, it's always bacteria. Like the legumes, you mentioned the nodules.

What the nodules do is they create the low-oxygen environment for the bacteria to thrive and fix their nitrogen.

There are programs to try to identify the set of genes necessary to make that happen so that we could then make fully nitrogen fixing crops like out of the ones that aren't, right?

Like including wheat. This is like a all right, they found an easy way to do a partial fix where, yeah, again, it's rather than these complicated nodules, it's going to create the biofilm

to create the lower oxygen environment. And we'll see remains to be seen how much they'd be able to whack back fertilizer, but even like reducing nitrogen fertilizer by 10%

could save a billion dollars a year. Yeah.

Yeah, and also, you know, this is the beginning of this study. It doesn't mean that they couldn't even make it more productive

with a lot more experiment and everything. But

this is the reason why I picked this news item is, you know, we talk a lot about these types of things. Like, hey, there's a cool thing that happened.
You know, it's in play.

They're studying it. But I think it's important for us to recognize here that CRISPR, which is a platform,

when you think about what it takes to create a platform like CRISPR and how many years and years and decades it takes to just get scientists to start using it.

So, CRISPR was discovered essentially in 1987. I think it took until

we're going like 2010 plus before it started to kind of hit the science scene and people were really using it to do stuff. And then, of course, since then, we've made tons of advancements on it.

It's incredibly powerful. It's incredibly useful.
And we need scientific funding to let scientists just experiment and try different things and see what we can do with it.

Now, particularly in a situation where we have global warming,

which is going to change so many things about

where arid land even is on the planet, let alone how productive it's going to be.

We do need advancements to help us continue to feed the 8 billion plus people that we have and more as the decades go by.

So, I think it's an important reminder that we show some respect to the scientists who created this and who are working with it now and realize that it is important that money is leveraged this way to find these discoveries.

Andrea, I noticed that you're quiet here, and I'll take that as disrespect. Do you not like CRISPR?

No, I like CRISPR. Some of my best friends are CRISPR, J.
No.

I literally was like, oh, I have two questions, but both of them might be things you've already covered on the podcast.

So one is a really simple one, which is, is there such thing as wild wheat or have we domesticated all of it? You know how, like, there's no more wild cows.

And then, my other one is, what sorts of things happening with CRISPR are you all covering on the show that, like, that you're super excited about? Because this seems awesome.

And I remember being super excited about CRISPR when it came to malaria and when it came to a couple of other sorts of like infectious disease-related things.

And this is, I'm not familiar with it in agriculture, so I was just kind of curious

what else I should be excited about. Well, there is wild wild wheat, and it's still out there.

Can we add it to the corner?

It's one of the ancestors of domesticated wheat.

You probably can.

Corn wheat is a little bit of a. Yeah, I'd have to take a look at it to

see what the head of it looks like, to see if there's actually anything there that you can turn into.

I only thought of it when you were talking about the importance of the agricultural revolution. I was like, oh, yeah, did we just find it and make it into something? And that's how we cultured.

That's a really dumb question.

Way more

nutritious and

bend high. And just like a lot of other things that we eat today, guys, like it can't survive.

All the wheat that we rely on, right? Yeah.

And corn and soy, probably. Yeah, it won't continue to exist if people don't grow it.
You know, like it has to be grown, it has to be fertilized, it has to be taken care of.

It's not the type of thing that, you know, just keep going on its own if we didn't tend to it. Like, it really does need us now.

We have a codependent relationship. I mean, I also was thinking about the various uproars from the non-GMO side of things.

And, Jay, you raise a good point, which is we've been modifying these things forever. Yeah,

thousands of years, right? Yeah, yeah. There's almost nothing that you eat that's not massively altered.

Carl Sagan said, just look around you. Everything you see is artificially enhanced.
Yeah.

All right. Andrea, this is interesting.
You're going to talk to us about our LLMs, large language models, are they changing how we think as a group or thinking? I'm going to look at this real quick.

Let's see. Yeah, so it's,

Evan, are you going to look up the answer? Yeah, yeah, I'm going to chat GPT that one. Yeah, perfect, perfect.
That's great. Yeah, I'll just read that out loud in my voice, and we'll call that good.

Yes,

Steve, so I have come to the show, and every now and again, I manage to muster some research that is not squarely political science.

This is social science, and I'm going to talk about political science briefly, but it's not exclusively political science by any stretch.

So, there's a paper that came out in late August that I'm super into.

Full disclosure, the author is my PhD advisor, Scott Page, at the University of Michigan, and I'm a big fan of his, and he and I work together, and so a lot of this research is stuff that I work on with him.

And so, big, big bias alarm bells going off, but it also means that it's something that I think is super interesting and wish more people knew about.

So, okay, so he published a paper at the end of the summer called Everyone, Everywhere, All at Once, LLMs and the New Physics of Collective Intelligence.

And if you didn't get excited at the movie reference, I hope you got excited by the phrase physics of collective intelligence, because that's just the coolest phrase I've heard in 2025.

Except for that word that Jay mispronounced 200 times.

I'm not even going to try to say it, so there's that. All right, so this is a paper about how we can use large language models to change the physics of collective intelligence.

I'm going to come back to the physics part, but first I want to ask you all,

are you familiar with the term collective intelligence? Is it a term,

or do you have a guess as to what it might mean?

I mean, I've seen it reference to just trying to accomplish things in groups, basically. Like the crowd is smarter than the individual.
Right. Right.

And sci-fi is like a kind of like a hive mind, which is a collection of

many minds.

That's right. Are you all watching Pluribus, by the way? Oh my God, yes.
Okay. No spoiler premise.
No spoiler premise. All right.
Small aside. But that is exactly the hive mind piece.

So yeah, so that's exactly the idea.

So it's this idea of the wisdom of crowds and how can we make it such that groups of people working together become smarter than, say, the sum of those people or ideally

smarter than and not just more knowledge, but more innovative, more adaptive, more creative,

all kinds of interesting things that an individual could never come up with.

Sort of the idea of like you're all sitting in a room and you're brainstorming, and you all come up with really great ideas because you've been bouncing those ideas off of one another, as opposed to if the five of us all sat in a separate room and thought of things by ourselves.

That said, you may also be thinking about every single time that the crowd is maybe not so wise, right?

You know, we think about herd mentality and we think about people all rushing to false information or to political views or religious views or whatever, or

homeopathy, or whatever it is that we would consider not so intelligent, but it's definitely part of that group think.

So, this is a whole area of research basically that crosses a number of social sciences that says, what can we do to make it so that when we put a bunch of people in a space together and they have some kind of common problem to solve, that they do it intelligently as opposed to not so intelligently.

And so, this is where I'm going to crowbar a teeny tiny bit of political science.

And so, if you're thinking about collective intelligence in economics, you're probably thinking about how do you design a market so that people can trade things and exchange value and innovate and all that stuff.

In political science, and Steve, you and I talked about this on the political reality podcast already, the idea of how do we get a group to make a good decision or come to some kind of agreement is pretty much what's happening every time we think about what does an election system look like?

How do jury deliberations work? What are the ways that Congress

deliberates with one another before they have a vote? And so all that stuff in politics is also collective intelligence.

And there's all kinds of other ways that a community might come to some kind of agreement as a group. I spent a lot of time with Scott working in the space of companies.

So companies are very interested in collective intelligence. So if I'm going to put together a team to solve some problem,

a tech problem or a CRISPR problem or whatever it is, it's not enough to just get a whole bunch of smart people and put them in the room and hope for the best, which is what most of us tend to do.

What you want to do is be thoughtful about how those people interact and what are the processes by which they bring different bits of information, deliberate over those pieces of information, and then select the best one.

And immediately, you should be thinking, and I don't know, maybe anyone listening who's ever been in a work meeting, I can certainly speak to faculty meetings at NYU and pretty much everywhere.

I don't feel like a lot of those meetings lead to collective intelligence. Usually, what happens is someone comes in with a strong opinion.

That person, if they're also the loudest person, will will dominate the conversation for a while.

Parish forbid, they're also the most senior person in the room, and so no one really feels comfortable pushing back.

A couple people might ask a few questions and then everyone nods and then we just do the thing that the guy who called the meeting wanted you to do.

Or you try to introduce new ideas and say, well, actually, you brought me in because I have a different perspective and I think there's X, Y, and Z problem. And then the rest of the room says, what?

No, that's silly and shuts it down. And so again, it's like, why even bring this outside expert into the room?

And so all of that is to say is that there's lots of things that we can do, and this can be in our own work, in our own lives, in our own communities, to make our groups smarter.

This article is about how we can use LLMs in a way that I at least, I'm curious if you all have seen this, but in a way that

I don't normally see people talk about the use of LLMs. So the idea here is picture a meeting that you're going into.
We got to solve this problem.

We got to go through, you know, whatever decision-making process that we have to go through. How are we going to get to collective intelligence? Well, there's three steps.

One, we want to have as many independent inputs as possible.

So if we're all sitting around trying to solve a problem, we need to hear from Bob and Evan and Jay and Steve and me, and we all need to talk it through and get that information out there.

We then need to also be able to consider all the options, and then we need to choose over those options. The problem is that those three things take forever.

If we were all sitting down to say, you know, in a meeting today, this is why we do it before the podcast, and said, well, what should we talk about on the podcast tonight?

And it's like, Evan will say his ideas for 10 minutes, and then Jay will say his ideas for 10 minutes, and I'll say my ideas for 10 minutes.

And before we know it, an hour has gone by, and we haven't even gotten to the deliberation.

So the idea in this paper is what if before the meeting, before anyone gets to the room where we're going to talk about whatever it is, each person talks to an, you can type it, you can talk it, talks to an LLM about what their ideas are.

So, you know, maybe it's us brainstorming for the political reality podcast. So I'll say 10 minutes into my phone.
I think we should do this, da-da-da-da-da.

Meanwhile, Steve is somewhere else doing that into his phone, Jay's doing it,

and we're doing it. And then while we're walking to the meeting and while we're sitting down and saying, hey, everyone, how are you? Get some coffee, blah, blah, blah, blah.

And LLM is summarizing all the key pieces. And so then by the time we start the meeting, we start from step two, which is let's deliberate over what the different ideas were.

And there's a bunch of benefits to that. One is obviously you're saving a ton of time of having to sit there and listen to everybody.
The trade-off is you don't hear all the nuances.

You hope that the LLM summary is correct. And so, obviously, depending on the importance of the conversation or the level of nuance and privacy, you wouldn't want to use it for everything.

But for pretty standard stuff, this could get you pretty far.

And the other great thing is that one of the best ways to generate collective intelligence is to have these independent inputs that are truly independent. It doesn't work.

And we've all seen these psychology studies where you sit in a room and you show a participant a circle on the wall, and everyone else in the room says, Ah, it's a square.

And then the person who's the subject kind of doubts themselves and says, Yeah, I guess it's a square, even though it's a circle.

Like, we are so quick to say, Oh, yeah, I agree with what this person said, or you all have been podcasting longer than I have, so let's do what you say, or just to agree with the majority that these independent inputs that we submit before I hear anyone else's ideas, and I just say it to an LLM, that's going to get much richer information from me and take more advantage of each of our individual perspectives as well.

So it's just a teeny tiny paper.

Oh, and the reason it's about the physics of collective intelligence is because if we think about physics, we think about the constraints on groups working together, and those constraints are often around space and time.

We all have to get to the same place to have the conversation, and we all have to be free at 4 p.m.

In this case, we saw during COVID and we've seen ever since that the constraints of space have really gone away, not completely, but largely, right?

We can do podcasts from Beirut and we can do podcasts from Tokyo and we can do podcasts from Connecticut. But the time issue has been a problem.
And so

this idea that we could use LLMs to speak simultaneously, to have meetings not at the same time, but then have a separate meeting where we actually talk about, you basically get to start meetings halfway through, but have the full amount of time you always had.

So I just thought it was a super cool idea. It's not, you know, a heavy-duty

experimental research paper in that sense. It's more of a thought piece that says, we've been thinking about LLMs as more of an individual tool.
I sit and it's my sidekick and it helps me.

But we could actually really get group minds working a lot better and faster and more creatively if we start to think about them at the group level as well.

But you're basically laundering your ideas through a chapter.

But so yeah, I would be interested in how that might distort

everyone's ideas.

Well, what you would want to do, and this is where you would hopefully, you know, work with an LLM that is either tailored to your organization, or if it's a bunch of doctors in the room, you know, I wouldn't just use mass available chat GPT or something like that.

But yeah, you would also, especially early on, you would also want to go through and make sure that the summaries of what you've said are correct.

There's certainly issues where if someone speaks with an accent or speaks in a slightly different style, that

their input would not be measured exactly the same way. And I certainly wouldn't use it for big decisions around, like, should we fire somebody or promote somebody?

But I don't know about you, but a lot of meetings that we go to are really

not that high stakes.

And so, you know, I have one tomorrow morning where it's like, yeah, rather than each of us recap, what if we did the recap separately and then read a bullet pointed summary right before we started?

And then, over the course of the meeting, you can then say, oh, is this what you meant?

And go back

and dig in. But it's certainly not perfect and certainly not for every conversation.

And there's also plenty of conversations where the murky middle of like saying a bunch of stuff and then saying, well, actually, maybe I disagree with myself. Or let me restate that.

That might spark an interesting idea and you would lose some of that. But, you know, think of your average insufferable corporate meeting.

There are some ways to improve them, I think.

Yeah, no, just you know, someone I know not too long ago had a meeting where they had to like approve bylaws or something, right? And

they have to have a public comment period. So literally, they had a meeting where 100 faculty members all could, every one of them, if they choose, could state their opinions about that.

And it took forever. They had to actually have a separate meeting because it took twice as long as they were planning on it taking.
And it was just interminable.

So I think that's the kind of thing where anything is better than that, you know. I mean, think about voting, right? In the United States, voting is largely simultaneous.

Yes, we have early voting and mail-in voting, but for the most part, within a very contained bit of time, we all show up and we turn in our ballots.

And if we did it sequentially, which is how most meetings and most public hearings are done, we would still be going through our first election. You know, like, okay, you can't go until Steve is done.

Like, okay, am I turned on? It can't happen that way. Yeah.
So it's like, so what are the sorts of things?

And I think it's a great question: is like, what sorts of meetings or gatherings or what types of collective intelligence are we trying to capture here?

If what we really want to do is just get a sense of what everyone's perspectives are, it could work well.

If we're trying to sit down and say, let's talk through our ideas for a new title, maybe we want to be in the room and hear all of the ideas, not just the shiny one we got to at the end.

I don't think this is strange at all. I think, you know, this is us interfacing with our latest and greatest technology.
I mean, this is what we're supposed to be doing.

Like, you know, it's worth even trying just to see if it works for your group or under what circumstances or whatever. But, you know, efficiencies are going to be found only if we look for them.

And I think, too, Jay, that maybe one thing that is a potential real upside for this is, again, thinking in the corporate world, it's like there's so many meetings where, or there's so much research that shows that women are more likely to be interrupted.

Someone who's a racial minority is more likely to be dismissed if their idea doesn't align with the ideas of the group. And I'm not trying to rattle, you know, woke slogans or anything like that.

This is like actual research that typically, you know, people from different backgrounds, different preferences, introvert, extrovert, different levels of comfort speaking up in a meeting, we're missing a lot of people's perspectives.

because the dynamics of the meeting are such that someone's interrupted or someone's talked over or someone doesn't feel comfortable speaking up in the first place.

Or like I said, you feel afraid to disagree with the majority.

But if you all have to say your thing ahead of time, like, I think this is a great idea, I think this is a terrible idea, I think we're missing X, Y, and Z, you're going to get everyone's information equally, as opposed to just hearing from the loudest or more confident

person in the room. So I think that is potentially really powerful.
That said, the opposite could be the case.

It could be if we're, if we use the wrong aggregator or we use the wrong type of tool, we could be replicating those biases. But I think sort of like you said, Jay,

it's worth giving it a shot to see if we could actually make meetings more inclusive this way.

I think it's worth mentioning that, you know, there is a lot of information out there about how, you know, working with LLMs is bad, right?

There's a lot of, you know, we get a lot of emails from people, and I've been having discussions with people. I think we should mention that, you know, they do consume a ton of energy.

We don't know from a big brother perspective, like how much of our information is private and everything. So, I mean, I think it's fine for us, of course, to talk about this stuff.

I I think those are legitimate concerns. You know, we were facing a genuine dilemma.
We have a useful tool that has some serious downsides.

You know, one of the biggest being its energy use, which is, you know, we've had to revise all of our projections about energy and climate and everything to account for the AI factor.

Yeah. There's no way around that.
Not by a little either.

But the thing is, I mean, it's a good point because if it produces only marginal advantages, it may not be worth it. But that's not typically how people think.

I think if it's even slightly more efficient or

people like doing it, a lot of people are going to do it, regardless. Yeah, I think, Steve, I

similarly,

usually when I think about LLMs and how I see people using it or how various outlets on the internet recommend, like, streamline your workflow with Gemini's, whatever, whatever. You're like, okay.

And it doesn't really, I'm like, is it really worth it, like the risks with the climate and the energy consumption to summarize my emails for me?

Like, this is shaving 30 seconds off my life, maybe, but not really. It's mostly just bothering me, and I'm going to go read the emails either way.

This one was one of the few instances of using LLMs that I was like, oh, this is a genuine change in how we do things.

And if it's possible that this generates more collective intelligence, it could be that it is the sort of thing that helps get us more quickly to more innovative solutions like the cool CRISPR stuff that Jay was just talking about and even what Bob was talking about is, you know, like greater collective intelligence could really get us there.

But, you know, I'm more excited about this than I am about like, oh, we're going to summarize, you know, emails. But yeah, no, it's a it's a huge concern.

And if I could put it back in the bag, I would strongly consider it. Is this a current concern? In other words,

it will likely become more energy efficient in the future. So, I mean, kind of like we bring about something now in 2025, but by 2035 it will be 90% more energy efficient, for example.

We hope so, but they've been working on it for a long time. Yeah.

So

it's one of the things that it's hard to say. But like a lot of things, if they do get efficient, they use that to make it more powerful.
So we never actually get the savings.

It's like as electricity gets cheaper, we get more electricity, we get more light. And it just uses more.
As you get more devices to clean our house, our house has to be cleaner.

Yeah, like multi-terabyte hard drives. We just, no matter how big they get, we fill them up.
My other concern here, though, is that, as you say, like, yeah, you should proofread the summary.

No one's going to do that.

Most people are going to do the lazy route and just. Read your terms and conditions, yeah, right.
Right, right, right. That's just,

you know, that's, we seek our lowest energy, right? We're not going to spend energy that we don't have to. Yeah.
You can ask ChatGPT to summarize it for you. Yeah, that's right.

Give me the 15-second terms and conditions. I've done that.

Well, everyone, we're going to take a a quick break from our show to talk about our sponsor this week, Aura Frames. Guys, we've all been there, right?

The holiday season, you got to buy a million gifts for a million people. You're running out of ideas.
What can you get that's really good, but also has a personal feel to it?

The answer to that conundrum is Aura Frames.

Yeah, this year I'm going to get Aura Frames for two of my friends, and I'm going to load a bunch of pictures on there of basically everything I have, which goes back, I don't know, 30 years at this point.

It's a really awesome thing to do because you're you're going to give them a bunch of pictures that they don't have and you're going to give them really awesome access to it.

And to do that, you just download the Aura app, connect to Wi-Fi, and then you can upload an unlimited number of photos or video that your friends will instantly see.

For a limited time, save on the perfect gift by visiting auraframes.com to get $35 off Aura's best-selling Carver Matt Frames, named number one by Wirecutter, by using promo code skeptics at checkout.

That's A-U-R-Aframes.com, Promo code skeptics. This deal is exclusive to listeners and frames sell out fast, so order yours now to get it in time for the holidays.

Support the show by mentioning us at checkout. Terms and conditions apply.
All right, everyone, let's get back to the show. All right, let's go on.
Do you guys remember Thea? Yeah.

Yeah. That's really.

That's the.

I don't remember. I wasn't there at the time, but I read about it.

The Mars-sized planet that crashed into the proto-Earth

creating four and a half billion years ago, creating our current Earth and moon system. Right.

So scientists have

scientists have a question, and the question is, where did Theia come from? Specifically, what part of the solar system did it come from?

Over Theia.

So how could they answer this question?

So do you guys know, like just generally, how do scientists know where something in the solar system is? Mineral composition

ratio, balance of minerals would be.

And what specifically?

You're right up to the point. Isotopes.
Isotopes. It's the isotope ratios, exactly.

Yep, that's right. Yeah, I mean, so apparently the isotope ratios tell you like so much about the history of

so much of stuff in the solar system, including where, like, how close to the sun did it form. Because there's different isotope ratios in different locations in the solar system.

So there's a recent study trying to address this question of where did theia come from by looking at isotope ratios. Smart.
But there's a problem. Oh.
Where do you look?

We have to look at Earth and Moon. Because when

it is believed that the Moon is made mostly of theia, right?

But the Earth is just a complete mishmash of Earth and theia, right, mixed together. Four billion years later, you know, all kind of, you know, yeah, where do you find something that happened?

Are we all made of theia?

There's a little bit of theia in each of us. That's right.
Andrew.

So what they did was they looked at a lot of mineral samples from different strata, and they also looked at samples of moon rocks. Sure.

So the Earth and the Moon pretty much look the same in terms of isotope ratios.

But by by doing an extensive survey,

you can't look at a piece of rock and say, this piece of rock is Theia, right? I mean, you just can't do that. But

what they could do is they said, okay,

if Theia came from the outer solar system,

what would we expect

a random survey of Earth rocks to look like? And we do have samples of those, right? Well, yeah, we have Earth rockets all over the world. No, no, no, but samples of non-solar system material.
Sure.

Not non-solar system material, non-earth material. So we have

as a reference. As a reference, we have samples of meteorites from all different places in the solar system.

So we have a catalog. We do.
We know what an outer solar system rock looks like in terms of its isotopic ratios. We know what an inner solar system rock looks like, et cetera, et cetera.

So the Earth and the Moon obviously look like inner solar system rocks. They look pretty much the same with each other.

So they sort of addressed this question by saying, what if Theia came from different places than the solar system?

What would

a survey of Earth rocks look like in terms of its ratios? And they concluded, what do you think they concluded? Where do you think Theia came from? It's a local boy. Yeah, I'm with

it would make sense, inner solar system, like all the other rocky. Yes, inner solar system.
It did not come screaming in from the outer solar system. It was probably our neighbor.

They think, and this is a little speculative, this is like a statistical kind of thing. They think it was probably a little closer to the sun than the Earth.

So you can imagine these two planets in very close orbits, but eventually they crossed, you know, and then Thea smacked into the Earth. Or Earth smacked into Theia, depending on your

perspective.

What a day that was.

I know, right? Can you imagine?

Would Theia have had to be smaller than the Earth, and that's why the Earth hung on?

I mean,

it was. It was like Mars-sized.
Yeah, okay. Again, I think Bob's correct in reality.
The best way to think of it, these two planets hit each other and merged and spit out the moon, basically.

Changed the orbit of the Earth, changed obviously the mass of the Earth,

gave us our satellite. What's really poignant, though, is that, I mean, it's possible.

We know how early life arose on the Earth. It's basically, as soon as the magma wasn't in a place, then life kind of started.
It's really, we push it back every millions of years all the time.

So Earth Mark I, right, the Earth that was destroyed, could have had life on it, and then it was utterly destroyed. And like, let's try this again, shall we? Reset.

Yeah, just to hit a reset. Imagine that.

Was there life on Thea? Could there have been? Sure. That's a good idea.
Sure, but we'll never know. Yeah, but we'll never know.
Oof. I didn't think of that one.
That's interesting.

And how long ago did all this? Four and a half billion.

Yeah.

Give or take. Give or take.

Long time ago. In the performance.
I'd love to go back. I'd love to go back right after the moon formed to see all the moon.
Oh, and see it 16 times the size of the sun.

Yeah, 15, 16 times. Imagine a moon gargantuan.
Just don't be near the water.

The tides

were as big as a mountain, mountainous tides.

Which actually, I've probably said this on the show four separate times over the years, but it's actually probably a good thing that happened because those immense tides went hundreds of miles onto land and just scoured everything and brought it back into the ocean.

And there's your classic, you know,

primordial soup.

That's probably a good thing.

All right. All right, Evan, tell us about some upcoming holiday scams.
Tis the season. Well, with Christmas coming in, oh, exactly, exactly.

And every year around this time of year, the holiday season, as it were, cybersecurity researchers publish their annual warnings about holiday scams.

And they say that 2025's holiday season is shaping up to be one of the worst yet. Of course, you're going to say that because they'll always say that in the current year.

But in any case, the company Malware Bytes

are famous in a way for many things, but also every year they release a report on how cyber criminals adapt to our habits from year to year.

And basically, it's not so much about the fancy hacking or anything that kind of goes on. It's just about how to take advantage of human weaknesses, you know, kind of the weaknesses of our brains.

They exploit predictable, measurable human behavior. Their report this year

has very common shopping habits that they point out that make people the ideal targets. For example, one of the strongest risk factors is being a last-minute shopper.

Those are the people who wait until the final days for the holidays are nearly

very close. And those people are nearly twice as likely to click on a fraudulent tracking link or fall for a fake delivery problem notification.
You guys have received those, right?

I mean, you know, emails.

Well, as a very last-minute shopper, I didn't know I was in such a high-risk category.

If you are. Because when the merchandise, you know, it's either supposed to be on its way and you're tracking it, there is this urgency or an anxiety spike.

And that's when people's decision-making becomes perhaps less than ideal. And they'll let their guard down sometimes.
And the scammers definitely know this.

And their phishing campaigns coincide with peak shipping windows.

So that's why we see these fake emails and texts claiming, hey, the package can't be delivered, or you need to update your payment information. That's a big one.

And then they're designed to look like, well, the big shippers, you know, Amazon, UPS, FedEx, United States Postal Service, even.

And these imitation pages, they're using what? AI-generated layouts. Yep.
So these things are becoming much harder to just detect upon glance. The wording is becoming neater, crisper, right?

You can't find the typos or the bad

verbiage that's going on in these things. They're cleaning it up and they're using AI to do it.
So yeah, here's another major risk factor. Impulse shopping from social media ads.

So, instant buy traps, they call it.

Look out for things like limited editions and

phrases like that that will get you to, hey, click on this because it's going to be better, easier for you and just more alluring. So, you see a list, you click it, you enter it, and then what?

The store really never existed, or you get some kind of knockoff of what you were going to be buying, kind of a bait and switch in that sense. And FOMO, right? Fear of missing out.

They definitely rely on that. Fear of losing a potential bargain is sometimes more motivating than the desire to avoid a possible scam.
So you have to be able to, you know, kind of check yourself

in those instances. This year, the criminals are also exploiting a behavior that didn't used to be risky, they say, price comparing across multiple tabs or apps.
Jay. Yeah.

I know that from our conversations, you have done price comparing shopping before. Extensively, yeah.
Right. I mean, you know, I thought of you immediately this.

Not that you would fall for it because you're a good skeptic and you have defenses against this. But this is where this is another weakness where they, where the scammers come in.

You know, you're flipping back and forth between what, Amazon, Walmart, Target, TikTok.

Timu, and whatever, looking for the lowest price, right? Nothing wrong with that.

But in that rapid-fire mode, they say that the consumer will become much more likely to mistake a fake storefront for a real one and then, boop, fall right into that trap.

Scammers are also capitalizing on this by cloning legitimate storefronts almost perfectly. Same fonts, same product listings, same color schemes.
It's becoming much more difficult to identify.

The only thing that changes is the URL. But how many people really look at the URL to make sure that they're going where they're supposed to be going?

I have gotten myself into that habit, especially when it comes to things like banking, among other things. You know, I never,

I have trained myself to get in the habit of looking at the URL to make sure and to use the built-in security features that are in a URL.

You know, if you cite information and those kinds of things, they're right there. You know, you just need to just go and click on them, take the extra second to protect yourself.

Tracking, they're talking about package tracking updates. That is where they're really apparently making good inroads because people are getting more tracking notifications more than ever.

And again, given it's the season, that gives people,

you know, just more activity in that area and it becomes an easy target for the scammers to exploit.

So they also make some suggestions as to what you can do, how to prevent this, some easy habits to get into or

easy steps you can take to minimize this kind of damage. First, they say, slow down.

even by two or three seconds, because studies have shown that when subjects deliberately pause before clicking a link, the rate of falling for a scam plummets.

And that micropause interrupts the automatic emotional response scammers depend on.

We have a friend, I won't name her, but we all, Andrea, you don't know her, but the four of us are. I was like, I'm right here, Evan.
Geez. And she, it's not you.
And

I've seen her do this in real time. She will absolutely go nuts ordering stuff while she's on her phone and

definitely not take those pauses.

That's her habit. I've seen her do it many, many times.
So

people,

this does happen. This does happen a lot.
Another tip: don't trust links in the tracking messages. Instead, go directly to the retailer or carry your website through your own bookmarker app.

And if there's really a problem, you'll see it there. They're also suggesting do not buy directly from social media ads, especially the ones that promise

rare items, scarcity,

items, exclusivity, or, you know, big discounts, 70, 80% off electronics. Find it through a standard Google search, not, or on a known retailer site, but not directly from social media ads.

They also give a tip

about

re-entering payment information. on an email notification.
Retailers will not ask for that through email or text.

So if a message that claims your payment failed, you know, if you get that, you have to really log into the retailer's website independently to verify it. Do not ever click the embedded link.

And that's something we've been talking about, I think, for years on all kinds of scams that they use through our electronic devices,

smartphones, and whatnot. So

they're pointing out this year that the biggest vulnerabilities

are not technical so much as they are behavioral, right? They're using our brains against us. And our brains make us an easy target to exploit sometimes in certain situations.

And this happens to be one of them. Yeah, there's just so much noise now,

especially with the SGU. Between Jay and I, we're ordering a lot of stuff for the studio, and sometimes he does it.
And so I'm getting constant notifications,

like two or three notifications per item that's ordered, you know. And so that creates a background noise that it would be hard to detect the scam one thrown in there.

That's why I just don't click anything. I don't click anything in any email ever.
It's just you have to have universal precautions.

Especially if it's a number you don't know or, you know, like, like Evan said, like, usually if it's a tracking update,

the store itself will send it

as well as the, you know, the text. So I just do the store one.
But I just really, you know, I mean, this is something we've probably been talking about with related to scams forever, but people who

are not particularly tech savvy or don't really realize that AI can replicate entire websites, like I just, I could really see, despite the fact, Evan, that you said that this happens, you know, every year claims to be the worst year, I could see these really being a problem.

I'm just thinking of people in my own life who wouldn't know to double-check the URL. Exactly.
Definitely. Yes, yes.
And as always,

the older population is always the most vulnerable of the population when it comes to

technology-related scams, definitely. So we got to watch out for

our friends, our family members, our parents, our grandparents, help keep an eye on them. Always.

Thanks, Evan.

Bob, tell us about hypervelocity white dwarves. Zombie bullet stars in the news.
I'm already with you. Yeah.

I'm already with you.

I heartily endorse this. That's it, yeah.
Yep, that's what I do. Try to suck you in and then.
Yeah, I'm in. Okay, you're in.
Let's see if I can sustain this. That's a good clickbait.
Probably not.

Yeah.

So a recent study and simulation offers what some consider the most compelling answer yet to what would cause cause hyper-velocity white dwarf stars traveling at 2,000 kilometers per second, which is fast enough to leave the galaxy.

What would cause that? And if it verified, if this theory is verified, this would all happen because a heavier companion white dwarf blows up twice.

So this was published recently in the journal Nature Astronomy. Now, we all know about white dwarf stars, right?

Stars around the mass of our Sun eventually will slough off their outer layers, right, leaving behind a massive but Earth-sized core, but with a stellar mass. So, this thing is quite a beefy dude.

And if it's solitary, just hanging out, it's going to cool very slowly over potentially trillions of years.

But then in 2018, the Gaia Space Observatory discovered a handful of these white dwarf stars traveling at insane speeds.

The very fastest that they detected were traveling at 2,000 kilometers per second. That would get from the Earth to the Moon in about three minutes.
That's a what, 239, 249,000 miles.

That's how long it takes me to microwave my burrito.

Right?

Remember, when we went to the moon,

it was like three days. So 4,300 minutes compared to three minutes.
So this thing is booking. Now, some of these white dwarfs, they weren't even just speedy.

They also were very unusually hot and puffy, kind of like these puffed up surfaces. And some of them also had heavy elements, which kind of shouldn't have been there.
So three mysteries here.

What did this? And no theory could explain all three of those unusual attributes until these researchers ran their simulations.

So in the simulation they had what they had they created two binary white dwarfs and these were helium carbon oxygen white dwarfs.

And those are white dwarfs that

can be together in orbit around each other for a very long time and kind of evolving together. So that's why they selected these helium-carbon-oxygen white dwarfs.

So what they are is essentially a helium skin over a carbon core, is a good way to think about them.

The primary, the biggest one in their simulation, was about right around just under 0.7 of a sun mass, so 70% of our sun's mass. The secondary white dwarf was a little bit smaller.

That was just a little bit more than 60% of our sun's mass. So

definitely similar to our sun, a little smaller. So this is what the simulation described

when they ran it. Imagine you've got these two white dwarf stars in orbit around each other,

maybe getting closer and closer. Some of the helium from the surface of the smaller white dwarf is transferred to the larger white dwarf.
So, that's number one. That's the first big step there.

The smaller one is losing mass. This extra helium is building up on

the bigger white dwarf. So, what happens is that causes a supersonic fusion detonation that races around the white dwarf, meeting at the other side.
So the surface essentially explodes.

So this is the first explosion. Then those shock waves, remember those shock waves that met on the other side from where the fusion started,

that converges, those shock waves converge at the core of this larger of, remember, this is the larger white dwarf that's been siphoning off helium from the smaller one.

So that detonation converges on the far side of the white dwarf and

it goes down to the core. And so, what you're essentially having is a deeper carbon detonation in the core of the larger one, and that annihilates the entire white dwarf.

So, that's the second explosion. So, the first explosion is the

helium skin igniting, and the second explosion is the rest of the carbon in the core of the white dwarf exploding the white dwarf. It's basically like no longer there.
Okay, so this is what

the simulation said is happening. So,

this is then the trigger. This is the trigger to turn the remaining core, remember the smaller white dwarf, into this zombie bullet.

It's the trigger. The forces that are created in this explosion of the larger one are on the scale of all the energy the sun releases in its entire existence.

So, we're talking about a tremendous amount of energy.

I have to assume that these white dwarf cores are pretty hardy because, in this simulation, it actually survives.

The smaller dwarf survives the detonation of the nearby larger dwarf star that just blew up. And this is what flings it at these ridiculous velocities, sometimes over 2,000 kilometers a second.

Okay, but as that is happening, now imagine you've got this large white dwarf that explodes, that flings the smaller one away. As it explodes, it's impacting the smaller white dwarf.
So

it's stripping away some of its outer layers, but it's also heating

the surface of the core that's exposed. And so that explains why we're seeing a very hot, puffy outer atmosphere to these hyper-velocity white dwarfs.
That explains that. And

the third, the final mystery seems to be solved because what's happening, you have a lot of fusion taking place on the larger white dwarf.

You've got the outer skin, the helium skin that detonated, you know, in fusion fire, but you also have the interior that also had some fusion taking place. So you've got lots of fusion taking place.

So it makes sense then that the simulation would put this freshly forged heavy elements into the mix of the outer layers of this smaller white dwarf.

And so that explains all three anomalies like no other theory has. This isn't a home run, of course.

This is just a simulation that they ran. But it's the single best explanation for these three big anomalies.

The extreme speeds, the puffed up heated state of these hypervelocity white dwarfs, and the odd compositions that all the other earlier models struggled to fit all of these at once, and none of them did.

But this theory does fit all of these anomalies at once. So I thought that was very, very interesting, pretty cool stuff.

In the future, these researchers are going to use wide-field surveys like the Verisi-Rubin Observatory to help put these theories to the test.

A best case scenario, I'm not even sure how achievable it is, but a best case scenario would be to actually observe this happening in real time, which would be a hell of a coup.

That'd be what, what, three 20s in a row rolled? Andrea? I think that's right, yes.

Okay. But in a sense, this whole talk has been a preamble, I think, to answer the question that a lot of you are thinking.

I hope some of you are thinking about it. What would happen to the Earth if it was hit by such a hyper-velocity white dwarf? So of course, I had to go down that rabbit hole.

So the question itself is actually technically wrong because a billiard ball-like impact would not happen. It would not be like billiard balls hitting.

If a white dwarf hit the earth at this velocity, or even

smaller velocities, just the fact that a white dwarf would be heading towards us, we would probably be spaghettified. But I think let me get a little bit ahead of myself here.

So as the white dwarf enters the solar system, our orbit would drastically change. It would become either more eccentric,

the Earth could be... ejected, we could spiral right towards the white dwarfs.
A lot of different things could happen, lots of different variables going on here.

We would not only have huge ocean tides, but we would also have crustal tides.

The crust of the Earth would form tides like it does right now, but they're super tiny. You can't even notice them.
But it does happen. Well, we would have huge crust tides.

It reminded me of the movie 2012 and what happens. You know, we'd have incredible volcanism, we'd have quakes.
Yeah, it would be a very, very bad day on the Earth as we were approaching.

It was a bad day. Yeah, as we were approaching each other.
And then we would hit the infamous Roche limit.

This is the limit, the distance from a larger celestial object where a smaller object, zone gravity, no longer holds together.

So once we approach the Roche limit of this white dwarf heading towards us, the Earth

could no longer hold itself together. Our gravity would kind of be like, oh, sorry, doing the best I can here, but I'm done.
And the Earth would just kind of start falling apart.

And this is all at about a million kilometers from a typical white dwarf. Depending on the size,

you know it could be a few hundred thousand kilometers it could be a million kilometers but when we were still at a solid distance uh comparable to the to the earth moon distance the earth would just start falling apart couldn't could no longer hold itself together and this is all because of the tidal disruption that's happening right it's all about the tides and tidal disruptions are are can be so powerful and so what's happening is the near side and far side of the earth would be would feel dramatically different gravitational forces and that's that's the basically the essence of tidal disruption.

The far side, because it's farther away, would be feeling significantly less gravity.

And because the gravity is so huge to begin with, it's a dramatic difference, and that would basically tear the Earth apart. And this is where the spaghettification starts happening.

It breaks apart into these glowing chunks, and in short order, all of those glowing chunks would then become plasma and intense radiation, like X-rays and ultraviolet.

The plasma could form a disk around the white dwarf and kind of slowly fall into it, or

it could form a stream directly spiraling right into it. You know, I'm not sure which one would happen, but luckily, luckily, this is ridiculously unlikely.

But it's a macabre and scientifically interesting at the same time, which, of course, appeals to me. It's a fun combination.

It's a lot of stuff slamming into Earth on this episode. Yeah, right? I mean, yeah, but

it's so interesting to think of what would happen. But man, imagine scientists are like, yeah, we see a hyper-velocity white dwarf heading towards the Earth area.

So basically, you've got about, I don't know how long, weeks or months before it's so close that it's just going to rip the Earth apart. So just so have fun

while you can.

Even if it didn't get that close, it probably would fling us out of the solar system or into the Sun or something. Oh, yeah.
If it wasn't just the outskirts of the solar system,

depending on where we were, you know, in our orbit around the Sun and compared to our proximity to the white dwarf.

Yeah, it could just fling us out of the entire solar system or send us right into its maw, like

the Star Trek episode, whatchamacallit, the Doomsday Machine, nasty stuff. So luckily, hopefully we'll never see that.

Star Trek, the original series, by the way. Would we live if Earth got flung out of the solar system? Like if we got flung into the sun, I feel like it's over instantly.

I think it would be pretty instant, right? Yeah.

Yeah, it would be bad. I recommend a short story called A Pale of Air.
Essentially, our atmosphere would rain down as snow

at different times because different gases in the atmosphere have different

freezing points. Then we would become a rogue planet, right, Steve? We'd have a rogue planet and

hopefully,

yeah, we'd probably, if you live near

an underground nuclear facility, you could probably last

the freezing. But yeah, things would be bad on the surface.
Yeah, being a rogue planet sounds a lot more fun than it actually is. Yeah.

I'm free to do what I want. Oh, no.
All right. Thanks, Bob.
Jay, it's who's at noisy time. All right, guys.
Last week I played This Noisy.

Pretty weird, huh? Yep.

Any guesses about what's going on here? I have a guess. Just three people talking.
Isn't it obvious? Go ahead, Eb.

It's George Robb trying out new voices for his Geologic podcast, which is that is so close. It's awesome.

That would not be a strange

correct answer. I like that.
I'll have to talk to George about that.

Andrea, I know you know what this is. Yeah, I know exactly what it is.
It's the background actors from Twin Peaks working out their lines. Oh, I like that.
That's good, too.

Oh, God.

All right, let's dig into this one.

Of course, you know, I love this one because it's so strange. Okay, so we got a listener named Matthew Cutler, and he said, I think this week's noisy is AI slop.
I love this answer.

Specifically, I think it's one of those audio-generating AIs that has been prompted to make up a scene from a comedy TV show.

We live in a world, guys, where that is an excellent answer. It's a solid guess, absolutely answered.
It's a very good answer. Not correct, but much appreciated.

Another listener named Matt Soskins, he said, said it's Vladimir Putin's duck. Yeah.

And yeah, it's a joke. I get it.
But there was a little bit of a step in the right direction there. So we'll keep going here.
Another email from Visto Tutti.

Visto says, I could say a bird like I always do, but the language is definitely Slavic. So the parrot would freeze in the winter.
I am convinced it is an animal talking.

So what animal can talk and survive the snow and ice? A walrus.

Walruses can talk? I think it was a sea lion that I had a recording of where he could mimic his handler.

And it sounded like a human voice, definitely. I don't know.
I mean, I wouldn't be surprised if walruses can vocalize. If somebody has a sample of that, send it to me.
Here's a sample of it, Jay.

Ready?

Okay. Okay.

So I got an incredible number of correct guesses. Now, every once in a while, someone, well, people send me in things all the time that are recent.
You know what I mean?

That have been in a lot of people's news feeds and stuff. This was one of them, but it was so good I had to play it anyway.
So I know a lot of people had recently seen this one.

But the first person who guessed correctly, this is absolutely the first person.

And only a couple of people admitted that they saw it recently, but I do know because it was out recent that a lot of people saw it. But I do believe this person guessed it.
This is Dennis S.

And Dennis says, hello, Jay. I'll skip the whole long time listener, first time caller, Spiel, because I'm too excited.

I actually recognize this one, as I'm sure all of your Russian-speaking listeners did too. It's Karlusha the Raven, who got super famous about two years ago.

I hope I guessed it before thousands of other Russians did. And thank you so much for your work and for the bestest podcast ever.

And that is Dennis Segat. Dennis, thanks so much for that.
Yeah, I did get a ton of emails from people that can speak Russian. I'm sure a lot of them are in Russia.

Yeah, this is exactly what Dennis said.

This is a raven that apparently was raised by people in Russia, and they taught it a bunch of different words and everything, and they're kind of having a conversation with it, and it's definitely entertaining them.

But it's really cool. You should look up the video to watch this because seeing a raven talk is pretty weird.

Yeah, yeah. And ravens are very different.

Yeah, did the raven seem unusually large? It's a big raven, yes. I mean, it's not.
Ravens are large. Within the normal range, but it's not within the normal range.

Yes, because I wasn't sure if it was somewhere. Ravens are bigger than crows.
They're very big birds. So I think it's more compelling because it's in a foreign language.
Yeah. Right?

We wouldn't pick up on the nuances.

Yeah,

it sounds like perfectly good Russian to us, so it sounds even more uncanny. But to a native speaker, they could tell that something was a little off.
Yeah, I'm sure. Yeah, absolutely, Steve.
I agree.

I thought of that in my own as well. And listen to it again.
It's the male voice.

It even interrupts, like the flow is very human-like.

Definitely just kind of sounds like a kind of grumpy old guy. It totally does.

I don't know if there's much difference between a drunk Russian and a raven.

You know what I mean?

I'm just kidding. Come on.
on.

All right, guys. I have a new noisy this week.
This was sent in by a listener named Aaron. Oh, and I'll warn you that this has some very high-pitched noises in it that may bother some listeners.

So this is your chance to turn down a little bit. And here we go.

That is Han Solo trying to get the Millennium

in episode five.

It goes on a lot longer than that. I mean, I thought extended as hell, man.
That is a lot. I thought it had such a cool series of noises, all sorts of different things going on.

So, anyway, anyway, calm down. I know you're excited.

But if you think you know this week's noisy or you heard something cool, you got to email me at WTN at the skepticsguide.org. Steve, as we speak,

the tickets for the Seattle show and the Wisconsin show, or shows, I should say, are up. Now, we have

three shows in both of these venues, right? So, Seattle and then Wisconsin. In Seattle, we have three shows.
Friday night, we're going to have a very low number, very high-profile meet-and-greet.

It's going to be a small number of people hanging out with the SGU. I think we're going to have like 20 tickets for it.

We're trying this out because we've gotten a ton of emails from people that just wanted something exactly like this.

So, we thought we'd give it a shot for a couple of different show weekends that we're doing. So, if you're interested, go to the website theskepticsguy.org.
You can see this Friday night show.

I have to formally name it. I came up with something fun a few weeks ago.
I can't remember. I'll look it up.
But anyway, it's the Friday night hangout. Then we have Saturday

starting sometime between 11 and 12 a.m., we're going to be doing an SGU Private Show Plus. That's a three-hour show.
This includes George Robb, of course. And it's a live recording of the SGU.

And then we have fun with the audience for about an hour. It's different every time.
You got to come check it out to know what it's all about.

And then then Saturday night, which is that night, we're going to have a VIP,

which is available if you're interested in buying tickets. This is for the extravaganza.
And then there's the extravaganza itself.

So honestly, there's four different things that we're doing in those two days. The extravaganza starts at 8 p.m.
All the details are on the ticketing sites, which are links are found on SGU's website.

Please come. We'd love to see you.
I'm getting tons of emails from people saying they missed it last time and are coming this time. We have a great series of shows for you guys, so please join us.

And then again, repeat everything in Wisconsin. The dates are up there.
And then, as a future mentioned, we're going to be doing all of this in New Haven at some point.

Hopefully, maybe March or April. We'll let you know when details come.
All right. Thank you, brother.
Just one quick correction. On last week's show, I said that cellulose was a protein.

I blame this the jet lag on this. It was just a brain farm.
I was thinking collagen.

I was like, in my mind, I was thinking collagen, collagen, which is a different, that's a more for you know animal structural protein. Cellulose is a polysaccharide, right?

It's a ribbon-shaped polymer of glucose molecules. It is

the most common, I think it's the most common structural molecule in plants, whereas collagen is in animicules. Sounds right.
So I just got the wires crossed.

I'm still massively jet-lagged, by the way. Still, right? You guys can tell through clever editing.
I kind of hide it as much as possible.

You'd think there'd be a fix for that, you know, or a routine, something they've figured out people can do, a hack of some kind that will help people. But I think

it's time.

Yeah.

And time. You know, melatonin may help a little bit, but

it's not so much my when I'm sleeping. I'm sleeping at night.
It's just that I just have

can't consolidate my sleep. I can't get enough sleep at once.
I'm waking up at 2 in the morning, you know. Thinking of things, I'm sure.
Right.

And then now I'm also, I have a colonoscopy tomorrow, so I'm prepping for that. Oh, wow.
And prepping means you're drinking bowel, you know, you're basically drinking motor oil.

That begins right after I get off the show. So, Steve, you went with the go use Bob's toilet.
Yeah.

I'll be over there, Bob. Right for the show.
I'm coming over there. Yeah, right.

Got my bowel prep and my iPad.

Are you going with? Which prep did you go with? The small liquid. All right.
Not the big liquid. There's a big liquid, there's a small liquid, and there's a pills.

This is a once-a-year thing?

Typically, once every five years. It's ten years if it's good, five years if they want to follow stuff.

For me, it was three years. Yeah, it depends on what they found the last time.
They say, that's a little suspicious, come back in three years, whatever.

But if it's totally clean, I think you can go 10 years between.

Bob rolled a one. Bob rolled a two.
Yeah. No, one would have been

real. One more than

worse.

It turns out the smaller liquid didn't work, didn't clean me up as much.

You got to chase it with a ton of clear liquids. That's the thing.
You can't just drink that and think you're done. You have to.

Anyway,

that's the night I have in store for me.

But in the meantime,

let's go on with science or fiction.

It's time for science or fiction.

Each week I come up with three science field items or facts, two real and one fake, and then I challenge my panel of skeptics to tell me which one is the fake. We have a theme this week.

The theme is scientific fraud. Scientific fraud.
Okay, you guys ready for this? Yes, here we go.

Item number one, the famous experiment in subliminal advertising in a movie theater to increase sales of Coke and popcorn never happened and was entirely fabricated.

Item number two, a recent analysis finds the number of fake publications in biomedicine was at least 5.8%,

with 15% being suspicious, amounting to over 100,000 fake papers published every year.

And item number three, researchers find that 40% of published peer-reviewed papers show signs of AI co-authorship, with 10% being fully authored by AI.

Jay, go first. Okay, the first one here: the famous experiment in subliminal advertising in a movie theater to increase sales of Coke and popcorn never happened and was entirely fabricated.
I mean,

that seems so likely that that sentence is true. You know, it's an urban legend.
I could see someone making that up. I could see it both ways, but I think that one is science.

The second one here is a recent analysis finds. The number of fake publications in biomedicine was at least 5.8%, with 15% being suspicious, amounting to over 100,000 fake papers published every year.

Okay, so this says it's a recent analysis,

and I mean, it's hard to know what the numbers would be,

but I absolutely believe that there are an increasing number of fake papers going out. You know, biomedicine definitely is a category I would expect a ton to be in.

You know, over 100,000 fake papers published. If anything, I would say that number is a lot more if this one isn't correct.

The last one here: researchers find that 40% of published peer-reviewed papers show signs of AI co-authorship, with 10% being fully authored by AI. Oh my God.

That one has got to be science. Oh man.

Wow, Steve. I will say that the last one is the one about the researchers that find that 40% of published peer-reviewed papers show signs of AI co-authorship.
I'll say that one is false. Okay, Evan.

The one about advertising in a movie theater, increasing sales of Coke and popcorn, that is classic urban legend, like urban myth.

kind of stuff right there and that it was entirely fabricated i i believe that is science Boy, I bet you there are even other examples of things like this.

We know the culture just gloms on to things, whether it's true or not. And if it has legs long enough, lasting throughout a generation or decades or whatever.
Yeah, stuff like that.

This would be a classic case of that, I think. So that one is science.

The second one about the number of fake publications in biomedicine was at least 5.8%, with 15% being suspicious, amounting to over 100,000 fake papers published every year. Holy moly.

So there's more than a million biomedicine papers published every year. So that 100,000 is based on the 5.8%.

So that's the lowest. Oh, that's the 5.8%.
That's the lowest.

Millions of biomedicine. Oh, my gosh.
Well, I guess, you know, it's happening all over the world, so then publishing all, but that's a lot. All that information.
Oh, my gosh.

It's this last one, though, that I think is going to wind up being the fiction. 40% of published peer-reviewed papers papers showing signs of AI co-authorship

and 10% being fully authored. How could you have a peer-reviewed process that would allow for that? That is just, why have it at all?

I mean, right, if you're not gatekeeping for things like that in 2025, what are you doing?

So I imagine they are really doing everything within their power to

stop this or detect it as best as possible. And I don't think 40% of this stuff is getting through.
I say that one's fiction. Okay, Bob.
I agree with you guys. I think you're pretty much spot on to

what I'm thinking as well. Subliminal, yeah, I've heard about it for literally decades.
It wouldn't surprise me that it's fake. It also wouldn't surprise me if Steve, you know, if you're whatever.

I'm trying not to medigame too much this time. I think the bottom line for me is that

the biomedicine seems reasonable.

5.8%, 15% sounds reasonable.

For this third one with peer-reviewed, 40% is just, I don't want to believe it. That would just,

it's such a dramatic number.

10% fully authored, even one in 10 sounds a little bit too dramatic at this point in 2025.

You know, geez, I hope I'm right here. So I'm going to say this one, the 40% published peer-reviewed,

showing signs of co-authorship. I'll say that one is fiction, and that puts me with Jay at this point, I think, right? And that's been waiting.
I know you guys were all full of them. Yep, we're in

the same boat. All right.
Okay, and Andrea, you get to go last. Yeah, so I'm going to go against the group here

in the spirit of collective intelligence. Someone's got to be wrong, so it might as well be me.
Touched.

So I'm going to go with, so I think the

5%, 5.8%

biomedicine publications being fake, I think that's science. If anything, I agree with others who said

it's probably higher, you know, and 15% are suspicious. I'm going to say that the researchers find that 40% of published peer-reviewed papers show signs of AI co-authorship.

I'm curious about what timeframe is, is that 2025 thus far? Is that the last time?

This is all very recent. Just recent.
Okay. I'm going to say that that is science because

I have

less confidence in the peer review process, perhaps.

Having been a part of it myself, you know, it's truly often a peer. And if you're working, you know, AI, the 40% is just AI co-authorship.

So that could mean very small segments are AI and faculty who are, they should be reviewing for these things, but I don't believe that they are necessarily. And I also, this is across all fields.

I bet there's some fields just pumping out some wild stuff.

So I'm persuaded that that number is real, which leaves very controversial, and I'm a bit torn on this, but I'm going to say that the experiment about popcorn and Coke being fabricated,

I'm going to say that that is the fiction. And I'm largely basing that on one thing, which is I've never heard of this experiment.

And so I'm not really like, oh, of course it's false because I don't know what the study actually was. I could see a version of it.

It does seem unlikely that we're in a world of subliminal advertising and that's not fiction. But I'm going to say that maybe the claim itself is that it never happened and was entirely fabricated.

I bet some version of it happened and it got way blown out of proportion. So I'm going to say that's the fiction.
Okay, so you guys all agree on number two, so we'll start there.

A recent analysis finds the number of fake publications of biomedicine was at least 5.8%, with 15% being suspicious, amounting to over 100,000 fake papers published every year.

You guys all think this one is science, and this one is

science. Great.

Yeah. I mean, or not great, but a lot of people are.
Oh, no, this is a huge problem. And this number is growing fast.

So,

one thing you have to change about how you may be thinking about

fake scientific papers, it's not just individual bad actors anymore.

There are actually now

systems of people, basically like organized crime,

syndicates,

their paper mills, cranking these things out.

They hook up researchers who are trying to buy, you know, to pad out their CV buy reputation with journals that will publish the article, with people who will write the fake papers.

And the numbers are increasing significantly. And the journals don't have a sufficient mechanism to

really prevent this.

Think about that.

One in 20 papers, at least, it could be

more like three in 20,

are fabricated or just are completely fake. You know, not that they tweak the numbers or something.

It's a paper mill. Part of the reason for this, too,

is that there are so many pay-to-play journals that are of very low quality, which also impacts the third one, too.

You guys got to keep in mind how many low-quality

they may be technically peer-reviewed, but it doesn't mean that there's somebody doing a good job there. They're just trying to publish as much as they can because they get paid per paper.

Some are straight up predatory, but other ones just have really low standards. Trevor Burrus, Jr.: Well, their industry is going to collapse if that's the case.
I mean, right? Oh, totally.

People lose faith in the published science, and if it's just you're buried with fake papers,

it's not sustainable.

Well, and the cost to generate papers is basically zero on an individual level, so you can just keep pumping all like that was the big barrier, was like, oh, at least a paper takes a while to write, but that's not the case anymore.

Right. Well, let's go to the number three then, since these are closely related.

Researchers find that 40% of published peer-reviewed papers show signs of AI co-authorship, with 10% being fully authored by AI. The boys think this one is a fiction.

Andrea, you think this one is science.

And again, it's the same kind of thing. It's like, you know, are those numbers reasonable? If you're thinking that there's a ton of poor quality

journals out there, this is another way to just pad out your CV. Just have AI write a paper, publish it in some pay-to-play rag, and there you go.

I hope nobody looks too close. I hope so.
Yeah, and if they do, you retract it and move on, which is what they do.

So, this one is

this is the fiction. You guys are correct.
You are right.

The numbers are.

The numbers are. Andrea, nice try.
I mean, I could absolutely agree with your angle as well, Andrea. Well, we may get that.
Well, it was the wrong angle, so don't agree too hard.

So, 10% are co-authored.

That's the upper. There's no

number for

fully authored by AI.

They found that recent studies show that 10% of published peer-reviewed papers show signs of being co-authored by AI.

So this is greatly exaggerating those numbers. Does AI have a role in

papers? Yeah.

Right. You could use it.

But not in the terms of public authority. But it shouldn't be authoring it, no.

Editing, analyzing, whatever. This means that the famous experiment in subliminal advertising in a movie theater to increase sales of Coke and popcorn never happened and was entirely fabricated.

Is science. So, yeah, this wasn't really an urban legend so much as this, an ad man, James Vickery, made it up and said it was real and sold it.

But then he later said it was all just a quote-unquote publicity stunt, which means he lied and got caught, right?

But the thing is that it kicked off a whole generation of people believing in subliminal advertising.

I've heard this since I was a boy. I've heard that many times, yeah.

And that's where the urban legends then take over, right, and then embellish it.

But the science never showed that it was true. And then eventually, people did do actual research on it, and it just was a complete bust.
It doesn't work. Yeah.

I was going to say, is there any evidence I would be shocked? Yeah. There is priming, right?

You can prime people, and you can prime people very subtly, but the idea of subliminal is that it's imperceptible consciously, but it's still affecting you. And there's no evidence for that.
Right.

One of my favorite priming studies, stop me if I've already said it, was a study where they had people come into little cubicles in an office and like draw a picture or solve a crossword puzzle and then they gave them like a little cookie or a snack.

And the treatment and control, they told them the treatment and control were something with like the puzzles, but the treatment and control were whether or not they were pumping a lemon scent into the room where they were doing the study.

And the rooms that got a lemon scent, the people were more likely to pick up the wrapper from their little snack and throw it away than the ones that didn't have a lemon scent.

Oh, because they wanted to maintain the perceived cleanliness.

There's like this, yeah, something about the smell like primes us to think about cleanliness or order or, oh, there's someone in here cleaning up. I don't know exactly what the mechanism is.

Is that why they lemon scent all those cleaners?

Right?

Yeah, because that's what I think about when I go, you know. Or that's why everyone did it is because

we had a religion, so we now associate it.

Yeah, but um, I mean, yeah, that's not subliminal, but it's pretty clever, and it's and you are perceiving it, you just don't really know the effect it's having on you.

I just thought it was super cool. Fascinating.
All right, Evan, give us a quote. This week's quote was submitted by a listener, Terry, from American Canyon, California.
Thank you, Terry.

It is easy to get international agreement in science. Scientists have all the same standards.
They are set not by beliefs, but by what works best.

Of necessity, there is therefore universal unity, and unity makes for goodwill.

And that was said by Bernd Heinrich, who is a professor emeritus of the biology department at the University of Vermont and an author of a number of books about nature and biology.

He has also made major contributions to the study of insect physiology and behavior, as well as bird behavior, including talking ravens. Thanks.
All right. Well, Andrea, thanks for joining us.

Of course. Thanks for having me.
Always. I love having you, Andrea.
You're always awesome, Andrea. This is so great.
Oh, you're always awesome. This was super fun.

I'm always happy to be here. And I'll be seeing you next week.
We'll be recording the next two episodes of the Political Reality Podcast.

And one day people besides us can listen to you. They'll actually be out there in the week.
We're getting ahead. It's more of a personal podcast.

It's a backlog while we do the, you know, we're having discussions with the editor and we're sorting all that stuff out. But once they hit, they'll be every week.
Yep. Yep.

That's very exciting and it's been a lot of fun. So I'll see you then.
And then you're committed. Then we're committed.
That's it. That's it.

Andrea, we haven't missed an episode in 20 effing years. So

your dance card is going to be filled for a long time. Wow.
Not a single episode. Not a single week in 20 years.
No. I'm impressed.

And we're not stressed out at all. Yeah, and it doesn't.

It's not taking a toll on anyone's health. But now it's like we don't want to break our record, so we've got to do it now.
Yeah. If you ever miss a week, we'll all know that everyone dies.

Yeah, something bad happens. No, or whatever.

If we miss a week, at least two of us are dead.

Yeah.

Or we're about to get launched out of the solar system as a planet. Exactly.

That's my prediction. Yeah.
Steve, if we get approached by a hyper-velocity white dwarf, I am not recording this show. Noted.

This is the ultimate show to record, Bob. You got to do at least a short one.

Quickie with Bob. I'm going to record.
Quickie with Bob. Yeah.
I think more quick than that. All right.
Well, thanks again for joining me this week, everyone.

Bye. Happy Thanksgiving to all.

Happy Thanksgiving. Thank you.
Bye-bye. And until next week, this is your Skeptics Guide to the Universe.

Skeptics Guide to the Universe is produced by SGU Productions, dedicated to promoting science and critical thinking. For more information, visit us at the skepticsguide.org.

Send your questions to info at the skepticsguide.org.

And if you would like to support the show and all the work that we do, go to patreon.com/slash skepticsguide and consider becoming a patron and becoming part of the SGU community.

Our listeners and supporters are what make SGU possible.