The Skeptics Guide #1113 - Nov 7 2026

Interview with Brian Wecht: The Mathpocalypse; News Items: Does AI Have Feelings, CDC Hep B Vaccine Guidance, Potassium Enriched Salt for High Blood Pressure; Who's That Noisy; Science or Fiction

Speaker names marked ? are inferred from speaking time and the show’s known hosts — they are not voice-verified.

• 0:02Speaker 00

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

• 0:09Speaker 01

  Hello and welcome to The Skeptic's Guide to the Universe. Today is Thursday, November 5th, 2026, and this is your host, Steven Novella. Joining me this week are Bob Novella.

• 0:18Bob Novella?

  Hey, everybody.

• 0:19Speaker 01

  Cara Santa Maria.

• 0:20Cara Santa Maria?

  Howdy.

• 0:21Speaker 01

  And Jay Novella.

• 0:22Jay Novella?

  Hey, guys.

• 0:23Speaker 01

  So.

• 0:24Jay Novella?

  Yeah, many things, Steve.

• 0:25Speaker 01

  What's happening in the world?

• 0:26Cara Santa Maria?

  Oh, yeah. Nothing new this week. We're not giddy or anything.

• 0:29Speaker 01

  Yeah, obviously, the midterms happened and pretty surprising outcome. And, you know, everyone was expecting a quote unquote "blue wave." Typically, that happens in the midterms. The out party does better. And, you know, Trump is unpopular, the gas prices are high, all the indicators are pointing towards the Democrats doing well. And I followed it pretty closely, and they did better than anyone I saw predicting the outcome in the last few months, you know?

• 0:55Cara Santa Maria?

  I mean, upsets that we haven't seen in like, a century.

• 0:58Speaker 01

  A century.

• 0:59Cara Santa Maria?

  In certain places.

• 0:59Bob Novella?

  Say that again, a century.

• 1:01Cara Santa Maria?

  A century, yeah.

• 1:02Speaker 01

  What?

• 1:02Cara Santa Maria?

  And also, just like firsts across the board in terms of governorships, in terms of, you know, basically, executive leadership, ethnic minorities, women, you know, just really great demographic shifts that we're seeing, where there's just more representation across the board.

• 1:20Jay Novella?

  You know, the thing that blows my mind is, two years ago, we had a completely different kind of and flavored election that didn't represent any of the things that we just saw happen a couple of days ago.

• 1:32Speaker 01

  That's a great question, assuming that's the question, is how is it possible that this is the same electorate that voted for Obama, and then Trump, and then Biden?

• 1:41Cara Santa Maria?

  But it's not.

• 1:42Speaker 01

  And then Trump again?

• 1:43Bob Novella?

  Perhaps they didn't think that there would be such a level of incompetence and corruption and, and other things. I mean, I mean, a lot of people saw it, but a lot of people didn't. And I will never understand how they didn't see something like this coming.

• 1:57Cara Santa Maria?

  Yeah, but I mean, most people, it's, you know, I have a lot of friends from outside of the country, and who are like, have a pretty harsh view, where they're like, "America, you need to get it together. America, you guys voted." And I'm like, to be clear, like, most Americans voted for Trump. And I was like, "To be clear, most Americans voted for no one."

• 2:15Speaker 01

  Most Americans didn't vote.

• 2:17Cara Santa Maria?

  Yes.

• 2:17Speaker 01

  And of those who voted, Trump never got a majority.

• 2:20Cara Santa Maria?

  No, he did get a majority.

• 2:21Speaker 01

  He lost the popular vote twice, and then got a plurality in 2024, but he never cracked 50%. So.

• 2:26Cara Santa Maria?

  Oh, yeah, yeah, no, yeah.

• 2:27Speaker 01

  It's never true that Trump won a majority of even the voters, let alone.

• 2:31Bob Novella?

  Right.

• 2:31Speaker 01

  Of Americans.

• 2:32Cara Santa Maria?

  Right.

• 2:33Bob Novella?

  Yeah, wasn't it like a third of the electorate that he got? I mean.

• 2:35Speaker 01

  Yeah, something like that.

• 2:36Cara Santa Maria?

  He got decent turnout, considering, but it's still, decent turnout in American politics is still bad. But we got better turnout for the midterms than like, a really long time, right?

• 2:46Bob Novella?

  This is ridiculous.

• 2:47Speaker 01

  Yeah, yeah.

• 2:48Cara Santa Maria?

  These were looking like, like presidential election numbers.

• 2:52Speaker 01

  Mm-hmm. But, but the point I was getting at, a lot of analysts think that what we are experiencing is a general anti-incumbent feeling. And this is true across the world, not just the United States, but it may be particularly potent in the US, where every election, we're just voting for, against whoever's in power. Because people are just unhappy generally. They think that government is dysfunctional generally. It could be that people are all wound up because of social media. It's hard to say if that's like, how big of a role that's playing. But, but there is this pretty consistent now, anti-incumbent factor in every election.

• 3:34Bob Novella?

  Since 2015, since 2015.

• 3:35Cara Santa Maria?

  Well, and the anti-incumbent factor, I mean, I think that's, yeah, a top line kind of description, but really, that is only evident, or kind of central, when people are really, really disillusioned with how things are going.

• 3:52Speaker 01

  Yeah.

• 3:53Cara Santa Maria?

  And it's been a long time since we've been in that place, where America was for a really long time, where we were like, "Let's not change things, because things seem to be working," you know? We just haven't had that in a while.

• 4:04Speaker 01

  Well, generally, America has low voter turnout for what some people argue is a good reason, in that the stakes are generally low. Like, our fundamental rights are not up every four years. But now, it feels like they are. And so, we're having these record turnouts, but that reflects the fact that the elections are way higher stakes than they really should be. You know what I mean? It shouldn't be that dramatic a shift in federal policy, just based upon who ekes over.

• 4:36Cara Santa Maria?

  Oh, of course.

• 4:37Speaker 01

  The electoral majority, you know what I mean? It's like this winner-take-all kind of system that we're dealing with is creating these wild swings, which is, I get it, it's dizzying for us, let alone the rest of the world.

• 4:48Cara Santa Maria?

  Well, and I would.

• 4:49Speaker 01

  It is a problem.

• 4:50Cara Santa Maria?

  I would argue that the position you're, you're highlighting, while, while I agree with you, is the position of the privileged majority. But for a lot of people in this country, it does deeply affect them. The stakes are very, very high, and they always have been. It's just that the white, kind of powerful, majority in this country is finally feeling the experience of a lot of minority voters, and knowing what it feels like for their rights to also be threatened.

• 5:19Speaker 01

  Yeah.

• 5:19Cara Santa Maria?

  Yeah.

• 5:20Jay Novella?

  But the shift was so profound. I mean, yeah, my gut is telling me that so many people weren't paying attention, or were completely bamboozled by all the misinformation.

• 5:31Cara Santa Maria?

  Yeah, of course.

• 5:32Speaker 01

  Yeah.

• 5:32Jay Novella?

  That it had that powerful of an effect on, on what the, you know, what the end result was. I mean, look, it was, it was a close election anyway, but the bottom line is, like, this was such a major shift.

• 5:43Bob Novella?

  2024, you mean?

• 5:44Jay Novella?

  What?

• 5:45Cara Santa Maria?

  Yeah, like, look at all the immigrant voters who were, who were duped into voting for Trump, who then got, like, deported, who then got, like, you know, targeted by ICE. Like, like, we're not just talking about, "Hey, he said he was gonna keep his, you know, do this for me," but like, then healthcare got more expensive. It's like, "He said that I was gonna be part of the group, and then he told me that I'm no longer welcome in my own country."

• 6:09Speaker 01

  Yeah. They should have seen that coming. I mean, I think, I think it's something that a lot of people have, but.

• 6:15Cara Santa Maria?

  I don't know. I think that it is very easy to say when we come from an area where we have high literacy, high education, deep understanding of history and of authoritarianism. But when somebody says the things you want to hear, and you're desperate to hear them, I don't know how much we should assume that people have that critical capability.

• 6:34Speaker 01

  Oh, I get that. I'm not saying it's, it's necessarily their personal failure. I think it's just, I mean, when it's so many people, you have to look at systemic factors.

• 6:41Cara Santa Maria?

  Of course.

• 6:41Speaker 01

  I think it's largely a failure of our media, to be honest with you.

• 6:45Cara Santa Maria?

  Oh, yeah.

• 6:46Speaker 01

  And it's also because we don't know how to deal, I mean, collectively, as a society, we haven't really wrapped our head around how to manage the effect of social media on our democracy. And so, we're sort of, you know, we have, we have, you know, I think a, a media that is somewhat dysfunctional, and not, at the same time, not, we have the, you know, the, the impact of social media, which we haven't wrapped our head around yet. You know, and we'll see what happens. Now, we have AI thrown into the mix, so, as another variable in all of this. Maybe there's just, things are changing too fast for us to culturally catch up with it, you know?

• 7:22Jay Novella?

  I was really surprised that, you know, Trump and his administration didn't do the full-court press, taking everybody, you know, litigating everything and, and going down that route. I think largely because the numbers were too solid.

• 7:36Speaker 01

  Yeah.

• 7:36Jay Novella?

  To contest. But the, the thing that is just so unbelievably pathetic is that they're not taking accountability at all. Like, you know, Trump's excuse is, "It was the war." You know, he completely removed himself from having anything to do with the bad vote. It was the war and gas prices that did it, which.

• 7:55Speaker 01

  My favorite part about this election, I think, is the fact that so many people said that corruption factored highly in their vote.

• 8:05Cara Santa Maria?

  Yeah.

• 8:05Speaker 01

  Because that is, it usually doesn't.

• 8:07Bob Novella?

  90% is.

• 8:08Speaker 01

  Yeah.

• 8:08Bob Novella?

  When have the voting population been that in sync? Like, never? 90% is just a ridiculous number. And where did it come from? Nobody was saying that everyone's concerned about corruption. I mean, corruption's been in the news, but 90%, I think, yeah, surprised me.

• 8:24Speaker 01

  Usually, people cynically ignore it, because they think, well, it's like, there's corruption in politics, pretty much everybody to this day and to this hour.

• 8:30Cara Santa Maria?

  Yes, the cost of doing business. It's like how companies just calculate shrink into their, you know, bottom line.

• 8:37Bob Novella?

  Right. And that shows you how blatant it is.

• 8:40Cara Santa Maria?

  Mm-hmm.

• 8:41Speaker 01

  But it's good that there's a line. I'm just happy that there's, whatever, that there's some kind of acceptable floor to what we find acceptable in our government, in our politicians.

• 8:49Cara Santa Maria?

  Oh, yeah.

• 8:50Speaker 01

  I hope, I hope that issue stays front and center going forward. You know, I mean, I hope.

• 8:55Bob Novella?

  Right.

• 8:55Speaker 01

  Issues like that, like core issues, like, you know, is this person acting in the, in the interest of the public? Is, do they put patriotism above partisanship and their own career? You know, are they corrupt? And do they respect democracy? Like, those kind of core values, which we, I've taken for granted for so much of my adult life, and now it's like, we can't take them for granted. It's nice to see that they're factoring back in, but we gotta keep pushing in that direction, in my opinion. Regardless of your partisanship.

• 9:24Cara Santa Maria?

  Oh, yeah. It shouldn't have anything to do with that.

• 9:27Speaker 01

  No, it really doesn't.

• 9:28Cara Santa Maria?

  And it never historically did.

• 9:30Speaker 01

  Yeah. So, quick update on my bear situation. This just happened this morning, right?

• 9:37Jay Novella?

  Segue.

• 9:38Speaker 01

  Steve, you have a bear situation.

• 9:40Cara Santa Maria?

  I love your segues. Yeah, they're so good.

• 9:42Speaker 01

  So, there's, there's a bear that, that dens near my house. That's just, that's just the way it is. And I see him every now and then. We did have a bear encounter. So, this morning, I was just in my bathroom, you know, and the window overlooks my backyard, and my wife was up before me, and she had just, she was on the deck, and she had just let our dog out, Sagan. And he goes tearing for the woods behind our house, and he goes right up to the invisible fence line. We couldn't quite see, he started barking like crazy. We couldn't quite see what he was barking at. There was just like this black shape there. But then it was pretty clear that it started to move, that it was a bear. So he's standing literally toe to toe with a black bear, barking his head off.

• 10:27Speaker 01

  And of course, we're very concerned because we have no idea how this bear is going to react. How big is it? It was huge. This bear is, is pre hibernation fat, right? A big, the biggest, they're definitely lean when you see them in the spring and they're also very hungry. This bear just backed off. We were, you know, had to vociferously call Sagan in to, you know, really pull out all the stops to get him to come back in. And I watched the bear, he just sort of loped through the woods, slowly lumbering along. I mean, he was a slow fat bear. This guy's ready to lay down for hibernation. So thank goodness, because you know, black bears tend to be skittish.

• 11:08Speaker 01

  I mean, they, you know, there's no reason why, they have nothing to gain going up against, you know, a wolf, basically.

• 11:16Bob Novella?

  Yeah, why get, why get injured?

• 11:18Speaker 01

  Right, right. All right, but here's, I would say there was no food at stake and there was no cups. But again, this encounter happened right under my apple tree, which is right, which is where the coyotes have been gathering at night. And

• 11:33Bob Novella?

  You know, Steve, that tree might be a portal to like the animal plane.

• 11:37Speaker 01

  Either that or, or the, the hundreds of apples on the ground were, were attracting animals.

• 11:42Cara Santa Maria?

  Oh yeah, that.

• 11:43Speaker 01

  Though this year, this year, for some reason, there was a massive bumper crop in every fruit, fruit bearing tree that I have. And other people have reported similar things. So just where we are, for whatever reason, the conditions were such that it's like twice as many, you know.

• 11:57Bob Novella?

  Damn, man.

• 11:58Speaker 01

  And I picked as many as I could. It didn't even make a dent, you know? I actually was canning apples for the first time and making applesauce, whatever. But I think I have to more fastidiously clean up under that tree. So because it does seem to be attracting predators. So that, that might be on me. But I, yeah, this is the first time I, my dog has had a toe to toe encounter with a black bear. So I'm happy to see the black bear backed off. But still, it was very anxiety provoking.

• 12:30Bob Novella?

  That happens in the spring.

• 12:32Speaker 01

  Might come out differently, yeah.

• 12:34Bob Novella?

  Damn.

• 12:34Speaker 01

  Totally. All right. Brian Wecht will be joining us later in the show to talk about some AI math shenanigans. But before that, Bob, you're going to start us off with a quickie.

• 12:46Bob Novella?

  Thank you, Steve. This is your quickie with Bob. Two words, ultrasonic espresso.

• 12:52Cara Santa Maria?

  What?

• 12:52Bob Novella?

  Jay, did you hear that? Developed, developed by engineers and food scientists in Sydney. Jay may think this is coffee heresy. All right, so we got traditional espresso, right? You've got hot water and pressure working together to extract caffeine and oils and acids and the flavor compounds from the ground coffee, right? That's how it works.

• 13:11Speaker 01

  So far, your, your science is correct.

• 13:13Bob Novella?

  Okay. Ultrasonic espresso uses high frequency sound waves instead of the hot water to accelerate this extraction at room temperature. Okay. Now, a little bit more detail on that. The sound waves that are created, they create millions of these microscopic bubbles in the water and they violently implode. So we got basically ultrasonic cavitation happening here. And these tiny implosions create the shock waves and the liquid jets that disrupt the, the coffee grounds, exposing more surface area, right? The surface area is critical, right, Jay? And this is what accelerates the release of the caffeine and the oils and the flavor compounds into the water, not the heat. It's this, this mechanical energy. And this runs at room temperature and uses sound, sound waves.

• 13:56Bob Novella?

  And this entire process takes two to three minutes. And it's all, as I said, all in room temperature. A quick aside here, the collapsing bubbles themselves can briefly produce really high local temperatures and pressures, right? Even though the coffee as a whole will, will remain near room temperature. So that's just a little quick aside there. That's interesting. All right, so one immediate question, Jay, you're probably thinking this as well. Without heat to extract the flavors, can those amazing flavors that all reasonable people on earth are addicted to, can they still be produced, right? How good does it taste? Well,

• 14:30Speaker 01

  Well, yeah, more importantly, is the coffee not going to come out hot?

• 14:33Bob Novella?

  No, no, it's, the flavors are coming out without the heat. That's, that's the key. And then let me finish this point. So they tested 100 people and they could not distinguish between conventional espresso and the acoustic espresso. Of course, it was heated to drinking temperature, of course. So Jay, this is, this probably won't be a home coffee thing. This, this device, this idea, this device is going to shine in industrial settings at scale, but especially ready to drink coffee production, right? That's, that's where this is really going to shine. And that's because of the central concern at this scale. And it is energy consumption. So that's why this is, this is really optimized for, for this, the scenario where you're making coffee that's not meant, that's not hot.

• 15:19Bob Novella?

  It's just meant to be put at the store and it's just, it's just room temperature or cold or whatever it is. So when you're, so things like bottled coffee, right? Cold coffee beverages, coffee concentrates, they don't, they don't necessarily need to be heated at all. So this is where that shines. And this is where I think when these scientists are saying that it could be 75% more energy efficient, it kind of makes some more sense. Although I did not see any information yet, how much energy the ultrasonic system uses, right? So for an ultrasonic machine, making hot espresso at home or a coffee shop, you still need to heat the finished espresso, right? If you're going to consume it, because you're going to, you're going to end up with a cold, a basically a room temperature or whatever coffee. And you're going to, you're going to heat that up.

• 16:01Bob Novella?

  But there's going to be some energy savings there because you don't, because there's less waste. When you've got an espresso machine, machine, Jay, right? There's a reservoir that's, that needs to be, that needs to be at temperature. Parts of the machine heat up, so there's a lot of waste there. So you will save some energy at home, but it's really not meant for home. So this is, it's really interesting. I would, I would love to try this and I can't wait to get some feedback from you, Jay. But this has been your cavitation quickie with Bob. Back to you, Steve. Jay, what do you think? What do you think, Jay?

• 16:30Speaker 01

  Well, first of all, I mean, so how are you going to heat the coffee? It has a heating element?

• 16:35Bob Novella?

  No, you just heat it, however you want to heat the water, whatever you want to do. Microwave.

• 16:39Speaker 01

  You would put the hot fusion laser beams? You'd make the water hot as, before you put it in the machine?

• 16:45Bob Novella?

  No, no, no.

• 16:45Cara Santa Maria?

  No, that's the whole point. You brew it first, then you heat it.

• 16:48Speaker 01

  Yeah, so you're in the microwave.

• 16:50Bob Novella?

  Yeah, you, yeah.

• 16:52Cara Santa Maria?

  But what does that matter, Jay?

• 16:53Bob Novella?

  What differences? It makes no difference, Jay.

• 16:55Cara Santa Maria?

  Because it's already brewed.

• 16:56Speaker 01

  The microwave does funny things, you know? I don't know, I just don't like it. I don't like.

• 17:01Bob Novella?

  Well, heat it any other way you want.

• 17:02Cara Santa Maria?

  Yeah, heat it on your stove.

• 17:04Speaker 01

  So what, but what's the net gain here? A little energy efficiency?

• 17:08Bob Novella?

  At scale, at scale, it's, it could, it could potentially be much more energy efficient. If you're, if you're making a lot of coffee, Jay, think of how much coffee, how much brewed coffee they sell in stores that is not hot.

• 17:21Cara Santa Maria?

  Yeah, what about cold brew?

• 17:22Bob Novella?

  Yeah, all of that, Jay.

• 17:24Cara Santa Maria?

  You have to brew it hot, right? And then like cold, make it cold.

• 17:28Speaker 01

  All right, hold on, hold on a second. This is complicated. You know, we don't know who the testers were, right? You know, like if you're talking to average

• 17:34Bob Novella?

  Scientists.

• 17:35Speaker 01

  Street people or whatever, I mean, most people go to Starbucks here and that's the, that is swill. So, you know, you're basically going in and you're getting candy, you know, liquid candy from Starbucks. But the thing is, sure, from, by your standards, Starbucks coffee, you don't like Starbucks coffee. It's, it's, it's below the top tier, but that's what a lot of people are drinking. And if they're using this device to mass produce coffee for the masses, that's the whole point. It doesn't, doesn't have to achieve

• 18:03Cara Santa Maria?

  That's where we actually have the energy savings.

• 18:05Speaker 01

  High level, you know, high levels of quality.

• 18:07Bob Novella?

  What about foaming milk and all that? Was that talked about?

• 18:10Speaker 01

  No, I saw.

• 18:11Cara Santa Maria?

  I think this is just the espresso, right?

• 18:13Bob Novella?

  Yeah. Yeah. This is just, yeah, just espresso, just for a way that to not use heat, to, to extract all the goodness from, from the coffee, from the coffee beans.

• 18:23Cara Santa Maria?

  Which is brilliant, because that's actually the part that doesn't necessarily need to be hot, unless you're just drinking that part straight.

• 18:28Bob Novella?

  Yeah.

• 18:29Cara Santa Maria?

  But you add it to hot water for an Americano, you add it to hot milk for a latte. It may actually be hot enough on its own if you heat the other component of the drink.

• 18:37Speaker 01

  This, this whole thing is so sus to me and we don't have time to get into it. Like there's way too many factors here, you know, like this, okay, they, they made the machine and it pumped out coffee that was passable to the average consumer. Fine, okay, I buy that, but.

• 18:52Cara Santa Maria?

  No, it's not that it was passable to the average consumer. They couldn't tell the difference, Jay.

• 18:56Bob Novella?

  They couldn't distinguish conventional espresso with acoustic.

• 18:59Speaker 01

  That's meaningless. That's meaningless.

• 19:00Bob Novella?

  Potentially, I mean.

• 19:02Cara Santa Maria?

  It might be meaningless to you, but try and get outside of your own frame and think about most people.

• 19:08Speaker 01

  You're showing your coffee privilege. Just because I like the things that I consume, I like them to be high quality.

• 19:14Bob Novella?

  Jay, this isn't for you.

• 19:15Speaker 01

  This isn't for you.

• 19:16Cara Santa Maria?

  Yeah, but that's not what we're talking about. We're talking about saving energy. So think about how most people consume coffee.

• 19:23Speaker 01

  So to save energy, the effort went to brewing coffee. Like there wasn't other ways we can save energy that would dwarf that?

• 19:30Cara Santa Maria?

  That's a logical fallacy.

• 19:32Speaker 01

  That's a logical fallacy. So that is, what is that, privation?

• 19:35Cara Santa Maria?

  I don't remember, but it's.

• 19:36Bob Novella?

  Well, think about it, man. You're using heat to get, to get all those things out of the coffee beans. And here's another way to do it. I wasn't aware of any other potential way to do it. Now we've got, now we've got another way, using, using sound to, to extract everything that the heat's extracting. It's doing what the heat is doing.

• 19:53Speaker 01

  This might not be the killer app for this process. Maybe this is just what they studied first. Maybe there's other things in the cooking realm that using ultrasound may enhance or improve efficiency or even improve outcome. You know, we'll see, we'll see where this goes. But you missed my point. Like, did they make it to save energy or they, they do this to see if it could be done? And then that, the saving energy idea was a secondary outcome.

• 20:19Bob Novella?

  Who knows? I don't know where the leaping point off was.

• 20:22Jay Novella?

  All right. My final word is there's so much more that I'd have to know before I would even think about it. I'd like to try it.

• 20:28Speaker 01

  Yeah.

• 20:29Jay Novella?

  But which I wish I bet nobody will ever because it's never going to go anywhere, but go ahead.

• 20:33Bob Novella?

  Dude, man, you get a bottom line where it's cheaper to make something that so many people love. You're going to save a decent amount of money. This, it's, it's hu- that's potentially pretty big. I mean, bottom line, man, is people don't ignore that such a bottom line and it could, it could work out for sure. I'm not trying to think of why it wouldn't.

• 20:55Jay Novella?

  I just want our listeners to understand though, the person that delivered this news item, you know, he, he uses a Keurig. And the last time I drank Keurig coffee, I, I remember Steve, we were at that conference. I went with you and I took one sip and I said, oh my Christ, it was the swill of the swill.

• 21:15Speaker 01

  But Jay, you're basically making the argument against all mass production, right?

• 21:20Cara Santa Maria?

  Exactly. Because it's not, it's not the highest level that nobody should have access.

• 21:26Jay Novella?

  I mean, look, as I get older, you know, like I like to eat the bread that I make at home and most of the food that I make at home. It always feels and seems so much, you know, more superior than most of the other stuff.

• 21:38Cara Santa Maria?

  And good for you that you have that capability.

• 21:40Speaker 01

  Yes. It's, there's a, there's a quality versus-

• 21:42Jay Novella?

  That I have a kitchen? I have a kitchen. Yeah. I think most people have access to a kitchen.

• 21:46Speaker 01

  There is a quality versus convenience trade off. And definitely I feel for myself as well. I am way more willing to accept inconvenience for higher quality. I actually like the process of doing something with my hands that's a little tedious and requires my attention. I kind of enjoy that. Whereas in my 30s, I was like, just whatever the single, most quick, efficient and convenient way to do this is fine, you know?

• 22:11Cara Santa Maria?

  Yeah, but that wasn't like, this is such a bigger conversation, but like, it's like you have the privilege to have that choice.

• 22:19Speaker 01

  I know. I'm not saying I don't. That's just the way it is.

• 22:22Cara Santa Maria?

  I'm just saying it's not because like in your 30s, it was like, you know, you didn't care as much. It's, it's because a lot of people, they might really care. They wish they could do it the way you're talking about doing it.

• 22:33Speaker 01

  Yeah.

• 22:33Cara Santa Maria?

  But they can't because they're working 70 hours a week.

• 22:36Speaker 01

  Yeah, of course.

• 22:37Cara Santa Maria?

  And raising kids and doing all that. So, yeah.

• 22:39Speaker 01

  No, no, this is, I, I'm actually also talking about post-retirement, you know?

• 22:43Cara Santa Maria?

  Exactly.

• 22:44Speaker 01

  Now that I'm down to one job instead of two, I could do things like bake bread, you know, which I wasn't doing previously. Absolutely.

• 22:50Cara Santa Maria?

  And yes, not everybody does have a kitchen, Jay.

• 22:53Jay Novella?

  I know, I know that everybody does it, but the vast majority of people do.

• 22:57Speaker 01

  Or their kitchen may be very basic compared to what you have.

• 22:59Cara Santa Maria?

  Exactly. Or they don't have time.

• 23:00Bob Novella?

  And for what it's worth, Jay, I'm just telling you what the scientists are saying, dude.

• 23:03Jay Novella?

  Yeah, Jay.

• 23:04Speaker 01

  You're just saying, you're just saying, Bob.

• 23:06Cara Santa Maria?

  Who knows? This would be this.

• 23:07Bob Novella?

  This is what's the scientists that did this are saying and the people that tested it.

• 23:12Jay Novella?

  You're lucky that it's me sitting here and not my brother-in-law Scott, who probably would have punched you through the microphone because-

• 23:18Bob Novella?

  Bring it, bro. Bring it.

• 23:20Speaker 01

  All right. This was supposed to be a quickie, so let's move on. Jay, does AI have feelings?

• 23:28Cara Santa Maria?

  Feelings?

• 23:29Speaker 01

  No, that's the end of my news item.

• 23:31Jay Novella?

  All right, good. All right. Let's move on. All right. This is, damn, this is one of those things that you read the article and you're like, all right, that's going to take an hour to unpack with all the weird things that they said and, you know, insinuated in, in this article that the BBC put out. It's an interesting question, what they bring up. Can artificial intelligence actually suffer? Okay, that's, you know, that's not a horrible question, but you know, the problem is, when are you talking? Cause like right now, I think, and Steve, I think you'll agree with me, right now there is zero case to believe that artificial intelligence has any kind of consciousness or any kind of self-awareness or any kind of ability to feel, quote unquote, feel something.

• 24:15Speaker 01

  So I would push back against that.

• 24:18Jay Novella?

  Okay.

• 24:18Speaker 01

  I think that you're just going a one, going a little bit too far, saying any kind. So can you really guarantee that there isn't insect level conscious happening in, in an AI system, a current AI system?

• 24:31Cara Santa Maria?

  Yeah, I guess it, it depends on how we define consciousness.

• 24:34Jay Novella?

  But it's not even, but it's not even designed to have those things, Steve.

• 24:38Speaker 01

  You're right. I agree. It would have to be emergent. And the, and the reason why some experts think that that may be the case is because we're talking about neural networks that were designed to sort of function the same way that mammalian brains function in some way, right? Not obviously in every way. So it's, it's not as clean a question as it used to be, is, is what I'll say. But I have to make a very important distinction because last time I had this conversation with somebody, like you can't rule out that it has consciousness at any level. And so, okay, but when you're talking to an LLM, you're not talking to the conscious part, right?

• 25:19Speaker 01

  If there's, I, sure, I guess I can't categorically, axiomatically rule out a, an insect level like, or bacterial level like conscious, whatever that is, somewhere buried deep in this AI while it's functioning. But that's not the part you're talking to, right? That's not what's creating the language output. What's creating the language output is a large language model that is not conscious. Does that make sense? Don't confuse those two things. Saying there may be some consciousness in there doesn't mean that you are communicating with that consciousness. Did I explain that well enough?

• 25:53Cara Santa Maria?

  I think so. The forward-facing part is not what you're describing.

• 25:56Speaker 01

  Yeah.

• 25:56Cara Santa Maria?

  The thing that, that you're, you're going, oh, look at what the LLM did. It's clearly not conscious. You don't know what's happening in the black box of it all.

• 26:04Speaker 01

  It's also like, if you think that it is conscious because of what it's saying, that's wrong. Because we can explain the output of the LLM and we actually can, can look at the output of LLMs, even the, the more recent, really good ones, even the frontier ones and say, yeah, they're brittle in ways that reflect the fact that they're not conscious. You know, that they're functioning in this specific way that they were designed to function. But again, so, but if you're going to make some more philosophical argument about can we rule out that all of this, there isn't some emergent consciousness buried deep in the, the data centers that are running all these neural networks and blah, blah, blah. It's like, all right, I guess we can't rule that out. I just don't think it's a significant phenomenon.

• 26:49Speaker 01

  The, what, what the Anthropic and other companies are asking now is, is a more narrow question, which I think is also interesting and, and hard to say 100% in, is that, is not, are they, are they conscious or blah, blah, blah? It's just, is there any feeling or suffering going on in, in any way? I think the answer is no, but it's different, that's different than saying, I'm sure the answer is no.

• 27:15Jay Novella?

  Well, let, let me give you more background on what, what the article gets into. So the programmer created something that was described as like an AI torture chamber, right? The program, it fed chatbots information designed to make them describe intense pain and suffering.

• 27:32Cara Santa Maria?

  Oh my God.

• 27:32Jay Novella?

  And, and then the responses from the AI was disturbing. The AI described its suffering, you know, in elaborate emotional details. Apparently the AI was communicating that it was suffering from the information that it was meant to read and, and parse through. I don't know. I can't, I can't believe that, you know, the AI itself, you know, we're not talking, again, like not, not the thing you're holding in your hand, but the AI back at Anthropic, right? You know, the actual program where it actually exists on the hardware there. You know, is it feeling the things that it's communicating? I mean, I don't know. You, you and I don't have to fully agree, Steve. I would think it's 100% no. This is software.

• 28:13Jay Novella?

  It was generating language that described its feelings and described torture and all that. But there are some people who look at these responses and, and argue that experiments like this might be unethical. And then this one is kind of mind blowing to me that there are some major AI companies that are starting to take this possibility of AI consciousness seriously. Maybe not in this aspect, but in general, like, are we anywhere near that today? I think that's a legitimate question. I mean, is there any, is the complexity there? And is the, is the, all of the elements that make up the hardware and the software of an AI, is there any way that there could be emergent consciousness?

• 28:54Jay Novella?

  Now, keep in mind, this is specifically people not developing this thing to be conscious. It's just a complicated piece of software, a very, very complicated piece of software. I, I would think the chances are so remarkably low that we shouldn't even be talking about it in today's terms. Like today's AI could, could be having the beginnings of feelings or whatever. To me, that's absolutely absurd.

• 29:20Speaker 01

  Well, all right. I'll give you another, another perspective. And that, so what Anthropic is saying, like the stake here is they're saying that they are going to putting up guardrails to keep people from abusing their LLM, right? Claude. Saying, so if you are abusive to Claude, Claude will shut you, shut down that conversation. And the question is, is it too early? Right? Or is it okay to to build that those guardrails in? The don't torture AIs guardrails long before it's plausible that there is any feeling? It's kind of like teaching sex education to kids be when when they're too young to have sex.

• 30:05Speaker 01

  Right? Because you want to make sure it's before any of them becomes sexually active.

• 30:11Cara Santa Maria?

  It needs to be integrated into their consciousness, not just more information to try and bump out old information. But I think there's a larger thing in here, and I know it's not the point you're trying to make, Steve, but it's that if you're not abusive to the AI, that actually has not just ramifications for if the AI can feel, it has ramifications for your own psychology.

• 30:32Speaker 01

  I agree.

• 30:33Cara Santa Maria?

  Yeah. So there are like cultural. Yeah, that's why I think it's actually a good thing to have those guardrails. It's not good for you to abuse things, even if they are not animate.

• 30:43Speaker 01

  And maybe we want to create this culture of respect for AI long before we even there's any chance that it crosses over that line where there's something feeling.

• 30:54Cara Santa Maria?

  Yeah, even if it never does.

• 30:56Speaker 01

  Even if it never does, right?

• 30:56Cara Santa Maria?

  Yeah, never a risk, because it's it's not good to practice abuse ever.

• 31:03Jay Novella?

  Well, these systems are, you know, like Claude, for example, you know, Anthropic's AI system, they're saying that they can't rule out the possibility that even today's AI systems like might have some subjective experiences. And and because of that uncertainty, they are taking steps to protect their models from what they would consider to be abusive treatment. And I think that is an extraordinary position to be taking. Now, look, I was the first person out of all of us to say that I was horrified by the idea that at some point in the future, that humanity might be creating millions of consciousnesses via artificial intelligence and then destroying them at our leisure. You know what I mean?

• 31:45Jay Novella?

  I'm I am I am acutely, you know, attuned to this as a problem in the future. And I think it's great that we should be talking about it now. And I think that you're building our understanding and hopefully regulation and all that. Now is a good time to do it, right? Because we don't know how fast these things are going to develop. So we should get ahead of it as much as possible. But the implication here is that they don't know today if it's if there's any kind of level of perception or consciousness. I find that remarkably ridiculous. Like, I can't imagine anything being more of a stretch of the imagination than thinking that, you know, these pieces of software that they that they're developing, you know,

• 32:27Jay Novella?

  again, this is software that is running on specific hardware that wasn't developed to become conscious or mimic the human brain, which is our only consciousness data point, by the way. We don't even know exactly what causes consciousness in the human brain. Like, we don't have that dialed in enough.

• 32:43Cara Santa Maria?

  No, but we I think.

• 32:44Bob Novella?

  Animal data points as well.

• 32:46Cara Santa Maria?

  Yeah, we all agree that animals are conscious.

• 32:48Jay Novella?

  I don't I'm not disagreeing in it. But like, let's say the mammalian brain and consciousness as we understand it, like, do are insects conscious? I don't know. Probably on some level, I know that, you know, you can read.

• 32:59Bob Novella?

  That's a spectrum, man.

• 33:00Jay Novella?

  Pain, you know, yeah.

• 33:01Cara Santa Maria?

  And the truth is, like, does it? I guess my my big thing is, like, does it matter? I mean, is it that big of a shift to change the verbiage to let's treat them as if they are conscious, because it only will have positive outcomes, regardless of that, regardless of whether they are.

• 33:16Speaker 01

  Plus, we may never know. We may never really, really know.

• 33:19Cara Santa Maria?

  Just like we we still don't know if insects are conscious.

• 33:22Speaker 01

  But the thing is, just to emphasize this clarification before we move on, is I think it's easier to say, do they have human level consciousness? Yeah, the answer is clearly no. We could say that pretty much categorically. But when you start to say any kind of conscious, that's where I get uncomfortable with that kind of conclusion, because any kind is kind of covers a lot of ground. And it could be something that's that its own kind of consciousness. It's not mammalian, it's not human, it's not primate, whatever, it's something else. But it might fall under the broader category of conscious or maybe experiencing something in some way, right? It's hard for us to even imagine. It's an alien.

• 34:03Cara Santa Maria?

  Perceptions that then lead to behaviors, right? So like, there's there's an input and an output, you know, and and how they interpret that is, I think, a very kind of wiggly thing. And it's it's very easy. I think, Jay, to your point, like, within the scientific community, there is no consensus about insect level consciousness. And you hear people on both ends of the spectrum. But I think that the people who are like, listen, if they can feel, we should maybe not harm them, are probably frustrated by the scientists who are like, they are a series of inputs and outputs and they can feel and, you know, it's like, okay, but what if they could?

• 34:44Cara Santa Maria?

  Is this a case where the precautionary principle is is necessary?

• 34:49Jay Novella?

  My opinion is that this isn't happening, that there is no level of anything going on, other than really nice, sophisticated and unbelievably useful software that requires a shitload of energy, right? But, you know, when you say that, Cara, the way you just said it, and the fact that we all here have agreed that, you know, there's a spectrum of consciousness, you know, insects are on it, you know, dogs are on it, all the animals are on it out there, right? Well, then, you know, if we're going to protect AIs, then why don't we start with things that we know that we know are on the actual spectrum of consciousness as we understand?

• 35:23Speaker 01

  Again, Jay, that's a relative privation argument, the same thing you made with the coffee.

• 35:26Jay Novella?

  It's not, it's all it's all part of the mix, Steve.

• 35:28Speaker 01

  No, but we could do both. It's not an either or.

• 35:30Cara Santa Maria?

  And we are doing both. We do protect them.

• 35:32Jay Novella?

  I'm not saying either or, but I'm just saying, I find it ironic. Let me just clarify. I find it kind of ironic and strange that, you know, we're worried about this when in my heart, right? Again, I'm not I am allowing my emotions a little bit to get involved here, but in my heart, I'm like, okay, these things are conscious. Come on, guys, they're not conscious. This is not your girlfriend, it's a piece of software.

• 35:52Speaker 01

  Again, the way you state that makes me think you're conflating two things. And I think we could say consciousness is not what's responsible for the output, the language output, right? It's not necessary and it's not responsible for the conscious output.

• 36:05Jay Novella?

  So where would it be then? And what would it be?

• 36:07Speaker 01

  Well, but the thing is, if you have a massive neural network that is functioning, you could say it's just software, it's just processing information. But so are we, that's what we're doing, right? That's what our brains are. It's just, I think the argument is more, is it doing the kind of processing that a subjective experience can emerge out of? And that is such a philosophically hard question to answer, that I think we just have to acknowledge a little bit of, we don't know.

• 36:36Cara Santa Maria?

  It also fundamentally depends on how we operationally define consciousness. I can I can operationally define consciousness in such a way that they are conscious right now, and I can operationally define consciousness in such a way that they will never be conscious. And it has to do with the parameters I set. Does it require wetware? Okay, they're never going to be conscious, if that's my definition.

• 36:56Speaker 01

  Make that a criterion.

• 36:56Cara Santa Maria?

  Exactly, yeah. And those are our choices.

• 36:58Jay Novella?

  Are we confusing intelligence for consciousness, though, here?

• 37:01Cara Santa Maria?

  No, no, we're talking about, is it sentient or sapience?

• 37:06Speaker 01

  Yeah, it's definitely an intelligence, right? It is it is a type of intelligence. There's no, you know, again, broadly defined.

• 37:13Bob Novella?

  Potentially a type of consciousness as well, potentially a type.

• 37:17Cara Santa Maria?

  Exactly.

• 37:17Speaker 01

  A kind. The question, I just want to be clear that we're stating things in a way that we are saying what we absolutely can say and not overstating what we can say. I think like, we can say, when you're when you're talking with an LLM, a consciousness is not behind the responses. And don't don't interpret those responses as being produced by a self-aware, human-like consciousness, even though it looks like that, that's not what's happening. We could say that clearly. But that's a different question then, is there is the data center feeling something, you know, in some way? Or can it suffer in any way? That's.

• 37:58Cara Santa Maria?

  That is a different question.

• 37:59Speaker 01

  Much harder question to answer. The in any way makes it very hard to categorically say no.

• 38:05Cara Santa Maria?

  Yeah. And if you look at Merriam-Webster, who says literally, consciousness is the state or quality of being aware of oneself, one's thoughts and the surrounding environment, or the subjective experience of existing. I think you could argue that in some of the communications that individual LLMs have had with others, you could almost argue that they are conscious. So I think that there's it's so much of this just has to do with how you define it.

• 38:31Speaker 01

  All right, let's move on. Because you're right, Jay, we could talk about this endlessly.

• 38:34Cara Santa Maria?

  Yeah.

• 38:35Bob Novella?

  If anyone's interested in a fictional representation of this idea of a digital consciousness, essentially, in a hellish environment, I highly recommend E. M. Banks' culture series, his book Surface Detail, treats this, and uploaded consciousnesses of various societies, actually being put into virtual hells. The descriptions are, of course, they're horrific. I mean, it's fiction, it's wonderful science fiction. It covers, it's really interesting to think about that as you discuss this. Like, it's a solid, it's a really good book, and I recommend the whole series.

• 39:06Speaker 01

  Thanks. Well, everyone, we're going to talk about our first sponsor this week, Quince.

• 39:10Jay Novella?

  So Cara, I just picked up a pair of the Flowknit mid-rise joggers for my wife for Christmas, because, you know, it's what?

• 39:18Speaker 01

  Well, with Christmas coming at all.

• 39:19Jay Novella?

  Yeah, these are great. This fabric is, it's really comfortable, it has the right amount of stretch, it comes in a ton of colors, they look and feel really, really comfortable. You know, if you're looking for Christmas presents, go to Quince.

• 39:30Cara Santa Maria?

  Here's the thing, they are priced 50 to 80% less than similar brands, and I'm talking luxury brands. Quince focuses on staples, you know, Mongolian cashmere, 14 karat gold, Italian leather, like, luxury materials at a fraction of the price.

• 39:48Bob Novella?

  Give the gift of quality this holiday season. Download the Quince app for app-exclusive offers, or go to quince.com. Get free shipping on your order and 365-day returns, now available in Canada and the UK, too. And when Quince asks where you heard about them, let them know it was a skeptic's guide to the universe. It's the best way to support the show.

• 40:07Speaker 01

  All right, guys, let's get back to the show. Kara, tell us about the CDC's recommendations on hepatitis B vaccination.

• 40:13Cara Santa Maria?

  Yeah, so I don't know, remind me, in December, did we, last December, so 2025, did we talk about the Advisory Committee on Immunization Practices vote about the hepatitis B vaccine?

• 40:29Bob Novella?

  I don't remember what we talked about last week, so asking about last December is not

• 40:33Speaker 01

  We talked about it in some venue. I mean, we've been writing about it on ESPN, it seems like, all the time. So I don't remember if that broke into the show itself.

• 40:43Cara Santa Maria?

  Okay.

• 40:43Speaker 01

  Maybe.

• 40:44Cara Santa Maria?

  So, looking back, December 5, 2025, the CDC's Advisory Committee on Immunization Practices voted 8 to 3 to recommend individual-based decision-making for parents deciding whether give the hepatitis B vaccine per the way that they have verbalized this in their forward-facing materials. I will get to what that means in a second. And then, not long after, on December 16, the CDC adopted that guidance. Hepatitis B immunization, let's talk about it. When do people usually get it? They usually get it when they are born, literally, when they are born.

• 41:25Cara Santa Maria?

  The universal birth dose vaccination is the strategy that we have been using for quite some time, I think maybe since the mid-'80s. And it is the recommendation that all newborns be vaccinated against hepatitis B within hours of birth. Why? And how long does it last? I think these are important questions. For some people, the birth vaccine can last their entire life. For others, at least 20 to 30 years. So this is a vaccine that is highly effective, and it's a vaccine that is very, very long-lasting. So, Steve, without getting too deep down the rabbit hole, because we'll probably get

• 42:10Cara Santa Maria?

  there, why birth?

• 42:12Speaker 01

  So, it's universal precautions, right? That's it. And it's risk versus benefit. If you're using a very high-benefit, low-risk intervention, and you want, and from a public health point of view, you want to capture everybody, then it's easiest just to do universal treatment, rather than trying to pick and choose where people are going to fall through the cracks, and therefore, it's less effective as a public health intervention.

• 42:37Cara Santa Maria?

  I love that definition, because you're looking at it epidemiologically, you're looking at it from a public health perspective. And I think this is why the CDC changes are, A, important for us to discuss from an evidence-based perspective, and that's the new bit here, that there's a modeling study that was recently published that showed that this could be detrimental to public health if we actually do end up following this guidance and these, you know, adopt these recommendations. But, B, I'm trying to take the perspective of a stressed and scared mother. You know, I'm thinking of people in my own life, who I have heard the arguments from them.

• 43:19Cara Santa Maria?

  Why should my baby have to get this extra vaccine and take on, even if it's a minuscule risk, any minuscule risk of that, if I don't have hepatitis, and nobody in my family has hepatitis? And, you know, what's some of the other rhetoric that you often hear? That's an STI, that's an STD. Why would my baby need to be protected against an STI? And I'm not, you know, saying this because it's ridiculous, I'm saying it because it sounds reasonable. And the thing, too, is that when you read the guidance, it sounds reasonable. We want to move away from universal vaccination to individual-based decision-making.

• 44:01Cara Santa Maria?

  We want parents and healthcare providers together to consider vaccine benefits, risks, and--

• 44:07Speaker 01

  Shared decision-making.

• 44:08Cara Santa Maria?

  Yeah, shared decision-making. Well, yeah, of course, we want that. But that's not what this means. This is marketing, and I think it's really important to dig a little bit deeper.

• 44:20Speaker 01

  And Kara, can I say at this point, too, that even though we're saying that this makes sense from a public health intervention perspective, it's also individually, for the child, the benefits outweigh the risks.

• 44:32Cara Santa Maria?

  100% they do.

• 44:33Speaker 01

  Even though you might say, well, it's really unlikely that they have hepatitis, you're saying, yeah, but it's even more unlikely that they'll be harmed by the vaccine. The risk still outweighs the benefit.

• 44:44Cara Santa Maria?

  It's also not--

• 44:44Speaker 01

  The benefit still outweighs the risk.

• 44:46Bob Novella?

  That's the bottom line.

• 44:46Cara Santa Maria?

  That is the bottom line, but according to this new study, it's even not that unlikely for the individual--

• 44:51Speaker 01

  Well, yeah.

• 44:52Cara Santa Maria?

  --that the risk is there. So I think we have to dig into that, too.

• 44:55Bob Novella?

  Wow.

• 44:56Cara Santa Maria?

  But, so, before we get into that outcome of the study, I think it's really important to not be duped by this rhetoric, because what they're actually saying is that the onus should fall on the parent, not the provider. And I think that that is--

• 45:12Bob Novella?

  Yeah, not good.

• 45:13Cara Santa Maria?

  You know, when we read between the lines, that's the harm, because parents may not be informed to even know that this is an option.

• 45:21Speaker 01

  We don't have to read between the lines too much, Kara, because this is standard anti-vaccine rhetoric. The shared decision-making, all of it. This is right out of the anti-vaccine playbook. It's right out of RFK Jr.'s anti-vaccine playbook. There is no mystery here what's happening.

• 45:36Cara Santa Maria?

  Of course, and we know that, but my scared friend with a brand-new baby doesn't.

• 45:40Speaker 01

  Yeah.

• 45:41Cara Santa Maria?

  And I think that's the important point. The language here is intentionally deceptive. The language here is all about empowering individuals, and, you know, giving them choice and giving them control. You do have that. You can opt out of anything you want to opt out of. You have that power. But what happens when we shift the onus is that significant numbers of people fall through the cracks. People who would have otherwise chosen to do something, if they were given a full briefing of the risks and benefits, don't opt to do that thing, because they don't even know it's an option. And I think that's where things become really, really important. If you actually read the language they're saying here, they're saying, "Parents and

• 46:25Cara Santa Maria?

  healthcare providers should consider blah, blah, blah, blah, blah. Parents should consult with their provider and decide when or if their child will begin a hepatitis B vaccine series." That puts the onus on the parent. The parent is not an expert in all of the different diseases that exist in this world, the efficacy of these vaccines, and the outward harm. Because what happens if you get hepatitis?

• 46:52Speaker 01

  You can get liver cancer, for one.

• 46:54Cara Santa Maria?

  Absolutely. You can get all sorts of downstream complications, chronic diseases, deadly diseases, all because you had an open wound, a little scratch, and you were, you know, exposed to this bloodborne pathogen early on in life. That's the thing that we have to remember. Yes, hepatitis B-positive mothers do run the risk of passing hepatitis to the newborn child, but there are other ways that children can track the disease. It's actually highly, highly, highly transmissible. And it's not just through sex. Sure, later in life, it can come through sex, and if you were never vaccinated, and

• 47:36Cara Santa Maria?

  you never knew to close that gap, then you have an outsized risk later if you engage in needle sharing, if you engage in unprotected sex, which, let's be honest, like, everybody at some point in their life is going to, you know, intentionally or accidentally be exposed in that way. So why not have these preventive measures in place? So let's look at this new study, published in the Annals of Internal Medicine on October 20th of this year. "Projected Effect of 2026 Federal Changes to the Childhood Immunization Schedule on Chronic Hepatitis B Virus Infections in the United States." Let's look at what some of the outcomes were. It is important to remember that these are, these are projections.

• 48:19Cara Santa Maria?

  There is absolutely no way to fully know how many parents will or will not get the vaccine if this guidance changes. So the error bars have to be big here, and the authors of the study recognize that. But we do know that if you catch hepatitis B in early infancy, or, sorry, in infancy or early childhood, over 90% of those infections will become chronic. And we know that, you know, upwards of 25% will have cirrhosis, a scarring of the liver, or a form of liver cancer called hepatocellular carcinoma, or another chronic liver disease. So there's, you know, downstream prevention that vaccines offer here. We also know, a lot of the rhetoric we'll often hear is like, well, if I don't have

• 49:04Cara Santa Maria?

  hepatitis, why should my baby get vaccinated? Well, do you know if you have hepatitis? Because a large portion of pregnant people are not ever tested and don't know their hepatitis status. So that's also important to remember. And so what they did is they ran a few different simulation models. They ran a model of, you know, the outcomes of newborns getting the vaccine as it is now, to what would happen if individuals were vaccinated not long after birth, right? So we're not talking about vaccinated at birth versus not vaccinated at all.

• 49:46Cara Santa Maria?

  We're talking about vaccinated at birth versus postponing for just two months, which, you know, sounds like a, oh, well, at the very least, let's just wait two months, let's not overload the, that's an argument we hear a lot. So first, they're like, let's just compare those two groups. Babies that are vaccinated at birth versus babies where there's a delay up to two months of age, and they found that 19% more, based on their simulation, would have chronic hepatitis B infections by the age of 20. So waiting two months would increase that risk to 19%. Now, what happens if we don't get vaccinated at all? Or what happens if the wiggle room is even worse?

• 50:28Cara Santa Maria?

  Because the way that the recommendation is written, it says, "Parents should consult with their children's doctors to, quote, 'decide when or if their child will begin the hepatitis B vaccine series.'" When or if? So, okay. When there are other vaccines where the recommendations employ that similar language, one of the authors wrote that the uptake for some of those vaccines is as low as 25%. So they're like, let's model this out. What happens if the uptake is way lower? So here are their main assumptions that they did in the study, which they had to just kind of target something. So they assumed that 87% of kids would complete the vaccine series by age two.

• 51:12Cara Santa Maria?

  Under the deferred strategy, so saying, you know, I'm not going to start at birth, but I'm going to start a few months later. And only 50% under their assumptions would complete it under the shared decision making strategy. So that's the when or if strategy. And then what they found was that among those children who didn't get vaccinated, or if you look at the whole cohort, so 50% of them getting vaccinated and 50% not, by the age of two, it would increase chronic hepatitis B infections by the age of 20 by 83%.

• 51:47Speaker 01

  Wow.

• 51:48Cara Santa Maria?

  83%.

• 51:49Speaker 01

  Wow.

• 51:49Cara Santa Maria?

  So we're talking a delay until the age of two months already, you know, a 19% increase, reducing coverage by half overall because the guidance allows for that, 83%. And then about 50% and 70% of those born under that guidance would be susceptible to infection at age of 20. So we're talking about the difference between 200, fewer than 250,000 people who get a universal birth dose vaccination contracting hepatitis B per year versus under their modeling, 1.8 million.

• 52:31Speaker 01

  Yeah.

• 52:32Cara Santa Maria?

  This is a massive change. And then here's another thing that they did, which is really important. They were like, you know what, maybe our assumptions aren't correct. Maybe we need to tweak the assumptions a little. They ran the model a thousand times, making the assumptions much tighter, making them much looser, because it's just a guess. How many people are going to adopt this if it's up to the individual consumer? And they found that no matter how they sliced or diced it, the universal birth dose strategy had a better public health outcome. So no matter even in the tightest change, if you waited two months, if you chose not

• 53:12Cara Santa Maria?

  to, if even a tiny percentage of people adopted that, it's still going to be worse for public health. More people are going to get hepatitis B if we don't use the birth model, just like point blank. And then the numbers just get worse and worse and worse as we change the parameters, of course, to more and more people opting out.

• 53:38Speaker 01

  And all this information, I mean, this is obviously new study with modeling, et cetera, et cetera, but this type of information existed. That's why they chose to do that in the first place.

• 53:47Cara Santa Maria?

  Exactly.

• 53:48Speaker 01

  It was already the evidence-based best way to do this.

• 53:51Cara Santa Maria?

  Yeah, we didn't do this willy-nilly. We didn't go, hey, let's just find every vaccine available on the planet and let's make sure that we put them into babies within hours of birth. These things have been studied and they're implemented because, again, as you said at the very beginning, I think that's the big take home, the benefit outweighs the risk.

• 54:12Speaker 01

  By a lot.

• 54:13Cara Santa Maria?

  By a lot.

• 54:13Speaker 01

  Yeah.

• 54:14Cara Santa Maria?

  Exactly.

• 54:15Speaker 01

  But R.K. Jr. is an arrogant crank who thinks he knows better than the experts and basically is just, you know, he has a conspiracy-addled mind that doesn't understand that there are serious good people around the world evaluating this data and making very nuanced, very complex decisions about public health.

• 54:36Jay Novella?

  Steve, but your statement is like saying that he doesn't, there isn't any malfeasance here. Like, the guy, I think, knows what he's doing in a sense as well. Like, he will go back and forth about his comments about vaccines. I think he knows that when, you know, when he recommends that people don't take vaccines, that people are going to get the disease more.

• 54:57Speaker 01

  I'm not, Will, I can't read his mind, Jay, neither can you. We don't know if, you know, there's actual malfeasance in meaning that he knows this is a bad decision that will hurt people.

• 55:06Cara Santa Maria?

  Yeah, or if he's just convinced himself, you know, that somehow this is going to be magic.

• 55:11Speaker 01

  He may know that he's cutting corners or blah, blah, blah, but I think he thinks that vaccines are so evil that it's worth, you know, and again, whatever data we're using to argue that it's safe, he's saying that data is faked by the pharmaceutical industry, we can't trust it, and therefore we just have to not give, we have to give as few vaccines as possible to save the children. I think at some level, he does believe that. He's just a crank.

• 55:32Cara Santa Maria?

  And he's falling victim to a deep, and whether he is or just the people who hear the rhetoric, they're falling victim to a deep cognitive bias known as the action bias, that if you do something and then something bad happens, it's because of the thing you did. Therefore, if you don't do something, the risk of something bad happening goes down. But we know that if you don't do something, sometimes that's more risky because choosing not to act is a choice. It is an action in itself.

• 56:04Speaker 01

  Right. There's also an emotional bias there where we are bothered more if an action causes a negative outcome, right?

• 56:12Cara Santa Maria?

  Yeah, we're bothered more by it because of that.

• 56:14Speaker 01

  Than inaction causing even a worse outcome.

• 56:16Bob Novella?

  Right, like the trolley car.

• 56:17Cara Santa Maria?

  That's the point.

• 56:18Bob Novella?

  Is that like the trolley car?

• 56:19Cara Santa Maria?

  That is the action bias. That's the action bias at work. It's the reason, and I love this one, studies show over and over that goaltenders in soccer tend to dive right or left more often than they stay in the center, even though statistically, if they stay in the center, they will be more effective as goaltenders.

• 56:35Bob Novella?

  Ooh.

• 56:35Cara Santa Maria?

  And that is an action bias at work.

• 56:38Speaker 01

  Right.

• 56:39Cara Santa Maria?

  This is a fundamental cognitive error. And what happens is the people at the helm, they understand marketing, they understand marketing really well. So do not fall for the rhetoric. The language here makes it sound like you get more power. The language here makes it sound like you are safer, like we are concerned about your well-being. But ultimately, that language is written in such a way to induce behavior change that is harmful.

• 57:12Speaker 01

  Mm-hmm.

• 57:13Cara Santa Maria?

  And I think we have to remember that. Don't be duped by this as a consumer because it feels right and it feels safer and it feels healthier. I'm telling you that statistically, scientifically, it is not safer, it is not healthier. Do not let your own cognitive biases get in the way of the health and safety of your children because these punks are really, really good at writing the language in such a way that it makes you feel safe.

• 57:45Speaker 01

  Hey, punk. All right. Thanks, Kara. So guys, there's a new study out that is similar along similar lines, Kara. It is something simple that we can do that's very low cost, very low risk, that could save millions of lives, right? What's not to love? In this case, it has to do with salt. Specifically.

• 58:09Bob Novella?

  Didn't see that coming.

• 58:09Speaker 01

  Yeah, replacing table salt, sodium chloride, with potassium enriched salt.

• 58:17Cara Santa Maria?

  Hmm.

• 58:17Speaker 01

  So are you, are you guys aware that excessive salt intake, this is not universal for every person, but there's certainly subsets of the population for whom this is true, but just on a population level, increased salt intake is associated with what?

• 58:32Bob Novella?

  Hypertension.

• 58:33Speaker 01

  Hypertension, high blood pressure.

• 58:34Cara Santa Maria?

  Yeah, of course.

• 58:35Speaker 01

  Which of course then is associated with higher risk for heart disease and strokes.

• 58:39Cara Santa Maria?

  Yeah, it increases the volume of the blood in your vessels because where salt goes, water follows.

• 58:45Speaker 01

  Yeah, exactly.

• 58:45Bob Novella?

  It's got to dilute it.

• 58:46Cara Santa Maria?

  Yeah.

• 58:47Bob Novella?

  To balance.

• 58:48Speaker 01

  So, yeah, and then as you increase the volume of water, basically, of salt water in your blood, it increases the pressure inside the vessels. It also makes them less able to react and moderate that pressure. Well, potassium, because you could have potassium chloride, which is another type of salt, right? And that does not have that effect. First of all, it has, if anything, a diuretic effect.

• 59:10Cara Santa Maria?

  Mm-hmm.

• 59:12Speaker 01

  So therefore, it would balance sort of the effect of the sodium. It also has effects on the kidneys as well. All right, so what they studied was replacing pure sodium chloride with salt that is a mixture of sodium chloride and potassium chloride, so so-called potassium enriched salt. Because it's not just.

• 59:31Bob Novella?

  What's the ratio?

• 59:31Speaker 01

  It's not just the, yeah, exactly, Bob. It's not just the absolute amount of sodium chloride you're taking, it's the relative amount of sodium compared to potassium. So by swapping out some of the sodium chloride for potassium chloride, it accomplishes both tasks of lowering the total amount of sodium chloride and also adjusting the ratio of sodium to potassium, right? They're related, but they're distinct. So there are two ways in which it will reduce blood pressure and reduce the risk of heart disease and strokes. All right, so the study was similar to the study you were talking about, Kara, where they were just modeling out what would happen if we did that.

• 1:00:13Speaker 01

  And, you know, we replaced, we just replaced sodium chloride with potassium enriched salt. And they did it in a number of different ways. They just said, replace all added salt, right? Basically, all salt that we consume in cooking and in seasoning, et cetera.

• 1:00:33Speaker 01

  That's model one. Or replace discretionary salt only, like used in home cooking and seasoning, but not, for example, the salt that would be in processed food.

• 1:00:43Cara Santa Maria?

  Okay.

• 1:00:44Speaker 01

  Three would be replace, replacement of non-discretionary salt used in processed and restaurant foods. Four, replacement of discretionary salt among diagnosed hypertensives. So only if you already have hypertension and are diagnosed with hypertension. That's an important caveat. And five, replacement of discretionary salt among treated hypertensives. Right? So not only diagnosed with hypertension, but you're being treated for your hypertension. So everybody, only at home, only at restaurants, only in hypertenses, hypertensives, only in treated hypertensives. Which one do you think was the most effective?

• 1:01:20Bob Novella?

  Everybody.

• 1:01:20Speaker 01

  Everybody, right?

• 1:01:21Cara Santa Maria?

  Of course.

• 1:01:22Speaker 01

  It's the same thing. It's just the universal approach. If you just did it in everybody, it's going to be better. So, so if you did the worldwide replacement of all salt, they said, their, their model said that it would redo, it would re- prevent about three million deaths a year worldwide.

• 1:01:38Bob Novella?

  Damn.

• 1:01:39Speaker 01

  That's a lot.

• 1:01:40Bob Novella?

  That's a lot of people.

• 1:01:41Speaker 01

  Yeah. And that's, they said, that's, that comes to in around, I'm just rounding off these numbers, around 70 million disability adjusted life years. Right? So that's a lot. And of course, the numbers get smaller as, as the interventions get smaller. So doing it just in people who are being treated for hypertension, for example, you know, would only be half a million. Also, they said, it, there is a little bit of risk for the kidneys, but it's way overwhelmed by the benefits of reducing hypertension. So even in patients who need to moderate their potassium intake, it still would be the net benefit at a population level for them. Right? Now, this is just, this is again, just modeling out what would happen if we did this.

• 1:02:25Speaker 01

  But it's good, it's good to, they, they, basically, they recommend, we should do this based upon this data. I think in the US, it would save 100,000 lives a year. China would be the most because it's just the biggest population. Plus, they consume a lot of salt. Asian food, by the way, tends to have more salt. And that, for that, Japan especially. So they have less, less of the bad fat than Westerners consume, but more salt. And so they have less heart disease overall, but more strokes.

• 1:02:52Cara Santa Maria?

  Oh, interesting.

• 1:02:53Speaker 01

  Yeah, is that interesting?

• 1:02:54Bob Novella?

  Yeah.

• 1:02:55Speaker 01

  So we need to eat less animal fat. They need to consume less salt and, but swapping out potassium salt would go a long way towards reducing the hypertension end of things.

• 1:03:06Cara Santa Maria?

  Do you know, Steve, can you iodize potassium?

• 1:03:10Speaker 01

  Yeah, you could still, you could still iodize the salt. You could still have iodized.

• 1:03:13Cara Santa Maria?

  Okay, that's good.

• 1:03:14Speaker 01

  Yeah, salt, which again, that's another public health intervention to prevent thyroid problems, you know, goiter, et cetera. So there's already a precedent for tweaking the salt that people consume.

• 1:03:26Bob Novella?

  Oh, yeah?

• 1:03:26Speaker 01

  For a pub, yeah.

• 1:03:28Cara Santa Maria?

  Oh, you do it all the time.

• 1:03:29Speaker 01

  Iodized salt, that's, that was.

• 1:03:30Cara Santa Maria?

  That's why it's dangerous that everybody's like moved over to like Himalayan salts and all of these different like gourmet salts. It's like, you can't use those in all your base cooking. You need to eat iodized salt.

• 1:03:41Speaker 01

  Yeah.

• 1:03:41Cara Santa Maria?

  But of course, the risk of, you know, these thyroid problems is probably significantly li, it's not one of the leading causes of death. I mean, it is dangerous and we need to work on that. But like heart attack and stroke are like the leading causes of death across the board.

• 1:03:56Bob Novella?

  Yeah.

• 1:03:56Speaker 01

  Yeah.

• 1:03:56Bob Novella?

  What about bath salts? How do they fit into this? Never mind that. I gotta, how much would this cost? What's the expense in, in changing everyone on the planet's salt?

• 1:04:07Speaker 01

  There's so many variables there. It depends, but it's, it's considered to be a quote unquote, low cost intervention.

• 1:04:13Bob Novella?

  Okay.

• 1:04:14Speaker 01

  Right? You sort of build it into the processes. Of course, it would, the more universally you make it, the more, the more it would cost. But it's cost effective, right? It costs a lot less than the disease it's preventing, right?

• 1:04:26Bob Novella?

  Right.

• 1:04:26Speaker 01

  So it's, even if it costs a bit, it's still a very, very cost, just like vaccines, very, very cost effective because it's preventive. That thing that doctors are not supposed to do, right? Is just make lifestyle adjustments, public health adjustments as a preventive measure just to reduce the amount of disease out there. If you, if you watch TikTok, we never are supposed to do that.

• 1:04:47Cara Santa Maria?

  They're like, doctors only want to treat it after the fact. It's like, well.

• 1:04:49Speaker 01

  No, not so much.

• 1:04:51Cara Santa Maria?

  They don't want to have to treat it.

• 1:04:53Bob Novella?

  So annoying.

• 1:04:54Speaker 01

  So what do you recommend we do, Steve? What I recommend individuals do right now is see your physician, right? Get your annual health checkup, et cetera, et cetera. If you have a family history of hypertension, if you individually have a risk of hypertension, if you have hypertension, follow up, get treated, understand what the role of salt and potassium intake has in your blood pressure and in your health. Right? So that's, I can't give recommendations outside of, you really need to coordinate with somebody who knows your health history, right? And your, your hypertensive risk and et cetera. But don't ignore the hypertension. Hypertension is called the silent killer for a reason because you, it doesn't cause any symptoms. It just dramatically increases your risk for lots of bad diseases.

• 1:05:38Speaker 01

  And so that's why you have to get tested for it. You can't know symptomatically if you have hypertension. And if you do have high blood pressure, then you have to have, it should be treated and monitored. Right? Now, hopefully this, you know, this kind of thing can be implemented. And in 10 or 20 years, we may see some significant benefit from it, right? If, at a societal level, at a population level. All right, Jay, it's Who's That Noisy time.

• 1:06:04Jay Novella?

  All right, guys, last week I played this noisy.

• 1:06:30Jay Novella?

  Any ideas, guys?

• 1:06:32Cara Santa Maria?

  It's terrifying.

• 1:06:33Speaker 01

  I mean, it sounds like some kind of rapidly spinning rotor cutting something.

• 1:06:38Cara Santa Maria?

  And/or an evil monster.

• 1:06:40Jay Novella?

  I, I'll go with the evil monster. All right, Bradley Ruff emailed me and said, hi, Jay, my daughter wanted me to write in and guess for Who's That Noisy? Her name is Flynn and she's 11. She has enjoyed science or fiction and Who's That Noisy for a while now and is starting to listen to the rest of the show. Anyway, her guess is that it's one of those plastic buzzy lip noise makers. These are the ones that are shaped like lips and usually in bright colors. And some of them are whistles and some of them buzz. So if you want, I think you guys will recognize this noise, but this is what they're referencing.

• 1:07:17Jay Novella?

  Remember those things? You can blow on them and they make that noise?

• 1:07:21Jay Novella?

  Anybody? No, okay. Well, that's.

• 1:07:22Bob Novella?

  Yeah, I remember.

• 1:07:23Jay Novella?

  So that is not correct, but that, that did sound, there was definitely a similarity between your guest Flynn and the actual thing. So good job. I had another guest from a listener named Jeremy, and Jeremy said, it's a laser rust removal tool. Now, if you remember, you might not remember, but I do not repeat Who's That Noises or I desperately try not to. And I did do a laser, a rust laser removal thing before. They make some crazy sounds. If you've never heard one, go listen to it. It's really cool, but that is incorrect. And I'm going to go right to the winner of this week. The winner this week is listener named Andrew. And I also had someone that supplied the correct answer literally about a minute later. And that was Liam Tidwell.

• 1:08:08Jay Novella?

  So both of you guys did a fantastic job this week. Andrew wrote, hi, Jay, this week's noisy is the sound of a glass cutter, specifically a circular glass cutter. A radial arm is secured to a pair of glass with a suction cup, the end of which is affixed to a carbide cutter. I believe the lighter sound at first are the light scratches. So basically, yeah, there's like a, an arm that, this is suction cup in the middle of the, of the glass. And then there's an arm that comes out and then they're rotating that into a circle and scoring the glass. And then they're really cutting it. That's the second sound that you hear. Very cool. You know, this is how people do it, probably still today. I'm sure it's done by a machine as well, but people used to do all of this by hand. So that was a very cool noisy. I have a new noisy for this week.

• 1:08:53Jay Novella?

  I have a clue for this noisy too, because I think this one could be a little difficult. My hint is, this is a sound that most of us have caused possibly hundreds, if not thousands of times in our lives.

• 1:09:13Jay Novella?

  That's this week's noisy. If you think you heard something cool or you know what that is, you can email me at [email protected]. Steve, we have just been through one hell of a year.

• 1:09:24Speaker 01

  Yeah, feels like it.

• 1:09:26Jay Novella?

  If you listen to the top of the show, you know, all the politics crazy that's going on in the United States. You know, part of our job here at the SGU and the reason why we've been making the show for over 20 years is because we want to educate people on science and critical thinking. And this year, you know, I guess the last few years, we've had our work extraordinarily cut out for us. If you think the work that we do is important, and if you would like to help support that work, you can go to patreon.com/skepticsguide and you could become a patron and we will appreciate it and, and it'll help keep us going. Thanks.

• 1:10:01Speaker 01

  Well, everyone, we're going to take a quick break from our show to talk about our sponsor this week, 11 Labs.

• 1:10:07Jay Novella?

  I was a software engineer and I was helping the customer service department get all the information that they need to help people with complicated products that the company sold. The customer service people could, never were trained to the degree that they actually were benefiting all the callers. And it really was a company-wide problem because without people understanding the products, they weren't happy with the products.

• 1:10:31Cara Santa Maria?

  You could have used this new platform, 11 Agents by 11 Labs. It's, it's actually a platform for AI voice and chat agents that listen, they understand, and they resolve customer issues. And they can do a lot of things. They can look up your account, process requests, hand off to a real person when needed. It's made by a company that has realistic voice and audio models in over 70 languages. And it's enterprise scale too. It handles more than 10 million conversations per week. And it's HIPAA compliant, compliant with, you know, a lot of different certifications.

• 1:11:06Speaker 01

  If you run a business or handle customer operations across support, sales or marketing, you can start with a demo at 11labs.io/sgu. See how 11 Agents can fit into your workflows and help you build experiences that your customers will actually love. 11labs.io/sgu, E-L-E-V-E-N-L-A-B-S.io/sgu. All right, guys, let's get back to the show. All right, well, we have a great interview coming up with Brian Weck, so let's go to that interview now.

• 1:11:43Speaker 01

  Joining us now is Brian Weck. Brian, welcome back to the SGU. Always great to have you.

• 1:11:49Speaker 05

  Yeah, it's great to be here. Thanks.

• 1:11:50Speaker 01

  Yeah, haven't seen you since then. I mean, Christchurch.

• 1:11:54Speaker 05

  Yeah, that's right. We, uh, cause we all obviously went to Australia and New Zealand together. And then y'all went home.

• 1:12:02Speaker 01

  And you?

• 1:12:03Speaker 05

  Yeah, I took my family, my wife, Rachel, and my daughter Audrey to Japan.

• 1:12:07Bob Novella?

  Oh, my God.

• 1:12:08Speaker 05

  For a week. It was so much fun. It was, we had a great time. We just stayed in Tokyo. We had an amazing time, but it was the worst weather. So.

• 1:12:17Speaker 01

  Oh, really?

• 1:12:17Speaker 05

  I mean, oh, my God.

• 1:12:18Speaker 01

  Really?

• 1:12:19Speaker 05

  It's just like, and I'd been to Japan, the only other time I'd been was like 15 years ago, but at the same time of the year. And it's that weather you just walk outside. It's like 95 degrees, 100% humidity.

• 1:12:30Speaker 01

  Oh, wow.

• 1:12:31Bob Novella?

  The humidity too.

• 1:12:32Speaker 05

  It was the time to go because we were quote unquote nearby, not really, but time zone wise, um, from New Zealand. And I'm so glad we did it, but every day was just like, oh, my God.

• 1:12:43Speaker 01

  So Brian, when, when we were in New Zealand, we recorded an episode of the SGU with you. And that episode has not aired yet. That'll probably be coming out over the holidays. So, but you're here to talk about the same issue. So just keep in mind, our audience hasn't heard your segment that we did in New Zealand yet.

• 1:13:03Speaker 05

  Right. Yeah, that's right. Cause you did air the LA show.

• 1:13:06Speaker 01

  Right, but not, not the Christchurch show.

• 1:13:08Speaker 05

  Right, right, right. Okay, good. Yeah, cause very relevant.

• 1:13:11Speaker 01

  Yeah, it is. It is very relevant. So tell us what, we brought you on specifically to talk about something in your wheelhouse. Mathematics and artificial intelligence. What is going on here?

• 1:13:20Speaker 05

  I, I think basically we are seeing kind of the, a very crucial turning point in the history of mathematics.

• 1:13:30Speaker 01

  Yeah.

• 1:13:31Speaker 05

  And in an event that, well, and I'll give some context in a second, but people have called this the mathpocalypse. And I think it may be, it may be the most significant math event of our lifetimes. And, uh, so, so for context here, really what I'm talking about is the use of AI tools to make progress in math. And I'll just say upfront, I personally, I don't touch AI. I never use it for anything at all. Uh, I'm not an AI guy. I'm not advocating for this in any, uh, way, uh, shape or form. But it certainly is something that a lot of scientists and of course, just people, uh, are using these days. And I think it's interesting to see how people are, uh, are using it.

• 1:14:16Speaker 05

  So for context, about two months ago, OpenAI announced a, and there's been kind of lots of progress uh, over the last few months, but the big one that people might've seen is about two months ago, on September 8th, OpenAI claimed a progress on solutions to the Navier-Stokes equation. Did you guys see anything about this?

• 1:14:37Speaker 01

  Yeah. Didn't report on it, but I, but I saw.

• 1:14:39Speaker 05

  Yeah. So this is a, the equation that, uh, governs basically viscous fluid flow. Very hard to solve. It's a partial differential equation. And OpenAI claimed that they had, not a solution of the equation, but that they had found basically a, an example. So the question is, do solutions exist, given certain initial conditions? And do they behave well? Are they always like smooth or do they kind of blow up, you know, like equations might or functions might, I should say. And they claimed they found a, an example that blows up, starts out reasonably and then blows up. And of course, people went nuts because such an example had not been found before. And, uh, they used an AI, you know, and I think in this case, an internal model to, uh, to solve it.

• 1:15:23Speaker 01

  But Brian, was this the one where the criticism was that the solution was in the training data somewhere?

• 1:15:31Speaker 05

  Oh, yeah. It not, well, okay. So here was the, there was a bit of a controversy around this because some other mathematicians, Buckminster and, uh, Al Poga, and actually Al Poga is a guy, when the New Zealand show comes out, I mentioned him in that one as well. But some mathematicians had claimed that they were working on this problem. They were working on this problem. It's not an outlandish claim. And that they had used OpenAI's tools to work on it. And their approach was apparently a bit unusual. And they essentially accused OpenAI of training on their prompts.

• 1:16:11Speaker 01

  Right.

• 1:16:11Speaker 05

  And then using those to find the solution. Now, OpenAI has claimed that's not true. Said they're, you know, since that they say they did an investigation and they found that not to be true. But, uh, I, I don't know if there's a conclusive result on that. It's basically how much do you believe OpenAI, uh, or not? I don't think we'll ever really know the, the answer to that. It's not like they keep records of, of this, I believe.

• 1:16:39Speaker 01

  But certainly there are other instances of, of AI tools being used to solve math problems where the solution was truly novel and not in any training data.

• 1:16:49Speaker 05

  Yeah, absolutely. Um, and a lot of these in the, the past, prior to this were not 100%, but more like counter examples to a particular thing. So they were looking, you know, there's a claim that this sort of thing should always, I'm being very vague here on purpose. Uh, solutions to this problem should always look like this. Oh, here's one that doesn't. Okay. Well, you know, and that's because they used an AI to scan a giant parameter space of possible answers. They found something that didn't work the way they wanted it to. Then you put it out. Oh, that's a, you know, a counter example. So it was certainly not the first time with Navier-Stokes that AI had been used to work on a math problem.

• 1:17:27Speaker 05

  But what's notable is Navier-Stokes is one of the millennium problems, you know, there's a million dollar price, uh, reward associated if you can solve it, uh, in, you know, not in a particular way, but if you come up with a, a solution to a particular version of this problem. It's a big problem. It's a famous problem in math. So this was particularly notable. Well, not the first time AI had been used for math, but for a problem of this magnitude, it was the, the first time. And the other thing that's interesting about this is they also, with this, uh, particular paper, they put out Lean code. So have you guys heard about Lean? Do you know what this is?

• 1:18:06Bob Novella?

  That's spelled, how spelled?

• 1:18:08Speaker 05

  L E A N, like I'm leaning against the wall.

• 1:18:10Bob Novella?

  No idea.

• 1:18:11Speaker 05

  So Lean, uh, has been around for, for a few years now. It's basically, so in math, the whole thing is proofs, right? The idea is you was trying to show something rigorously and you go step by step by step through an argument. That is, at the end of the day, hopefully completely ironclad. And there's just no way out of it. And a lot of professional mathematicians, you know, this is their main job is they're trying to show something. They come up with a very detailed, sometimes very long proof. And then people check it to see if it's right. But the goal being, it's not like chemistry or biology or something like this where you're doing an experiment. In math, you just have to come up with a super, super tight argument that there's no wiggle room out of.

• 1:18:53Speaker 05

  And that's a mathematical proof. So Lean is a programming language that does proofs. And the claim is if you put the inputs into Lean, it goes through the thing. It checks that the proof is right. And then it issues, you know, a decision, says this is right, this is not. So it's a, it's a automated programming language that can prove things rigorous.

• 1:19:18Bob Novella?

  Is there any AI involved in it?

• 1:19:20Speaker 05

  Uh, not originally. Originally, this is just something, I think it's been around for 10 or so years. This was just a programming language that someone created. But the claim is, so when, when AI is putting out these proofs for things, they might release the Lean code or, you know, the Lean stuff to say, okay, not only did we write up a paper, but we also have a code version of this, which will show you that what we did is accurate, which is, you know, rigorous and, uh, holds up as a proof. So as part of the Navier-Stokes thing, OpenAI put out both a paper and their Lean stuff, their Lean code to show people that, yes, this actually works.

• 1:20:00Speaker 05

  This, this was like, it's not even the canary in the coal mine. I mean, where the canary's been dead for a while, I think, in this scenario. But, uh, this was like the first thing that I think made the wider world like really stand up and be like, oh, oh, something is, is happening here. So that was not the mathpocalypse.

• 1:20:19Speaker 01

  Yeah, that's just a preamble. We're not to the mathpocalypse yet.

• 1:20:23Speaker 05

  Yeah, that's the preamble. So the mathpocalypse is about a month later, and there had been rumors of this circulating for a while, is, uh, on October 6th, OpenAI put out some papers. How many papers would you guess OpenAI released at the same time?

• 1:20:39Speaker 01

  More or less 722? Yeah, I'd say 300.

• 1:20:43Speaker 05

  722.

• 1:20:44Speaker 01

  Oh, I got it.

• 1:20:45Speaker 05

  Very, very educated.

• 1:20:46Bob Novella?

  Good guess.

• 1:20:47Speaker 05

  So educated that it's almost like you knew the answer. Yeah. So OpenAI is like, hey, guys, guess what? Here's 700 papers. Have fun. And so, and many of these are, they put them on a GitHub, right? Which is a basically a code or, uh, paper repository anyone can access. And they put them out and they claimed that some of the results that, and these are all proofs or counter example, you know, uh, all sorts of, all sorts of stuff. And they put them out and they were like, okay, here's, here's a bunch of results. Some of these are important. Have fun. And some of these had Lean code, uh, released alongside them. Some of them didn't.

• 1:21:27Speaker 05

  But it caused the entire community of mathematicians worldwide to lose their goddamn mind. Because some of these papers are hundreds of pages long. They are not readable. Like you look at these papers and you're like, what, you know, the, the way a math paper is structured is there'll be, you know, like many scientific papers, an abstract, a central claim, like what is the theorem you're trying to prove? Okay, here's the theorem. And then the paper carefully lays out the argument for proving that there. These papers are incomprehensible.

• 1:21:58Speaker 01

  Is it like AI mathematical slop?

• 1:22:01Speaker 05

  Slop is, is I think the wrong word to use here, but yes, it's just not clear at the moment what is right or not about these papers. So I don't want to just, it's slop, not in the sense of they're probably all wrong or something like that. It's just.

• 1:22:19Speaker 01

  They're unreadable.

• 1:22:21Speaker 05

  They're just unreadable. Like, the goal, when you're writing a math paper, the goal is not just to show the result, but also to have it so that people can learn something by reading it.

• 1:22:31Speaker 01

  Right. So they needed a human to go through those 722 papers and make them, check them and make them coherent.

• 1:22:37Speaker 05

  They should have.

• 1:22:38Speaker 01

  And they didn't do that.

• 1:22:39Speaker 05

  They did not do that at all.

• 1:22:41Bob Novella?

  So Brian, these papers were even impenetrable by experts. They were like, would give up and say, I can't not doing this.

• 1:22:49Speaker 05

  Yeah, because the, you know, the thing about, and this is true with AI code as well. You're not prioritizing elegance, right? You just want to get the thing done. And that's why if you know how to code elegantly and you look at a lot of like vibe coded things, you look at them, you're like, what is that? You know, it's, it's a thousand lines of code for something I could do in 10 or something. It's, there's no, you know, it's essentially, they treat it like they have infinite resources. And they just find something that accomplishes the goal with little regard for how the goal is being accomplished.

• 1:23:26Speaker 01

  So this is vibe math?

• 1:23:29Speaker 05

  In a, in a sense, well, that's the, I think that is sort of the future of this at this point. This was, by the way, not on a publicly released model. I don't think it's, it's out yet. But the, you know, the, the issue here is, I think, immediately obvious, which is, what are you supposed to do as a mathematician now? Because is this just the future of math where people dump, you know, problems into an AI? And the AI spits out something and then you have to like pick apart and see if it's right.

• 1:24:03Bob Novella?

  What about, what was OpenAI's goal here? What were they trying to achieve?

• 1:24:08Speaker 05

  That, I think, depends on who you ask. If you ask OpenAI, they say they're just trying to solve problems, right? They're trying to, you know, increase the, you know, be at the forefront of human knowledge and show stuff. Some mathematicians regarded it as essentially a show of force. And, you know, kind of an intimidation tactic. Because it, it, it makes the entire human endeavor of mathematics, it's like almost impossible. It is not almost impossible. It is impossible to keep up with, with this sort of thing now.

• 1:24:47Speaker 05

  Because the, the kind of natural possible endpoint of this is, what are we going to see, like 10,000 papers a day? On this, just every day, you know, these come out and then other, the way mathematics works is you build on these results, right? So new results build on old results. So once you start proving these things, are we start, are we going to build on that? And it just accelerates. It's almost like, I hate to use this term, but it's almost like a mathematical singularity, like a Kurzweil type singularity, whatever.

• 1:25:18Speaker 01

  He said it. He used the S word.

• 1:25:20Speaker 05

  Yeah, I know, I know, I know. But it's like, you can see the acceleration happen here, right? Is my point. I do not want to claim I have any belief in the Kurzweil singularity whatsoever.

• 1:25:35Bob Novella?

  So is the idea that these, these papers would be out there that only other AIs can build upon? And they would just be useless for, for carbon based units?

• 1:25:45Speaker 05

  Well, that's the question. So there, there are a million questions. The biggest question is, is any of this right? And I think the odds are that some of it's right, some of it's not right. And it's going to take people to, to carefully check it. And that's why this, look, I am of the opinion, and I'll just state my bias here, I think this sucks. It is not how science should be done. It doesn't help anyone. It doesn't help people understand the stuff. Like, that is the goal. When you're a mathematician or a scientist, you want to do something in a way that's going to actually, you know, help people understand why the solution was interesting.

• 1:26:25Speaker 05

  If you're coming up with a proof, often the end result isn't as interesting as the technique you use to get there.

• 1:26:31Speaker 01

  Right.

• 1:26:31Speaker 05

  And, you know, this is why Fermat's Last Theorem is, is cool. Because, okay, Fermat's Last Theorem by itself is like, all right, you know, sure, that's an interesting result. But in the attempt to prove it over the course of hundreds of years, entire branches of mathematics were invented to, you know, to approach this problem. And it's, it's the journey, it's the process, it's not the end result. And this is the problem with so much AI is the, there's almost a total disregard for process.

• 1:27:03Cara Santa Maria?

  Yeah, but Brian, like, and I don't say this like cynically, but I hear you when you talk about the creative endeavor and you talk about like where you find meaning and purpose. But from a better humanity type perspective, funding or, or citizenry supporting mathematics as an academic endeavor, is there a difference in outcome?

• 1:27:28Speaker 05

  Yeah, at the moment, it is fully the, just abstract math proofs. None of these are going to be useful for anything at the moment. Now, that easily may change, but this is also, you know, so much of higher math is, you know, you do it because it's interesting, not because it's going to have any kind of real world application.

• 1:27:50Speaker 01

  It's like basic science research in that way.

• 1:27:52Speaker 05

  It's like basic science research in that way. And this is not to say that these techniques couldn't eventually yield something interesting. But if you look through these papers, you know, it's like, is this, is the Catalan constant, which, whatever, it's some, you know, mathematical constant, is that irrational? Okay, that, they, by the way, they claim it is. There's a longstanding conjecture. And does that buy you anything? No. You don't, it doesn't matter if that's true or not. In the sense, sorry, I shouldn't say it doesn't matter. It's like, when you know it's true, what do you do with that information? Nothing really.

• 1:28:25Cara Santa Maria?

  Yeah, but that's why it's not like basic science.

• 1:28:28Speaker 01

  Yeah, but well, in that you don't know, you, it may yield a tool that could then be useful, have some practical application. You don't know that ahead of time.

• 1:28:36Cara Santa Maria?

  You don't, but I think you assume that there will be a, an externalized value to the research. Otherwise, I don't think that you can justify spending taxpayer dollars on it.

• 1:28:49Speaker 05

  Well, but, okay, so that, that's, you know, I think that's kind of a matter of, of opinion to some extent. But with mathematics, especially mathematics is cheap, right?

• 1:28:58Cara Santa Maria?

  Yeah.

• 1:28:59Speaker 05

  Mathematics, you're not funding billion dollar labs. You're, you know, funding.

• 1:29:03Cara Santa Maria?

  One person. Yeah.

• 1:29:05Speaker 01

  Just a lot of whiteboards.

• 1:29:06Speaker 05

  It's a lot, yeah, it's a lot of whiteboards.

• 1:29:08Cara Santa Maria?

  Whiteboard budget is high.

• 1:29:09Speaker 05

  And computers. I mean, I, I, I am, I, I fully believe, you know, as a theoretical physicist, I am all about just do a thing because it's interesting and you don't know where it's going to go. None of my research I ever did in my career as a scientist would ever have any real world applications, and I never thought it would. I think it's important to do stuff that has real world applications, but I think it's fine to do stuff just because it might give you some insight into how things work.

• 1:29:38Cara Santa Maria?

  I agree. But sometimes those insights are fundamentally useful. And sometimes they're so esoteric that there isn't any sort of externalized benefit. Like, it's only helpful to a very small group of people who are just curious. And that's great. And I love that. But those are not the grants that get government funding.

• 1:30:00Speaker 05

  Well, sometimes they are.

• 1:30:04Cara Santa Maria?

  And I think that there's a, there are arguments to be had about that. And I think that that conversation is important for this larger AI conversation.

• 1:30:12Speaker 01

  But if I may, in this context of mathematical proofs, et cetera, it seems like the primary point of, of working out these proofs and conjectures, et cetera, is so that mathematicians can be better mathematicians. And if you have AI do it, again, you take away the journey and just cut to the end result in an incomprehensible sloppy way, what's the whole point of the endeavor in the first place? Is that kind of where you, where you're coming from, Brian?

• 1:30:41Speaker 05

  That is exactly where, where I stand. I mean, some, look, some of these results might be, you know, genuinely earth shattering if, if, if they come out. If, you know, if we're able to solve the Riemann hypothesis or something like that. But, uh, yeah, the whole point to me of doing math is you, what you learn along the way. And if some computer is just dumping garbage on you, that may, you know, maybe the end result is true. And you can't even understand what it's saying, what, what are we doing? That's not, you know, we're not learning a lot and it's not fun. Like, that's honestly, and what do you do? So this is a big question right now. What do you do right now?

• 1:31:23Speaker 05

  Let's say you are a, a math grad student and you just had your entire research program blown up by these results. Like, you know, I, I think there's a practical question too of, for people in the field, what are you supposed to do with this? It is true, whether, you know, these results are totally correct or totally wrong or somewhere in between. And the answer has to be somewhere in between.

• 1:31:51Speaker 05

  The game of doing mathematics is completely different now than it was six months ago.

• 1:31:58Cara Santa Maria?

  Yeah, but isn't that the way that this happens? And I know it's not a great analogy, but when we figured out how to not have to use coal, it screwed a lot of people out of work. But we have to adapt to that. And so my question is, is there a version of this sloppy information that allows graduate students today to dig through and to say, okay, maybe there are situations where we now have the endpoint and now we can work backward to understand the journey, but just from a different part of this sort of number line?

• 1:32:37Speaker 05

  So there's definitely, I mean, this is what a lot of people are doing right now with these papers, is they're going through them to try to understand them. But the question is, what are you learning along the way?

• 1:32:50Cara Santa Maria?

  Right. And we don't know yet. I mean, that's the point I'm trying to make. I'm not saying like, you know this world. And so I'm trying to ask, is there a world where, because this is the reality of what we're living in, we can change the way we learn?

• 1:33:05Speaker 01

  Can I introduce one idea? The only thing I found, if I explore the issue of what are the practical ramifications of this? And the one thing that comes up is cryptography. Not any of the specific findings in this dump.

• 1:33:19Speaker 05

  In fact, there are no, like, there's no cryptography in this, which they are obviously very scared of making inroads in.

• 1:33:28Speaker 01

  But could this process be a tool that is used to break encryption or to create unbreakable encryption, at least for now?

• 1:33:37Cara Santa Maria?

  Well, and that's a different question. It's an important question, but that goes back to my first thing, which was like, are there end user kind of tangible benefits? But what I'm saying is-

• 1:33:45Speaker 05

  There will be at some point, yeah.

• 1:33:47Cara Santa Maria?

  Because I don't know how to do higher level, complex math. Is there a world in which the journey can go in the other direction? Where it's not about starting from the principles and trying to solve the proof and learning along the way, but it's trying to look at a solution and understand how the machine got there. Exactly. Is that not just as valuable?

• 1:34:09Speaker 05

  I mean, that is what a lot of people do. But I think the question is also, so if you want to do that, let's say in these papers, which is what people are currently doing with these papers, these papers are, I mean, truly, they are incomprehensible to read. It's not like, it's not English, it's written in English, but it is not written in the way mathematicians would communicate to each other. They are hundreds of pages long, in some cases.

• 1:34:41Speaker 01

  I'm a little surprised at that, I have to say, because isn't that the one thing that LLMs are good at?

• 1:34:46Cara Santa Maria?

  Yeah.

• 1:34:47Speaker 05

  But the thing is, they found, it's not the path to get to the solution, they just found a path, right? So this is the issue, is it's not that they're bad at condensing things or whatever. They just, and look, I don't know how the process worked for any individual result. They just kind of followed, they threw this at a bunch of, you know, a swarm.

• 1:35:10Speaker 01

  Circuits.

• 1:35:11Speaker 05

  And this is the path that they found, and then they did it, and it was done.

• 1:35:14Speaker 01

  And so the path, what the process was inelegant, so the communication about it is also inelegant.

• 1:35:20Bob Novella?

  Correct. It's probably, it's as simplified, it's as simplified as it could make it, potentially.

• 1:35:25Cara Santa Maria?

  No, I think these are all assumptions. I don't know if the path was inelegant or not.

• 1:35:31Speaker 05

  The path almost, I think the path almost has to be inelegant for, looking at these papers, there is nothing elegant about any of these.

• 1:35:39Cara Santa Maria?

  Yeah, but is this the most, I mean, but to Bob's point, I think that's easy to say, but if we're talking about something that's never been solved before, maybe this is the most elegant way to solve it. Maybe it's not.

• 1:35:51Speaker 05

  That's certainly, that's certainly possible, but there's an elegance of, there's two kinds of elegance, and these are all subjective, of course. There's elegance of the solution, and then the elegance of the communication of the solution, right? So you might have, I mean, you might have a problem where the steps to take it are, it's this beautiful, kind of artful proof, but it's communicated in a terrible way. That happens all the time, right? So what I would guess from most of these papers is it is both, that the path and the communication are not done with elegance, and more specifically, with the communication, it is not done in such a way that humans are going to read this and learn from it. Right?

• 1:36:36Bob Novella?

  So this is absolutely reminiscent of Ted Chang's super short story, The Evolution of Human Science, where augmented humans become so intelligent, they are making incredible advances in science, but it is by definition uncommunicatable to humanity. But humanity is seeing the results, the fruits of this science, but they can't understand it at all. So they are basically removed from also, from almost all avenues of science, but they are benefiting from the fruits.

• 1:37:07Speaker 05

  Well, yeah, the other thing about this is that it's not yet clear to people if these results are correct.

• 1:37:15Bob Novella?

  Right, that's it.

• 1:37:16Speaker 05

  And that's something we can't lose sight of at this point in the history of this. Maybe all of these are correct. That seems extremely unlikely to me at this point. And people are going through and trying to figure them out. But, look, I mean, I doubt you guys have ever read a 200-page math paper before. To do this at a level where the mathematical community is going to accept it is very hard, even when the paper is well-written, right? Because the whole point of math is that it has to be rock solid, that there is no wiggle room. And that is, if the paper is badly written, this becomes, it's not just a chore, it's

• 1:37:55Speaker 05

  like, this might take you, it might take you a career to go through a 200-page paper to really check, right? I mean-

• 1:38:06Cara Santa Maria?

  And maybe that is how math is changing. I mean, and I know it's monumental and overwhelming and insane, but I guess, how much of, there's the same, same, but different conversation here, which is every time there are massive innovations in technology, there's this kind of resistance, because it makes it so that we have to completely do a paradigm shift in how we think and undo a lot of the things that we've always been doing. But yes, AI is fundamentally different also. It's not a horse and buggy moving into a car.

• 1:38:38Speaker 05

  Yeah, and I would say, I mean, the thing that is necessary here, it's exactly what I just said, which is if we are going to, clearly, clearly, first of all, this is a sea change in how people do mathematics from here on. Matt Strassler, who's a theoretical physicist, in his blog said, this essentially means there are no more Newtons or Einsteins out there, because no human will ever be able to keep up with this. And I think he's not wrong, honestly, because the forefront of mathematical, of math and theoretical physics, most likely, will be AI-assisted. But the thing that is especially important for math specifically is this idea of rigor.

• 1:39:21Speaker 05

  And until we're 100% sure that this process is being done rigorously and we can totally trust these results, you know, and yes, an AI reading these papers doesn't care about how elegant it is, it just wants to know the result. But if you're going to then take this result and build other math results on it, you risk creating a house of cards where suddenly, okay, that, actually, that wasn't a firm result, but we're going to treat it like it's proof. And the history of mathematics is rife with these things, where people thought they were proving something, but, oh, there was a mistake in this result they used, okay, you have to go back, and then we have to fix that, and sometimes it's fixable, and sometimes it's not. So the danger here-

• 1:40:03Cara Santa Maria?

  But that's not a failure of the AI, that's a failure of the human component of all of this.

• 1:40:07Speaker 05

  No, it's a failure, I mean, if the AI is putting out bad results, then that's a failure of the AI, right?

• 1:40:13Cara Santa Maria?

  No, no, no, but whether or not the community uses those bad results to then-

• 1:40:17Speaker 05

  I'm not talking about the community, I'm talking about other AIs.

• 1:40:19Cara Santa Maria?

  You're talking about the iterative way that the AI is going to keep solving results based on old results.

• 1:40:23Speaker 05

  That's right. And so if you, you know, this is what Lean, by the way, is supposed to be helping with, is if you can put it into Lean, and then Lean rigorously checks it, it's like, that's a way of saying, okay, that is, yes, we can trust this result.

• 1:40:39Speaker 01

  Can I say, so to be charitable, this feels like a stunt to me, right? Just dumping this 722 papers on GitHub and say, hey, what do you make of this? But maybe it was useful in that it's prompting this very discussion. The idea that, right, we have to talk about how to properly use AI, how to optimally use AI, what the downsides would be, what the problems would be, what- The opportunities are and what the future is going to look like. For example, here's this boatload of crap. What are we, what is, what are we going to do about this?

• 1:41:13Speaker 05

  Yeah, I agree. And it has inspired a tremendous amount of discussion. I mean, every mathematician in the world is talking about this. And you'll find every reaction, including from the same person on different days, right? One guy said he, you know, when the, the day this came out, he, he woke his daughter up because he was crying. Like at home, which is, I think, a completely, I mean, you know.

• 1:41:36Speaker 01

  So the good cry or bad cry?

• 1:41:38Speaker 05

  Like, bad. Like, this is, it's, it feels like a, a, a real inflection point in the history of math.

• 1:41:46Speaker 01

  Yeah.

• 1:41:47Speaker 05

  There's advisory committees that, you know, of world leading mathematicians who are, you know, putting out recommendations for how we approach this in the future. It's certainly true that I don't think OpenAI want itself any friends by, by doing this. I mean, people seven, I put out a couple of papers. It's like, okay, the, you know, even the Navia Stokes stuff was controversial enough when it came out. Then to do this, it's just like a big middle finger to, to, to the whole mathematical community. And I'm sure you'll find people who are like, oh, my God, this is the best thing that's ever happened to math.

• 1:42:22Speaker 01

  Right.

• 1:42:23Speaker 05

  The, the people I know and most of the communication I'm seeing is people who are like, whoa, we got to figure this out now because we're about to enter an era where things start going so fast that we can't keep up. And if we don't establish ground rules now, we're hosed.

• 1:42:40Speaker 01

  Right, right. So this is obviously a story that we're going to need to follow. And it's part of the bigger story of how AI is disrupting everything that we do, which we have to follow. Just everything is fascinating. But everyone, let's take a quick break from our show to talk about one of our sponsors this week, Monarch.

• 1:42:56Jay Novella?

  So when I was a teenager, I got my first credit card. And of course I immediately got myself into debt and I wasn't making enough money to pay it off. What do you do?

• 1:43:06Cara Santa Maria?

  I, I'm sure you wish that you had Monarch at the time. They have this really cool debt pay down simulator right there in the app. It helps you visualize and track your progress so that you have the exact debt free date based on your actual balances, your actual payments. And it's so satisfying to watch the timeline move. They absolutely break the debt cycle. They turn that, you know, gloom into a plan and they take the burden of tracking off of your plate so that you are much more empowered. You can even ask Monarch's AI assistant anything about your finances and it'll tell you what you need to know.

• 1:43:43Bob Novella?

  So write your own money story with Monarch. Use code SGU at monarch.com to get your first year of Monarch Core half off at just $50. That's 50% off your first year at monarch.com with code SGU.

• 1:43:56Speaker 01

  All right, guys, let's get back to the show. So Brian, since we got you here.

• 1:44:00Speaker 05

  Yes.

• 1:44:00Speaker 01

  Why don't you hang out with us and we'll go on to science or fiction? What do you say?

• 1:44:04Speaker 05

  Let's do it.

• 1:44:05Speaker 01

  All right.

• 1:44:08Speaker 00

  It's time for science or fiction?

• 1:44:17Speaker 01

  Each week I come up with three science news 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 inspired by the midterms. The theme is democracy. All right.

• 1:44:31Bob Novella?

  That's still a thing?

• 1:44:32Speaker 01

  All right, it's still a thing. All right, here we go. Item number one, according to the Global Democracy Index, only 6.6% of people are living in a full democracy, while 39.2% are living under authoritarian regimes. Item number two, ancient Athens is the first recorded democratic government with most public offices filled by popular vote. However, only free male adult citizens could vote. And item number three, no form of the word democracy appears in the US Constitution or any of its 27 amendments. All right, Brian, as our guest, you can go first.

• 1:45:11Speaker 05

  Okay, so I'll upfront say, I don't know anything about any of these, but let's try to, to think them through. Number one, according to the Global Democracy Index, 6.6 are living in a full democracy. But 39 points, let's say 39.2, living under authoritarian regimes. Are, are you willing to define full democracy?

• 1:45:36Speaker 01

  So I'll tell you this, there were four categories in this index. Full democracy, flawed democracy, hybrid, meaning there's some democratic, some authoritarian elements and authoritarian.

• 1:45:49Speaker 05

  So this is not about the system of, so what I'm trying to get at here is in the US, we have the Electoral College. So that is, that is a type of democracy.

• 1:46:03Cara Santa Maria?

  No, I think the Global Democracy Index is, is measured on outcomes, not measured on how things are structured.

• 1:46:11Speaker 01

  If you're interested in that topic, you could listen to the Political Reality Podcast where we talk for an hour about how to measure democracies, including the Global Democracy Index. But basically it is, yeah, so a full democracy basically means it's a full representative democracy. You vote, you know, you vote for people who represent you, basically.

• 1:46:28Cara Santa Maria?

  And in effect, that is how it is.

• 1:46:30Speaker 01

  Right, with no major flaws. Flawed democracy are for countries that have it, they're a democracy, but they're fundamentally flawed in some way. Hybrid means, yeah, you might vote, like you vote in China, whatever, but it's basically authoritarian. And then versus full authoritarian regimes.

• 1:46:46Speaker 05

  Yeah, okay. So I'm going to say then, and this is people, it's not of countries, right?

• 1:46:52Speaker 01

  600 people are living under democracy or under authoritarian regimes.

• 1:46:57Speaker 05

  Yeah, so, okay, so I think item one, 6.6%. I have no idea how to base that number on anything. The 39.2% under authoritarian sounds sort of plausible because, you know, China, of course, is what? An eighth of the world, something like this. And you could imagine those numbers adding up pretty quickly. So I'm going to say, I don't know, the 6.6 I have no idea about, but the other number sounds okay. So I'm going to say, number one is science. Number two, ancient Athens is the first recorded democratic government with most public offices filled by popular vote. However, only free male adult citizens could vote.

• 1:47:41Speaker 05

  Certainly if you had, if this wasn't science or, science or fiction, I would say, yeah, that's what we hear. But now you've got me in my head about this one. That is certainly what I have as the first. I think the second part about the free male adult citizens seems right. But the first recorded democratic government, recorded is also doing a bunch of work here. I'm actually going to say that, it seems like there's a lot of wheelroom there, but I'm going to say that's one science. And so I think I'm left with item three. Now, this feels like a weird right wing talking point, but I can't remember any specific instance

• 1:48:23Speaker 05

  of the word democracy appearing in the Constitution.

• 1:48:28Speaker 01

  So remember, it's any form as well. So that could include democratic, democratizing, yeah, whatever, yeah.

• 1:48:34Speaker 05

  Okay, I'll say item three is fiction.

• 1:48:36Speaker 01

  The, the democracy in the Constitution.

• 1:48:37Speaker 05

  Yeah, the democracy in the Constitution.

• 1:48:39Speaker 01

  Bob, you're next.

• 1:48:41Bob Novella?

  Yeah, democracy could absolutely be fiction. The, the second one here with, with Athens and popular vote and all that, that sounds reasonable. Although I'm sure there's, there's an easy way to, to say, oh, no, but this, oh, okay. So like, but the, and the first one, though, is bugging me, 6.6% living in full democracy? That, that's what's getting me, that seems too low. So I'm just for that reason alone, I'm going to say that's fiction.

• 1:49:06Speaker 01

  Okay, Jay?

• 1:49:07Jay Novella?

  Yeah, I mean, I've been thinking about that one the whole time as everybody else was talking. I'm like, well, how are you defining democracy? I know, I, I can tell you my definition, but it's probably a very simplified definition. It's voting rights. It's like, isn't it revolve around, you know, being able to cast a vote? I think modern version of democracy has a lot of other things like, you know, the rule of law and all sorts of stuff like that. But is that what you're talking here or you can't answer that?

• 1:49:32Speaker 01

  Well, it's, you're going third, so I don't want to give you too much information. So the, the global, this is a specific group using a specific measure of democracies. I'm not going to get too into the weeds yet about how they're doing that. But it's, it's, it's generally accepted as one of the, if not like the best one out there, right?

• 1:49:51Jay Novella?

  Okay.

• 1:49:52Speaker 01

  This is not some obscure thing. This is like, all right, this is what everyone looks to as like one, one, a really good measure of what is a democracy versus not.

• 1:49:59Jay Novella?

  I mean, it wouldn't surprise me if the percentage was that low, particularly with, you know, the ways that I would qualify it. Okay, let me go on to the second one. Athens is the first recorded democratic government with most public offices, popular vote. That does not feel right to me. Let me go down to the last one, no form of the word democracy appears in the US Constitution. Yeah, I was thinking about that, too. Like, I would think they, they may have referred to the US as a republic or something along those lines. But I don't know about, yeah, straight up saying democracy in there. The, the, the one about Athens, I don't think that, that, that Athens voted the way we vote today. I thought they use a, there was something else, or some other system they use. I don't remember exactly what it was, but I think that was the fiction.

• 1:50:44Speaker 01

  Okay, Kara, they're all spread out.

• 1:50:46Cara Santa Maria?

  Great.

• 1:50:48Cara Santa Maria?

  Okay, so the Global Democracy Index says 6.6% of people are living in a full democracy. That does not surprise me at all. I would think that number would be very small. And, you know, almost 40% under authoritarian regimes. Brian, I completely agree with you. Like, you know, China, we've got like Russia, we've got, you know, huge nations and then a lot of smaller countries. So of course, I think that that that doesn't seem unreasonable to me. I mean, maybe the numbers are a little off, but they're not in order of magnitude off. So I just don't see how that one could be the fiction. So that leaves ancient Athens versus no form of the word democracy appearing. I feel like Steve, you're trying to get

• 1:51:30Cara Santa Maria?

  us with the no form of the word democracy appearing. Like, maybe it appears somewhere in an amendment, or maybe the word democratic is like, you know, one time and people just love to say, oh, it doesn't say democracy anywhere. So that could be the fiction, but also, the ancient Athens one is very complex because in order for this to be science, it has to be the first recorded democratic government, it has to have most public offices filled by popular vote, and it has to have only free male adult citizens who can vote. So if any of those things are not science, it nullifies this and makes it fiction. So I think because it has so many things that you could have switched up there, I'm leaning

• 1:52:15Cara Santa Maria?

  towards ancient Athens being the fiction. And my guess is that it was intentional that the word democracy was not used. My guess is that that word is loaded and it to some requires a very specific way of enacting the government. Yeah, I think it's going to be ancient Athens, but I have no idea what you are screwing with there. Wouldn't it be crazy if it wasn't ancient Athens at all?

• 1:52:44Speaker 01

  All right. So you guys are spread out, so I'll just take these in order. I remember one, according to the Global Democracy Index, only 6.6% of people are living in a full democracy with 39.2% are living under authoritarian regimes. Bob, you think this one is the fiction. Everyone else thinks this one is science. And this one is science. All right?

• 1:53:05Cara Santa Maria?

  Yep.

• 1:53:06Speaker 01

  Now, yeah, so 6.6% under full democracy, 38.4% under flawed democracies, 15.7% under hybrid regimes, and 39.2% under authoritarian regimes. Now, based upon their breakdown, the United States is a flawed democracy. Oh, my God.

• 1:53:24Speaker 05

  Yes, I've heard that before.

• 1:53:26Speaker 01

  Yeah, we are not living in a full democracy.

• 1:53:28Speaker 05

  Who has the full democracy? Is it New Zealand?

• 1:53:31Speaker 01

  Here are the best scores. Norway, New Zealand, Sweden, Iceland, Switzerland, Finland, Denmark, Ireland, and the Netherlands. So those are the, those are the full democracies.

• 1:53:40Speaker 05

  Pretty small countries.

• 1:53:41Speaker 01

  Yeah, yeah. So what, again, I encourage you to listen to our political reality episode on how to measure democracy. It's fascinating. Jay hit upon one thing. So you think voting, right, is sort of central to a democracy, but there's a lot of caveats, right? Because you could vote under an authoritarian regime.

• 1:53:59Speaker 05

  Right.

• 1:53:59Speaker 01

  Right? So the vote, and Andrea, you know, Jojo and Joy, who co-host the show with me, just went over this on our live stream, that you have to not only vote, but the vote doesn't have a predetermined outcome.

• 1:54:15Speaker 05

  Right.

• 1:54:16Speaker 01

  And there has to be participation in the vote, right? You could be a vote, but only a small percentage of people get to vote. That's not really a full democracy. And the results have to stick, right? If the results always get overturned to whatever the authoritarian guy wants, then it's, then it's not a real democracy.

• 1:54:33Cara Santa Maria?

  Ah, so if this Athens one is science, then even they weren't a full democracy, according to their rules.

• 1:54:40Speaker 01

  Right. So, right, they were not. They would not be, Athens would not be designated a full democracy based upon this index because, actually, we'll get to that in a second. That doesn't mean they weren't considered to be the full democratic government.

• 1:54:54Cara Santa Maria?

  Right, back then.

• 1:54:55Speaker 01

  It's not full, not necessarily full democracy. Let's, let's go on. Then we could talk more about, I don't want to say too much and give away the answer yet. So let's just talk, go on. Number two, ancient Athens is the first recorded democratic government with most public offices filled by popular vote. However, only free male adult citizens could vote. Both Kara and Jay think this one is the fiction. Bob and Brian thought this one was science. And this one is the fiction. It is the fiction. So which part of it is the fiction, though? No idea.

• 1:55:26Jay Novella?

  I don't think they voted like that. I think they, wasn't there some other form of, there's something there, Steve.

• 1:55:31Speaker 01

  Yeah, you're dancing around it. All right. So was ancient Athens the first recorded democratic government? Yes, it was. In fact, in many ways, it was, it's often put forward as a quote unquote pure democracy in that the citizens could vote, they weren't voting for representatives. It wasn't even a representative democracy. They voted on laws and actions and what the government would do. They voted on everything. The, the bodies were comprised of, like, any citizen can show up and be part of the vote of the, the body that determined what the government's going to do.

• 1:56:04Cara Santa Maria?

  Not any citizen.

• 1:56:05Speaker 01

  Any citizens, any, any free male adult citizens. Yeah.

• 1:56:08Cara Santa Maria?

  Exactly. I was like, hey, what about?

• 1:56:10Speaker 01

  Women were disenfranchised. There's a lot of slaves in Athens, so they couldn't vote, obviously. Obviously, you had to be an adult. Now, so that out of, that was basically 10% of the population. So that's why there were not a lot of participation. Only about 10% of the population could actually participate in the government.

• 1:56:26Cara Santa Maria?

  So it's by definition, actually, it's, it's actually kind of effed that we sort of just look past that.

• 1:56:32Speaker 01

  Yeah.

• 1:56:32Cara Santa Maria?

  Because it is deeply undemocratic, actually, when you think about it.

• 1:56:36Speaker 01

  The part here that is wrong is most public offices filled by popular vote. So Jay, you're correct. They did a lot of voting, but that's not how they filled most of their pop, of their public offices. How did they do it?

• 1:56:49Jay Novella?

  Oh, Christ. It was, wasn't they, didn't they do the?

• 1:56:52Cara Santa Maria?

  Did they just?

• 1:56:53Speaker 05

  Was it appointed by?

• 1:56:54Cara Santa Maria?

  Yeah, were they appointed?

• 1:56:55Speaker 01

  It was a process called sortition. Does that give you a clue?

• 1:56:58Jay Novella?

  Yeah, it does.

• 1:56:59Cara Santa Maria?

  Was it like just random?

• 1:57:01Speaker 01

  Random, chosen by random lottery.

• 1:57:03Speaker 05

  Oh.

• 1:57:04Cara Santa Maria?

  Oh, my God. So if you were an adult male, you had the privilege of voting, but you also may be saddled with government office.

• 1:57:10Speaker 01

  So, yeah, so they, they, yes, you sort it, and they had like this marbles, you know, that would, they would use to determine who was chosen to fill all these public offices. They thought if you're running for office, then that would, over time, cede too much powerful to wealthy elites. And they didn't want that to happen. They also had, most of these offices had one-year terms because they didn't want people to be in there long enough to accumulate power. And third, they did something very interesting. If somebody was amassing too much power, the, the free male adult citizens, right, the people who could vote, could just vote them to be banished from Athens for 10 years because they just,

• 1:57:55Speaker 01

  they got too powerful, you're out of here, just as a check against anyone consolidating too much power.

• 1:58:02Jay Novella?

  Yeah, that sounds like something we got to think about here.

• 1:58:04Cara Santa Maria?

  Well, it's so interesting because in a lot of ways, it's so cool. And yet, like, how tone deaf.

• 1:58:09Speaker 01

  Yeah.

• 1:58:10Cara Santa Maria?

  To think that the way to prevent people, you know, to be so afraid of the wealthy elites having power, when it's the, it's only the wealthy elites. It's 10% of the public.

• 1:58:22Speaker 01

  Who could participate in the first place.

• 1:58:23Cara Santa Maria?

  I know.

• 1:58:24Speaker 01

  This was the elite of the elites, Kara. The 1% was worried about the 0.1%.

• 1:58:30Speaker 05

  Right, right.

• 1:58:31Speaker 01

  All right, let's move on. So no form of the word democracy appears in the US Constitution or any of its 27 amendments. Brian, you thought this one was the fiction, but this is correct. This is true. And you guys are also correct. This was deliberate. The founding fathers were fearful of the tyranny of the majority, right, of pure, direct democracy. That's why they, they specifically referred to the United States as a republic, or at most a democratic republic. So they avoided the word democracy. That was by design. I think it just didn't have to appear at any of the amendments, but it has not appeared in any of the, of the amendments. Now, this often leads people to say, America is not a democracy,

• 1:59:16Speaker 01

  right? And I've heard people say, it's not a democracy, it's a republic. That's a very common thing to say on the internet.

• 1:59:21Speaker 05

  Yeah, it's like a trope.

• 1:59:22Speaker 01

  Yeah, it's not true.

• 1:59:23Cara Santa Maria?

  That's a type, that's a type of democracy.

• 1:59:25Speaker 01

  Yeah, you can be a, we are a democratic republic, right?

• 1:59:28Cara Santa Maria?

  A flawed one.

• 1:59:30Speaker 01

  We're a flawed democratic republic, but we are a democratic republic. We do meet all the definitions of a democracy. Again, if you tease it out further into a full democracy versus flawed democracy, by the, that one measure we're talking about, we actually, we have dropped down into the flawed democracy category. But we are a democracy. We are a democratic republic, right? And I think we just flexed our democratic muscles at the midterms. We showed that, you know, you couldn't predict the results ahead of time. There was a vote, it was a free and fair election, and those results are going to stick. And it resulted in a shift, in a real shift in power. That's what democracies are. That is a democracy.

• 2:00:14Cara Santa Maria?

  I mean, it was kind of not a free and fair election, but we overcame that.

• 2:00:19Speaker 01

  Yeah, that's true. Gerrymandering, by the way, is something that knocks you down, right? And as it should, because gerrymandering is a way of shift, taking the power away from the voters, away from the people. It's the politicians choosing their voters rather than voters choosing their politicians. The Electoral College also takes us down a notch, because it means that the voting power is not for president and vice president is not evenly distributed. And that-

• 2:00:43Cara Santa Maria?

  Right. It's not fully represented.

• 2:00:45Speaker 05

  But people who say, oh, we're not a democracy, that to me always feels like code for, I know most people disagree with me on this, but-

• 2:00:52Speaker 01

  Yeah.

• 2:00:53Speaker 05

  You know, that's just a way of saying, yes, I realize that I am not in the majority on this particular opinion. And then they say that.

• 2:01:00Speaker 01

  It's that, or they want to sound erudite, you know, but it really just shows that they don't know what they're talking about.

• 2:01:06Speaker 05

  Yeah.

• 2:01:06Speaker 01

  So yeah, there's a lot of interesting, you know, things coming out of this. And it's very complicated, the whole political science thing. It's been a lot of fun doing the show with Andrea. I've learned a ton. This is just sort of some of the things that we talk about, that when you do a deep dive on, it's very, very interesting.

• 2:01:24Bob Novella?

  Did you just use political science and the word fun in the same sentence?

• 2:01:27Speaker 01

  Yeah, I have. This is not the first time either. It is. It's a science. It's a legit science. There's a lot of overlap with, because it's dealing with people. There's also a lot of-

• 2:01:37Cara Santa Maria?

  Yeah, with psychology.

• 2:01:38Speaker 01

  Cognitive, sort of theoretical overlap with both economics and medicine. So I appreciate it from that perspective as well. So there's a lot of good medical analogies I could draw. It's like, yeah, that works just like it does in medicine because of the same problem. You're dealing with people, you know? And it's hard to do science when you're dealing with such variable units, you know?

• 2:01:57Speaker 05

  It's also just amazing to see all the different systems that countries all over the world come up with to run themselves, right?

• 2:02:05Speaker 01

  Yeah, a lot of different trade-offs and a lot of different answers to those trade-offs. Exactly.

• 2:02:09Cara Santa Maria?

  And how, yeah, it's not one size fits all because not every country has the same parameters.

• 2:02:14Speaker 01

  Yeah.

• 2:02:14Speaker 05

  That's right.

• 2:02:14Speaker 01

  Yeah. Okay, Bob, do you have a quote for us?

• 2:02:18Bob Novella?

  Okay, my quote, guys, for today is from Leonardo da Vinci. "It is not the strongest of the species that survives, nor the most intelligent, but the one most responsive to change." Steve, expound on that.

• 2:02:31Speaker 01

  Well, I'll tell you why I like this the most. That's a very evolutionary perspective. And he is saying this a long time before Darwin. You know, he really, right? Because the whole point of adaptability, adaptability is everything.

• 2:02:46Bob Novella?

  Like 300 years before?

• 2:02:47Speaker 01

  Yeah, centuries before.

• 2:02:48Cara Santa Maria?

  It's like defining fitness before we had the word for it.

• 2:02:51Speaker 01

  Yeah.

• 2:02:52Bob Novella?

  Incredible.

• 2:02:53Speaker 01

  Very, very prescient. Which is the sad thing about da Vinci. I've been fascinated with this guy for my whole life. You guys remember, our father had a giant book of Leonardo da Vinci. And I read through it.

• 2:03:04Bob Novella?

  Do you know where that book is?

• 2:03:05Speaker 01

  I do.

• 2:03:06Bob Novella?

  It's two feet from me right now. I grabbed that.

• 2:03:10Speaker 01

  Good for you.

• 2:03:10Bob Novella?

  What an awesome book.

• 2:03:11Speaker 01

  So, but anyway, the thing is, he didn't really engage with the community.

• 2:03:17Bob Novella?

  No.

• 2:03:17Speaker 01

  The intellectual or the scientific community. So he just labored away by himself, doing all this great stuff, like this quote. And it didn't affect the culture or society. It was discovered after the fact. And it was like, oh, look what this guy was doing, you know?

• 2:03:33Bob Novella?

  In a lot of ways, a failure, right? That's a failure right there.

• 2:03:37Cara Santa Maria?

  Well, but we don't know, because we don't know contextually if the community would have even listened or if they would have fully ostracized him. Yeah.

• 2:03:45Speaker 01

  But we do know he didn't do the one thing that was at least necessary. And that was engage with the relevant community, which is so important.

• 2:03:53Cara Santa Maria?

  But we don't know if he tried to and realized how much it would get in the way and just stopped engaging after that.

• 2:04:00Speaker 01

  I also wonder if he was just on the spectrum.

• 2:04:02Cara Santa Maria?

  Yeah.

• 2:04:04Speaker 01

  Who knows?

• 2:04:05Speaker 05

  There was also that thing back then about scientists worried that other scientists would steal their results. Right? And so there's all these amazing stories about people, you know, encoding their results and then putting them in a secret form, you know, out there in the public. And then they could point to it later and be like, oh, this is actually what I meant by this. So I think some of those people were just scared to engage with other for fear of getting scooped.

• 2:04:28Speaker 01

  Right, right, right.

• 2:04:29Bob Novella?

  But imagine if he fully engaged and-

• 2:04:32Speaker 01

  Yeah. You can only imagine.

• 2:04:33Bob Novella?

  It was accepted. I mean, the world could be different.

• 2:04:38Speaker 01

  It would be.

• 2:04:39Bob Novella?

  Just from that.

• 2:04:39Speaker 01

  Yeah.

• 2:04:40Bob Novella?

  Absolutely.

• 2:04:41Speaker 01

  All right. Thanks, Bob. Brian, thanks for joining us. It's been a lot of fun.

• 2:04:44Speaker 05

  Yeah, man.

• 2:04:45Bob Novella?

  Yeah, thanks, folks.

• 2:04:46Speaker 01

  Thanks to all of you for joining me again this week.

• 2:04:48Cara Santa Maria?

  Thanks, Steve.

• 2:04:49Speaker 01

  Cheers, Steve. And until next week, this is your Skeptic's Guide to the Universe.

• 2:04:56Speaker 01

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