This week’s line-up is a perfect example of the universe having a weird sense of timing. A river drops to record lows and suddenly Europe starts coughing up skeletons, war relics, and Roman engineering like it’s doing a historical clear-out.
Then we jump straight into AI designing viruses, and finish with Viagra quietly auditioning for a second career as a cancer helper.
None of these stories are comforting. All of them are fascinating.
When rivers shrink and history pops up
Europe’s Danube has hit record low levels, which is obviously not a fun climate headline. But it does mean the river is revealing things it normally keeps hidden, and some of those things are… a lot.
In Hungary, the falling water exposed German military remains, including a Wehrmacht motorcycle with its rider still there, still holding on. It’s the kind of discovery that feels less like archaeology and more like the river briefly turning into a crime scene.
Over in Bulgaria, the Danube’s retreat revealed the foundations of the Constantine Bridge, built nearly 1,700 years ago under Emperor Constantine. It didn’t last long in its day, but the fact it’s reappearing now is a reminder that big human projects can vanish for centuries, then casually return when the environment shifts.
It’s a strange trade. Climate change is doing real damage, and it’s not a “silver lining” situation. But it is forcing the past back into view, whether we’re ready for it or not.
AI-made viruses are making us nervous
Then we move from ancient engineering to modern science fiction. Researchers used AI to design viral genomes, specifically viruses that infect bacteria, with E. coli as the main target.
They generated hundreds of designs. A small number actually worked, meaning the AI didn’t just make plausible-looking code; it produced functional viruses that could kill bacteria.
On one hand, that’s exciting. If we can design bacteriophages on purpose, it opens up new ways to study viruses, fight bacterial infections, and maybe build more precise tools for medicine.
On the other hand, the phrase “AI-designed virus” makes everyone’s shoulders tense for good reason. Even when the target is bacteria, it raises the obvious questions about biosecurity, oversight, and what happens when powerful tools become easier to use.
This is one of those moments where the science is impressive, and the governance needs to keep up.
Viagra, and the surprise side quest into cancer
And then we get to Viagra, which most people file under “one job, very famous”. Except researchers keep finding that drugs with a clear original purpose sometimes have unexpected effects elsewhere.
Recent studies suggest Viagra might be linked to better survival outcomes in men who later develop cancer, especially when combined with statins. The proposed mechanism is that it may help limit cholesterol availability, which matters because cancer cells use cholesterol to spread.
Important note before anyone gets ideas: these findings are correlative. Correlation is not “everyone start self-prescribing blue pills and hoping for the best”.
Still, it’s a good reminder that medicine is messy in a useful way. Sometimes the drugs we already have end up pointing toward new pathways we didn’t see coming.
So that’s the week: rivers dropping low enough to expose war relics and Roman bridges, AI successfully designing viruses that kill bacteria, and Viagra turning up in conversations it was never meant to be part of.
It’s all the same theme, really. The world keeps changing, and the things we uncover, invent, and repurpose don’t always arrive gently.
CHAPTERS:
00:00 Introduction
00:32 The Danube and Nuclear Plants
04:04 Low River Reveals Mammoth
04:31 Wartime Wrecks and Remains
07:41 Constantine Bridge Unearthed
11:31 Mohenjo-daro Inequality Math
17:56 AI Designed Viruses
23:53 Science Upside vs Risk
28:37 Viagra and Cancer Link
SOURCES:
https://www.sciencealert.com/worlds-richest-10-are-costing-earth-trillions-study-finds
https://www.nytimes.com/2026/08/06/science/ai-viruses-bacteria-arc.html
https://futurism.com/science-energy/european-drought-heat-wave-climate-change-danube-roman-ruins
https://edition.cnn.com/2026/07/31/climate/shrinking-rivers-reveal-past-relics-in-europe
https://edition.cnn.com/2026/07/31/climate/shrinking-rivers-reveal-past-relics-in-europe
https://archaeologymag.com/2026/08/roman-altar-dedicated-to-jupiter-danube/
-
Rod: [00:00:00] So it's 2026, and that beautiful European river, the Danube, is, as I think environmental scientists call it, fucked. It flows through many countries, and a lot of different countries are noticing the problems in different ways. The Hungarians, the National Water Authority there, they measured the levels a few days ago in Budapest.
Rod: Sorry, Budapest. Nine inches. Now, that sounds shocking to me. The previous record low was 13 inches in 2013, so it's a lot lower than that, and 13 inches sounds outrageous. There are many ungood flow-on effects from this. So the Hungarian Prime Minister, whose name is Peter Magyar, he says, "For the first time in 44 years, the entire Paks Nuclear Power Plant will be completely shut down due to the record low water level of the Danube, and an energy crisis situation may arise because they use water from the Danube to cool the nuclear facility."
Rod: And that plant produces nearly half the country's electricity. No biggie. In Romania, August the 3rd, the Cernavodă, [00:01:00] you can take that to the bank, nuclear plant, also uses Danube dupes to cool itself. So they thought, "We've got to address this problem," and they did the smart thing. They got their navy to do a controlled explosion of 180 kilos of explosives to move a whole bunch of large rocks, restricting the flow of water, to move it towards I know you're not allowed to talk about fuel, to move the water towards their nuclear power plant.
Rod: And they said, "Oh, the operation was absolutely a success. It will help the cooling of the plant and water transport on the river." Of course, the plant may still have to temporarily at least or occasionally suspend its operations, but their acting defense minister said, "Any extra day of it working is worth more than the cost of the explosion."
Rod: One day. On top of this, cargo ships are forced to significantly reduce their loads, and they're slower, and you can get less of them on the rivers, so it's more expensive and slow. Tourism effects, riverboat cruises are disappearing. One riverboat in Bulgaria on the Danube ran aground. None of this sounds good, unless you're a historian or an archeologist[00:02:00]
Will: This is your song This
Rod: is my song and your song Yeah. I'm just glad we found a gnome to write it
Will: It's time for a little bit of science.
Rod: Hmm.
Will: I'm Will Grant, an associate professor in science communication at the Australian National University.
Rod: That sounds important. I'm Rod Lamberts, a 30-year science communication veteran with the brain, mind, spirit, soul, and steely gaze of a teenage boy.
Will: And today, as well as the effects of the Danube being super low, we've got, "Hey, scientists, gonna need to work on your messaging."
Rod: A, and hopefully not too I.
Will: We've also got some antiquity updates.
Rod: Ooh. I'm also gonna do a Viagra update.
Will: I, you know, I don't like nuclear power plants that don't have sufficient cooling.
Rod: Yeah, 'cause some people, the fussy ones, say you should cool them sufficiently- Yeah ... otherwise things go awry.
Will: You know.
Rod: But the Romanian solution- You know ... blow shit up.
Will: I get it.
Rod: [00:03:00] Blow a hole in the river- ... so we get more of the water. It's like-
Will: Yeah, it smells like you get more water for, like, four seconds, and then- Yeah
Will: then you're back to the flow that was there before. I don't know the geography of the situation of the geography there, so.
Rod: There is a lay of the land.
Will: Yeah, yeah. Well, indeed.
Rod: And there is some topography.
Will: One's gotta know the lay of the land.
Rod: Yes.
Will: Um-
Rod: With the topography.
Will: But, oh, my God. Ah.
Rod: You know- I, I, I love it.
Rod: I wish there was a whole story in that, but I was just like, "No, it's just funny," 'cause, like-
Will: Oh, th- but this goes... I mean, same thing, same thing happened. You know, we talked a couple of weeks ago about the cruel heat in Japan. Hmm. And it's like, if you rely- Hmm ... on electricity, which, which the world does, th- but that electricity re- relies on cooling from water, and then suddenly you're too hot for that, like that's a, that's a vortex of I don't like.
Rod: Implications.
Will: Like, like, that's a torment nexus to me.
Rod: Repercushtions.
Will: Yeah.
Rod: Yeah. Yeah. I, I, I agree. But I'm, I'm not talking about it. I'm gonna talk about the fun stuff. I'm gonna talk about the salad days of history.
Will: Okay. Okay, so we're not worried about the climate here. No,
Rod: this is- We're focused on- ... this is a good news story
Rod: what do, what have
Will: we got?
Rod: This is a good news story about the Danube being reduced to basically the feeble little stream of a [00:04:00] dude with a very large prostate trying to have a wee-wee. Oh, like
Will: that.
Rod: Yep. Yeah,
Will: okay.
Rod: So in Bulgaria, in July, the bit that flows through there, very, very small and trickly. But it revealed the remains of what is believed to be an ancient mammoth skeleton.
Will: Oh.
Rod: Which is cool. They're like, "Ha."
Will: A mammoth.
Rod: What do you
Will: know? A mammoth.
Rod: If, if we didn't have almost no river left- Yeah. ... we never would've known that.
Will: That's awesome.
Rod: So a local villager in Ryahovo Village went, "Ugh, that's not a dog or a horse." And then they got people in. Big horse. It's like, "Ugh, we think it's a mammoth," so that's cool.
Rod: Croatia, the receding water revealed, as they put it, a skeleton of a cargo ship called the Fulton, which is a transport ship for coal going down the river in 1937, and it sank.
Will: Okay.
Rod: I don't know if it's got a lot of historical value other than- Yeah ... oh, look at that, a ship from- There's a thing.
Will: There's a thing.
Rod: Yeah, yeah ... a ship from 90 years ago that carried coal. Near the Croatian town of Opatovac. Perfect pronunciation. Opatovac. A Nazi warship, the remains of that basi- didn't float to the surface, became
Will: revealed. [00:05:00]
Rod: There were Nazi skeletons inside. Back in Hungary, perfectly preserved Wehrmacht DKW NZ350-1 motorcycle from World War II.
Will: Oh.
Rod: Oh. DKW, of course, stands for-
Will: Deutsche A Kloppenheimer
Rod: Yes. Or the other version, Drumpf Kraftwagen Kraftwagen
Will: Your Drumpf
Rod: Kraftwagen
Will: Oh, I love that
Rod: And human remains were found beneath the motorcycle
Will: Well, indeed
Rod: Not above, beneath
Will: I mean, if the skeleton was still holding onto the motorcycle-
Rod: Of course they were
Rod: like The Wehrmacht.
Will: Like
Rod: it's- World War II Wehrmacht, you do not release- It's
Will: a ghost rider,
Rod: you know ... Gromerzen je nicht die motorcycle And so they, they said there are identity tags that are well preserved on these soldiers so they could be identified Okay,
Will: you can go back to the family. That's cool
Rod: And say, "Oh, ha, we found
Will: Grandpa" And, and who gets to keep the mo- motorbike?
Will: 'Cause I feel like restoring a World War II motorbike- Mm. Mm ... that has been in a river to working, that would be, that'd be a craft project. That's
Rod: just great Well, look, I mean, that's why they say, oh, like, perfectly preserved. I'm thinking perfect is probably doing a lot of heavy lifting there, but-
Will: Yeah, indeed
Rod: identifiably [00:06:00] preserved structurally familiar Known
Will: to be motorbike. Yeah, exactly Once was motorbike. That's
Rod: That seems not a mammoth
Will: Its, its noun-ish-ness is still there
Rod: Exactly. And the two dead men, they say they're gonna be buried in the German war cemetery west of Budapest, and I feel like there's a lot going on there in a German war cemetery.
Rod: I believe there are 16,000 people buried in that cemetery. Yeah Could you come and take your corpses away, please?
Will: Yeah, just some of them, not all of them
Rod: Anyway, this is good, so they can identify them and they can bury them. They also found in the similar region German Telemine 35 anti-tank mines. Cool Telemine I know.
Rod: Uh, remnants of hand grenades, artillery shells, you know, all that good stuff that kids love to play with. And it seems that this, this stretch of the Danube, according to the Hungarian Institute of Military History, was probably a war debris dump. So these things, I don't think the motorcycle-
Will: Yeah
Rod: You, you would hope they got in there by accident It
Will: wasn't a Steve McQueen situation of, of doing a jump into the river
Rod: Well, it might have been, 'cause there are people under it, but- It's like, "I can jump this."
Rod: "No you can't, Fritz." "I certainly can, Heinrich." [00:07:00] "Okay, I'll come with you"
Will: What are you doing under my motorbike?
Rod: And sploosh, and then well, hopefully one day the Danube will, like, reduce to a trickle and will be able to be discovered. So more exciting Exceptionally low water levels on the Danube exposed, again in Hungary, a Roman altar dedicated to Jupiter, which is pretty cool.
Rod: Jupiter was the king of the gods for the Romans. Archeologists hoped they were gonna find a medieval ruins of a Pentelic monastery. They found that and they're like, "That's pretty cool." It wasn't a Pentelic monastery. So they found a fragment of an altar, and a few days later they discovered other fragments in limestone, on tablets.
Rod: So basically they're going, "Oh, there's a whole bunch of Roman shit in this region," and that helped the Danube being in a terrible condition. Coolest find of all though, this is my favorite. So in Bulgaria, or Bulgaria if you would prefer, the enlarged prostate trickle of the Danube revealed foundations of the Constantine Bridge.
Rod: So it's nearly 1,700 years old. It was opened back in 328 AD by Roman Emperor Constantine the Great. Wow.
Will: [00:08:00] Constantine- Wow ...
Rod: the month. The number one.
Will: The Constantine, he's, he's a big dog of emperors, you know?
Rod: Yeah. The fir- Well, the first. I don't- The second was
Will: a- He's not the first emperor. Like-
Rod: No, no, he was the first Constantine.
Rod: But the Constantine the second-
Will: He, as, as- ... pale shadow ... uh, Edward Gibbon says is, is, uh, responsible for the decline and fall of the Roman Empire, but it took a long time. Uh-
Rod: Is that what Edward said?
Will: Yeah, Gibbon. Yeah.
Rod: He's always been against Constantine.
Will: Well, he, well, he blames the Christians. But it took a long time.
Rod: In the timeline of history, the 200 years is the blink of an eye.
Will: Yeah, but it's al- almost like saying, you know... Oh, my analogy's gone.
Rod: It's like saying stuff- ... about things that are relevant to the situation. Yeah. There's no question.
Will: Like, it's unfair.
Rod: Far be it for me to pick on Christianity. I pick on all religions equally.
Rod: Good. But not here. So yeah, uh, that was, yeah, opened in 328 AD, and it linked the ancient cities of Ulpia Oescus- I've
Will: gotta... I'm just, I just wanna pause.
Rod: Okay.
Will: Like, Constantine himself opened it.
Rod: Yep, he had the giant scissors and the ribbon. No,
Will: well, that's what I'm wondering. Like, are we saying... I g- I get that the Roman Empi- Empire built it.
Rod: Yeah.
Will: Did [00:09:00] Constantine go and- Yep ... open it?
Rod: Yep.
Will: Like-
Rod: And he personally bit the throat out
Will: of- No, but I just- ... 14,000
Rod: slaves ...
Will: I just, I do like the concept of, of Roman emperor doing like tr- "Here's what we bring for you" "Good
Rod: day, I'm here" . "Here's a bridge" "
Will: Have bridge" Like, and, and, and there's a ribbon-cutting cere- But- "
Rod: It's a nicer bridge"
Will: But there must have been some sort of political ceremonies where people are looking forward to it.
Rod: Surely.
Will: And, and there would've been local dignitaries that go, "All right-" I think there
Rod: would've been ... "we're gonna celebrate this moment" Th- this bridge is a big deal, is like- Of course ... more than two kilometers long too. Like it's- Are you
Will: serious? ...
Rod: big-ass bridge. That's
Will: a big-ass bridge.
Rod: Yeah, it's like a shitload of bridge
Will: Yeah, okay.
Will: Uh, I, I think, I think you get the dignitaries, so yeah.
Rod: I, I would have thought so. Um, I don't think they found any remains of dignitaries though.
Will: And also we sacrifice someone.
Rod: One of you noble folk-
Will: One of you-
Rod: You're gonna go on to glory ...
Will: it's gotta be the most noble person who is not powerful enough to get out of it.
Rod: So that's pretty impressive. Yeah, the two clicks long linked a city in Romania to a city in Bulgaria, but it's, it's also regarded possibly one of the longest bridges in the [00:10:00] antiquity times, but it didn't last very long. So by 367 AD it had been basically destroyed. So it only lasted 40 odd years, about, uh, yeah, about 40 years before it stopped and then got buried by the water.
Rod: But luckily, you know, climate change fixed it. So my takeaway from this is when life gives you catastrophic climate change induced environmental horror lemons, you make beautiful, fascinating historical archeological lemonade.
Will: Thank you. It's all good. I love that. I
love
Rod: that. Yeah
Will: So I have some antiquity news for you as well.
Rod: Yes, you do.
Will: This one comes from the journal Antiquity. And I love this because, well, it pushes against some ideas that we have both about the ancient world and about, and about the modern world.
Rod: Yeah.
Will: And it's, it's just, it's got a nice overlap of a couple of different types of research going in here.
Rod: I gotta say, whenever I hear someone go, "Oh, actually, this thing that you've been taught for 2,000 years is bullshit," and I think, "That's awesome."
Rod: And also, in 10 years you're gonna tell me that too is bullshit because we found something else.
Will: Look, look, I think there's a chunk of- It's fun ... we have to work with what we've got. Absolutely. Of course. In, in any- Yeah ... uh, ancient [00:11:00] history- Yep ... it's like you might have partial written records- Yeah ... or you've got archeological records, and it's like you've gotta try and work out what you've got.
Rod: And your LIDAR scans.
Will: Yeah, all of that. Yeah. But also, I think the thing we do have to remember is so much of ancient history of different periods- Yeah ... and, and I'm not saying it's not now, but of different periods, has been self-serving. And this story- Oh, yeah ... might be self-serving as well. Yeah,
Rod: yeah, yeah.
Will: But they wrote with a certain sort of view, the idea of old forms of economy that basically emerged and, and, and it comes towards capitalism. Like, that's serving a certain sort of agenda. Yeah. Anyway, so, so this one, the ancient city of Mohenjo-daro, Indus Valley.
Rod: Oh, the Indus,
Will: yeah. So i- if, if you go back to y- your ancient river civilizations-
Rod: Mm-hmm
Will: the Indus is, is one of the good ones. You got... Well, y- you got u- your Nile civilizations, y- Euphrates civilizations, your Indus River. But Mohenjo-daro is, um, one of... Well, it's the largest city i- of the Indus Valley Civilization.
Rod: Oh.
Will: It achieved a peak sort of 2600 BC, [00:12:00] so BCE, to 1900 BCE, so about 4,000 to 5,000 years ago.
Will: Ugh. Yeah. It was city-like. Yeah. So this is sort of contemporary with different periods in Egypt.
Rod: Yeah.
Will: Certainly some of the, the early Egypt stuff. But there's a flourishing city there.
Rod: It fascinates me that there were these full-on cities at a time almost too remote to imagine, really.
Will: Mm. Mm.
Rod: The functional day-to-day.
Rod: I mean, yeah, people had to eat and sleep and do poos, but- We've
Will: always done that, yeah ...
Rod: it's just, it blows my mind. Like, we had a functional, full-on func- like a big-ass city.
Will: Well,
Rod: well- Many of them ...
Will: come to that in a bit, because there's some awesome things. Now, a lot has been known about Mohenjo-daro for quite a while in terms of what sort of civilization it was.
Will: I just love this. Uh, some researchers, what they've done is they've looked at house size in Mohenjo-daro. Okay. Because you can, you can get maps... Yeah, well, the... You can see a lot of the houses. Yeah. You can see what would be the, the outline of the houses. There's, there's a lot of stuff still there. Yeah, yeah.
Will: But also you can see the layers of the different houses, so you know the dates roughly [00:13:00] of this house- Oh, yeah, yeah, yeah, yeah, yeah ... and then this house had been built on top, and this house- Yeah ... had been built on top.
Rod: Yeah.
Will: And so using that you can construct a picture of the wealth distribution in the city- Right
Will: over time.
Rod: Love it.
Will: The fascinating thing here is by... Basically, what they're arguing is- If you have, like, giant palaces and then small hovels-
Rod: Mm ...
Will: high inequality. You know, you got your, you got emperors like the Pharaohs-
Rod: Then nothing ...
Will: and then nothing And then
Rod: shitholes
Will: And everyone lives in shithole.
Rod: Yeah.
Will: Common in, uh, Egypt- Yeah ... uh, at the time.
Rod: Yeah.
Will: Mohenjo-daro, nothing like that.
Rod: Really?
Will: Like n- they've observed before that there's no palaces really. Things are a little bit weird. There's noth- nothing really big. Huh. And what they do is they convert this to a, a Gini coefficient, G-I-N-I. This is the coefficient we still use to look at inequality around the world.
Will: It's a, a number between zero and one. It basically, you put every, every country or every city or every civilization on that scale.
Rod: Yeah.
Will: Zero means perfect equality. That means everyone has exactly the same amount of money. I,
Rod: I love they've made zero the perfect number. [00:14:00]
Will: But one means perfect inequality, means one person has all of the money- Oh
Will: and, and no one else has any. Now, neither of those are, a- at all possible. I guess you could theoretically go, you make a micro society, like you've got 10 people living on an island, and you could go, "Okay, we, we formally divide all wealth equally" Yeah You could say that that is a highly high equality society.
Will: So zero-
Rod: So 10 minutes after we set this up, totally high equality.
Will: Yeah, yeah Or, or flipping around, you could have a micro society again where one person is like some total dictator- Yeah ... and they control everything- Yeah, yeah, yeah ... and everyone else is their slave.
Rod: Yeah, yeah.
Will: So in a micro society, maybe- Yeah
Will: but, but anything larger than 10 people, it's, it's kind of im- implausible to imagine anything. No. Yeah. But you could put every society on there. I would argue, and this is a value judgment, that- What? ... high inequality societies are probably gonna be a bit problematic.
Rod: A bit? Oh.
Will: So what they've done is by looking at household size, and this is a pretty reasonable thing- Mm
Will: like still the same for us now.
Rod: Not household, house.
Will: House size. [00:15:00] Yeah, yeah. Yeah, literally house size. Yeah, yeah. Literally the- Like the square meterage- Yeah, yeah ... of house. It's pretty plausible now. Like you could, you could look at a, a house chart and go, "Okay, if you have some people that have airport hangar sized houses-
Rod: Ah
Will: and some people have the hovels, you've, you've got a fair bit of inequality." Back then, okay, so they look at Mohenjo-daro- Yeah ... and they compared it with a bunch of others. So they've got from similar periods, the Bronze Age Greek city of Knossos- Yep ... and the Maya site of- I
Rod: think it's pronounced- ... Palenque
Rod: Knossos.
Will: Knossos, with a K? I don't know. But they're like .86 or .75. Okay. High inequality.
Rod: Way up, yeah, yeah.
Will: What that means is they've got a king who's got a lot of power, and, uh-
Rod: A few courtiers and then all poor
Will: bastards And then, and then all poor slaves down the bottom. Yeah, yeah, yeah,
Rod: yeah.
Will: But Mohenjo-daro is way down the bottom, .44.
Will: So they can't find any palaces, so it's a highly i- A, a much more equal society
Rod: Interesting there that .44 is very low, but it's still, there's still a long way to go- Oh,
Will: it's, it's not
Rod: perfect ... from zero to... But what I mean is it's, that, that to me is interesting. Like this is, this is looking like a shining light.
Will: Yeah. [00:16:00]
Rod: Oh And it's still
Will: only .44. Look, look, what it probably means is there's- Hmm ... like solid middle class and solid working class- Yeah ... and maybe they have slaves or something. I don't know that.
Rod: Pretty much Canberra.
Will: But it, it means there, there probably isn't your uber overlord- Yeah, yeah ... overlord sort of situation.
Will: Yeah, yeah, yeah. And, and what they're saying is, you know, instead of gold-filled tombs and huge temples- ... Mohenjo-daro focused on sophisticated brick-lined drains, organized street layouts. Ah. As well as this, you've got better urban infrastructure. They have- Every
Rod: time better ...
Will: they have- Every time ... systems of, uh, a credit and debt system that is decentralized, so everyone sort of- Ah
Will: you've got deliberative spaces. So-
Rod: Tick, tick, tick, tick ...
Will: it's just a nice piece of like archeological work using mathematics to, to- Yeah ... analyze it here, that says not all things were you just had, you know, pharaohs at the top. You know, that it's like god, god emperors- Yeah, not everything's a
Rod: 1930s Hollywood take on
Will: ancient societies.
Will: Absolutely. Yeah, yeah. You know, like that ancient societies organized themselves in so many different ways. Hmm. And this seems to be one where they chose to go, [00:17:00] "Hey, let's be a little bit more equal. Let's make the drains a little bit better and not have a god emperor"
Rod: Look, A, yay. B, you know someone's gonna come along within the next two to 15 minutes and say, "House size, blunt indicator of garbage" Ah.
Rod: Let's get into, but for now, as you say, you know, you gotta, you gotta go with what you got right now. I'm very curious to see what happens in a town like this one where we live right now, because house size is going off.
Will: Australia has the biggest houses in the world now.
Rod: Uh- We don't need them. We don't need them
Will: No, no, agreed, agreed.
Will: But-
Rod: My house is too big and I don't need that, but I've got a big block, so there's space around. People don't understand space. But
Will: also, if you've got a- Space is a luxury ... if you've got a two-dimensional house, as in like three-dimensional house, you've got, uh- I
Rod: do have a three-dimensional
Will: house ... a three-dimensional...
Will: Well, what I mean is-
Rod: We paid extra
Will: for that ... if you've got, if you've got multi stories- Oh, yeah,
Rod: yeah ...
Will: not gonna be visible in the archeological record, I'm afraid. It'll just get you walls.
Rod: They'll be able to tell from the depth of the foundation that it took extra weight, therefore- Oh, okay ... probably had more than one, one layer.
Rod: Okay So Samuel King and his colleagues are, they're pretty excited. They work at [00:18:00] Stanford, the Arc Institute as well, California Research Organization, and they're authors of a new study published in Science, which- Is this- ... we've, we've never had.
Will: Is this one, one where people should work on their messaging a little bit?
Rod: Look, I, I'm interested in your take 'cause this, this is one of these rare moments where, and we, we might as well leave this in, where we've both done the same story and don't realize. But it'll be interesting to see what the takes are. 'Cause here's my take. What I enjoyed was how, for starters, they're manufacturing new viruses.
Rod: Which is tops, or rather new viral genomes. In itself, no big deal, and again, as you know, dive in at any point. Scientists have been doing this kind of stuff for decades, manufacturing viruses.
Will: Well, ah, depending on how you define it. Like-
Rod: Depending on your messaging.
Will: Yeah. Well, well it depends on what you're doing.
Will: So- Yeah ... I wouldn't quite say decades. Scientists may have been- Two. Okay. 2,000. Using CR- CRISPR and things- Yeah, yeah ... we certainly have the capability to edit viruses. We do. Edit, edit genomes- We
Rod: do ...
Will: to make things that are new. However, to design from scratch, you know, I don't know how much that-
Rod: Who [00:19:00] said from...
Rod: No, you know. We do this to learn how viruses work.
Will: Yes. Good.
Rod: Fair enough. And to investigate antiviral drugs and vaccines. Again, fair enough. And the language put here was, "Biologists have uncovered, uh, some of nature's grammatical rules for what makes a virus, but many remain a mystery."
Will: Uh-huh.
Rod: So these folks used an AI model from OpenAI called Evo.
Will: Wow.
Rod: This is a piece of irrelevant trivia that I knew you needed to know. So I wanted to see if they could get the AI- Oof ... to find more of these grammatical rules that they haven't found yet.
Will: AI designs new viruses. There's the big balls take.
Rod: Yeah, it's, no, it's fine. The good thing about this is it's all fucking fine.
Will: So, okay, so these researchers-
Rod: Yeah, yeah,
Will: yeah ... you know, they, they got AI that was, uh, not ChatGPT, but made by OpenAI.
Rod: Very ChatGPT-ish allegedly.
Will: Yeah, a large language model-
Rod: Yeah ...
Will: trained on genetic codes from a variety of viruses, bacteria, plants, and people.
Rod: Nine trillion nucleotides-
Will: Well, there you go
Rod: stuffed through the system. That is
Will: a lot.
Rod: Nine fucking trillion.
Will: How, how many nucleotides have [00:20:00] you?
Rod: Six.
Will: Well, okay. All right. You know, so trained... And then said, "Okay, produce a particular type of virus that doesn't infect humans" so a bacteriophage. It only, only infects bacteria,
Rod: so. It only affects apparently E.
Rod: coli.
Will: Which is great. So they got 302 AI designs, and then synthesized. So AI designed in computer-
Rod: Yep ...
Will: and then synthesized them in the lab, as in made-
Rod: Yeah ...
Will: made these, made these using the virus-making machine.
Rod: Most didn't work.
Will: Yeah.
Rod: Most didn't work.
Will: 16 did- They did ... proved effective at killing, uh, E. coli bacteria.
Rod: We... I love the... And here's, here's some of the language I read from this study. They said, um, how did they put it? "In one dish, at least, clear dots began to appear in the cloudy film of bacteria-
Will: Yep ...
Rod: which is an indicator of them multiplying," which is apparently really good news. "The viruses burst out of the cells, leaving behind the ruptured husks of their dead hosts."
Will: Yeah.
Rod: Which is great because it means they were viable.
Will: Yes. AKA- Yes ...
Rod: can murder other bacteria. So,
Will: okay. Uh, aside from the benefits, I just wanna talk about [00:21:00] the security- Mm ... of the situation. Mm, mm. So this is, this is research being done in a secure laboratory that's like a... I don- I don't have the, uh, the biosecurity protocol.
Will: It's nine
Rod: miles underground.
Will: Yeah, I don't know. 17 doors. But, but, but- Yeah ... but so this is not bursting out. But this is the messaging point where it's like designing new viruses. Mm-hmm. A, the world has viruses already. Did we need more?
Rod: Probably many. Enough, perhaps.
Will: Using AI to design virus. Now, I totally buy into the, the argument that, uh, we should learn how viruses work, and-
Rod: Yes
Will: and computational synthetic biology, AI techniques can absolutely help us. 100%,
Rod: 100%. No argument.
Will: But there's a bit of me that's like, "So you design some new viruses, and yeah, they can't hurt us"
Rod: No, no, it's okay. It's okay. 'Cause we did it on something that only touches E. coli, so don't worry about it.
Will: Yeah, I,
Rod: I- It's fine
Will: look. But, but I think w- you know, the other version here is, is we can learn a lot of this, and this is also a demonstration of a proof of concept that clearly it is possible- To use the same technology to maliciously create new diseases
Rod: What are you talking about?
Will: I, it, it-
Rod: [00:22:00] What are you talking about?
Rod: Didn't you read it? They just did it on one that doesn't affect humans, so it's fine. It's fine.
Will: Now, I, I know that, uh, it's not easy to do this. Like, let's assume that the lab is well-funded. Sure. It's, it's chock-a-block full of the right equipment, both- Yeah ... the computational equipment- Yeah, yeah ... the data they need to put into it, the genome synthesizing s- uh, systems, and then the testing facility.
Will: Yes. So, so all of those stuff, and you need a bag of PhDes, grads, other people all the way through it to know all of these things.
Rod: Yes.
Will: So I'm not saying it's simple, but-
Rod: But they do exist, and in more than one place
Will: But it has been done now. Yeah. It's not like some lone backyard guy could go- No ... "Hey, ChatGPT, make me a virus."
Will: Uh- It is not that ...
Rod: you forgot a pivotal word: yet.
Will: Well, but the-
Rod: Yet ...
Will: well, there are other th- like, as in you need different machines than ChatGPT can do, but it does- Sure ... it does point to a direction where it's like, oh, some of these are possible. And it's not like your terrorists- Mm ... are zero-funded [00:23:00] institutions.
Rod: The article I was reading said they got a bunch of non-involved people, experts. A guy from Oxford, protein chemistry, says, "Uh, these are not just sickly versions of the stuff that already exists." So that's a good start.
Will: Yeah.
Rod: So we're not talking interesting experimental. "Though they were not radically new creations, they tend to be very similar to natural species.
Rod: They rely on the same underlying biology." But, but- We don't know exactly what they are- No ... 'cause they are truly unique-
Will: Yep ...
Rod: and not found in nature. And what I enjoy is this bit. In the US at least, computer-based research, like generating viruses and DNA with AI, is not prohibited by policy unless it involves an entity of concern.
Will: Entity of concern. Right.
Rod: So like smallpox or something. Okay, like- So if you try and fuck with smallpox, no, no, no, naughty, naughty, naughty. But if you fuck with something else, it's like-
Will: Yeah ...
Rod: we don't know that. And this, this other research says, "So it's easy to tell if a natural virus like smallpox is an entity of concern."
Rod: But- "What about everything else?"
Will: I think actually it is worth putting in context of what the other potential benefits are, you know?
Rod: Sure.
Will: Yeah, yeah, yeah. But, uh- Yeah ... you [00:24:00] know, we could look at, uh, designing virus that, uh, get rid of malaria.
Rod: The cure viruses.
Will: But might solve some significant problems out there.
Will: Absolutely. Or you could look at, you could look at designing viruses or genomes that are, that are drug delivery pathways. So- 100%. 100% ... there, there are absolutely some awesome things here, but wow, does it open up a new world here.
Rod: Yeah. I mean, look, the history of science has taught me one thing. What could possibly go wrong?
Will: Look, we'll probably learn, and,
Rod: uh- I bet we will. And if we're lucky, we get to keep on living to do something with the knowledge that we got. It's just like, ah, it sounds so nice, right? You've got these people who are very clever, and I think the, the main author is a grad student. It's impressive work. 9 trillion nucleotides.
Rod: Fuck me, that's amazing. Oh, yeah,
Will: yeah.
Rod: But-
Will: But this, I mean, this is also... I know we on here have poo-pooed, I know, have laughed at the stupid mistakes of AI.
Rod: Of course we have.
Will: But, but this is actually one of the places where computational AI-assisted research is actually doing things you literally couldn't do otherwise.
Will: Oh, hell yeah. What [00:25:00] you are doing is- Hell yeah ... is quite a targeted approach. You are learning from- Yep ... a whole bunch of data sets that do e- these things- Yeah ... and then saying, "Okay, what can you do, do in this space?" Yeah. And it doesn't, and, and it might be random. Some might, might be errors, but it, but building here- Yeah
Will: you literally couldn't do this without the AI computational technique. So this is-
Rod: Oh, ethics aside and concerns about the future, I love it. I fucking love it. I love this shit happening. I think it's so interesting. I think it's so sci-fi. Love every second of it. Then I go, "Eh, but what if..." But the, my, my gut- Yeah
Rod: my gut goes, "This is fantastic." Look,
Will: look, I think the thing is- "This is so much fun" ... we, we say the what if.
Rod: Yeah.
Will: I don't think there is sufficient security protocols that can actually stop all of this. The lone actor- Yeah ... the, the well-funded terrorist group- Yeah ... or, or the malicious state.
Rod: Yep.
Will: Malicious states absolutely have the capability to one day do something like this.
Will: They can, they can see this- The
Rod: upside is most of them- ... and they can go- ... have a tenth of a brain where they go, "Yeah, but if we release this-
Will: Yeah ...
Rod: it could come back and bite us too."
Will: Yeah.
Rod: Like, as a rule, if [00:26:00] you're smart enough to do this, you're probably smart enough to go, "Ha, implication."
Will: Yeah.
Rod: My fingers are crossed.
Will: Let's hope.
Rod: My fingers are crossed. PDE5 inhibitors. I know you know what they are. PDE- PDE5. They're medicines. They relax smooth muscle cells, and they widen blood vessels to increase blood flow.
Will: What do we use them for?
Rod: Well, the most famous one is sildenafil.
Will: Oh.
Rod: We'll get to that. Started as a potential angina treatment or to treat coronary artery disease, and at first it wasn't that promising, so they fucked with the, um, the doses, and they actually accidentally discovered there were stiffening effects of this drug.
Will: Erections. So we're talking Viagra, essentially.
Rod: We are talking Viagra. So 1998, it became the first widely approved treatment for erectile dysfunction. Viagra was born Since then, other uses have been found for Viagra, but also other versions of these inhibitors. So there's a, another brand name called-
Will: Potato soup
Rod: Potato soup?
Will: Uh, I, I, d- I thought you wanted me to guess some other uses for [00:27:00] Viagra.
Rod: What I can't... And, and I'd, I'd-
Will: Gardening? Like is it, is it-
Rod: Well, look, d- how much do you like gardening?
Will: I didn't mean as literal as that.
Rod: How much do you wish you liked gardening- ... or would, would your partner prefer you liked gardening?
Rod: So Revatio, it's licensed to treat high blood pressure, i- particularly in the arteries of the lungs, 'cause it relaxes blood vessels in the lungs and makes it easy for the heart to pump blood and improves exercise capacity. So I'm surprised, um, athletes don't dope on this. Maybe they do. Other PDE5 inhibitors, they've been investigated or used to treat cardiovascular conditions and urinary symptoms.
Will: Mm.
Rod: So there's, you know, there's a broad brush.
Will: Okay.
Rod: Cialis, which a lot of Americans in particular have heard of, which is Tadalafil, which is a similar drug, been licensed to treat enlarged prostates.
Will: Okay.
Rod: There's also research suggesting Viagra could have other benefits for men who are living with prostate cancer, like for side effects, et cetera.
Rod: And a couple of years ago, researchers from Oxford, they found Viagra could potentially help treatments of, and even prevention of, vascular dementia. So [00:28:00] what that is, is- What's
Will: vascular? Yeah ...
Rod: well, it's, it, it, basically Sildenafil gets into the blood vessels of the brain, and it can improve blood flow and how responsive the blood vessels are to changes in blood pressure, et cetera.
Rod: So what that means is you get improved blood vessel function in people, and particularly those who've had cerebral small vessel disease- Okay ... after a minor stroke.
Will: Yeah,
Rod: okay. So it can help sort of open it up, smooth things out, and get things flowing.
Will: Yeah. Always like things flowing. Like, that's, that's me.
Rod: We like flow. So basically this sort of stuff demonstrates the potential of this, and this is important, well-tolerated, widely available drug to prevent certain kinds, at least, of dementia. So there's ongoing research into all these things. And there's more, and currently the most recent one for cancer.
Will: Are you serious?
Rod: Yeah, I'm very serious.
Will: Viagra for cancer?
Rod: So apparently men, of course it's men, 'cause it's Viagra, men who are taking Viagra before being diagnosed with lung, prostate, or colon cancer had better survival rates than men who weren't.
Will: Really?
Rod: That's the kickoff.
Will: Is there, like, you gotta take it a lot? Is this like-
Rod: Just if they were, [00:29:00] if they were, you know- If they
Will: were using
Rod: I've got, I've got lung, prostate, and/or colon cancer- And
Will: they didn't know, so this is not, like, like-
Rod: Didn't know, no. They were just taking it anyway, 'cause they like to, they like to have a little bit of the inter- Yeah ... intercourses. So they take the pills, and then they got diagnosed with lung, prostate, or colon cancer.
Rod: It's like you guys are surviving better than the dudes who weren't-
Will: Wow ...
Rod: taking the little blue pill, which is interesting. Then it, there's more. Those who were taking Viagra and statins- Yeah ... so for high cholesterol, even higher survival rates.
Will: Hmm.
Rod: So they're going like, "What the fuck?" And cholesterol is key here, and I didn't know this.
Rod: So from the journal Cancer Research, I know you were reading the latest issue when I came in to start recording. Apparently, cancer cells need to get their grubby little cancer fingers on cholesterol to help them metastasize. Do they? To spread. Yeah. Apparently cholesterol helps metastasis.
Will: Oh.
Rod: Spreading of cancer through the body.
Rod: It's so delicious. No, like, they're the stuff of good times. So anyway, sildenafil helps prevent cancer cells from getting hold of the cholesterol it needs, and the quote was, "An unexpected mechanism by which the drug appears to [00:30:00] disrupt tumor metabolism and limit metastatic spread." So it basically disrupts their ability to mobilize and use cholesterol to spread through the body, which is pretty cool.
Rod: So it doesn't actually kill the cancer cell directly.
Will: Yeah.
Rod: But what it does is-
Will: Disrupts how it goes around ...
Rod: yeah. It, it consistently, they say, reduced the ability of cancer cells to migrate.
Will: This is weird. This is so weird.
Rod: Very weird. And in combination with statins, so cholesterol effect plus the blood vessel, et cetera- So,
Will: so what you're saying-
Will: complication ... be on some old people drugs, and you'll be better off.
Rod: Yeah. If you don't- If you get old people diseases ... wanna get cancer, already be on something that you didn't know about and might not be good for you.
Will: This, this is a great treatment plan.
Rod: It's not a plan yet. Yet. Science, as you know, sometimes happens in stages.
Will: No, I get it. I get it.
Rod: Sometimes. So they say after they adjusted for multiple confounds, potential things that would confuse the results, sildenafil use was associated with improved survival among patients with lung, prostate, colon cancer. So there have been experiments done on mouse and human cancer cells, not full bodies- Mm-hmm
Rod: just cells, [00:31:00] including breast cancer cells.
Will: Yeah
Rod: And lung, lung, colon, et cetera. They show that sildenafil reduces the ability of those cells to migrate, to metastasize, whatever. So, like cool as hell in, in theory. But you gotta be careful. Here comes the warnings. The findings are observational, so correlational.
Rod: Analysis-
Will: So don't just go and treat yourself
Rod: Don't just go and start- treat yourself to some Viagra.
Will: I mean, treat yourself, but don't treat yourself.
Rod: So to be clear, as we said, this is not a, a controlled trial, you know, perfect study. The analysis is retrospective too, so that kind of shit happened. People look back and went, "Huh."
Will: Yeah. "
Rod: There's something going on." So you can't determine whether sildenafil was responsible for the outcomes. Also, the study only examined whether there was people who were giving Viagra to patients after they were diagnosed. So it was not s- talking about people who were given it afterwards. So they're not saying, "If you get these cancers, start taking Viagra."
Rod: This is all about people who are already on it- Already doing it ... anyway. But it really did make a difference- Hmm ... it would appear. So we don't know. So my takeaway is simple. One, anything Ozempic and those sorts [00:32:00] of drugs can do or don't or can't do, Viagra will fix. 'Cause Ozempic can do anything. So
Will: you're saying, you're saying a combination of peptides and Viagra-
Rod: Will fix everything.
Rod: Yeah. Shit you didn't even know you had, shit you didn't even know you need to fix. And my other question is, if Viagra ends up helping women with non-erectile dysfunction issues, what could the pill do for us?
Will: Wow.
Rod: They should take Viagra and we should take the pill. Wow,
Will: that is so woke/bro podcast.
Rod: Such science.
Will: There you go.
Rod: I'm all about the science.
Will: Well, listener, if you wanna know about a particular bit of science, send us a question. Hmm. Uh, you can send it to our email address, which is cheers@alittlebitofscience.com.au.
Rod: Slash nothing. There's no slash.
Will: Don't, don't do the slash. '
Rod: Cause that would be a website, not an email address.
Rod: I just learnt that. Technologifier.
Will: So basically Steve, Steve-
Rod: Irwin ...
Will: Gates Musk.
Rod: Steve Gates Irwin Musk.
Will: Yeah, yeah. Irwin.
