This week’s episode is a reminder that the world is held together by three things: ancient engineering that still embarrasses us, human psychology that refuses to be tidy, and animals that behave like they’re actively trying to ruin your data set.
Roman Concrete Is The 2,000-Year-Old Flex We Still Can’t Beat
Let’s start with the Pantheon. It’s been standing there for roughly 2,000 years with a massive dome made of unreinforced concrete, and it’s still doing better than half the stuff built last decade. The Romans weren’t just lucky. They had a recipe.
The mix involved volcanic ash (pozzolan) and quicklime, which kicked off chemical reactions that helped the concrete keep strengthening over time. Recent research suggests the structure can even “self-heal” in a way, with minerals like calcite forming and filling small cracks as water moves through. It’s the kind of discovery that makes modern builders both excited and slightly offended, because the takeaway is basically: the ancient guys figured out something we’re still trying to reverse engineer.
The Dark Factor
Then we move from ancient materials to modern minds. Psychologists have been looking at whether there’s a single underlying “dark factor” that connects traits like narcissism, psychopathy, and Machiavellianism. Not as separate little quirks, but as different expressions of the same core tendency.
The unsettling part is that studies suggest this D factor can be pretty stable over time. It’s not just a ‘bad week’, it can be a consistent pattern, shaped by both personality and environment. Which raises the obvious question: if you can measure it, can you predict behaviour better - and then actually do anything useful with that knowledge - or do you just end up with a neat label for something humans have always known exists?
Pigeons And The Joy Of Not Following The Rules
And finally, pigeons. The birds we all pretend not to notice until they’re directly in our path. Pigeons have been used in psychology research for ages because they’re great for studying reward-based learning and decision-making. But what makes them especially interesting is that they don’t always behave the way you’d expect.
Put pigeons in a chaotic, unpredictable environment and they don’t melt down. They keep going. They’ll try patterns, abandon them, try new ones, and generally operate right on the edge of randomness. It’s frustrating if you want clean results, but it’s also a pretty good model for real life, where the rules change constantly and the “right answer” isn’t always obvious.
So that’s the week: concrete that heals itself, a dark personality thread that might explain a lot of human behaviour, and pigeons casually demonstrating how to function in a messy world. It’s three very different stories, but they all point to the same thing: durability, behaviour, and adaptation are rarely simple, and the best lessons often come from places you wouldn’t think to look.
CHAPTERS:
00:00 Intro
01:58 Dark Triad Explained
02:51 Measuring the D Factor
08:17 Ethics and Pre-Crime
09:12 Roman Concrete Mystery
11:56 Self-Healing Calcite
13:02 Roman Self-Healing Concrete
14:10 Plant Based Diet Confusion
17:25 Pigeons In Psychology
19:32 The Edge Of Chaos
-
Will: [00:00:00] It is time for a little bit of science. I'm will grant an associate professor in science communication at the Australian National University.
Rod: and I'm Rod Lambert. I'm a 30 year science communication veteran with a mind of a teenage boy.
Will: and today. Well, we've been at the Beach Large Hadron Collider.
Rod: Yeah. Drinking in swimming cocktails. In fact, I'm still there. I dunno where you
Will: are of, uh, isotopes and science and stuff like that. So rather than giving you our regular pile of science, we've
Rod: saved you something delicious. We've been scrolling away little snippets throughout the millennium and you're gonna get a bunch of those right now.
Rod: It's gonna be fab.
Will: Enjoy.
Rod: So 100 plus years ago, psychologist Charles Spearman, he was really into trying to measure intelligence, and he did in fact measure intelligence in a bunch of ways. You've probably heard of him.
Will: Spearman's ...
Rod: G.
Will: Yeah, that was it.
Rod: So he tested people on a range of types of intelligence. So he broke it into, ooh, one, two, [00:01:00] three, a bunch.
Rod: Things like fluid reasoning, crystallized intelligence, working memory- Okay ... processing speed. There's all these different ways to consider notions of intelligence.
Will: Why didn't they ever go with, like, streetwise?
Rod: Oh, yeah, yeah. They got, have you got your shit together? Could you-
Will: Like, no, I feel like that's, that's a known form of
Will: Yeah, streetwise.
Rod: I, I think that's not wrong. Like cultural or social-
Will: Book smarts. Book smarts. You got book smarts.
Rod: Your book smarts and your street dudes.
Will: That's the, that's the number ... They're, they're big categories of intelligence.
Rod: So Charles Spearman, he did all these tests, and he, having looked at all these different kinds or subsets of intelligence, he kind of did some statistical jiggery-pokery and came up with a g factor.
Will: Ooh.
Rod: So basically, as a rule, r- roughly speaking, if you did fairly well on a couple of these, you probably did better on others as well.
Will: You've probably got good book smarts.
Rod: Probably, but you don't necessarily have to be really good at all of them to have a high g factor. Yeah,
Will: you might be lacking the street smarts.
Rod: Yeah, but you've got all the others, say, like high g.
Will: Yeah.
Rod: You've got a good G spot.
Will: Yeah,
Rod: okay. Had, had to do it. Had to do it. Flash forward to this century, some [00:02:00] psychologists found a factor for something quite different.
Will: Oh.
Rod: So you've got a guy like, uh, I think it's a guy, Zetler, a guy from, uh, Copenhagen in Denmark, University of Copenhagen.
Rod: You
Will: know you can look these things up.
Rod: Yeah, but I was busy. I was busy looking at the pig sperm and stuff. September 2018, he and his colleagues, they said there are other aspects of human personality that also have a common denominator, which means that just like intelligence, one can say that they are all an expression of the same dispositional tendencies, but it's got all these subcategories, and they were focused along with some friends from Germany on the dark triad.
Rod: So that's the, uh, the interplay between psychopathology or psychopathy, narcissism, and Machiavellianism.
Will: Mm-hmm.
Rod: So that's the classic three things if you're high on those, you're a jerk human, basically. Uh. A nasty dark and wind. I
Will: think that's judgmental.
Rod: No, no, it's, that's science. It's classification.
Rod: You're a jerk.
Will: I
Rod: You're a nasty hates people jerk. Technical term. It's in the, uh, DSM-7.
Will: Mm.
Rod: Or it will be in 15 to 50 years. But they wanted to know more about darkness, so these three were not enough, so they went deeper. They went beyond the three. So they did four studies involving over two and a half thousand participants.
Will: Okay.
Rod: A [00:03:00] bunch of survey questions to measure nine distinct dark personality traits. So they took the first three, Machiavellianism, narcissism, psychopath- psychopathy. God, it's hard to say. Psychopathy. And they added egoism, moral disengagement, entitlement.
Will: Oh, okay.
Rod: And one of my favorites, sadism. You don't mind a bit of sadism?
Rod: Yeah. Just a little.
Will: Yeah.
Rod: Torture a few ... Oh, you're, you're a teacher.
Will: Oh.
Rod: I- By, by definition.
Will: No, I, indeed. Indeed. I think a teacher needs a, just a little streak.
Rod: A hint of the sadist.
Will: Just a hint.
Rod: I agree. It's like no, no, no-
Will: With love. With love ...
Rod: this is good suffering.
Will: Like indeed. Indeed. Like in the Buddha himself.
Rod: Said-
Will: And, uh- ... suffer
Rod: will ...
Will: Tripitaka, as I get all my Buddhism from.
Rod: Do the accent.
Will: No.
Rod: Sadism, self-interest, obviously they disagree.
Will: Yeah.
Rod: And spitefulness.
Will: Oh.
Rod: So these are the nine. There's quite a few. And they were asked to rate themselves, these two and a half thousand people, on their extent of agreement with statements.
Rod: Classic way to see what kinds of people you are. Yeah. You know, on a scale of X to Y, how much do you agree with these statements? Things like, "I know that I am [00:04:00] special because everyone keeps telling me so." Tick.
Will: Blaze, yeah. Bec- "
Rod: I'll say anything to get what I want."
Will: Yeah, okay.
Rod: Tick. "It's hard to get ahead without cutting corners here and there."
Rod: And this is just example, "Hurting people would be exciting."
Will: Yeah. Okay.
Rod: I see that he's-- For the people who are just listening-
Will: Are we getting back to intelligence here? Like what, what is the intelligence?
Rod: No, this is, this is about a different factor this. So it started- Ah ... with, you know, the G-spot, but now we're into the dark factor.
Rod: So all these dark traits, they're distinct, but they also overlap a bit- Mm ... just like the intelligence factors. There's some overlap here. So they ran the stats- Yeah ... having measured all these people, and they did the little, you know, the numbers, and they, they made the D factor, which is the dark factor of personality.
Will: Okay.
Rod: The dark factor of personality. So D expresses itself in different ways in different people.
Will: So we can get a, a number here? Are we gonna get a number that will tell how much of an asshole someone is?
Rod: Ish, it's just more like high D, low D, mid D- Oh ... you know. Yeah. But n-
Will: I wonder, like you're a 79.
Rod: Well, you-- I, I-- We'll get to that.
Rod: The-- You can get a percentile.
Will: Oh, good. Good.
Rod: You [00:05:00] know, are you higher or lower compared to others?
Will: Yeah.
Rod: They said, look, yeah, it reveals itself in different ways in different people, but they say we map the common denominator of the various dark personality traits, and you can basically say someone's higher or lower generally on their D factor because it indicates how a person is likely or how likely a person is to engage in behavior associated with one or more of these traits.
Rod: So in theory, you could be really low on sadism and entitlement, as you are.
Will: Yeah.
Rod: But you could still have a whole, a very high D factor because you're high on the other ones, machiavellisms, egoism, et cetera. Sure. Sure. And if you're curious, if you go to darkfactor.org- Dark factor ... you can-- Darkfactor.org.
Rod: Did you
Will: do your own?
Rod: I did.
Will: What'd you get?
Rod: Really low, and I was like, "Oh, I thought I was more mysterious than that," but no, I'm actually quite pleasant. I'm like bottom five percent or something like that. Oh,
Will: look at you.
Rod: I know.
Will: You're a nice guy pretending to not be.
Rod: I tried. Like I was like, "Ah, tell the truth," though just out of curiosity.
Rod: It was like, "No, I'm not very dark." I thought I was quite dark, but no, I'm not very interesting.
Will: Oh, we should have done it together.
Rod: It's quite long.
Will: Oh.
Rod: But [00:06:00] you're contributing to research, which obviously we all agree with. So yeah, um, it, it'll be in the show notes too, but yeah, darkfactor.org. So Zettler and his friends, or her friends, I can't remember, they continued to study the D factor since twenty eighteen.
Rod: So there's a twenty twenty-one longitudinal study. They tracked more than twelve hundred adults over four years, and what they found was the D factor was the most stable personality measure-
Will: Are you serious? ...
Rod: full stop.
Will: Stable?
Rod: Stable. It was more consistent over time than any individual dark trait. So you could change your egoism or whatever- But you, you-
Rod: but the underlying-
Will: You've got some trait that, that you're born with it, you, you-
Rod: Maybe you're born with it, but they are-
Will: It's consistent.
Rod: Mm.
Will: Wow.
Rod: Yeah, which I think is fascinating. So even though other things might vary and, and also other personality measures may vary, but they found the D factor was remarkably stable.
Will: It's like your, your choice to how you engage with other people. That's, that's- Mm ... always the same.
Rod: Hmm. Well, how you consider yourself in relation to. But they went forward. 2026 study, because the consistency over time is interesting. 2026, more than 8,000 [00:07:00] Germans, uh, United States, and Denmark folks, people with high D scores, significantly less likely to work in social professions like teaching, nursing, and therapy.
Will: What a surprise.
Rod: I know.
Will: What a surprise. And they work in more asshole professions like CEOs.
Rod: For example. But they're also-- And they say they're even less likely to even want to. I'm like, "Well, of course."
Will: Well, sure.
Rod: Yeah. Like, do you want to work with people? Fuck no, unless I'm getting something. But my favorite was 2025 in, uh, PNAS, which is the- You
Will: mean the Proceedings of the National Academy of Science, an
Rod: august journal
Rod: an august journal. It is very august. Just, you would have thought-- I still, every time I read it, I think when you were sitting around the table, didn't anyone go, "Sounds like dick, dude. Like maybe we should call it something else"?
Will: No, they didn't.
Rod: The National Academy of Sciences, Proceedings, like NASP. How about that?
Will: Stuff from the National Academy of Science.
Rod: Yeah. Anyway, in PNAS, the D factor doesn't just vary between individuals, it varies between societies. So there are societal differences. Countries and US states in particular that have more toxic social conditions tend to have higher D factors.
Will: Toxic in what sense?
Rod: Like nasty, [00:08:00] self-interested- So they are
Will: nasty s-
Rod: Yeah. So they, they're saying it's, it's a stable disposition, but the environment does definitely, it would appear to them anyway, shape how dark people become. And I'm not quite surprised by this, but I'm glad to see studies on it to find out. Yeah. So if, if you're in a jerk environment, you're more likely to be a jerk-
Will: Yeah
Rod: I think is the summary. But the wrap-up is the thing that concerned me. There is a bunch of stuff about this is interesting for further science, et cetera. But what I love is here, knowledge about a person's D factor may be a useful tool, for example, to assess the likelihood that the person will reoffend or engage in more harmful behaviors.
Will: Ooh.
Rod: So we're gonna measure and go, "Oh, you're more likely to be naughty."
Will: Ooh, that's a, that's an interesting one.
Rod: Isn't it? And it's like, uh-huh, we're gonna use it for thought crimes and pre-crimes.
Will: Yeah, it's very pre-crime.
Rod: A little bit. And then I thought, you know, what I'm waiting for now, is there gonna be more, let's call it demographically targeted?
Will: Hmm.
Rod: And you know, on people who may be more, let's call them visibly identifiable. So are they, are they, are
Will: they stable if we might be able to train people away from them? Like can we, can we give them a sort of empathy?
Rod: I think my, my suspicion is for many [00:09:00] people you could nudge them away- Yeah ... but certain folks are down that line.
Rod: I'm just waiting for the racist side. If, if it's not out there already, this will be applied to racism for sure.
Will: Mm.
Rod: I have zero doubt. So that's a bit of fun.
Rod: The Pantheon in Rome
Will: Love it ...
Rod: it's a 2,000-year-old temple. It's capped with an enormous dome of unreinforced concrete, and it's the largest kind of structure like this in the world. So unreinforced, huge dome of concrete. It's still up and honking today where many other structures have crumbled into the dust, including ones that are centuries-
Will: Not the pyramids
Rod: younger. So their astounding durability wasn't just reserved for concrete and mega structures and temples and stuff like, like things of grand cultural importance. Heaps of Roman structures, big and small, still exist today, and they're in comparatively speaking, crazy good condition. You can
Will: see 'em.
Rod: Comparatively speaking, they do way better than most other engineering. And so this is why engineer Paolo Montieri at, uh, UC Berkeley says, "Look, exploring ancient engineering techniques can lead to important revelations."
Will: Okay. Yeah.
Rod: So fair enough. So research up to this point, he [00:10:00] was saying-
Will: Is it using slaves?
Will: 'Cause I think, I think that was a big part of, of Roman engineering
Rod: They put enslaved people.
Will: Well, indeed.
Rod: And orphans.
Will: I don't know if it's in Journal of, Journal of Engineering: Orphans and Underused Resource.
Rod: Probably is. It's in most journals if you go back far enough. So research to date found things like, uh, an extraordinary durability was largely due, the, in the concrete, to a reaction between volcanic ash, called pozzolan And quicklime, and it produced an exceptionally resilient mineral within the concrete
Will: Oh, cool
Rod: Part of the reason to do it.
Rod: And they call these, uh, pozzolanic reactions, and it's fundamentally important to the, the strength of these-
Will: So Roman concrete, better concrete because of
Rod: Pozzolanic reactions Indeed So Paolo and his buddies from Beijing University of Technology thought maybe there's something more going on than just your pozzolanic reactions.
Rod: So the problem of course with this is your fuddy-duddies in charge of pantheons and stuff don't want you getting up there and chipping off bits of concrete- What? ... to test it out. So it seems like the powers that be who protect that kind of thing don't care [00:11:00] about lesser structures like the toilet at Emperor Hadrian's villa in Tivoli.
Will: Oh my God
Rod: Which no surprise, is also still standing- Oh my God ... because pozzolanic reactions. Now
Will: what do we mean by toilet here? Is this, is this like- You- Is it the general bathroom area?
Rod: No, no Or is this literally the- Literally
Will: the shitter And is it the emperor's one?
Rod: Yeah
Will: Okay
Rod: The... Although the word communal is involved
Will: Yeah.
Will: Well, I, I- But
Rod: that's how they
Will: pooed I've, I've always, I've always pondered that. You know, I'm like-
Rod: You'd rather poo with people?
Will: I...
Rod: No You're a people
Will: pooer I'm not
Rod: So yes, the, uh, the toilet at, uh, Hadrian's Villa was apparently, yeah, the com- as I say, the word communal came up more than once. So apparently that was okay to take a bit of concrete out of to test.
Will: Really?
Rod: Interesting, isn't it? The
Will: worker's cottage toilet? No, Hadrian It had to be
Rod: the emperor's toilet No, an emperor's dunny
Will: Hadrian in particular
Rod: Yep, definitely Hadrian. Ooh, the wall, like the wall was impressive Not Marcus
Will: Aurelius, he was a good emperor. Hadrian,
Rod: you know Well, Marcus Aurelius just shat outside.
Rod: He was a monster. Very good politically, but terrible with personal hygiene. So they extracted a small piece of concrete from the communal toilet. [00:12:00] They ran a bunch of these high-resolution imaging techniques on it and, and examined it down to the nano scale-
Will: Of course ...
Rod: which is apparently quite new. So they saw evidence of your pozzolanic reactions- Yeah, yeah
Rod: no surprise. But they also found over centuries, carbon dioxide from the atmosphere reacted with leftover lime-
Will: Yeah ...
Rod: and it produced calcite-
Will: Oh, like in shells ...
Rod: which you find in, in limestone and stuff.
Will: Okay
Rod: Yeah. And they'd found calcite before, but they'd never studied or mapped this, the architecture of the concrete in 3D before, so they looked into it.
Rod: Turns out the effects of calcite And its presence, in fact, aren't just some byproduct of aging that it just sort of appear and, you know, happened to be there. It works alongside your pozzolanic reactions. It crystallizes within tiny pores and cracks, and it makes the concrete denser, gradually sealing weaknesses that would have otherwise spread.
Will: Are you saying it's like it's, it's improving itself over time?
Rod: Yes.
Will: Wow.
Rod: I mean, fuck off. Seriously, I'm like, what-
Will: [00:13:00] Man, those Romans.
Rod: He, he-
Will: How, how did they know this?
Rod: According to the-- Paolo, the engineer, "The findings suggest carbonation over a long period also enhances the durability of concrete and can help it seal cracks as it ages," which is fricking wild.
Rod: That's awesome. It's absolutely wild. So they're basically saying, "If we really get our heads around this, we can provide new insights into- We
Will: too could have concrete that lasts ...
Rod: Longer.
Will: Yeah.
Rod: And I thought, "Well, why, why don't you just do that then? Just do it now, and their concrete will last longer." But apparently modern buildings put a lot more demands on the materials, and they're larger than the dome, and they get reinforced with rebar and stuff, the metal that reinforces concrete, and it decays and reacts with stuff and causes trouble.
Rod: So it's not as simple as just-
Will: Oh. Adding- So you can't just do it.
Rod: No, we need to fiddle with it. But they're very optimistic that by unlocking these secrets they'll find more-
Will: I love it ...
Rod: stuff. But yeah, essentially the Romans kind of accidentally had self-healing concrete is my summary, which I just think is cool as hell.
Rod: I'm
Will: gonna make some.
Rod: Go on.
Will: I reckon I could buy these ingredients on the internet. I'll make something out of Roman [00:14:00] concrete.
Rod: Go on.
Will: Maybe I will Well, I got one that I don't know what to do with, but, uh, I'm billing this as I guess I had assumed here
Rod: I'm gonna help you know what to do with it
Will: Okay. What does the term plant-based diet mean to you?
Rod: Vegan. I'm gonna go with the things you eat only grow in the ground
Will: Yeah. They're, they're plant.
Rod: Yes
Will: They're plant-based And
Rod: I don't...
Rod: By plant, I don't mean came from an industrial facility
Will: Well, this is another one I had not thought to ask-
Rod: Yeah ...
Will: but I guess my assumptions were not precise because, uh, latest study in the American Journal of Lifestyle Medicine-
Rod: Oh, classic. Life- the Journal of Lifestyle Medicine I know, right? Hey Is my velour velour enough-
Rod: on this smoking jacket?
Will: I feel maybe it's something about how much a smoking jacket improves your health. Maybe they're very legit. Like I'm gonna, I'm gonna go and investigate the American Journal of Lifestyle Medicine and find some more stuff. Do you know
Rod: their listing, they're like, "Those monsters, they ruined our reputation."
Will: No, it sounds lovely. Sounds lovely. But I had not thought to investigate whether the term [00:15:00] plant-based diet was ambiguous or not, but apparently it is.
Rod: If you want it to be, I'm sure it is.
Will: Uh, in this survey-
Rod: Uh-huh ...
Will: 35%-
Rod: This is Americans?
Will: Of Americans-
Rod: Okay ...
Will: said, "No, plant-based means there's plants and meat and dairy and eggs."
Will: And I'm like
Rod: Oh, it's true. Like there's, there's broccoli.
Will: I put the broccoli on the bottom of the plate-
Rod: Yeah ...
Will: and then I put my meat- And then half a cow ... I put my meat and cheese. And a kilo
Rod: of cheese.
Will: You know, everything I've heard plant-based is like you're eating, you're eating plants. There's a plant at the
Rod: bottom.
Will: The, and, and so at least 35% plus 7% that they didn't know, and 2% gave another answer. And I'm like, I gotta go investigate more. What do they,
Rod: they- They, they wanted to. That's willful. It's like- They- ... yeah, it is plant-based. They put it on a lettuce leaf.
Will: But we're at like 40% who are giving an answer that is-
Rod: Bullshit
Will: is not sure about. I, I'm
Rod: like- KFC, this is a full chicken bucket. There's 95 pieces of chicken, potato and gravy and chips. I put it on a bed of lettuce. It's plant-based.
Will: But this is the thing. I
Rod: mean- Bullshit. [00:16:00] Bullshit. And this is con- confirmation bias and bullshit.
Will: I'm loving someone, someone who's like, "Yeah, I eat plant-based.
Will: That means I have some plants in there." Like- Well, quite ... I mean, you could go, okay, is it plants at the bottom of the plate or more plants than meat? Is that plant-based? Nope. Or It's not. I really thought
Rod: it was obvious. Oh, no, no, no. Actually, no. Exclusively plant-based or includes plant-based.
Will: No, I, I think exclusively plants is what you're saying.
Rod: You need a qualifier though. Whereas
Will: exclusively plant-based still says based and then on top. Like- So
Rod: I don't- So only, only plants.
Will: Actually, the authors of the study said, uh, "Maybe we should move away from the term plant-based." Okay. Um, uh, provide a consistent definition so people actually know what we're talking about, or we could move away from the term and call it something different, you know.
Rod: Zero animal product.
Will: Yeah, exactly. Something like that. But I really thought plant-based was pretty freaking clear.
Rod: I did
Will: too. Turns out it is not. There are at least, well, 35% of people who believe plant-based could be
Rod: The [00:17:00] follow-up question has to be, do that 35% A, just being a dickhead. You've gotta work hard to not know what that means.
Will: That's what I would've thought. I had assumed. Of course. And so-
Rod: And you know what they say.
Will: But yeah, but I do love the concept of someone going out there going, "Yeah, I eat a plant-based diet."
Rod: Chomp, chomp. Bullshit. Yeah. They're lying. They're fibbing. There, there's a very small percentage of that 35 who are actually like, "Oh, wait, what?"
Rod: The rest of them are just being shitheads.
Will: Look, I'll be racist and say it is Americans Where would we be without pigeons? Well, mm, there are plenty of places we wouldn't be, or pla- maybe some places we are that we shouldn't. But the annals of your psychology is stuffed full of pigeons. Pigeons, pigeons, pigeons everywhere. So pigeons everywhere. Classic is your Skinner box. People, they put pigeons in boxes.
Rod: They give them different stimuli, get them to tap a lever and-
Will: Yeah, yeah,
Rod: yeah, yeah ... in response to certain things. They learn stuff about them.
Will: Why not rats? I mean-
Rod: I don't... Oh, no, rats have been given that sort of thing.
Will: No, I know. No, but why, why would you use a pigeon rather than a rat? Like-
Rod: I don't know.
Rod: Maybe they wanted to do more than just rats.
Will: They wanted to add flying.
Rod: Ex- exactly. [00:18:00] Exactly.
Will: See if the feathers made a difference.
Rod: Yeah, indeed. Or a beak. So they're particularly prolific in studies of learning, memory, and award, reward associated behaviors like gambling and so forth.
Will: Okay. Yeah.
Rod: Or tr- or, you know, research about it.
Will: Pigeons, those gamblers.
Rod: Well, no, they helped us see the- All the time on the
Will: prediction markets ...
Rod: the value of your intermittent rewards like a slot machine, like a poker machine.
Will: Yeah, I get it. I get it.
Rod: They've also trained pigeons to do things like tell the difference between a Picasso and a Monet.
Will: Well, indeed.
Rod: Important work It
Will: is interesting though
Rod: It is interesting
Will: That, that there is enough of pat- enough pattern recognition that they can spot-
Rod: Mm ...
Will: some sort of category. I mean-
Rod: Yeah ...
Will: they, they are, they are, you know, of similar times, but, uh- Yeah ... they're quite different artists
Rod: They can discriminate between malignant and benign tissue samples and radiology images, so probably in some ways similar.
Rod: Are you serious? Yeah. Apparently. Are
Will: you,
Rod: you- With astounding accuracy.
Will: Wow.
Rod: I don't know if they're better at
Will: it than the AI. Hang on. So these are tissue samples or pictures?
Rod: Images. Yeah.
Will: Images. Images. So they're seeing. I always thought this was some sort of sniffer technique.
Rod: No. Well, they don't have noses.
Rod: Haven't you looked? They've only got beaks. [00:19:00]
Will: Beaks have noses in them.
Rod: Yes, there are noses on the top. But yes, they, um, can do that. They can apparently, and I can't go into details because it was too much so I didn't read it, count like primates count, telling numbers. For which-
Will: That's sad ...
Rod: for pigeons.
Will: No, f- I'm on pr- team primate here, obviously.
Rod: What, what ob- what's obvious about it?
Will: Obviously, uh, you're not on team primate?
Rod: I'm in no teams. I'm on all teams.
Will: Oh, my God.
Rod: They can apparently also, again, there's a lot of detail here, but tell... They can get trained to tell the difference between real words and sort of jumbles of letters. I'd like to know more about that, but I didn't have the, uh, patience for this.
Rod: Yeah. Anyway, a core premise from pigeon research is from, back from 1898, Edward Thorndike. Pigeons, and people, and other animals generally do things they are rewarded for again.
Will: Yes.
Rod: You're, you're rewarded for it-
Will: Yes. I know we, we would go, "Wow," but obviously, it, it, it required- Yeah ... some demonstration at some point.
Rod: It did. Yeah. I mean, you know, come here for the mind fucks that are-
Will: Yeah ...
Rod: 150 years old. There's also, though, an associated assumption with this, which is rewards also make the rewarded behavior more [00:20:00] consistent. What specifically about this will give me- Yeah ... the reward the most efficiently or quickly? So the way a specific behavior is performed over time should become quite stable.
Rod: Mm.
Will: Mm.
Rod: The way in which it's done. So scientists apparently have repeatedly tested whether rewards increase the frequency of the behavior, but not their effect on the consistency of the behaviors so much. Yeah,
Will: okay.
Rod: That's less common. So a bunch of people, University of Iowa, experimental psychology people, they wanted to check it out.
Rod: So of course, they went to pigeons.
Will: Yes.
Rod: As you do. So anyway, they gave pigeons a series of five different colored buttons to peck at.
Will: Yep.
Rod: You needed to peck five buttons to get the reward.
Will: Okay.
Rod: The order didn't matter, as long as it was five buttons.
Will: Okay.
Rod: As expected, the pigeons learned to do this quite quickly.
Rod: No surprise. But what they didn't expect was they never fell into a routine order of pecking. Oh.
Will: Oh, so they didn't just go one, two, three, f- like left to right? Like-
Rod: Or whatever the first five they hit, they kind of go, "Well, I'll just do that again." They just did five all over the shop.
Will: Ha. 'Cause I would absolutely go, "All right."
Will: Yeah. Yes, maybe, maybe my brain [00:21:00] is thinking there's a sequence, or I would just do the same thing, just because, you know- Because it
Rod: worked ...
Will: because it worked. I would just go left to right. Yeah.
Rod: Yeah.
Will: Wow.
Rod: Yeah, apparently
Will: not. Any o- any order?
Rod: No.
Will: Those monsters.
Rod: I know.
Will: I could not live with that. I could not live with that- No, it's not for you
Will: level of chaos. Like, that's ridiculous.
Rod: Your place on the spectrum does not cope with that part. So the lead researcher, a guy called, uh, Wassermann, he says, "Look, there'd be no reason not to expect that the animals would converge on a single favorite pattern."
Will: But it's not being rewarded, so why would it?
Rod: Well, you'd think just because they did it once, given they tend to remember what to do, they would probably start to just do the same thing again. Or they'd work out, "It doesn't matter, so I'll just go the same button five times."
Will: Yeah, it doesn't matter.
Rod: Why move around? So I'll do it, I'll do it easy. Just go dong, dong.
Rod: Yeah, exactly. But apparently not. He says, "Now you could argue the birds are just utterly resistant," this is his words, "The birds are utterly resistant to locking in anything stable." Now, that's a big call.
Will: Well, they've clearly locked in so- I mean, they're getting the reward here.
Rod: Yeah. No, but that's his call.
Will: Yeah, okay.
Rod: Utterly resistant to locking in anything stable. So their current explanation: pigeons have adopted or adapted to an [00:22:00] ever-shifting environment. They're talking your city pigeons, your piazza pigeons. Yeah. Their surroundings are constantly changing, and then here's a quote I really enjoy, too. "They seem to avoid stability in their decision-making.
Rod: Instead," quote, "choosing to live at the edge of chaos."
Will: Nice. Which
Rod: is-
Will: That's a, that's a great postal address ...
Rod: isn't it? That's, "Where do you live?" "At the edge of chaos." And that was, of course, in the title of one of the articles about this, and I'm like, "Okay, the pigeons live at the edge of chaos." So of course they say it's time to check this with other animals, fiddle with some of the conditions, 'cause maybe it would illuminate how the brain makes behavioral decisions while learning.
Rod: Sure. UCLA professor, a psych guy who wasn't part of the study, says, look, he wasn't particularly surprised by the results, but he says, "This paper leaves open many questions about the," brackets, "neurological mechanisms for future scientists to explore," which to me sounds a bit like, "Well done, you did something.
Rod: We could do other stuff." Yeah, look, I'm a, I'm a huge fan of all bird-based research. You know, if it involves a bird, I want to do it. That's the only reason I studied psych. I just want to use birds.
Will: Really?
Rod: Cassowaries.
Will: [00:23:00] Cassowaries are a great bird.
Rod: It's like, "Did they peck the buttons?" "No, but they destroyed the equipment."
Rod: Smashed it. Everything in the room's gone and they killed two
Will: lab assistants. Like, I feel like being hit by a, hit by a cassowary in a siege, like in a trebuchet. That's the thing.
Rod: So I don't know what to do with this result other than kind of go, "Oh, that's kind of interesting," and it is, but the whole choosing to live at the edge of chaos I think is maybe a little dramatic.
Will: Bit of hyperbole. A
Rod: lit- exactly. Your hyperbs.
Will: Well, that was your little bit of science for the week.
Rod: holiday edition. You're special by the pool wearing a bikini edition.
Will: But because you're on holiday, you know that you still have the power to give us the rating that you need to give us. Yeah,
Rod: seven stars on every app. Even things that don't do podcasts.
Rod: Yeah.
Will: Go out there and write it on like a recipe app
Rod: an Uber and Yelp. Is it Yelp still a thing?
Will: I think so. I'm
Rod: I don't know. I'm at a restaurant where Don't ye
Will: listener, if you've got some topics that you want us to explore,
Rod: tell Will.
Will: How would you tell Will his
Rod: number is? 0 4 0 5 oh. Uh, cheers. At a little bit of science Do com
Will: au.
Rod: au[00:24:00]
Will: Do that. We want your stories.
Rod: we wanna hear from you.
Will: Lovely listener. Enjoy the pina colada.
Rod: Oh and the
Will: col. Pin colada.
Rod: Pini Kaia Pina Pia
Will: Pina Kaia of
Rod: of the Clade
Will: Penai. Cate.
