SGU Episode 923

From SGUTranscripts
Jump to navigation Jump to search
  Emblem-pen-orange.png This episode needs: proofreading, formatting, links, 'Today I Learned' list, categories, segment redirects.
Please help out by contributing!
How to Contribute

SGU Episode 923
March 18th 2023
923 terminator zone.jpg

"Extraterrestrial life has the potential to exist on distant exoplanets inside a special area called the 'terminator zone' — a ring on planets with one side that always faces its star and one side that's always dark." [1]

SGU 922                      SGU 924

Skeptical Rogues
S: Steven Novella

B: Bob Novella

C: Cara Santa Maria

J: Jay Novella

E: Evan Bernstein


DM: Derek Muller, Australian-Canadian

science communicator

Quote of the Week

The Scientific Revolution has not been a revolution of knowledge. It has been above all a revolution of ignorance. The great discovery that launched the Scientific Revolution was the discovery that humans do not know the answers to their most important questions.

Yuval Noah Harari, Israeli public intellectual

Download Podcast
Show Notes
Forum Discussion

Introduction, Raccoon Dogs, "Last of Us" show[edit]

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

S: Hello and welcome to the Skeptics' Guide to the Universe. Today is Thursday, March 16th, 2023, and this is your host, Steven Novella. Joining me this week are Bob Novella...

B: Hey, everybody!

S: Cara Santa Maria...

C: Howdy.

S: Jay Novella...

J: Hey guys.

S: ...and Evan Bernstein.

E: Good evening everyone!

S: Happy day after the Ides of March everyone.

C: Oh yeah.

B: Ha!

E: All right, yes.

J: Sure if you want that, that's fine.

E: Yeah, that's what [inaudible].

B: Should've watched that scene from Rome yesterday.

S: I like the Ides of March-is one of my favorite days of the year.

C: No, it's not.

B: Really?

C: Why?

S: It is.

B: Wow, that's a [inaudible]

E: What did you do to celebrate?

S: Because March is the transition from winter to spring and the middle of the month is the halfway point. So we're basically in spring now, even though spring's still a week away.

E: It has a ground hog-ish theme sort of to it in a way, but okay.

B: That's not even in my top five.

C: Yeah, why would that be your favorite? Couldn't just like the equinox be your favorite then?

S: Oh yeah, yeah. But we're almost at the equinox. That's the point.

C: I see.

E: Cara, everybody celebrates the equinox.

S: No, mentally for me-

C: [inaudible] Ides of March.

S: March is always that transition. And to put a point on this earlier this week before the Ides of March, we had the worst winter storm of the year just this week.

E: Oh my gosh.

S: And today it's 60 degrees and spring like, so there you go.

B: Right.

C: I see.

B: That wasn't that bad at all.

S: But it was still the worst.

E: Exactly what we got to.

S: It's the first time we had any significant snow on the ground was this weekend. It's already gone. It's a melted, it's beautiful out spring.

J: Beautiful.

C: It's beautiful. Oh, okay. Two things. One, just saw a pretty interesting news alert reading it as I talk, so bear with me, that there's some new evidence and this Catherine Wu of the Atlantic says the strongest evidence yet that an animal started the pandemic. So there's new genetic samples from China, from Wuhan, from that market that appear to link the early COVID genes to raccoon dogs. Do you guys know about raccoon dogs?

S: I'm just learning about their existence, which shocks me. It's amazing. They look, they look-

B: It's that like Texarkana?

C: Kind of. They look like raccoon and dogs.

S: Yeah, they do. They really do look like raccoon dogs. That's the perfect name for them.

E: It's a hybrid.

C: They're really cute.

S: It's a hybrid.

C: But apparently they were also being illegally sold at the market and they may have been shedding the virus. I don't know. There's obviously we got to dig more into this, but it kind of coming on the heels of all this conversation about lab leak and it's just so fascinating that now three years into the pandemic, we're still discovering new things about how it started. I personally only just now got COVID, which is fascinating to me as well. I'm on the tail of it-

B: How's that going?

C: -feeling better. It lasted three days. It's amazing. I started testing positive Wednesday night. I got symptoms Wednesday morning, maybe even Tuesday night. Tested positive Wednesday night took Paxlovid the minute I had a positive test. It was negative by Saturday. I was negative before I finished my Paxlovid, which was Monday.

J: Oh wow.

E: Interesting.

J: That's powerful stuff, man.

C: Yeah, I also think that whatever strain this is is just a quick and dirty. Like it didn't make me that sick, granted I'm like fully vaccinated and boosted-

S: That's what it is.

C: -but a lot of other people I know who had it were also not that sick, but they're all vaccinated too. But it was like kind of quick, just feels like a cold. But something else that I find really fascinating is that, in my sickness, I watched the Last of Us and I'm sort of obsessed.

S: That's awesome.

C: Speaking of.

B: Oh, fantastic show.

C: It's so good, Bob.

B: Oh, my God. On so many levels.

C: So many. Like, and I could kind of care less about like the zombie component of it, even though I get it.

B: Blah!

C: But to me, that's the background.

S: You're right, that's total background.

C: Like that's the vehicle.

B: Yeah, it is. It's a nice background.

C: It is. It works. But like, and of course, I didn't play the video game. I don't, I didn't already know about the lore, but the story and the psychology behind the story. I just, I love when something is ostensibly like a fun, zombie or like horror or suspense kind of show, but really at its depth it's a deep-

E: Rom-com.

C: -question about humanity.

S: Where are you in the series?

C: Oh, I finished it.

S: Oh, yeah. We did a full review on AQ6. It was glowing reviews all around. It was really the character development and everything was fantastic. Definitely recommend it. But getting back to the raccoon dog that I just want to talk about.

B: Oh my.

S: Because this is, we've obviously been following the lab leak theory, both here and on Science-Based Medicine, since the beginning of the pandemic. And what's going, it's interesting, this is, exploding again. What's interesting is that you have like the two communities, which take their different approach to the question are getting so polarized. Like they're both now entrenching if anything. So essentially you could take two approaches to the question of where did the virus come from. There's the scientific approach or the epidemiology, virology approach, right, which is to say, what does the virus look like? What does it look similar to? What's the epidemiology of the early detections of the virus? And then the other approach is the investigatory approach, which is just what happened in that part of the world at the time. So recently the FBI and the Department of Defense both came out saying, yeah, we're pretty sure that the Chinese did this, right, that it was leaked from a lab and they were responsible and they're covering it up. But we can't tell you why we think that. It's secret, but we're-

C: Yeah, that's really helpful.

S: Trust us, we're really sure about this. Scientists on the other hand are going, no, that's not what the scientific evidence shows. It shows the epidemiology. It zeroes in on the Wuhan wet market. This is an animal spillover. There's no signs of engineering. That's one thing. That's why they have to back off from the engineered virus to the lab leak virus because there's no evidence that it was engineered. It's a natural virus. It just accidentally got out of the lab. Therefore we could avoid the evidence from biology. But the epidemiological evidence still points to no way. It was just a spillover from the wet market. So this new little piece of evidence is huge.

C: Right. Because we didn't have that before.

S: They're the one thing. The one thing we didn't have was the animal of origin. But they're saying that they have connections between the virus and viruses in raccoon dogs. That's the last piece.

E: I'm sorry. Who's they? And who's in this context?

C: So it's an international team.

S: Scientists.

E: Who's they?

C: It's an international team of virologist, genomisists and evolutionary biologists. So a big collaboration.

E: Okay. So it's not the-

C: No, this is published-

E: It's not Chinese government.

S: No.

C: No, no, no, no, no. This is international collaboration of individuals basically saying that exactly what Steve said. Like everything pointed to the fact that this basically wholesale seafood market seemed like the origin, but nobody could point to the actual genetic data saying this is the animal that it came from. And now they've basically said, okay, looking at these sequences, the raccoon dogs that were being sold there could have been carrying and shedding the virus in that site at that time.

E: And how will they be able to confirm this for sure?

C: I don't know if we'll ever fully confirm it.

S: Yeah, I mean, it's in the past, right? But all the again epidemiological, neurological data is pointing to the wet market. And now we have a much one more layer of detail, a possible species of origin. We're getting closer to the viruses of origin. This is as close to a smoking gun I think as we're going to get again scientifically. Just doesn't jive with the lab leak theory. And I think my, my sense is that the investigators, right, not the scientists, but the government people, they're basing their opinion on the fact that China is hiding something.

C: And they are.

S: And they are.

C: I mean, China's not been very forthcoming.

S: It's not a compelling argument that there was a lab leak. Because remember, Evan, that China's position is that the virus came from outside of China. They're not saying it came from the wet market in Wuhan. They're saying it was some other country that imported it with frozen meats. It wasn't even China's fault. So they don't like either of the theories, right? And so they are being squirrely. They are hiding things. But that doesn't mean that it, that doesn't support the lab leak theory per se. Because they would also do that to try to deflect interest from the Wuhan wet market theory.

C: Right. But unfortunately, we have swabs. There are swabs of the wet market around when we know the pandemic started. And that individuals outside of China do have access to these swabs. The swabs are positive for COVID-19. But we didn't yet have any indication that there was an animal of origin also implicated in the swabs. Now, after new analysis of these samples, they are saying this animal host could be the host. Whereas previously, the Chinese researchers, when they first uploaded the data and shared it said, we've looked at it. No host can be deduced.

S: Right. So unless the FBI decides to declassify some information that really is like a smoking gun, this is, I think-

E: They won't.

S: -this is the best evidence that we have. This is again, I don't think the lab leak theories ever going to die to be honest with you.

E: No, it'll be a fight for a long time.

C: Yeah. But ultimately, what is our level? Where do we say that the proof is proof enough? So the argument here is, yeah, if we could find a raccoon dog with COVID at that point in time and watch it biting a person, like, oh, there, we saw the spillove event. But what they're saying is, in a single swab, if you've got genetic analysis or genetic material from a raccoon dog, you've got proof that a raccoon dog can carry this. And then you've got in that same swab COVID-19 virus. That's kind of as close as we're probably going to get.

S: Yeah, definitely strengthens that theory.

Quickie with Steve: Mask-wearing update (10:27)[edit]

  • [url_from_show_notes _article_title_] [2]

S: One of the quick follow up before we get on to the news items. Remember I reported a few weeks ago on the meta-analysis looking at studies, asking the question of what physical measures work in reducing the spread of respiratory virus and they concluded that, "masks don't work". And I pretty much tore apart.

E: Cochrane.

S: Yeah, that was a Cochrane review. I pretty much tore it apart. And specifically, the way the bottom line conclusion that was being presented by the lead author and by a lot of people who work for whatever reason, want to believe that masks are not effective in preventing COVID. Since then, and I did get some pushback on email from some of our listeners on that thing, I was whatever trying to dismiss the evidence, although I did a very detailed analysis of why it was crap. And I wrote about it also on Science-Based Medicine if you want to get into even more detail. Well, since then, there have been a number of other medical science communicators who have basically said the same thing that I did, although with sometimes different subsets of details that they focus on in terms of why the study didn't show, did not show that masks do not work. And in fact, the best evidence shows that masks do work at reducing the risk of spreading and contracting COVID in high-risk situations. If you wear the mask properly, like is the one big caveat. If you're not wearing a properly, it doesn't work. Since then, here's the new bit, right? The new bit is, Cochrane now has walked back their conclusion because they, here's what they said, they felt compelled to put out a statement. They said: "Many commentators have claimed that a recently updated Cochrane review shows that masks don't work, which is an inaccurate and misleading interpretation. The review examined whether interventions to promote mask wearing help to slow the spread of respiratory viruses. Given the limitations in the primary evidence, the review is not able to address the question of whether mask wearing itself reduces people's risk of contracting or spreading respiratory viruses. This wording was open to misinterpretation for which we apologize." It's kind of a soft apology, but they're basically trying to walk back the way the study has been misrepresented. They specifically dinged the lead author. They said he specifically went out of his way during an interview to misrepresent the study, did not show what he said, which again was the main point of the criticism.

E: Biased.

S: Yeah, it was clearly biased. So if you are interested in that, David Gorski did a follow-up post on Science-Based Medicine, talking about Cochrane's walk back, also talking, going into more detail on the things that I didn't write about when I discussed it, namely the history of Tom Jefferson, the lead author, who is sort of a chronic denialist when it comes to, like, these old vaccines don't work for this, they don't work for the flu and things like that. So it's not surprising that he would come out with this and he would, basically misrepresent the findings of the study in the media. So he does his usual deep dive going into that angle of things. So I would recommend you could read that for follow-up.

News Items[edit]

New Lunar Space Suits (13:45)[edit]

S: All right, Jay, you're going to start us off with the news items with the new prototype lunar spacesuit.

B: Oh, boy.

E: With tie and matching cumber buns.

J: Yeah, this is a milestone because we really needed these spacesuits in order for the Artemis missions to continue. So how did NASA get it done? So it turns out that NASA and a company called Axiom Space, who is a NASA contractor, they revealed their new space suit on March 15th, 2023. So the version of the spacesuits they displayed, these are prototypes, but they're very close to the final product. And they're going to be used for training purposes later this year. So they're going to finalize these suits for the most part, very soon. These new spacesuits are intended for EVAs or these are for outside of the spacecraft. Anything outside of a spacecraft that's what they're intended for. And the two main things are spacewalks, right? You have to repair the ship, go outside, do something, install something, or you're going to be walking on the moon's surface and eventually on maybe even Mars's surface.

S: Yeah, they hint that they have like multi-mission, and they hint that eventually this could also be used on Mars.

J: Yeah, I'm sure they're going to iterate them. It's not like the old days where they make it and they don't really do anything. We have the technology to make materials and things much faster than we ever did. So we definitely be continuously improving these suits as they go. So aesthetically, they look really cool. If you haven't seen them, I recommend that you just take a quick look. You can Google it in a second and see it for yourself. But the version I saw, there was like a, mostly black with orange detailing. There's a large dome for the wearer's head that gives them a really, really good field of view, much better than any other field of view I've seen. They have a great range of motion. They're easier to get into than previous suits. For instance, these have a hinged back where you can open up the back of the suit. And then the astronaut will enter the suit feet first and kind of to slip into the suit that way instead of having it be done in pieces or like, the upper part, on the older suits, like you had to have the upper part like lower it onto you and stuff like that. It was very awkward. It's also much less bulky than the current EVA suits that are being used on the space station. And check this out. The current space suits that we are using today on the International Space Station are from the space shuttle and they are 40 year old technology.

C: Not surprised.

J: 40 years old.

C: Wow.

B: Wow.

J: Like let that sink in. That most people that will hear my voice were not born. You know what I mean?

S: But Jay, there's space age technology. (laughter)

J: It's just interesting to think, when NASA, they made their last spacesuits, they worked really well. They were able to do what they needed them to do and they just didn't have the time or money to do what's happening today, which is, redo the entire thing from the ground up. Now, let me give you a little backstory here. So NASA was developing a space suit for the Artemis Moon missions on their own. You might remember that NASA showed us a prototype of their space suit back in 2019. We talked about it on the show for sure. Now, that suit was called the XEMU, which stands for exploration, extra vehicular mobility unit. NASA was on track to spend a lot of money, like a billion plus US dollars to complete the space suit project. But the problem was, the main problem was that they were years away from being able to complete it. They weren't going to be able to have them ready for the Artemis III launch, which was absolutely necessary. They had to be ready for Artemis III or Artemis III wasn't going to happen. So NASA realized that they needed to choose a contractor to finish a usable space suit. So they had mostly time constraints, but money was also a factor in these decisions as well. So Axiom was chosen in September of 2022. Now, think about what I just said. Axiom was chosen in September of 2022 and they have their prototype done today, like, as of a couple of days ago. So NASA-

S: And Jay, so NASA already dumped $420 million into their failed suit, which they're now, I guess, going to abandon.

J: No, no, they didn't abandon it. They handed that suit over to Axiom and Axiom took it from there.

S: Is it Axiom or Axiom?

J: Whatever.

S: It's Axiom.

J: Depends on what part of the country you're in, I guess.

S: Yeah. Axiom sounds stupid. All right.

J: Axiom. All right so what happened?

S: They completed it for 228 million, but it looks completely different than the NASA suit.

J: Yeah, but there's a lot of technology in there that was ported over. I mean, they had a starting point.

S: Yeah.

J: So NASA got into a $3.5 billion contract with them. And now this contract is pretty interesting. So it means that NASA will not own any of the new spacesuits. That they will engage in essentially a service contract with that company. And the company will be responsible to provide NASA the suits that they need whenever they need them. And, I guess, however many that they need. It's not a contract that's going to last forever, of course, there is some time limit to it. I didn't, it couldn't find any more details about the contract, but what I do know-

S: You know what it's called? I've got some details. You know what it's called, first of all?

J: Yeah, go ahead, say it.

S: Well, Axiom is going to be providing, "moonwalking services". That's what they're calling it. Moon-walking services. I also found out that NASA, even though they're just contracting out these services, in the contract, NASA gets, whether licensure or ownership or whatever, to all of the knowledge that develops from using these suits on their missions. So they own the tech, basically.

J: Axiom could also sell their suit technology to anybody that they want as well, which is a two-way street.

S: And also they contracted with Collins Aerospace. So it's a competition between two companies. And that's, again, also going to play a NASA's favor.

J: But this is all part of the budding industry, which is going to be space travel. And NASA is encouraging companies to invest time and money into developing technology for this, because all ships rise in the rising tide. They want more space exploration. They want more technology that goes into this so everything gets better. So I totally see what they're doing here, and it makes a lot of sense. I mean, they're going to be outsourcing a lot of future work to companies that normally they would do in-house. They're not going to do it that way anymore.

S: I think it's a good thing.

J: So Axiom ended up taking the XEMU suit, like I said, and they started developing their version of it from that starting point. And they used NASA's suit design and technology, but they made a lot of improvements, and they did complete the build, which is pretty shocking thinking that they've only been doing it since September of last year. They were able to do this for less money than it would have cost NASA, and they were definitely able to do it in a time frame that makes all these things line up correctly. Now the final design of their suit has increased flexibility, better technology to protect the astronauts from outer space in the lunar surface. This includes dramatic changes in temperature that astronauts have to go through. Like, let's say you're on the surface of the moon, and you walk from a lit area of the moon to shadow. We're talking about hundreds of degrees temperature change, just like that, from moment to moment. So it was really important that the boots are insulated to handle really frigid temperatures on the moon's south pole, which spacesuits have not been used that way before. So they were aware that the boots were going to be coming in contact with incredibly cold temperatures on the moon's south pole, so they built that technology in there to handle that.

S: I like the fact that they have an HD camera built into the helmet.

E: Hell yeah, why wouldn't they?

S: Right?

B: What kind of HD?

S: I don't know [inaudible].

E: For a AK probably.

J: So yeah, the-

B: Doubt that.

J: That's true, they also have lights mounted to the helmet. I mean, very much like this is getting very close to all the science fiction spacesuits that we've seen.

E: Yeah, the Martian kind of feel to it.

J: Yeah, definitely. I mean, they have improved flexibility. Like they're able to walk, bend, twist, squat, crouch. Like, none of these body movements were really, you were not able to do that in previous spacesuits. They're very much built for low gravity moon missions. And just so you know, Axioms' new suit is called the AXEMU. Right? So they kind of combined the NASA name with their name. Another substantial upgrade that Axioms made was to make the suits modular so they can fit 90% of males and female body types. Which if you've read into anything about spacesuits, historically, they have a big problem with today's spacesuits because when they want to fit one to a person that might have a smaller body type, they've been having a lot of trouble doing it. There's only so many larges and mediums that they have. But these spacesuits are completely swappable. There's lots of components like you could change the length of the legs and the arms very quickly with the main body of the suit so they could fit basically anybody. And they also have the ability to alter the suits so they fit each individual even better. So inside the suits, there's things that they can do to make them really, fit like a glove and personalize the way that the suits fit. Just one more thing that I found out that when you do look at the pictures of these spacesuits and that you're going to see that they're black and orange, they're actually going to be white when they're put out on the mission because that's to reflect solar radiation because you don't want the suits to heat up so they have to, they're actually trying to keep the suits as cool as possible.

S: Yeah, but just the training suits will be black. This is like one of the last pieces of the puzzle that NASA needed to get to the moon, right? Because it means that as the less is working, that's probably going to all be work out. If they needed the spacesuits and the lunar lander. So now they got the spacesuit. So now they just need like an actual functioning lunar lander.

J: Right. That's coming.

E: Coming soon. Yeah.

S: Hopefully. All right. Thanks, Jay.

End of Life Care (23:58)[edit]

S: All right. Cara, tell us about end of life care. How are we taking care of people who are dying?

C: Right. So anytime there is a news article that sort of, I guess, coalesces is with my fundamental interests and research and clinical practice I'm always excited. So I'm glad when I put this in the list, you were like, yeah, yeah, you can do that. I'm like, thanks, Steve. Because this one I think was especially interesting to me. It's actually about aggressive end of life care. And of course, that's kind of a construct. How do we define what aggressive end of life care is? And I'll kind of get into how the authors did it. And so it's a collaborative group of individuals who published in JAMA Network Open. And the title of this study is "Incidents of aggressive end of life care among older adults with metastatic cancer, living in nursing homes and community settings." So basically, they posed an interesting question, which was, do older adults with metastatic cancer get more aggressive end of life care if they're living in a nursing home versus living in the community? Ultimately, the answer to that question is probably not the most interesting finding in this study. There's a lot of interesting findings, but how they decided to go about answering that question is they looked at basically a large data set. It's a cohort study that used a data set called the surveillance epidemiology and end results database, which is linked with the Medicare database and the minimum data set. And they were able to basically pull out 146,329 older adult patients. I think the average age, which was around 72, no, 78. So even older than that. Okay, the average age was around 78. It was just over half men, 52% men. So it's a pretty robust sample of individuals covering deaths that occurred between January 1st, 2013 to December 31st, 2017. All of these patients, the full 146 plus thousand patients, had metastatic breast, colorectal, lung, pancreatic or prostate cancer. And so they were like, okay, we're going to look at all of this data and we're going to say, what is the incidence rate of those who are living in nursing facilities versus those who are living in the community? What is the incidence rate of what we would consider to be aggressive end of life care? And they operationally defined that a handful of ways. They said they're continuing to get their cancer treated. They are admitted to the intensive care unit. They have more than one emergency department visit or more than one hospitalization within the last 30 days of life. They died in the hospital or they were enrolled in hospice but only within three days of death. So clearly not early enough to get the care that's necessary from hospice. So they saw all of those different parameters as being indices or different constructs that they thought were an appropriate way to sort of signal aggressive end of life care. So I guess before I go on, I'm curious, do you guys agree with those different markers? Do you think that that's appropriate?

S: I mean, they're not unreasonable. But they're just markers.

C: They're just markers and they make that point a lot because it's really hard to know when you're looking at a huge cohort study like this, whether some of these people wanted aggressive end of life care or not. Some people, it is their culture, it is their ideology and it is their deepest wish to have every option made available to them regardless of their age, regardless of their outcome risk. For some people, that's not the, actually for most people that's not the case. But there are some people who will always want the whole book thrown at them. And so it is important to remember that they couldn't tease those people out of this study and that probably did bias the results a little bit. And also sometimes things happen and people may not have been declining as quickly and then they might have died suddenly. That said, part of the smart thing that they do is they looked specifically at people with metastatic cancer. So they weren't just looking at older adults who were generally healthy. They were looking at people who specifically probably had poorer prognoses. And they wanted to know, community dwelling versus nursing homes, what are we looking at here? And they found some pretty interesting findings. There wasn't a huge difference, although there was a significant difference. It just wasn't huge between nursing home residents and community dwelling residents. Who do you think was receiving more aggressive end of life care?

S: I'm going to say nursing home.

C: Yep, nursing home residents. But here's the thing that I find more interesting. 64% versus 58%. So basically still most people are getting aggressive end of life care in the last month of their life. We're talking about within 30 days, that's where those markers were measured. Something is wrong here, when we know that over and over and over, the literature show, and there's so much literature to back this up, that most people want to die at home. Most people who have a end stage, at least metastatic cancer, have things like a DNAR in place. I shouldn't say most people, but many people have things like a DNAR in place, [inaudible] in place, different sort of measures to try and prevent aggressive end of life care. And yet we still see that the general bias is toward fighting really, really hard and utilizing very often expensive, unwanted, poorer outcome, aggressive treatments. When somebody is very often going to die anyway. And here comes this really complicated question about end of life care and how we can do things differently, because advocates have been working on this for a really long time. What do you think is kind of the number one thing? This study points to it, but it's not what this study is about. What do you think is the number one thing that most experts within the medical professions point to that they say is protective against this?

S: Protective against doing too much at the end of life?

C: Yep.

S: Is advanced directives.

C: Advanced directives are really helpful, but I would say it's probably not the number one thing. It's super important though, but the number one thing is early palliative intervention. It's like when somebody is diagnosed with a life limiting illness, start palliative treatment early. And there's so much stigma against palliative care in this country because I think people equate the word palliative with death. But that's not what palliation is. Palliation by definition is easing symptoms. That's all it is. A palliative care physician is an expert in symptom management. Many people who get palliative care live long, full lives.

S: I think they confuse palliative care with only palliative care, which is not necessarily the same thing.

C: Exactly. Which is a component of hospice treatment. And that's why things get a little bit confusing. There's another problem that they talk about in this New York Times article, which I find really fascinating, which is the question of palliative interventions that could be seen as life extending interventions, but they're used in a palliative way. And the complicated rules around hospice. So I'm going to do a quick and dirty on hospice. I'm not a hospice expert, but I've attended a lot of trainings in this area. And I've worked near and around hospice quite a lot. Hospice in this country, I'm very specifically saying in America because there are hospice benefits in other countries, is a Medicare benefit that is basically free for anybody on Medicare to get six months of a supportive treatment as their nearing end of life. They have to have six months or less of a prognosis. And then they have a lot of really incredible benefits that are available to them. But basically the trade off there is that they can no longer receive life extending treatment, if that makes sense. So somebody's dying of cancer. They decide, I no longer want to get my chemotherapy. I'm ready to enter into hospice. Then they move over into a treatment paradigm, which is about comfort. The issue is some people might get palliative cancer treatment. I had a very patient who had a tumor wrapping around her trachea. She was getting palliative radiation. There was no intention that this was going to cure her. But it was making it hard to breathe. And so the radiation was making it so that in her last days, weeks, months, she was more comfortable. And in the article, they talked quite a bit about palliative dialysis, which is a different model than dialysis for like curative purposes. Yet there's still some confusion about what treatments can people have when they're in hospice care. And that may be one reason, in addition to cultural reasons, in addition to stigma reasons, and in addition to just low information, why people often choose not to go into hospice. They see it as a harbinger of death. They fear what hospice means when ultimately anybody who works in hospice sees that actually the outcomes are much better. People tend to live longer. People tend to have better outcomes when they have less aggressive medical intervention and more palliatively focused intervention, which is sort of ironic. It's like paradoxical, but when you really start to think about it, it makes sense. And it's unfortunate because this is, I think, just such a taboo subject still in our society and our culture. We don't talk about it enough. And because we don't talk about it, there's a lot of misunderstandings. There's a lot of stigma. And what ultimately ends up happening as we see time and time again is that people wait way too long to do any decision making around it, and then they make decisions under duress. And that's obviously going to be the least conducive way to have the best potential outcomes. And outcomes are variable. For some people, an outcome is a longer life, sure, but for many people, a positive outcome is better quality of life. But ultimately, the interesting takeaway of this study is that aggressive medical care is really, really common in this country at the end of life. And there are very few situations in which aggressive medical care makes sense at the end of life. Invasive intensive surgeries where it's very unlikely that somebody would be able to recover from that are very often done probably more for psychological purposes than for actual medical purposes. And that ultimately becomes a massive waste of money, and it becomes a very sadly clear way for somebody to suffer needlessly in their final days, weeks, months.

E: Is there an industry that's pushing it?

C: Well, I mean, there is a question around whether medical care, I mean, what industry, like "the medical industry", but like what really does that look like, right?

E: Right. I know a lot of people spend a lot of their dollars at the end of their life for this kind of thing. So, I hate to be so cynical, but there's money to be made here. And therefore, you got to figure some people have an interest sort of in pushing this kind of these kinds of things.

C: But ultimately, the people who make those decisions, if you think about it are the families and their physicians, it's not the insurance companies or the company, like the medical supply manufacturers, the people who stand to make money from the utilization of these things are not the same people that are making the decisions about whether or not these things are appropriate. They are the actual boots on the ground physicians and the families. And I think that yes, that may be a small variable. I don't think we can say 100% that never happens because there's a million variables that we have to tease out. But I think more often, the reason that aggressive medical care happens is because there is a culture of well, we have to do everything we can. Otherwise, we didn't love you enough. Otherwise, we didn't try hard enough. Otherwise, we didn't do our job so well enough. And there's this really intense sense of failure and shame around not doing, "everything". When ultimately doing everything is actually more harmful than doing less. And that's the culture shift that I think we're starting to move toward, the paradigm. But what's required to make that paradigm shift is experts in this field. It's palliative care physicians, it's hospice experts, it's nurses who have expertise in this area. We can't expect that all physicians are trained in this paradigm because they're just not.

S: Evan, do address your question a little bit too. It's a bit of a myth that there's a disproportionate amount of healthcare expenses that are at the end of life. About 11% of healthcare costs are in the last year of life, which doesn't seem out of proportion to me, given that, of course it is, that's when you're going to be sickest, right? When you're dying. And you also have to remember that we don't always know when you're in the last year of your life. A lot of the times we're trying to extend your life and we fail. And you die regardless of aggressive treatment. And so that ends up being a lot of aggressive treatment towards the end of your life. It doesn't mean it was worthless or had no chance of working. Cara is talking about when we know we're in the last year of life because you have a terminal diagnosis.

C: Specifically this study.

S: Yeah, this study, that's a little bit of a different situation. And I agree with what you're saying. My personal experience, this obviously biased by my neurological background, is that overly aggressive care in my experience is almost always driven by the family. Physicians are usually relieved when we come back off on a case that we know has a really bad prognosis. Again, it's biased because in my line of work, a bad prognosis means like your brain is damaged, it's not.

C: Right. So you have, yeah, you sort of have had to go through that kind of deep training.

S: It's also about quality of life. It's their, what's their functionality going to be? If they do manage to somehow survive, they're not going to be neurologically intact. It's supposed to somebody that's a bad heart and an intact brain. If they could survive, whatever condition their heart has, they can have mentally a full quality of life. Neurologically, if you're struggling to survive, you're probably not going to get back to your neurological baseline. But in any case, it's very complicated because you don't, first of all, you don't want to fall for ageism. You're like, oh, you're too old to treat, meant to put a lot of resources into it. And, I think people, you do have a right to aggressive care when it's appropriate.

C: And if it's what they want.

S: And it's what they want. So I think the things we absolutely have to correct are people getting aggressive care they don't want. That's a lose-lose. You have to make sure that doesn't happen,

C: And I think some of it is just the culture of America, it is our fundamental culture.

S: A lot of it is culture, a lot of it is trust, unfortunately. But also there isn't a mechanism to efficiently resolve any conflict. And the default is to do everything. And that may not be the default.

C: Yeah, the default should not. I agree.

S: I think that if, there should be a very efficient mechanism by which if you have, whatever, two physicians say this is futile care, that becomes the default. The problem, but unfortunately, the problem is like if they say, well, this is futile care. Basically, we're going to pull the plug, whatever that means technically in that particular case, you can't take back that decision, right? So like you have to go through the whole process of resolving any conflicts surrounding withdrawal of care before you do it. And I've seen that take a really long time. So I was like, oh, okay, one month and one million dollars later, we're doing the inevitable. That was completely-

C: And that person was potentially suffering, their families were suffering.

S: Exactly.

C: We're suffering that entire time. And honestly, the medical providers are suffering and dealing with a vigorous trauma that whole time.

S: It's very stressful to watch that situation. I have to take care of a patient that you know it's futile and it's not pleasant, you know?

C: And really, this shouldn't be on the shoulders only of the physicians. This should be a cultural shift and a cultural conversation that we're having. And we should all be starting to accept more responsibility around what happens at the end of life. But unfortunately, in America, we are very death denying and this stuff, we tend to sweep it into the shadows. We don't bring it out in the open.

S: All right. Thanks, Cara.

GPT-4 Is Here (40:27)[edit]

S: Bob, I understand that GPT-4 dropped.

C: Oh, no. I mean, yay.

B: Yeah, right? This is earlier than I anticipated. So yeah, it happened, generative, pretrained, transformer version 4 or more simply GPT-4 has been released. OpenAI's latest version of the software that underpins the now infamous ChatGPT, text generating tool is out. So what's the bottom line? Yeah, I know we talked about ChatGPT recently, but this latest update to a full version, full 4.0 from 3.5 has been greatly anticipated. There's been rumors and discussions for months about what it would be like and when it would be released. So I think it deserves definitely deserves a discussion. So if you haven't been paying attention to GPT, well, now's a good time, I think. Trust me. This is only going to get bigger. And if you're, if you read a lot of science news, tech news, AI news, especially, it's all over the place, man. GPT again stands for open AI's generative, pretrained, transformers, which are deep learning AI models that process language, large language models, LLMs, you'll probably see that initialism a lot. So we've trained on huge data sets of human-generated text. There are many applications, including language modeling and there's also translation, of course. And obviously generating very human-like text for applications like chatbots. The previous version 3.5, which was used for chat GPT, pretty much caused an unprecedented AI uproar. Right? Was that a good way to classify that?

S: I thought it was more of a kerfuffle.

B: No, much more than a kerfuffle. Oh my goodness. And much closer to uproar. The art generators like Mid-Journey, that was more of a kerfuffle. This was more of an uproar. Now students are using it to write their papers, right? And countless articles and reddit comments, oh my god, are out there claiming that all our jobs are in jeopardy. And some even audaciously claim that full-blown AI's and the singularity, if they haven't finally arrived, are probably only months away. Oh my god. So yeah, that's, no. That last bit is not true but people, a lot of people are saying it and believing it. So now chat GPT-4 has been released. And I think we're not prepared for what will ensue. It's going to be a hell of a ride. So all right. So what do we know about this release? OpenAI said the following about it in their blog Tuesday, March 14th, the day before the Ides of March, Steve. They said: "GPT-4 is a large multimodal model, accepting image and text inputs, emitting text outputs. That while less capable than humans in many real world scenarios exhibits human level performance on various professional and academic benchmarks. Over the past two years, we rebuilt our entire deep learning stack and co-designed a super computer from the ground up for our workload." I recommend their full OpenAI's full blog on that lot of more details in their blog. So let's get to the meat. Then the question of the day, I think, is what's the difference between four and three point five, right? OpenAI says: "In casual conversation, the distinction can be subtle. The difference comes out when the complexity of the task reaches a sufficient threshold. GPT-4 is more reliable, creative, and able to handle much more nuanced instructions than GPT-3.5." OK. So now to determine quantitatively what this difference is, they tested four and three point five on different benchmarks. Like I said, one representative example was a simulated bar exam. 3.5 was in the bottom 10% when they tested it. 4.0 was in the top 10% of test takers. So dramatic, 90th percentile, exceeds the average score for real humans taking the bar. That's impressive. And a lot of the many other test results that I saw were similar where 4.0 was much more impressive than 3.5. And just many of them were just generally impressive overall. Let's say GPT-3 had a limit of 1,500 words in terms of what you could ask and what it would come back with. 4.O's limit is 25,000 words of text, something like 50 pages. So you can just throw huge, like a huge science article at it and ask good questions about it. And that's what I will certainly be using it for just as a preliminary, like, get an assessment for the research, especially if it's very technical. So the most dramatic changes, though, seems to be this multimodal aspect to it, meaning that you could feed it not only text, but images, which it can analyze based on your questions about it. So if you give it an image, it will accurately describe not only what's in the image, but it'll describe the scene as well, apparently. So for example, Jay is in the picture with meatballs. He's making dinner. I'm assuming it could come out with something similar to that.

J: That picture is on my wall right now.

B: Nice. Now, I can't, when I heard that, where it could actually generate descriptions of what's in the scene, the first thing I thought of was the 30,000 iPhone pictures I have that are always a pain in the ass to search, because of course I haven't sorted them properly, because now it's just way too daunting of a task to even contemplate. But I love the idea of like, yeah, go through all my pictures and then if I do a search on, show me all my zombie figures. It would just bam. All just come back.

E: 29,000.

B: I'm really looking forward to that. Okay. So the image analysis goes even another level deeper. There was a demo where they showed, ChatGPT-4, they showed it in image of a boxing glove hanging over a seesaw, right? With an object on the other end of the seesaw that was that that was lowered to the ground. And it asked, what would happen if the glove dropped? And it correctly predicted that the object would, on the other end, would be launched upwards. So it kind of like understands the physics, he knows what the thing was, kind of, can explain what the object is and anticipate what would happen given a specific scenario. So, all right, how about this one? You can show it a picture of a website, maybe one that you drew on a napkin and it could output the code for a functional version of that website.

J: Wow.

B: So that's, that's pretty cool. That's fairly impressive. I like that idea. At least for the, at the very least, you can get a start on a website. I'm not sure, what the prompt would have to be, you'd have to add many more prompts, obviously, to really get something complicated, I would think. Oh, how about this one? This was even more impressive in a lot of ways. They showed it a joke in the form of a series of images and it showed a cell phone with the wrong charger. And it correctly explained why it was funny. And, being able to deconstruct a joke and express why, correctly express why a person would find it funny. And that's pretty impressive. I kind of liked that, that ability. That said, I didn't see that. I didn't think it would be able to do that, even necessarily with version 4.0, but that's pretty slick. Okay. So some of the obvious quick take downsides is that the image analysis piece is not available to the general public and the explanation that they gave is that they need to more fully assess the potential abuses, like using it for mass surveillance. I mean, if it could recognize right people and things and objects in a scene, it, that would be a hell of a tool for nefarious purposes. Although we have similar tools already. I mean, you walk into any good casino, probably, almost in the world. And they'll, they know that who you are before you walk in the place, generally speaking, but I'm sure this, this has even, more capability and utility. The other thing is that OpenAI for now, at least is very secretive. If more of a closed AI, they not disclosing details about the training data or other information. And that's not good. I mean, it's good for them because they can then build on it and they're not communicating it to the community, though. So the community in general would suffer from this lack of transparency. So I'm not sure, what the plan is for that. I mean, maybe that's just the no brain or business decision that they have to make at this point and to not share all the details I could then be copied. I don't know. So a lot of that is very impressive. And I think we will be discovering even more capabilities that even open AI is not aware of. They're scientists that are using it that were generally impressed, but then they would feed it a lot of technical articles and then they were, they became much more impressed with its ability and also discovering new abilities that they didn't even anticipate. So that's interesting. There's also this idea of, emergent properties as things become more complex. That's something that, that was discovered even in 3.5. So who knows what kind of emergent properties might come out of 4.0? Something that would then, they really don't even, aren't even aware of right now. So but the big but here is at the downsides that have been identified, clearly identified in 3.5, the fears and the abuses of the system, GPT 3.5, it was prone to make stuff up, right? I mean, something that's being called hallucinations. So there's a lot of hallucinating going on coming out with these apparent facts with extreme confidence that are just flat out wrong. And then also, ChatGPT could get like weird and even scary. There's examples of kind of like scary quotes that are like, whoa, what's going on there? There's a lot of, there's a lot of potential issues that have been raised. So what is open AI doing about that? And I got to say that I'm kind of happy with at least initially in my research with the approach that they're taking on dealing with that stuff. They, these OpenAI scientists classify much of this bad behavior as an alignment issue. And if you're familiar with the artificial general intelligence discussion groups, there's a lot of talk about the alignment problem. And alignment in the sense that AI is not aligned with our human values, for example, right? So OpenAI is saying things that sound good from what, from what I could tell so far. They seem to be taking these issues very seriously. For example, they say we've trained language models that are much better at following user intentions than GPT-3 while also making them more truthful and less toxic using techniques developed through our alignment research. The result they claim is that GPT-4 is our, they say our best ever results, though far from perfect on factuality, steerability and refusing to go outside of the guard rails. So it's better than GPT-3 or 3.5 in those areas. I think it's going to be very hard though to, I don't think we could really reasonably expect them to do away with all of it. There's always going to be probably some sort of fact problems and kind of like going over the guard rails from time to time. I think they can lessen the incidences happening, but I don't think they're going to really get rid of, you're not going to get rid of it 100%. I think because then at that point if they did, it'd be changing the fundamental aspects of GPT too much. So I think it's going to be there to some degree, even with GPT five and six or whatever. The Bing version 4.0 has been available for the past month, kind of like secretly, swapped out 3.5 with four for the past month, which was interesting. So, you can use it now. There's a waiting list and also a waiting list to use it for APIs, but people are using it now. Like I said, the image analysis aspect is going to be delayed while they look at it. So a lot of this is interesting and fascinating and has amazing potential. But as I said, the downside, the potential of this amazing tool to weaponize this information and misinformation, it just gets scarier and scarier the more I think about it. This tool has passed the conversational uncanny valley. That hill is done. And with humans and their overactive agency detection, it's just way too easy to fool people into thinking, to be fooled by these tools.

S: So its passed a Turing test.

B: Oh, yeah. Turing test is so antiquated at this point. It's not what we thought it was even just a few years ago. People are just too easy to be fooled and that makes it very dangerous. And I think we really need to put in some very thoughtful safeguards. I read some of the comments made by philosopher and cognitive scientist Daniel Dennett, Steve, making interesting points. He's made some interesting points. Let me do some of his quotes here that I found he's very interesting. He said regarding the technology's creators: "They should be held accountable. They're on the verge if they haven't already done it of creating very serious weapons of destruction against the stability and security of society. They should take that as seriously as the molecular biologists have taken the prospect of biological warfare." That was really interesting and I think there's a lot to that statement that I think a lot of people need to be thinking seriously about. He says that this is the real code red. "We need to institute new attitudes, new laws and spread them rapidly and remove the valorization of fooling people, the anthropomorphization. We want smart machines, not artificial colleagues." I really like that statement. He's right. We want the smart machines and the artificial colleagues, at least the way the technology is now and the way it's going to be in the near future, artificial colleagues can be very dangerous, very, very dangerous. As much as I hate to potentially slow down some of the advances in artificial intelligence and all the benefits it can bring to society, this is one of those disruptive technologies that I've talked about for a long time, although this specific application involving large language models was less obvious, I think, and predictable, than the easier low-hanging fruit, like say nanotech gray goo. Yeah, of course, that's something that's easier to spot. Not too many people really saw this specific scenario coming. We lost out on potentially having more years of discussing it and preparing for it, but it's here and it's now, and we got to reap its benefits, but we also have to be very, very careful because this can be very dangerous. And Daniel Dendett has some interesting things to say about it.

S: I think that this is going to be a story we're going to be talking about all year. And we'll end it to the next few years. New iterations keep going. We're still on the steep part of the curve here. New iterations are going to come out. By the time we sort of settle into what one version could do, the next one's going to be here. It feels like that's what happened with GPT-4.

B: And I'm sure there's going to be revelations. Even over the next few weeks there could be many revelations, and we may even need another update on four and what four is about and what's been discovered that isn't really in the news even now.

S: All right.

Terminator Zones (55:45)[edit]

S: You guys know what terminator habitability is? Habitable?

C: What? Habitable.

S: Terminator habitability.

C: Like the Terminator?

B: The way robots, and where they live and what kind of environments they like, that kind of stuff.

E: Hasta la vista.

S: This is the idea. We've talked about this before. The idea that there could be a habitable belt around gravitationally locked planets, or-

B: Tidally locked.

S: -dwarfed stars. Yeah, they're tidally locked. So that one side of the planet always faces their parent star. Yeah, the side always faces away.

B: So one super hot, one super cold, and in between it can be potentially habitable?

S: In between which is the terminator. Right?

E: The Goldilock's part.

S: That's the terminator in between. So that could be habitable. So the news is that astronomers have for the first time used climate modeling software to model what the climate on such a planet might be.

B: What took them so long to do that?

S: I don't know. But here, previous research they say in the study, previous research focused on eye region habitability. Have you ever heard of that Bob?

B: Eye region?

E: E-Y-E?

B: Spell it, what kind of eye?

S: E-Y-E.

B: What are talking about?

S: In other words, just the-

B: The storm type of thing?

S: -just the part of the planet facing the star is habitable because it's at that, it's at it, towards the outer part of the habitable zone. So that, the part of the planet that's always getting light-

B: Just warm enough, it's just warm enough.

S: -it's warm enough to keep water in liquid state. And then the rest of the planet is frozen. This pushes it in between a hot and a cold zone, again, sort of this sort of equatorial terminator zone.

E: The planet's kind of locked to the star at that distance though? I thought they always had to be close to the star.

S: It's still close. That's still close. Because M dwarfs are, they're red dwarves. They don't give off a lot of energy, a lot of heat. So you have to be relatively close. In order to be close enough that water could be liquid, you probably will be tidily locked. All right. So this is what they found. Now, here's another reason why previous studies suggested that this may not be possible. Because they were focusing on water rich planets. Right? Like if you look at the earth, which is a water rich planet, 75, whatever percent of the surface covered by water, if we were tidally locked around an M dwarf, then the ocean on the side of the planet facing the sun, facing the dwarf star would evaporate and would cover the planet in a thick cloud of water vapor, which is not good.

B: Wouldn't it freeze and drop on the other side on the far side?

S: Yeah, but what would happen though? I mean, it would be a constant cycle of that happening. And it would be the shroud of water vapor that that's what they, that's what the model showed, right? It would be a cloud of water vapor that would not be good for habitability. But they said, all right, but what if this is the new bit, what if we model a planet that's tidally locked around an M dwarf that's not water rich, right? That's a little bit drier. And so that the side of, so what would you say happens, Bob, there's very little water on the side of the planet facing the star. And that water evaporates and eventually whatever it's a desert. So you have a desert on the side of the planet facing the star. You have glaciers on the other side and you have the terminator zone of habitability. What would, could that work? What would that be like? The modeling that they published said, yes, that could work. You could have a terminaor habitability on those planets.

B: It's awesome.

S: Yeah, it depends on a couple of variables. They had to, they said again, they took climate models used for Earth. They had to be able to say, okay, but we have to change the rotation rate, right? To match a tidally locked planet. But otherwise they used the models that are used for Earth's climate. And they found that there would be limited energy transfer, right? If you don't have an ocean on the front side of the planet that really limits the heat transfer, so that you do end up with that terminator zone with hot and cold separated. Because other previous models also suggested that with a lot of water, there would also be a lot of heat transfer, which would cool the hot side and warm the cold side. And so there would be actually a broad zone of habitability, but then that was complicated by the problem of too much water vapor in that model. So maybe this might be the model that works, right? The terminator habitability. Now the reason for the interest in this is because M dwarfs are the most common stars in the galaxy, right? In the universe, there is like 75, 80% of the stars out there are M dwarfs.

B: Yeah, man a lot.

S: And so if they could potentially have planets with habitable zones, with life on them, that dramatically increases the amount of life that they would be on average, the density of life-sustaining planets. There are still some problems, though, right? The big problem is that the red dwarfs are unstable. And early on in their life, they're massively unstable. They would definitely rip away the atmosphere of any planets that were in the habitable zone. They do settle down relatively, but they're even in their mature settled down state, they're still more active than say a yellow sun.

B: And they're close to that. It might be pretty bad.

S: Yeah, so it's still an unresolved, I think, debate if there is a sweet spot where you could have a planet that either migrates in later in the age of the star or reconstitutes an atmosphere somehow, through outgassing or whatever. And would the star be settled down enough that the planet could hold on to the atmosphere? Maybe once it's tidally locked, where most of the atmosphere is away from the star, it's on the other side of the planet, would that be enough to hold on to enough of an atmosphere? I don't know. I think that's still an open question, but clearly astronomers have not given up on this idea. And this is the latest research in that. And still, I think we're left with the conclusion that the most likely candidate for life are orange stars because that's the sweet spot. It's probably orange stars. Because again, the cooler the star, the longer they last, but the cooler they are. Red dwarfs are too unstable. Yellow stars, like our own, obviously can have life on it, but they live for six to eight billion years, something like that. Orange stars live for tens of billions of years, 20, 30, 40 billion. And they're stable and they have a nice habitable zone where you don't have to be tidally locked. So they might be in the sweet spot for life. In terms of the number of stars out there, right? So most stars with life may be orange stars.

B: Giant stars? No, they go, even though they have much, much more fuel, they go through it.

S: Blue stars, no.

B: Hundreds of millions of years. They live like hundreds of millions of years, not very long for life to evolve. They have the biggest ones.

J: So what makes a star have a different color light?

S: It's just their mass. That's it.

J: Oh, it's all about the size.

S: The bigger they are, the hotter they burn and the, yeah, the, and temperature is color, right? So they go from red to orange to yellow to, to blue. And there's no green stars.

C: What about the white ones?

S: Yeah.

C: Are the white ones weird? Sorry, aren't there white dwarfs?

S: Yeah, white dwarfs, but they're not.

C: Oh, you're right. They're not really stars.

S: They're not burning hydrogen. They're not white because they're hot. And just so some of like, there's no green stars, because of the physics of how it works, it's, I know, I remember Phil Plait had a good explanation for why they're on one of them.

C: And white dwarfs are like remnants, right?

S: Yeah, they're remnants. I think hey go from yellow to blue, basically, is what happens. Yeah, as you get hotter, just as a feature of, of the temperature. You can get bright yellow, almost white from what I understand. But then then it converts into blue as they get hotter.

B: Yeah, I'm seeing here that stars, harder than the sun also look white. It isn't until you reach temperature around 11,000 Kelvin before it starts to look blue from our perspective.

S: Yeah, yeah.

Ohio Chemical Spill (1:04:53)[edit]

S: All right. Evan, give us a quick update on the Ohio chemical spill hubbub.

B: Oh my god.

E: Yeah. And we've had some requests to give maybe a little update on this. It happened a little while ago, but I'll get you up to speed. Here's the background. February 3rd, train derailed in Ohio in a town called East Palestine. There were total of 50 cars on this train. 38 of them derailed. All of them were eventually damaged by a fire and subsequent controlled burn that followed. There were chemicals being transported and nearly a million pounds of vinyl chloride, along with nine other substances that are deemed hazardous to people, animals, and the environment. The company that runs the rail or at least this particular train, Norfolk Southern Corporation, their officials say that some of the cars carrying the vinyl chloride were not breached during the crash, but they were at risk for exploding. So a few days later, February 6th, the crews released and burned those contents and created a massive smoke plume above the town for several hours. This is all within a few days of obviously everything having happened, it made enormous news. That's the very basic background information. Some of the recent developments have been concerning the health and well-being of the residents inside and on the outskirts of East Palestine. So this particular update comes from a joint report from the Ohio Department of Health and the US Department of Health and Human Services. Some residents participated in a health assessment survey and they released the result of those surveys or the survey. It showed that the most common symptoms being experienced by residents include headaches, coughing and anxiety. So 168 surveys were completed and of those 125 people reported headaches, 108 reported anxiety, 103 of them reported coughing. Others were fatigue and tiredness, 97 of the 168 participants. And irritation/burning of the skin, 88 of them, almost-that's about half. As part of the same report, they say that the water sample results from private water systems of East Palestine homes continued to show no harmful levels of contaminants. There were 157 private systems tested, none have shown harmful levels of the contaminants. So that's a good sign. The derailment and subsequent control burn required that evacuation orders had to go out for people in a two mile radius. But they were allowed back shortly after the February 6th control burn. But anyone not heating the orders at the time were told to remain indoors a few days later by the 8th, by February 8th people said that it was safe to return to their homes. This recent report basically reemphasized that for anybody who for whatever reason is not gone back at this point, it is safe to go back to your homes. They're still testing the water. They're still testing the air and they've not detected any concerning levels of contaminants. This particular report also addressed the issue of animal necropsies. Necropsies? Necropsies.

S: Necropsies.

C: Necropsy.

E: Animal necropsy which is an autopsy of animals that had died at the time of the derailment. They said the preliminary reports on six deceased wildlife animals submitted to the Ohio Department of Natural Resources and Ohio's Department of Agricultural Animal Disease Diagnostic Laboratory they showed no evidence to support chemical toxicity as a cause of death. Because one of the first things when this was all new and occurring is that there were wide reports of wildlife having died within hours in first 24-48 hours of the spill having occurred. But preliminary and everybody's upset about everything and you're fast blame the spill on it, but they've not been able to find the evidence to support that the toxicity is the cause of the death of these animals.

C: Who is they in all of these equations? Who's actually doing all of these studies?

E: So these are, again, the report came from the Ohio Department of Health and the US Department of Health and Human Services. So they're affiliates or whomever the departments that report to them are work in concert with them are the ones. State level agencies and federal level agencies. But there's more. There was another report, an independent assessment that came out. This one came from researchers from Texas A&M University and Carnegie Mellon University. So what they found is they did air sample of East Palestine and they said that there are detectable levels, higher than normal levels of certain chemicals in the air. Acrylene, benzene, vinyl chloride, butadiene, naphthalene, trichloroethylene and trichloroethane. That's a lot. So these are, yes, so these are among the chemicals that the train was obviously carrying and there are some detectable levels in the air. However, the results also showed normal levels of what are called the BTEX. BTEX chemicals. Benzene, toluene, ethylbenzines and xylenes. Those are associated with petrol products. Those are known to have carcinogenic effects, but they said those levels were nominal. So that's some more good news. But they said the acrylene is perhaps the one that had the highest concentration. And what they basically said is that it's as if you were in a major city. That's the kind of level of acrylene that you're experiencing in these air samples in the East Palestine. So it's unusual for East Palestine and it's up by quite a bit of numbers. But I suppose if people can live in cities with these kinds of contaminants in the air, it can cause problems with people with pulmonary issues, breathing matters and things that people have to deal with on a daily life. It's sort of like living in a major city at that point. But they said if it's prolonged exposure and it does continue, there is definitely need for people to be aware of that. And it can definitely cause some damage. They're continuing to monitor the situation and to see if the levels will be dropping. Hopefully they will be over time. But as of right now, that is what's going on. For health reasons, obviously, they're going to continue monitoring this for a very long time. They're going to continue to test water, continue to test the air. But for the moment, for the moment, it looks like people are not in any immediate danger just because they are residing there. But stay vigilant on this and get your water tested if you're on a private water system.

S: Yeah, if you have your own well, definitely get your well water tested. Do you know how many derailments there were trained derailments in the US in 2022?

E: Over 1000, I think.

S: Yeah. 1164.

B: That many?

S: Three a day. Three every single day. On average.

E: Yeah, it's part of a larger conversation about train safety.

S: Yeah, so these are mostly minor. Obviously, it's not this kind of derailment every day. Most of them are in trail yards. They're mostly caused by human error. People screw up and they go off the track. And I think 14 people died. I'm sorry, 14 people were injured, one died in 2022 as a result of derailments. So that's why it's not major news all the time. But the issue has been that, especially since COVID, but also just to, because of deregulation and maximizing profits, trained cars are getting longer and longer, the larger number of cars being pulled at a time. And they haven't been forced to update some of the breaking and safety measures. So this is definitely, this is an opportunity to rethink the safety issues.

C: And is the infrastructure of the actual tracks like good? Was it being maintained or is it like super old?

S: Apparently, it's good. Apparently, the train tracks are good and they've been improved since like the 70s, 80s, et cetera. So that's really not the issue. Again, the number one reason for derailments is user error, right? It's human error. Number two is broken rails, is physical error on the tracks themselves. And then, obviously, the trained cars can metal can fatigue. And I think that's what happened here, probably there was fatigue on one of the cars. And it caused something to break. And that's obviously, once one goes, the whole thing goes.

E: Yes, yes. There's video of a fire that occurred underneath one of the cars. And this was like 20 miles before the actual derailment.

S: Yeah.

E: So, a ring system caught video of it as the train was passing by. And one of the undercarriages of one of the cars was on fire.

S: Right. And that was the other way. Why wasn't that picked up? Why couldn't break in time?

E: The detectors failed to, right.

S: Detectors, they didn't have enough detectors. They're only required to have so many and like every so many miles, et cetera. And anyway, so there's definitely going to be a pulse, I think, a regulation to try to nail this down until there's the next major accident and then we'll be hand ringing all over again. But overall, statistically, things are getting better. Obviously, there's still some issues. Okay, thanks Evan.

Who's That Noisy? (1:14:52)[edit]

S: Jay, it's Who's That Noisy time.

J: All right, guys, last week I played this Noisy:

[Repetitive, whirring, buzzing, and beeping of various intervals]


B: It's an otter.

J: It's a sea creature.

C: Talking sea mammal.

S: It sounds like one of those old car alarms, where they cycle through a bunch of different annoying noises.

J: Yes, a hundred people wrote that, that one and answer.

C: But it's clearly not a car alarm. It's like a dot matrix printer sounding like a car alarm. What is it? I'm so confused.

J: William Steel wrote and he says: "Hi, Jay, finally won a Who's That Noisy? My wife suggested that this week's noisy sounds like a home security alarm. Maybe one that is a bit older." So yes, that, as I said, the alarm, a lot of people wrote in car alarms and house alarms and things like that. It's not correct, but there's nothing there either. There's nothing you're doing is wrong, but it's okay. Because things sound like other things. I've been, this is what I'm saying about Who's That Noisy. So many things sound like so many other things. Another listener named Ed Mason wrote in and said: "Hey Jay, this sounded to me to be similar to the encrypted messages that missile silo commanders would receive that would contain messages, the weather or even the orders to launch the missiles." So basically saying, is it military, some form of military communication? That's a very interesting guess, but that is not not correct. I'll move on to the next one. Benjamin Greenberg wrote in said: "Hi Jay, this week's noisy reminds me a bit of laser etching for metal items." He says: "However, I'm pulling a hard turn and saying it's military targeting systems." Not correct, but very interesting guesses this week. And another person named Mike Todorovic said: "Jay, you legend, hope you are well. And this is this week's noisy, a dot matrix printer." Oh my god. If a dot matrix printer made that noise, they would not have survived the six days or seven days or whatever the hell people had you.

C: I mean, it made a pretty horrible noise. Let's be honest.

J: But it wasn't that horrible.

C: It just didn't cycle like that. It was like one fifth of that cycle. It was like the first.

J: Yeah, it was just, yeah, you're right. It wasn't like changing the horrible noise to another horrible noise. So the winner for this week, Stephen Liss, he writes in: "Having spent too many hours in one myself, that sounds very much like an MRI machine." And indeed, that is an MRI machine. Listen again [plays Noisy] it's almost as if the person who designed the sounds for this thing said, let's make it as annoying as possible when people are stuck in this thing for hours.

C: I think the reason it didn't resonate for me is it sounds like that was recorded, I could be wrong from outside the machine, not from inside the machine. And I don't think I've ever administered an MRI of only ever had one.

J: Yeah.

E: And what did you hear inside the machine?

C: It's the same. It's just way freaking louder. It's like in your brain.

E: Geez.

C: It's so loud you usually have to wear headphones in there. Or not headphones on - earplugs.

J: Yeah, they're loud.

C: They're so loud.

J: So many people wrote in basically telling me like why they recently had an MRI and what the experience was like. But yeah, I mean, this is a piece of technology that when you need it, you need it. And most of us don't experience it, but many of us do.

New Noisy (1:18:21)[edit]

J: I have a new Noisy this week and this Noisy was sent in by a listener named Benjamin Treblue. And here is the Noisy:

[Background hissing then light squawking/squeaking honks]

J: Okay. So it's that shrieking noise that you hear. That's the one I'd like to know. What is it? What the hell is that? Steve, do you have any idea?

S: No.

J: Okay. If you have an idea you can email me at

S: All right. Thanks, Jay. Well, we have a great interview coming up with Derek Muller. Veritasium. So let's go to that interview now.


Interview with Derek Muller (1:19:02)[edit]

  • Derek Muller, an Australian-Canadian science communicator, filmmaker, and television personality, who is best known for his YouTube channel Veritasium

S: We are joined now by Derek Muller. Derek, welcome to the Skeptics Guide.

DM: Hey, thanks for having me.

S: The listeners out there may know Derek as Veritasium, a very popular science and technology communicator on the internet, the YouTube's and we're just great to having you on the show. We've been watching your stuff for a long time. So tell us a little bit about your journey to becoming Veritasium. What how did that end up being your career path?

DM: Well, I think I always wanted to be some sort of creative person making videos or films or something like that, but it seemed like a terrible life choice, especially when I graduated back in the year 2000. There were no careers in social media at that point. So I feel like I did the smart thing and I studied engineering and physics at college and then I went on to do a PhD at the University of Sydney. I was trying to merge these passions in creative fields, performance and stuff like that and the sort of film side. I was trying to merge that with my interest in science. So the PhD was actually like how do you make a film that will teach someone about physics? So that's what the PhD was, which you know now seems pretty relevant, but at the time definitely seemed weird. Like I was sort of painting myself into a corner there because it didn't really seem like that would be an employable skill set. But yeah, I got to a certain point in my life when I was like financially stable enough to just say, I've been essentially working on backup plans up until the age of like 27 or 28. And that's when I said like, I can take a chance right now and I can start making YouTube videos. So yeah, that's kind of how it got started was just as this outlet for a desire to be creative and to, I guess make videos that I thought were sort of meaningful or maybe helpful for people. So that's the goal. That's where it started.

S: So you actually have a PhD in making YouTube videos about science.

DM: Well, at the time YouTube didn't really exist. So it's just a PhD in how to make films to teach physics. Yeah, that's where it started.

C: What department was that in? Like how did they do it at the University of Sydney?

DM: Yeah, interestingly, there was a movement that started, I want to say like maybe 70s or 80s where physics departments started taking on education. I think there's a couple of reasons for that. One of them is because physicists are just like we can do everything better than everyone else. So, it's like, well, what do these education departments have that we don't? So the physicists started looking at their own teaching and like claiming that physics was like this unique domain where you needed unique skills and unique approaches to teach it. And so there was a whole field of research that began and it's called physics education research. So when I saw that that was a possibility to join up with a physics education research group, it felt like the perfect option for me to merge those interests in physics and films. So, yeah, I was in the School of Physics. I was also co-supervised in like psychology and education. So yeah, cross-disciplinary.

S: And in your videos, of course, span a wide range of topics, beyond physics and engineering. So can you walk us through a little bit of the process you go through and when you decide on, this is my next video and then what do you do?

DM: Yeah, I mean, every video is sort of a torturous process to figure out exactly how we're going to do it. I think like I've avoided just being a formulaic channel where it's like, well, in every video, I just do a script and that's it or something like, I hope that keeps things interesting for the audience. For me, it's just, I think it reflects my personal feeling that I don't like formulaic shows. I don't like formulaic movies. I don't like anything where there's like a single goal and like, we want to throw the ring into the mountain or something and these guys want to stop us. Anyway, like, the stuff like that just doesn't appeal to me. So I'm always about doing something that sort of feels new and it's sort of an adventure and hopefully the audience doesn't really know where we're going and I don't really know where we're going either. So I would say each video is like, it's a big battle and I think the ambition is to make the best video on that topic that is out there. That's always our goal, whether we achieve it or not is another question. But it requires a lot of research, writing and thinking and talking to experts and just trying to get a handle on the topic in the best way that we possibly can and then, yeah, turn it into a YouTube video. Like it's, I feel like the process is just a mess. Like it's a, I feel like every video is a new battle and it's probably not the best way to go about making a YouTube channel, but that's what we do.

S: You do what it takes, whatever that is. And you sort of have to go.

DM: Basically, yeah, basically. I'm trying to think like we wanted to make a video on, on this like rods from God idea, like the US military idea of dropping tungsten size telephone poles from space.

B: Oh, yeah.

DM: And they, they would come in and, and hit the earth at like mach three and we kind of wanted to get an idea or get a sense of like how much energy is, they're in kinetic energy. And so our plan was to go out to the desert and load on as heavy steel weights as we could onto like a helicopter and drop them from as high as we could, safely do so. That turned out to be a massive challenge. So, feel like I learn a lot, I guess with every episode, but yeah, it's often about doing weird stuff like that.

S: Yeah, that was fun. I like the fact that, because you could talk about that one specific thing, like how much energy is there in kinetic energy. But just that simple idea of we're going to drop rods from space onto stuff to damage it. There's so much physics involved, like why tungsten, how do things heat up when they fall into the atmosphere and, how do you target all these, there's a lot of physics on the way to looking at just the basic question of dropping stuff.

B: It's fascinating.

DM: Totally. Yeah, no, absolutely.

S: Yeah, that also seemed like the most mythbuster inspired episode that you did.

DM: Yeah, well, I mean, we were doing this episode and someone on my team was like, should we contact Adam Savage and see if he wants to come. And so we did and he was game to come out. So, that was, another fun aspect to make in that video. And of course, like, he had so many good ideas for what we should have done and how we should have planned it. I still feel like, like we're rookies in a lot of ways. So we make a lot of rookie mistakes. And I think in that video in particular, we got roasted in the comments for like not being better prepared. But simply like pulling off, what we were able to pull off was, was a real challenge. So yeah, everything's hard. And of course, everyone wants to be a Monday morning quarterback and say, like, oh, you should have.

S: Of course.

DM: Yeah, we should have. We should have. But it's not common for us.

S: Yeah, you try making videos, buddy.

E: Let's see your channel. Seven million views. Have you started any of these projects or explored anything, but you had to abandon it. You couldn't go through with it for whatever reason. Couldn't get to the end.

DM: Yeah. I mean, there's all sorts of ideas, some of which, I won't share because maybe they'll still come to fruition at some point. But, some videos we get down the track and it's just not quite what we thought it was going to be or it's not quite the, it's not not hitting the way we were. We felt it would, I would say in the early days of making Veritasium, like the videos were short and easy and, you make it in a day and it's no big deal. And so, if you have one that's like, not as good as you hoped it would be, you can still put it out there. It's no big deal. But now that the channel's got, like 13.5 million subscribers, I feel like there's a certain pressure to make sure that the videos are all great. And so if you get to a point, even if we've invested, say, a month in research and writing and doing some animations and things and recently I looked at one of these videos and I was like, well, it just doesn't hit, doesn't feel as strong as I think it needs to. And so we sort of, we can them sometimes, which is painful, but, I think it's necessary.

C: When you approach a video, maybe now versus when you started, how much of your approach is trying to hack the algorithm, is trying to figure out exactly how you're going to make something that is going to, generate clicks and generate revenue? And how much of it is this passionate, like, I am who I am, I'm making this for me and for who I think my core audience is. And you don't like it. Like, how do you find the sweet spot between those two things?

DM: Yeah. I mean, that's a great question. I feel like I've lived through a lot of shifts on YouTube. There's some like overarching shifts. Like, YouTube has pushed us essentially algorithmically to make longer videos. So, you can plot out the length of my videos from the early days and the average length goes from like two minutes up to over 20 minutes now. So, that's one clear shift that we've made. And I would say, how do we embrace that? Well, now we have to pick a video topic that deserves 20 minutes. So back in the day, you might just do something quick. I guess these days we would turn that into a short. So like, this is the new phase. I feel like the career of being a YouTuber is kind of like living in on sort of like shifting sands and you have to keep moving one way or the other, to keep your balance and to keep moving forward. I'd say there's been, there's been so many shifts over the years, the shift to watch time was a really important shift, which again drove that sort of longer content times. But I would say that other shifts like, people scrolling on mobile, like mobile becoming the main platform where people are watching the videos and the fact that YouTube pre-rolls the first, five, ten seconds of the video or whatever, just as you're scrolling. So now we have to find ways of utilizing that first five, ten seconds in a really effective way as opposed to just saying, like, putting up a sponsor card or doing something like that. So, these days, I think we're very conscious about building out that first, probably 30 seconds and making it like a sort of a sizzle for the whole video so that if someone wants to get a sense of like, what is this video going to be? We try to like put that in the first 30 seconds. I think it's interesting how in that way, we've shifted back to a TV type model, where in a lot of TV shows will like tease the whole show up front, right? So it's like we're rediscovering the same techniques to hook people on YouTube now. So definitely that's gone that way. I would say, Cara, you raised the question of like, am I making these videos for me or am I making it for the algorithm? I would say, I think that trying to work within the YouTube system has made the videos better in a sense. I think it has forced me to grapple with some things. I mean, not always better, but I think like because YouTube hopefully figures out what the audience wants and what audience like watching, if you try to make things that are good for the algorithm, hopefully they are also good for the audience. And that's not again, it's not universally true. But I'd say like my pacing is picked up, my sense of storytelling I think has improved and just the ability to like, really cut out the chaff and stuff like that. So I would say like a lot of those skills have been refined over 12 years of doing this. At the same time, I feel like sometimes it's just really hard to make videos. I mean, I think the hardest part of YouTube is maybe the title and thumbnail side of it because we kind of have to regard that as equally important to the video, which is unfortunate from our perspective because we just want to be great video makers and we want to make a just a killer video. But the problem is if no one clicks on that title and thumbnail, you're never going to get the view on the video. So it's like, because that is the gateway, yeah, we really struggle.

C: Yeah, it's interesting because as you mentioned, like adjusting and in some ways bending to the algorithm is ultimately helping you refine your craft. But at the same time, you are still constrained by your foundational ethic of science content. So it's not like you're saying, okay, well, what does the best? I'm just going to do like a, I don't know, like a gaming video or a room review of like all of my stuff in my room or an unboxing video. Like you're not switching your content. You're just refining within those parameters.

DM: Yeah, I mean, that's a good point. I would say like we also are drawn to topics that will make a good title and thumbnail or something that's really clear. For example, the video we just published was like World's Strongest Magnet. So I've done a lot of these sort of superlative titles over the years of like World's Roundest Objects, World's Heaviest Weight, World's Highest Jumping Robot. And there's just something about those titles which is simple, clear, obvious and compelling to. And there's a large, a large audience. So yeah, we are drawn to some topics and maybe we struggle with others. Like I would say a good example of one that we struggle with is like the Fast Fourier Transform, which is like, if we can talk about why that's so important, I do think it's fascinating. And just how ubiquitous this sort of piece of mathematics is. But like the story of how it came to be is even more fascinating. But now how do you encapsulate that in a title and a thumbnail? That is a significantly more difficult challenge. And so yeah, I don't know. Like that video I made just because I wanted to make the video and then later we were like, how do we package this? And later we're scrambling to try to make it palatable in 50 characters and an image.

E: You can't just say you won't believe this and leave it up to the imagination of the viewer.

DM: Yeah. I mean, I think sometimes the clickbait, like becomes anti-clickbait. If you're so far in that space of like, oh, you wouldn't believe what happens. Or like, there's no way, it's like just because it sounds so over the top, you actually lose people. And I think particularly this audience, which is like, I think my audience is smart, intelligent, like sophisticated. I don't think, some of the people who are watching other channels, I won't name names. But like a lot of people who are watching that are like 13 or maybe not even not even 13 years old. They're just like appealing to kids and appealing to that kind of click level. That's not our audience. And so we have to be a little bit more sophisticated and not too, too, sort of exaggerating or sensationalist in the title and thumbnail to really, get people in.

C: There's this side of psychology that I find so fascinating, which is like, why are people interested in the things they're interested in? And the other side of YouTube is exactly that. It's these bananas views on things that like, to me, I just don't get why people are into it. But you have some insight into that because this is your world.

DM: Yeah. I think to a degree, I would also say like, and this is one of those really strange things of like living on this platform for 12 years is like, I've seen people come and go. I've seen channels rise and fall. And partly, I attribute that to quirks in the algorithm. I think that it's never going to be perfect, the algorithm. So in a way, you think of like, how is it working five years ago? Well, it was working kind of wrong five years ago. And it was strangely advantaging in certain kinds of content and suppressing other kinds of content. So it's really strange to me to see how things have gone that way. One of the earliest weird things about YouTube was like, once you had subscribers, then views became kind of guaranteed. And your growth was like, there was a real big positive feedback loop. So YouTube has had to combat that. And then sometimes I think they've gone too far. And so having like a big channel with a lot of subscribers actually kind of counts against you and acts is a weight. So I don't know, they're constantly trying to balance all these factors. I don't know how you could make an algorithm that isn't subject to weird circular feedbacks. People who run YouTube will say, the algorithm is the audience. We are just representing what the audience wants. The problem is there's nearly infinite content and the audience is oblivious to almost all of it. So the algorithm has to show them something on that limited screen real estate. So what do you see? I don't know, 10, 20 videos? Okay. So now they've seen those things and maybe they want to watch some of them. kay. So now they want to watch those things, but they didn't know about the virtual infinity of stuff that was never shown to them. So immediately you get sucked into weird positive feedbacks like that. And I don't know that there's a way to break the cycle. From the creator perspective, what I can say is that if you want to grow and maintain an audience, it's sort of essential to have some viral hits. And the reason is you just need to become discoverable to people outside of your core sphere. Otherwise, your audience will kind of like dwindle and shrink down to nothing over time. So in my view, like some viral hits are essential to sort of a long term career on the platform.

J: And that's the trick, isn't it? Because like, and what is that formula? Nobody knows it.

DM: About virality? I actually think there is a certain, I mean, like some things are unexpected for sure. But I actually think there's some sort of clear principles that you can put in place to have a better chance of a viral hit than not. I think like from our perspective, videos that are really well researched and they're on topics that will surprise and delight people and really change the way they see the world. If we can do that in a really nice way with great storytelling, I think, and a good title and thumbnail, we'll have a pretty decent shot at getting a ton of views. So yeah, it's not 100%. But there's definitely things you can do to improve your chances.

S: Or you could say there are things that you don't do, you have no chance. There's at least a minimum criteria that you need to hit before you could have a really popular video. But I think, and I wonder if we're getting to the point now where social media platforms like YouTube are so well established, they really are more like TV in a way as they were saying. It's kind of merging towards that end of the spectrum. And having not all popular videos are necessarily viral. You know what I mean? It's not like just this quirky thing that goes out of control and nobody can understand why. It's that, yeah, you have a solid audience because you have good content and you put it out regularly and you develop your regular audience. That's not really going viral. That's just having a really popular video, right?

DM: That's true.

S: That exists now. It's not really just all random shit against the wall and some of it goes viral. And most of it doesn't.

DM: When I think about virality, I often think about sort of exceeding your normal audience by a large margin. So, for us, a viral hit would have to be, well over 10 million, maybe 20 million, 30 million. That's sort of viral territory for us. Whereas if you're a channel with 50,000 views, a video that gets 500,000 is viral. So, yeah, it all depends on like the base of subscribership and, regular viewers.

J: Derek, thank you so much, God. It was great talking to you.

B: Yeah man.

DM: Yeah, that was a real pleasure. Thanks for having me.

S: Yeah, thanks. That was Derek Muller, Veritasium, maker of the Veritasium, Science and Engineering YouTube videos. If you haven't seen them, please do check them out. That is also an excerpt from a much longer interview that we did with Derek, which will become available soon to our premium members. So keep an eye out for that. All right. So on with science or fiction.


Science or Fiction (1:39:26)[edit]

Theme: Improvements in technology

Item #1: Engineers have developed concrete made from lunar regolith and ordinary potato starch and requiring only low temperature heating that is almost three times as strong as regular concrete.[7]
Item #2: Aerospace engineers have made an airplane propeller blade design that significantly reduces noise while increasing thrust and efficiency by 20% over a traditional design.[8]
Item #3: Scientists report a new process for electrosynthesis of multicarbon products from CO2 that is twice as efficient as existing methods.[9]

Answer Item
Fiction Propeller blade design
Science Lunar regolith concrete
New process for electrosynthesis
Host Result
Steve sweep
Rogue Guess
Lunar regolith concrete
Lunar regolith concrete
Lunar regolith concrete
Lunar regolith concrete

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

S: Each week I come up with three science news items or facts, two real and one fake. And I challenge my panel of skeptics to to tell me which one is the fake. It's kind of a soft theme this week. There's a three news items, but they all involve improvements in technology. Are you ready?

E: Okay.

S: All right. Here we go. Item number one, engineers have developed concrete made from lunar regolith and ordinary potato starch and requiring only low temperature heating that is almost three times as strong as regular concrete. Item number two, aerospace engineers have made an airplane propeller blade design that significantly reduces noise while increasing thrust and efficiency by 20% over a traditional design. Item number three, scientists report a new process for electrosynthesis of multicarbon products from CO2 that is twice as efficient as existing methods. Jay go first.

Jay's Response[edit]

J: All right. This first one engineers have developed concrete made from lunar regolith and ordinary potato starch and it requires low temperature heating.

S: So in other words, you can cook it like in an oven. You don't need super high temperatures.

J: It's three times as strong as regular concrete. Right out of the gate, this thing seems like fiction to me.

B: Is it stronger on the moon compared to the earth?

S: No. Objectively, it's stronger.

J: I just don't see how it's three times stronger. Why would it be three times stronger? Why would potato starch? That's just very odd. The second one aerospace engineers have made an airplane propeller blade design that significantly reduces noise and increased thrust efficiency by 20%. You'd think that more thrust would mean more noise.

S: Yeah, that's [inaudible].

J: But I could see this being science though. The third one scientists report a new process for electro synthesis of multi-carbon products from CO2. That's twice as efficient. Yeah, so either the two and three don't seem that crazy to me.

S: Multi-carbon products are basically like making fuel. From making high energy, hydrocarbons out of CO2 from the atmosphere. That's the idea.

C: Oh, this is cool. Okay.

J: Yeah, that's cool. But again, with it ever scale up and all the things that make these types of advancements not go anywhere. But I'm sure some scientists report that there's this new process that they can make fuel from. This happens all the time. I'm not surprised about that. I think I'm just going to go with my gut and say that the first one, the potato starch and lunar regolith are not going to be made with low temperature and they're not going to be three times as strong as regular concrete. That's got to be fiction.

S: Okay, Bob.

Bob's Response[edit]

B: Yeah, the last two sound reasonable in the regolith, I agree with Jay, just tweaks me a little bit more than the other ones although nothing would surprise me with these. But yeah, I'll go with the regolith as well. Fiction.

S: Okay, Evan.

Evan's Response[edit]

E: I suppose so. It's just so bizarre. And how, so how much lunar regolith do we have and how much can, are you using to test this kind of stuff? I mean, it must be such a small amount. Super tiny. So I'm wondering-

B: Or they made an analog.

S: Yeah, they may have used simulated regolith in this study.

E: What? Wait a minute. At the concrete made from lunar regolith, doesn't it? You're saying possibly synthetic regolith?

B: Close enough.

S: Yeah, simulated. I'm saying it could be simulated regolith. But the point is it predicts that that's what it would be. That's what the properties would be with actual regolith.

E: Well, yeah, where is the, I mean, the other two have a patina of plausibility. This one's got nothing to it. This regolith and potato starch. All right, I'll go with the guys. I'll say that one's fiction.

C: Okay. And Cara.

Cara's Response[edit]

C: I have no idea why I last, I think that was a fluke because I know nothing about any of this shit.

S: I was just, because you were first last time.

C: Oh, okay. Cool. Yeah, I'm like, I have no knowledge of any of this. I'm just, I got to go with the guys. Because otherwise I would not be, that would be a bad gamble. So go with the guys.

Steve Explains Item #3[edit]

S: All right. So you're all going with the first one. So we'll take this in reverse order. I'll start with number three. Scientists report a new process for electrosynthesis of multicarbon products from CO2 that is twice as efficient as existing methods. You all think that is science. And that one is science. This is a nice solid incremental advance. Yes. So generally, the process they use catalysts in a, either a neutral or alkaline environment. So they said, let's try it in a strong acid environment. And they were able to get twice the efficiency. So it's, the overall energy efficiency is 25%, which is twofold improvement over the previous benchmark. And they're able to make these multi-carbon molecules. Right? Basically, this is the catalyst causes carbon-carbon coupling. Right? Once you have that, then you could build sort of these hydrocarbons. And also molecules that feed into many industries are basically high energy compounds. Right? That feed into many different industries. So it is sort of the, this is the holy grail of carbon capture. You make it into, because carbon dioxide is a low energy molecule. You get energy out of it by combining it with oxygen. And then we want to reverse that process in order to put it back into a high energy molecule. And the more efficiently we can do that, the better. Because obviously, it takes a lot of energy to capture carbon. It's kind of counterproductive, right? So any improvement is good. So yeah, I would consider this an incremental technological improvement, but solid, twice, doubling of efficiency. Very nice.

Steve Explains Item #2[edit]

S: Let's go back to number two, aerospace engineers have made an airplane propeller blade design that significantly reduces noise while increasing thrust and efficiency by 20% over a traditional design. You guys all think this one is science. And this one is the fiction.

E: Dammit.

C: Damn.

S: Sorry guys.

B: Sweepy sweep. You bastard.

S: This would be massive if true. Like this is just really exaggerated. I mean, the 20% efficiency, thrust efficiency, that's massive. But here's the situation, a little bit of background. There's traditionally a trade-off between noisiness and efficiency when it comes to propeller design. The more efficient you make them, the more noisy they are. And the less noisy they are, the more efficient they can be. And that seems counterintuitive because you would think that noisiness is a sort of any efficiency in and of itself. But it has to do with the fact that if you know how propellers could have two blades?

E: "To blave"

S: Or it might have three blades, or it could have four, or it could have six blades, right, as long as they're [inaudible]. So the more blades you have, the less noisy the propeller, but the less efficient it is as well. And so there's this trade-off in blade design for propeller planes between efficiency and noisiness. And so what there was a reason to study it, they were trying to say, all right, is there a way, is there a sweet spot? Can we do anything to sort of maximize efficiency without having to have like super noisy planes? Now that the noise level is significant because obviously for passengers, and I don't know, I'm sure you guys at some point in your lives have rode in a commercial propeller plane, a prop plane, right? They're freaking noisy. It's like, baaaa, the whole time you're riding the plane. Yeah, they're significantly noisy. And that's even, making the, that's the compromise noise level. That's not even the maximum noise level, because you know what I'm saying? But what they found was that when you have more blades, it disrupts the vortices at the end of the blades, and that's where a lot of the noise comes from. It's the blade tip vortices. So more blades, because just the flow of the air around and reduces the blade tip vortices, and so more blades are there, that's why they're less noisy, but they're also less efficient. So what they found was that they could alter, they basically did a study to see if we alter the design of the tips of these propeller blades. Can we reduce the noise and also stay within sort of the sweet spot for efficiency? And so they were able to create a blade that has a significant like multi decibel reduction in noisiness with only a small hit and efficiency. So it was actually, it still reduced the efficiency. It didn't even increase it at all, let alone by 20%, which is like mind boggling, but it was only by a little bit. So like, always, we can have, we could maintain our efficiency, basically, with like a 3% hit, but, but significantly reduce the noise. That was the breakthrough. And so you got a 20% I thought was big enough to justify, a fiction. The reason why there were still studying, propellers, in 2023-

B: Good question. Jet, jet engines?

S: Because, no, no these are propellers for propeller planes.

E: Because of the battery electric planes.

S: Yes, because of electric planes, because if we want to have electric propeller planes from, for short to medium distance commercial routes, we want them to be as efficient as possible and just quiet as possible. Because they also will, it's not just for the passengers, they could be flying low over residential areas, et cetera. There could be an, we want to reduce the environmental noise that they would generate as well. So it's still, it's very important to make sure we're optimizing blade design for those two features.

Steve Explains Item #1[edit]

S: Okay, let's go back to number one. Engineers have developed concrete made from lunar regolith and ordinary potato starch and requiring only low temperature heating that is almost three times as strong as regular concrete. This one is science. Now I'm actually not sure if they used simulated regolith from the moon or not. I know that they used simulated Mars regolith because we don't have any.

E: Yet.

S: And they specifically say they had to simulate the Mars regolith. But they just said they used lunar regolith and without specifically saying if it was simulated or it was sample from the moon. So I don't not sure. But it could have been, that's why I said it could be simulated. I don't know. But what they, what they found was when they, they went through a process where they combine starch from, it could be from many different sources, but potatoes apparently work the best. They combine that with the simulated Mars regolith or the lunar regolith and whether or not it simulated. And then they compress it, they heat it and they compress it some more. And you get a really good concrete. The starch becomes a very effective binder. The regolith is a good substrate. You get this sort of biocomposite, and they tested it and it had the Mars regolith had about twice the compression strength. The lunar regolith about a little bit less than three times the compression strength. And they also had greater, they call it flexural strength. So they had greater compression strength and flexural strength for both. Now, the same people, their previous study, they looked at making, making concrete from the blood and piss of astronauts. That worked because because they wanted the nitrites or something from the urine and some kind of protein from the blood as a binder. Maybe the albumine, I don't know. And it worked. It worked, but they said, but that's probably the limit, the supply of astronaut blood may be limited. (laughter)

E: Well.

C: One would hope actually.

S: So we need an alternative binder. So that's what they did the spa-lips that he, for using, basically, food starch. Now, they call their product StarCreat, which is interesting, StarCreat. Obviously, the idea would be, if we like say we have a settlement on the moon and we're growing our own food there, that you could have like the waste products of the food could be used as a source for starch, which could then be used as a local building material. And so that would really strong concrete, which would be nice.

C: Right. So people watched the Martian or read the Martian, clearly.

S: Totally. Yeah. And it would work on Mars as well. Which is awesome. It would make, they said, if you had a kilogram of the starch, a kilogram, you could make about 250 bricks. So that's not a lot for a kilogram. So it's, it's not the kind of thing where we're going to be shipping the starch to the moon. So you could do it for some starter, I guess, for some of it. It's really intended for you're growing it locally, right? You're using it from, as a byproduct of the food production that's supporting the settlement, you can also have starch to build stuff out of it, which is obviously a good idea. We definitely want to build as much as we can out of local material when you're on the moon or Mars, because shipping stuff there is ridiculously expensive and limited. So yeah, it's pretty amazing. And the thing is it works so well. Why not use it on Earth? It's just potato starch and even like the simulated Martian regolith, which is obviously made of stuff from Earth, was still stronger than regular concrete, so twice as strong, you know.

C: So what was so special about the regolith? Was it like, would it work with dirt?

S: I don't know. I mean, I guess so, but I mean, just the physical characteristics of it.

C: Right.

S: Would lent itself to this. They want to, and it wasn't, they weren't looking for the best substrate. They wanted to know, will this work on the moon? Will this work on Mars? So, so interestingly, I guess they could, somebody could say, okay, let's take the same concept and just try to find the best substrate on Earth. Whatever that would be, abundant, cheap and also strong. And it, and the, I think what they really discovered was, was that food starch is a really good concrete binder. If you have the right substrate with it. And the key is, and this is not to be glossed over, is that you can make this concrete with oven temperature heating, with low temperature heating. You don't need like these thousands of degrees that are necessary to make concrete. So, we make the concrete that align and you have to heat that, which releases the CO2. And also you have the energy that you need to heat it to whatever it is like, two, three thousand degrees. So, this was just, cooking at oven temperature. They also said that you add a little bit of salt that, that makes it even, it makes it a little bit stronger, the salt. And they said we could source, get this, we could source the salt from the tears of astronauts.

E: Come on. What are we doing here?

S: Because it just takes a pinch of salt.

E: Oh, my gosh.

S: So, you have to make the astronauts cry.

E: Blood, tears and sweat.

S: Right. Yeah. I guess sweat might work too.

E: Oh, funny.

C: I think they're crying. They're crying because they came all this way just to build a bunch of shit at a concrete.

S: Right.

C: Very depressing to me. I don't want us to build things out of concrete on the moon.

S: Sure we should. Why not?

C: Oh, my God.

S: All right, Evan, give us a quote.

Skeptical Quote of the Week (1:55:48)[edit]

The Scientific Revolution has not been a revolution of knowledge. It has been above all a revolution of ignorance. The great discovery that launched the Scientific Revolution was the discovery that humans do not know the answers to their most important questions.

 – Sapiens: A Brief History of Humankind, by Yuval Noah Harari (1976-present), Israeli public intellectual and historian

E: This week's quote was offered up by listener Jason from Sydney. Thank you so much. Here is the quote. "The scientific revolution has not been a revolution of knowledge. It has been above all a revolution of ignorance. The great discovery that launched the scientific revolution was the discovery that humans do not know the answers to their most important questions." And that was written by Yuval Noah Harari, who wrote the book Sapiens, A Brief History of Humankind, which he says we probably already know and read and sadly know. But now it adds that to my list of books to read. And basically he says he's making the point that the scientific revolution was one where cultures had a revelation and admission of our own ignorance of nature. And we had to discover it that it wasn't already in the texts that were already written.

S: I agree with that.

E: Yeah.

S: Right? Admitting your ignorance is the first step to discovery and knowledge.

C: First comes insight then comes action.

S: Yeah. And not only that, the more we, you always have to separate knowledge into three places. There's the known knowns, the known unknowns and the unknown unknowns. And what happens is as our knowledge grows, the known unknowns category grows. And so our ignorance grows as we get more knowledge in a way. Right? And that's a good thing because that's-

C: Or at least our awareness of it.

S: Yeah. Yeah. And the known unknowns is where most advances are happening. But the way to look at this is that the bigger shift is from the unknown unknowns to the known unknowns. Knowing what we don't know is the more important shift than to the known known. Do you know what I mean? Because you're not aware that you don't know something. That's like that's ignorance is not even knowing that you don't know something.

C: Right. Because you can't do much about that.

S: And it gives you the illusion of knowledge, which is the worst. Which is dangerous.

C: Dunning–Kruger the shit out of everything.

S: Exactly. All right. Well, thank you all for joining me this week.

J: Surely.

B: Thanks.

C: Thanks Steve.

E: Thank you.

Signoff/Announcements ()[edit]

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

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


Today I Learned[edit]

  • Fact/Description, possibly with an article reference[10]
  • Fact/Description
  • Fact/Description




Navi-previous.png Back to top of page Navi-next.png