SGU Episode 948: Difference between revisions

From SGUTranscripts
Jump to navigation Jump to search
No edit summary
(news items done)
Line 186: Line 186:


== News Items ==
== News Items ==
<!--


** We recommend adding section anchors above any news items that are referenced in later episodes (or even hinted in prior episodes as upcoming). See the anchor directly above News Item #1 below, which you would change to {{anchor|news1}}
'''J:''' I'm at the stage now where I am fully intimidated by artificial intelligence, because--
-->
'''S:'''
 
'''B:'''
 
'''C:'''
 
'''J:'''
 
'''E:'''
<!-- ** the triple quotes are how you get the initials to be bolded. Remember to use double quotes with parentheses for non-speech sounds like (laughter) and (applause). It's a good practice to use brackets for comments like [inaudible] and [sarcasm]. -->
 
''(laughs)''
''(laughter)''
''(applause)''
[inaudible]
 
'''J:''' ...fully intimidated by artificial intelligence, because--


'''S:''' --Well, let's talk about that, Jay, since that's my news item. Let's just slide right into that.
'''S:''' --Well, let's talk about that, Jay, since that's my news item. Let's just slide right into that.
Line 219: Line 200:
|publication = nn
|publication = nn
}}
}}
'''S:''' We've talked about the latest crop of artificial intelligence, basically the generative pre-trained transformers. Transformers is a type of AI technology. Generative means that it's generating content, and pre-trained means that it's already been trained on a lot of data. Another way to refer to these applications like ChatGPT, again, the GPT stands for generative pre-trained transformer, is these are large language models. They are modeling language, so the AI itself doesn't necessarily have to understand anything about the words that it's generating. It's just predicting the next word chunk, and remarkably, remarkably effective at generating natural sounding, real sounding text, poetry, prose, answers to questions
'''B:''' It's striking, yeah.
'''S:''' Yeah, it is pretty striking.
'''B:''' Really blowing away deterring tests at this point, right?
'''S:''' Totally. There's been a number of studies looking at how ChatGPT specifically does when given certain standardized tests, or how ChatGPT does on academic performance, right? For example, recently, one study showed that ChatGPT was able to pass the medical boards, there you go, with a just passing score, but still, just passing is passing. All right, so there's one recent study, which I thought was worth updating this, and then we could discuss. This is looking at how ChatGPT performs at a university level, right? So now, essentially what they did was they took essay type test questions in 32 different fields, and they gave students the questions, and they gave ChatGPT the questions. And they had blinded evaluators, people who took that course.
'''E:''' Sure, they didn't know what they were.
'''S:''' Yeah, just grade them, like how did they do? And the results were that ChatGPT did pretty well. So in 9 of the 32 courses, it did as well or better than the students.
'''B:''' This is ChatGPT 4.0, right?
'''S:''' Yeah, this is the latest ChatGPT.
'''E:''' Wow.
'''S:''' And most of the rest, it was in the range. It did pretty well. Yeah, it wasn't quite statistically as good as the students, but it was up there. There was a few courses where it did poorly.
'''E:''' Shame.
'''B:''' Poorly, flat out poorly?
'''S:''' Flat out poorly, like significantly less than the students. One was mathematics. So we already know ChatGPT, not so good at math. The second one was economics, because it's very math-based. And the third one was coding, also because of the math-based component of that. So basically, that's just telling us ChatGPT isn't good at math. But other than that, as long as there wasn't a huge math component to the course material, it did as well or better than the university students at basically doing a test or homework type essay-type assignments. The obvious concern here is that students will be using this to do their assignments, right? And they did a couple other elements of the test. They put the ChatGPT's answers through two standard bits of software whose job is to detect ChatGPT-generated answers.
'''B:''' Oh, give me a break. Does that really...
'''S:''' So the two tools, the two ChatGPT detecting tools, failed to detect the AI's work in 32% and 49% of the time. So in one of them, it was a coin flip, right?
'''B:''' Yeah, to roll the dice at this point.
'''S:''' The other one was only a third of the time they didn't detect it. So that's obviously not good.
'''B:''' Yeah, at this point in time, I think it's pretty, people generally agree that those kind of tools to detect it are just not even good. Don't even use them at this point.
'''S:''' There's a lot of false positives too, claiming that a student's work is an AI's work. So yeah.
'''E:''' Yeah, that's true too.
'''S:''' All right. So a little bit of nuance here, and this is similar to other research, because I've been following this type of research pretty closely. And so you could start to ask, what kind of content is ChatGPT good at and what kind is it not so good at? And we could break it down into a few categories. So one is factual knowledge, another is cognitive analysis, and a third is creativity. So in general, as you would probably guess this, ChatGPT does best at factual knowledge.
'''E:''' Sure.
'''S:''' It's just knowing stuff. If the test is just, do you know facts, ChatGPT does really well.
'''B:''' Well, is it so much facts as do you know your training data?
'''S:''' Well, yeah, that's the same thing.
'''B:''' Yeah, it is. Just throwing that out there.
'''S:''' It obviously doesn't understand anything, but it can reproduce factually correct answers to fact-based questions. Creativity and cognitive analysis, it did not do as well as the students. It did better in the fact-based questions. But this is interesting. If you evaluate the questions based upon the difficulty, as the questions got more difficult, the gap narrowed.
'''E:''' I see.
'''S:''' Meaning that the students got harder, quicker for the students than it did for ChatGPT. I don't know if that pattern was going to hold up and be reducible, but that's interesting to think about that. You would think that, oh, for the really challenging questions, ChatGPT faltered more than the ChatGPT did. And in fact, it started to become harder to separate them the harder the questions became, the more challenging they were in terms of abstract thought and creativity.
'''E:''' Creativity is a little more subjective, though, so harder to measure.
'''S:''' Yeah, maybe that's an artifact of that, you're right. The authors do mention that, the more subjective measures, it's harder to make firm conclusions about. There's a couple of things we could chat about here. One is, this is a snapshot today. This is going to be different in six months, in a year, in five years.
'''E:''' Six months, yeah.
'''S:''' Back to 10 years from now, forget about it.
'''J:''' It's unpredictable.
'''E:''' Oh, we're going to look back and laugh at this news item.
'''S:''' I would say we shouldn't assume that we're going to keep the same pace of improvement as we've been seeing over the last few years.
'''E:''' No {{w|Moore's law}} here?
'''S:''' No. And there are some experts who are saying, and I've read their essays where they argue specifically that we're going to plateau. We're already getting diminishing returns. And so you have to think, why are they so much better than at doing this kind of task than previous AIs? One is that it's being trained on massive amounts of data, but there's only so much headroom there. There's only so much more data you could train it on. And again, you get diminishing returns. Once you're training on billions of samples of data, do you really get better if you do another billion?
'''B:''' Right. You would need orders of magnitude more, I would think, and that's not going to happen.
'''J:''' Steve, I was under the assumption that we were going to be moving into expert systems where it's trained very, very specifically, like to be a lawyer or to be able to solve medical issues.
'''S:''' Yeah. So that's a question. How much room for improvement is there on specially trained GPTs? And also, how much more room for improvement is there just on fixing things like the hallucinations, where it sort of makes up stuff? So you might have to add some subroutines in there that reduces the errors in the hallucinations. You could have specially trained versions of ChatGPT that's not just trained on the internet, but that's trained on all the medical stuff or all legal knowledge, legal textbooks and cases and everything.
'''B:''' I've heard ideas, Steve, about taking those, making many, many of those like little expert systems and then linking them together, which sounded intriguing. Who knows if it's a viable path forward, but it does sound intriguing. But yeah, clearly the low-hanging fruit has been picked, but I still think we may be a little bit surprised as to all the different ways we can apply it and use it and tweak it. Like typically, right? It's the classic sci-fi scenario where it's like no one's really seeing how it's going to be used and how it's going to be applied, so we may see more of that. But yeah, we definitely have gotten all this low-hanging fruit and who knows, we may need a paradigm change to really have that next leap.
'''S:''' But I think we could say, it's safe to say, like according to this and other studies, ChatGPT is up to the task of doing any student's homework, right? At this point, and it's only going to get better. It's already at the point where it's so hard to detect there's no reliable detection method. Plus the other thing is, students can easily subvert any detection method by just lightly editing the output of the ChatGPT, you know what I mean? It would be kind of silly to just use it completely unaltered, but you could have it do 95% of your work for you and turn a several-hour task into a 10- or 20-minute task just by, it might take you a few tries to get the good prompt, and then you get the result that you're happy with, and then you sort of lightly edit it so it sounds like your prose. And that's basically undetectable at that point. No one could detect who did the actual work. So what are teachers or professors to do, right? That's the question.
'''B:''' I think it boils down to the testing, because the kind of testing where you can't use AI to help, and at some point it may come to the point where you have to be in a specific room where technology won't even function. Like you have some super slick kind of ear insert, inner ear insert, and computerized glasses, whatever. You would have to even try to make sure that wasn't in there, and to just be like, it's all on you. It's just you and a pen and a paper and no technology to help you. That's how they may really assess how much are you actually learning, because homework is going to be worthless in terms of assessing.
'''S:''' Or oral exams.


'''S:''' We've talked about...
'''B:''' Exactly. Same idea.
 
'''E:''' Yeah. But isn't that – that presents a whole other range of difficulties, right? Because some people just would do much better in a written test rather than an oral test on the same material.
 
'''B:''' Well, write it in the class where you can't have any help. I mean–
 
'''E:''' Yeah, I mean what I suppose you would have is a database where you would log in to do your assignment, which the school or the institution controls, and it's impervious to being able to be manipulated by the ChatGPT. I mean–
 
'''S:''' So there's another approach. So first let me back up a little bit and say that if you read commentary about this, I've three basic approaches to this problem. One is to say if kids want to cheat, screw them. They're cheating themselves out of an education. That doesn't work, I think, because teachers need a way to assess the effectiveness of their own teaching. And if that is taken away from them, they may not realize that what they're doing isn't working. And also grade sometimes – you know, some classes, some schools grade on a curve. There's a certain competitiveness in terms of going to the next level of your education or whatever. So there has to be a level playing field, and you don't want to punish the students who are not cheating, basically. So I don't think that works. The other one is really aggressive detection methods. I think we agree that's going to fail. That is failing. So we're left with basically methods of evaluating the effectiveness of the teaching and how much the students have learned that are ChatGPT proof. This may require a significant reorganization of the entire process. This may not be a quick fix. So for example – and I think the good news is I think we're already moving in this direction. And I think I've said on the show at my medical school, we don't lecture anymore. We don't give lectures. Students consume the material on their own time via video, podcasts, audio, and written material. And then when we have class time, we're doing workshops. So I think that's what we need to – we need to go to that model where it's like, okay, on Tuesday's class, we're going to be talking about chapter three. Here's all your material. Here's the chapter. Here's a video about it. Here's lectures about it, whatever. By Tuesday, you should have consumed all this material. And then during that classroom time, you're talking about what do people think about it and analysing the content or solving problems with the skills that they were supposed to have learned. And so that not only becomes a superior way of teaching the material, but also a way of evaluating the students. And I think any testing, as Bob said, needs to be live in the classroom, free of technology.
 
'''E:''' And the student would know that if they are kind of more limited than the other students in their participation, that their grade may suffer as a result of that, right?
 
'''S:''' Yeah. I mean, this exists. I mean, I had classes where even in college 30 years ago, it was like classroom participation will be part of your grade. That was just baked into some of the classes because they were so – like we had one class on the Canterbury Tales. And the discussions, you had to have read it. And if you didn't read it, you weren't going to be able to participate in the classroom discussion. And that would have negatively affected your grade. That was part of your evaluation. So that just needs to come forward as the primary way of evaluating students, not like here write an essay because that's just useless now as a way of evaluating students. And so the final thought I'll leave you with is that I think it's pretty clear that this needs to happen or something like this. We need to brainstorm other ways to sort of make ChatGPT-proof classwork. But universities and colleges and schools can't just put this on the teachers. They can't just say, like, do all this additional work to sort of reimagine education. They need support and infrastructure. And this is a disruptive technology. It has successfully disrupted institutions of learning. And there needs to be, I think, a significant change to how we evaluate students. And that's going to require resources. You can't just dump that on teachers.
 
'''E:''' Should there be punishment if students are caught cheating?
 
'''S:''' Well, that's the second approach is like trying to detect and punish cheating. I think because there's so many false positives, false negatives, it's problematic. You end up punishing students who didn't cheat.
 
'''E:''' Yeah, it would have to go through-
 
'''S:''' I would say no, I don't think we're not effective enough to do that.
 
'''E:''' Let's ask ChatGPT what's the best way to get around this problem. It has some ideas.


=== Geothermal Energy <small>(21:59)</small> ===
=== Geothermal Energy <small>(21:59)</small> ===
Line 229: Line 323:
|refname = geothermal
|refname = geothermal
}}
}}
'''S:''' Right. All right, Jay, tell us about geothermal energy.
'''J:''' Well, Steve, you and I are fans of geothermal. We've talked about this in the past. We've discussed briefly where you could get a geothermal heating and air conditioner put into your house, right? And this is a similar concept, except on a huge, huge, like an order of magnitude, larger scale or even more. So the fact is the United States, basically every country is sitting on a significant and largely untapped source of clean renewable energy. And this is geothermal energy. So this energy source is capable of potentially powering the entire world, since heat energy is everywhere. If you drill deep enough it gets warmer and warmer.
'''S:''' That's the key. That's the key. Drilling deep enough.Of course.
'''J:''' As a renewable energy, it could potentially be a game changer if things go well now compared to wind and solar geothermal would be consistent. There would always be that heat there that we could use to produce steam, to turn turbines, to produce energy. Wind and solar as you guys know, it's unpredictable. We have external factors, weather conditions. Is the wind blowing? Yes or no. Is the sun shining? Yes or no. Sometimes it's yes. And sometimes it's no. Well, geothermal energy, it's always there. The heat is always there. So this makes it a consistent and reliable source of clean power. Now, historically, it's been difficult to collect geothermal energy efficiently because we've had to find unique geological conditions that are required for traditional geothermal plants to take advantage of it. So today's geothermal power plants tap into naturally occurring underground hot water reservoirs and things like that that can drive turbines that generate electricity. But there's only a certain number of these out there that would work by putting a fact putting a an energy generator on top of. So only a small fraction of America's electricity is produced through this method. And it's zero point four percent. It's not a lot. It's relatively insignificant today how much we're getting from it. Existing drilling and fracking technologies, though, have been steadily getting better as the years go by. There's been a significant technological boom for drilling and fracking since the early 2000s. And these advancements now make geothermal energy collection feasible. Like we are we are at that point where the technology has progressed far enough, where we have dozens of new energy companies popping up and are all trying to take advantage of this new these new fracking and drilling machinery and techniques that we have. And they want to get in on the geothermal potential here because it's likely to happen. It's just a matter of time. We have companies out of out of all the companies I pick out a couple here like Fervo Energy. They're using fracking techniques to try and crack open hard rock formations and then create artificial geothermal reservoirs. Basically they drill deep enough where there's a constant store of heat and then they inject water into the fractured rock. Then the heat would create steam from the water. And that again, that's all you need, right? You need a way to collect that steam energy. And they have different techniques to do that. I can get into a little detail. Like in this instance, they would create they would drill two holes very, very deep vertical hole and then a horizontal hole and then do it and do it again like a kind of like a mirror image. And then you have basically a way in and a way out, right? And then when the water is injected into it, the steam would come out like the other hole not the hole that's injecting the water. And then there's another approach that this company, Evor, is doing. It's called a closed loop approach. And this is essentially exactly like what geothermal is in the home on the home level, except it's much deeper and much, much, much more dangerous and complicated. But this closed loop approach is it involves like drilling sealed pipes into the hot rocks, right? So the water is not touching the rocks, the water is inside the pipes, but you pump the water down into the pipes and then it becomes it turns into steam it picks up an incredible amount of energy from the heat. And then on the other side you have the heat energy coming out as steam and then again you have a generator there, but it's not not just putting water out into the rockets all stays inside the pipes. So the main factor that's driving interest in expanding geothermal energy is this constant need, of course, for more clean energy solutions than we have today. And we need them to operate consistently meaning 24 hours a day now, particularly during times when wind and solar energy production fluctuates. So when wind and solar have made huge technological strides, they're doing really, really good. They're they're kicking ass. Every time you look at the charts and read what they're capable of doing, how much energy they collect and everything, the numbers just keep going up, the costs keep going down. But they still and always will need backup energy sources or at least we need to have this amazingly aggressive batteries connected to our grid to store the energy. And a lot of experts are saying it's very unlikely that we're going to get to that, especially in the short term. So geothermal energy could potentially fill this gap that we need to fill when we're using solar and wind. And look, it's a reliable carbon free alternative. There's a lot of untapped potential here. There's challenges and risks, of course, that come with this. Fracking could cause earthquakes. And they have to be very careful about how they're handling that. Allowing these kinds of projects takes a lot of time. And it has a complex process, especially if it's happening on federal property. There's a lot of regulatory reviews that take years and years. You know how slow governments can move on stuff like this. And again the government, the U.S. government in particular, isn't funding geothermal anywhere near its funding wind, solar and hydrogen. So I personally have a lot of hope for this technology. It seems almost like a no brainer. It's clean. We're using technologies that we already have. We just need them to get more refined and we need them to basically get to the finish line now. There's these companies are trying to do it. They're looking for funding and who knows what the future will bring. But I am guessing I think that that this is very likely to happen. It's not going to happen quickly again, because the money isn't isn't there yet. But I do think that in the within the next 10 to 20 years, like geothermal could be a huge, huge supply of energy as long as things go well.
'''S:''' Yeah, I think there's no question that there's going to be increased geothermal going forward. The only question is how much and how much really depends on how deep you have to dig. So like all all forms of energy, we need to pick the low hanging fruit first. There are locations where you don't have to drill that deep to get basically to to build a geothermal plant so we could do all of that. Then there are ones where you have to drill deeper, but within our current drilling technology. And then there's the deep geothermal where we're basically like, yeah, we've got to like, we've got to develop the drilling technology itself to be able to drill deeper and economically. And that's the big question mark. So it's really just a matter of how much are we going to have based on how deep we need to dig. And I've read estimates that, yeah, 10 to 15 percent is kind of the low hanging fruit in terms of like an existing energy demand in the U.S. Other countries will have a lot more. This is very region specific. It just depends on your country. And I agree that also, like the main advantage here you glossed over kind of a lot of debate and controversy about like how much wind and solar can we really get to? And it really, it really depends on grid storage. With enough grid storage, we can have 100 percent wind and solar. There's no question about that. But that but enough grid storage is a lot of freaking grid storage. And it's still questionable how quickly and how much we're going to be able to develop things like closed loop hydro. If that turns out to be something that we can do in the next 20 years, then that would give a huge boon to wind and solar. But the big advantage to like hydro geothermal and nuclear right now is that they just require a single connection to the to the grid and it could be an existing connection. And they can produce a lot of electricity and you just plug into an existing grid infrastructure. The biggest the biggest hurdle with wind and solar is that it's more than the intermittency is that it's widely distributed. We have a backlog of years of new connections to the grid and we need lots of grid upgrade if we're going to be able to move like solar power from one place to another or wind power from one place to another. And so that may be a limiting factor. So things that can just like swap out coal or natural gas for another large base load source, there's a certain there's a huge advantage to that in terms of how quickly we could decarbonize our energy infrastructure. If it didn't matter how long it took, then we could get go we could say, OK, we'll just do all wind and solar however long it takes. But we're not we don't have that luxury. We got to do it as quickly as possible. Which means we've got to pick the low hanging fruit of every type of energy and every type of low carbon energy. And then the final thing about geothermal, I think, is that it's one technological breakthrough away from being an order of magnitude, more important than these more modest projections.


=== Failed Star <small>(31:55)</small> ===
=== Failed Star <small>(31:55)</small> ===
Line 236: Line 339:
|publication = Space.com
|publication = Space.com
}}
}}
'''S:''' All right, Bob, tell us about this failed star. Why are astronomers excited about that?
'''B:''' Radiation belts, radiation belts like Earth and Jupiter's have for the first time been found outside our solar system and they're around a brown dwarf. Why should we even care about this? We shouldn't. Back to you, Steve. Oh, wait, I got a few more things to say.
'''E:''' It does not impact my immediate.
'''B:''' I totally lied.
'''E:''' The leads on this paper are Melodie Kao, fellow at the University of California, Santa Cruz and Professor Evgenya Shkolnik of ASU’s School of Earth and Space Exploration, published in Nature on May 15th. Now, it's easy not to think about radiation belts, even though one begins, look straight up a thousand kilometers straight up. There's the beginning of a radiation belt. I personally, I don't see a lot of news items about them. I haven't come across them in a while and I'm always kind of looking around for something. But they are fascinating regardless. And it looks like they just got way more fascinating. Radiation belts are donut shaped magnetic structures found around the planets and moons in our solar system that have essentially large scale magnetic fields around them. So Earth, Jupiter, other gas giants, some moons as well. These are and all of these radiation belts are filled with charged particles, high energy electrons and protons. That's what you find in them. Sometimes you find more protons in one and more electrons in the other. But it's essentially high energy electrons and protons. Now, it's believed that the electrons and protons come from our sun's solar wind and they just get permanently trapped in our own magnetosphere. You may be thinking about auroras right now. And if you are, I send you a polite but earnest golf clap right now. The aurora borealis and the aurora australis, otherwise known as the northern and southern lights, are probably caused by those same electrons and protons trapped in the radiation belts. But some of them get channeled by the magnetosphere down to the Earth's polar regions when the solar activity warrants it. And then there they interact with the atmosphere and they produce the beautiful glow that we know as the northern and southern lights. Now, our own inner and outer radiation belts were discovered in 1958 by the first American satellite called, come on, people, the first American satellite.
'''S:''' {{w|Telstar}}?
'''B:''' What? Closed. Not bad, Steve, but it was {{w|Explorer 1}}.
'''E:''' Oh, that's good trivia right there.
'''B:'''  It is. It is. And I didn't think of it either.
'''E:''' Everyone knows. It's weird. Everyone knows Sputnik.
'''B:''' Oh, my God. Right. Nobody knows Explorer. Right. Explorer 1. But this was very an important thing, an important satellite for many reasons. Space scientist James Van Allen and his team were behind this discovery. And the belts are still called the Van Allen belts to this day. You may come across that term, but they are dangerous areas for sure. They interfere with electronics and satellites. They can harm or even kill astronauts. But limited exposure is safe. It's safe enough. You can go. You can fly through it without really worrying about it. I mean, if you if you went outside of your ship, you wouldn't even notice it if you were just even floating right in the middle of one. But your instruments would certainly notice it. Now, Jupiter, as you might imagine, has a monster radiation belts, many, many times more powerful than the Earth's. They're actually recognized as one of the most hazardous areas of the solar system, besides Jay's bathroom, of course. So charged particles from the sun and interestingly, vulcanism on Jupiter's moon Io are what we believe cause a good chunk of these radiation belts in Jupiter. These charged particles are accelerated within the belt, like accelerators on Earth, like the LHC, Large Hadron Collider. So if you landed on Jupiter's moon Europa within one of these belts, some estimates say that you would get a lethal dose in hours, hours. Like you got a lethal dose. It's a deadly place. And of course, nearby, if you had electronics nearby you on Europa it would meet a similar fate. So yeah, when the aliens in Arthur C. Clarke's sequel to 2001 is based on Odyssey called 2010, when those aliens said all these worlds are yours, except Europa, attempt no landing there, maybe they had maybe they had our safety in mind. Dudes, don't go to Europa. That's a deadly place. Stay away. I don't think that's what they had in mind though, at all, Steve. So don't worry. Okay. So we have identified these radiation belts in our solar system. Okay. That's fine. How common are they? Well, we don't know, but we now have one new massive new data point. And it doesn't evolve a planet or a moon, but a brown dwarf. Now, brown dwarfs are, these are oddball astronomical objects. They are really are bizarre. They're at a huge 13 to 80 Jupiter masses. They're heavy. They could be up to 80 times the mass of Jupiter, but they're too big. So they're too big to be a planet. And they're not big enough to be to sustain hydrogen fusion in their cores. So they're not really considered true stars either. You'll often hear them referred to as a failed star or a substellar objects, which is fair enough. But they do fuse deuterium and even lithium if you're over 65 Jupiter masses. I say, if you fuse new elements in your core, you're a star. I call you a star. That's just me. All right. So the nearest brown dwarf to us is called LSR J1A35, blah, blah, blah, whatever. I'm going to call it LSR. It's only 18.5 years away. To be fair, this is a brown dwarf binary. It's a binary pair, but we're only talking about one of them. LSR, I'll call it. It's only 18.5 light years away.It's right in our backyard. In fact, the term backyard in this case is inadequate. That's so close. It's basically touching the deck that's attached to the back of our house. It's like right there. It's the third closest system to us after Alpha Centauri and Bernard's star. I mean, it's just stupid close. 18.5 light years is nothing, relatively speaking.
'''E:''' Let's send a probe there with a solar sail. Let's get there.
'''B:''' Yeah. Because LSR is so close though, that's why we are able to see its radio emissions in such great detail and make this breakthrough that they have made. In fact, the image resolution they got would surprise the hell out of me. They got, it was 50 times better than even the James Webb Space Telescope, 50 times better than JWST. So how could that even be? They did this by using very long baseline interferometry, VLBI. So they coordinated 39 radio dishes all over the Earth from Hawaii to Germany to create in a real sense a telescope the size of the Earth. Not in a full sense, but in a realistic sense, the telescope for some purposes is the size of the Earth when you spread it out over many, many radio dishes like that. So this is how they imaged the black hole recently that was all over the news was it last year, a year or two ago. They use this very similar technique. And without this technique, that brown dwarf would have just looked like a point of light, just like pretty much any other star. It really wouldn't have been able to infer really much at all. So this very long baseline interferometry technique was critical. So they were able to actually resolve the dynamic magnetic field around the brown dwarf, its magnetosphere. This is the first time this has been done and its shape and nature was recognizably similar to Earth's. And even more so, it was more similar to Jupiter's, in fact. Now, from what they can infer, as you might imagine, the radiation belts around this brown dwarf belied its name. Some estimates put it at 10 million times more intense than Jupiter's. 10 million times more intense of a radiation belt than Jupiter. So how long would an astronaut last near this? Regardless of gravity or anything else, just the radiation belt. So it's hard to say how long an astronaut would last. But if you extrapolate from Europa, if a lethal dose in Europa, you would get in a few hours. So what's one 10 millionth of three hours? I figured a thousandth of a second you would last and you would have a deadly dose in just a sub-second. That's going to be a very, very nasty place to get close to. All right, so that's the radiation belt. But what does the aurora look like on this brown dwarf? Now, imagine right now a classic aurora on the Earth. We've all seen the pictures, right?
'''E:''' That nice greenish glow.
'''B:''' Amazing, among the most beautiful images, the most beautiful things you could see in the night sky, arguably.
'''E:''' Everyone but Bob can't see.
'''B:''' I saw one decades ago. It went way down into Connecticut. It was light. It was not a dramatic one, but you could still get the sense of the beauty of an aurora borealis. So imagine that. Okay, now multiply that by about a million. And that's the aurora you would probably see on this brown dwarf. Maybe you might need radio eyes to see it. I'm not sure what the actual wavelength would be, but that's a million times the intensity of a classic aurora on the Earth. Amazing. But it gets even better. That's not even the really good stuff. The researchers speculate that there may be an exoplanet in this system, an exoplanet. Now, they think it perhaps if it exists, there's no real hard evidence, but if it exists, it would exist perhaps like Jupiter in the sense that there's a body in a close orbit that's supplying many of the charged particles that compose its radiation belts and auroras. So there may be some volcanism on this exoplanet that is spewing out the charged particles that the brown dwarf then accelerates and causes such a nasty environment in creating the radiation belt and creating the occasional auroras. Now, I love the lead scientist on this Melodie Kao's analogy of a radiation belt. She compares it to a yard, but she said the yard doesn't have flowers, it has glowing energetic particles. And then I'll quote her directly here to finish this. She said, "The particular properties of each radiation belt tells us something about that planet's energetic, magnetic, and particle resources. How quickly it's spinning? How strong is the magnetic field? How close it lives to the sun? Does it have moons that can supply more particles? Or does it have rings like Saturn that will absorb them and more? For the first time, we're able to see what sorts of yards, brown dwarfs, and low mass stars have. I'm excited for the day that we can learn about the magnetospheric yards inhabited by exoplanets." So really interesting the fact that we could use these radio signals to infer what the radiation belt and auroras look like on these distant brown dwarfs that eventually, hopefully in the future, we could then fine tune this technique so that we could look at them for exoplanets. And then by observing this radiation belt on the exoplanet, we can determine all of these details about the exoplanet, about its spin, its magnetic field, how close to the sun, all these things we can infer just by interpreting these radio signals that were emitted by the radiation belt. Fascinating stuff. Thank you for your time. Back to you, Steve.
'''S:''' So that's why astronomers are excited.
'''B:''' Yes.
'''S:''' All right.


=== Doctor and NDEs <small>(43:55)</small> ===
=== Doctor and NDEs <small>(43:55)</small> ===
Line 243: Line 386:
|publication = Yahoo News
|publication = Yahoo News
}}
}}
'''S:''' All right. Evan, tell us about this doctor who thinks that near-death experiences are real.
'''B:''' Oh, God. ''(Evan laughs)'' Real what, though?
'''E:''' Here you go, Bob. Put aside the Van Allen belt for now. Pick this one up. I read this at a website called inside.com. It was featured on yahoo.com's website. And seeing as how I have a Yahoo-based email, I get all kinds of updates and news flashed at me. Comes at me a million miles an hour. So this was one of them. An article about a doctor named Jeffrey Long. Jeffrey Long is a medical doctor specializing in the practice of radiation oncology. That's, yep, radiation to treat cancer. Dr. Long is also the founder of something called the Near-Death Experience Research Foundation. And I went to their website and I looked at their mission statement, and here's what it says. "To research and study consciousness experiences and to spread the message of love, unity, and peace around the world." The banner on their homepage reads, "Where science and spirituality come together." Okay, all right. As a scientist, Jeff founded this organization in 1998. He wanted to know if near-death experiences, NDEs, I'm going to call them going forward, if they were real by directly asking the people who have experienced these things. And the answer is a resounding yes, as a result of his research he was able to author a New York Times bestseller book called Evidence of the Afterlife, the Science of Near-Death Experiences. And he's written some other stuff. He's been on all the major networks all the TV shows, Oprah, Dr. Oz. The only place I didn't see him was Goop. But he's pretty much been everywhere talking about this stuff for a long time. He says that studying near-death experiences, sorry, NDEs, has made him a better cancer doctor. Do I need to explain what a near-death experience is? Do you want to hear it?
'''J:''' Migh as well.
'''S:''' Give a bullet.
'''E:''' Yeah, well, here's his definition. I mean, these are his words. "A near death experience is someone who is either comatose or clinically dead without a heartbeat, having a lucid experience where they see, hear, feel emotions, and interact with other beings." That definition, I think, needs some clarity, right, Steve? I mean, clinically dead?
'''S:''' Yeah, that's how you define it. In most studies of NDEs, it's just defined as the heart is stopped. A period of time when there is no spontaneous heartbeat. Often, people are undergoing CPR, cardiopulmonary resuscitation, during that time. That's considered "clinically dead".
'''E:''' Right, clinically dead. OK, all right. So yeah, so there's the hard definition of it. Now, he came across, so he was in his practice. He was a resident at the time, the late 1980s. And he was in the Journal of the American Medical Association. He came across an article describing NDEs. And here's what he said, it stopped me in my tracks. All my medical training told me you were either dead or alive. There was no in-between. But suddenly-
'''S:''' Really? What course are you on?
'''E:''' Right. All his medical training told him this, Steve. But suddenly, he was reading about a cardiologist describing patients who had died and then came back to life, reporting very distinct and almost unbelievable experiences. And from that moment, he was fascinated with NDEs. So yeah, his organization collects stories from people who have these experiences. And then he, these are his words. "Evaluates them in the mind of a scientist and a doctor. I've made opinions based on evidence and came to this as a skeptic. But in the face of overwhelming evidence, I've come to believe there's certainly an afterlife."
'''S:''' Bullshit.
'''E:''' I'm a medical doctor. Steve, he's...
'''S:''' He's already thrown that authority away with the hyperbole that he's been using. You're not either alive or dead. There's a pretty big continuum between those two things, right? Your heart could be stopped, but your brain can still be functioning, for example. That's a pretty big point if we're talking about experiences that are remembered after somebody revives. He's clearly not coming at this from a skeptic scientific point of view, medical background point of view. He wants to believe this.
'''E:''' Oh yeah, absolutely. It's amazing that he's able to say that and has been saying this for decades with basically a straight face.
'''S:''' Or he's been bedazzled by the idea of it and is just engaging in motivated reasoning. But what's inexcusable is that he's promoting basically a religious belief by citing some kind of skeptical scientific authority, and that's BS.
'''E:''' That is exactly right. He's an oncologist, as I said before. He says, here, "I'm a medical doctor, but I've read brain research and considered every possible explanation for NDEs. The bottom line is none of them hold water. There isn't even a remotely plausible physical explanation for this phenomenon." Steve, those are his words exactly I pulled that.
'''S:''' Yeah, which also, I mean, right there, again, he's totally full of it because there are absolutely plausible explanations. He's not even addressing the most plausible explanation, which is that the memories that people are recalling didn't happen when they were "clinically dead". And research to prove that it's happening during a period of time with no brain activity have failed. They have failed to demonstrate that anything is happening at a time when you can document no brain activity. So wrong. That statement is a straight up wrong. Either he is just ignorant of the topic he professes expertise on, or he is being intellectually-
'''E:''' This is argument from authority, Steve, is it not? I mean, this oncologist basically-
'''S:''' Partly, I mean, partly, but he's also making specific factual claims, which are just not true.
'''B:''' He's a punk.
'''E:''' And he goes on. I mean, there's parts in here in which he talks about, what it boils down to is that he says his research, and he has studied apparently more than 5,400 cases, okay, through his organization for whatever that's worth. And he said, here's the thing. He said, "My day job still revolves around helping patients fight cancer. I don't tell my patients about my NDE research, and yet my work with NDE has made me a more compassionate and loving doctor. I'm able to help my patients face life-threatening diseases with increased courage and passion. My goal is to help them have more healthy days here on Earth, but I firmly believe that if and when they pass, they will be at peace."
'''S:''' Yeah, this is all subjective.
'''E:''' Steve, will he be writing for Science-Based Medicine anytime soon? Submitting an article?
'''S:''' Yeah, I doubt it. I'd be happy to engage in a written debate with him on SBM or elsewhere.
'''E:''' You recall, Steve, it was, oh my gosh, it's coming up on 10 years in which you had your debate, you and Sean Carroll.
'''S:''' That's right.
'''E:''' Against proponents of near-death experiences.
'''S:''' Yeah, where we kicked ass.
'''E:''' You did. You moved the needle, right?
'''S:''' By the rules of the debate, we won.
'''E:''' Absolutely. Yeah, no, it was significant, too. It wasn't just a small blip. You convinced a good chunk of that audience to think differently about this after it was all said and done.
'''S:''' Yeah, but the bottom line is, just as a scientific question, they haven't proven that NDEs are a phenomenon, that they are anything other than just brain activity occurring sometime around the entire experience, right? Because they haven't been able to demonstrate any mental activity when the brain is not functional. That's it. Until they do that, they have nothing, nothing. All right, let's move on.


{{anchor|futureWTN}} <!-- keep right above the following sub-section. this is the anchor used by the "wtnAnswer" template, which links the previous "new noisy" segment to its future WTN, here.
{{anchor|futureWTN}} <!-- keep right above the following sub-section. this is the anchor used by the "wtnAnswer" template, which links the previous "new noisy" segment to its future WTN, here.
-->
-->
== Who's That Noisy? <small>(51:54)</small> ==
== Who's That Noisy? <small>(51:54)</small> ==
{{wtnHiddenAnswer
{{wtnHiddenAnswer

Revision as of 05:56, 25 October 2023

  Emblem-pen.png This episode is in the middle of being transcribed by Hearmepurr (talk) as of 2023-10-24, 19:25 GMT.
To help avoid duplication, please do not transcribe this episode while this message is displayed.
  Emblem-pen-orange.png This episode needs: transcription, time stamps, formatting, links, 'Today I Learned' list, categories, segment redirects.
Please help out by contributing!
How to Contribute

You can use this outline to help structure the transcription. Click "Edit" above to begin.

SGU Episode 948
September 9th 2023
948 geothermal energy.png

"Some are trying to unlock geothermal energy by using techniques from the fracking boom." [1]

SGU 947                      SGU 949

Skeptical Rogues
S: Steven Novella

B: Bob Novella

J: Jay Novella

E: Evan Bernstein

Guest

AB: Anna Blakney,
professor of bioengineering

Quote of the Week

Not bound by facts, the hoax is free to fabricate feelings & the genres associated with them; it's this artfulness and ambiguity that help explain [its] popularity.

Kevin Young, American poet

Links
Download Podcast
Show Notes
Forum Discussion

Introduction, past Dragon Cons; AI brain interpretation of song

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 Tuesday, August 29th, 2023, and this is your host, Steven Novella. Joining me this week are Bob Novella...

B: Hey, everybody!

S: Jay Novella...

J: Hey guys.

S: ...and Evan Bernstein.

E: Good evening everyone.

S: Cara is not joining us this week. She's too busy. Just finishing up her work and whatnot.

J: Now that she's a doctor and all, she's very busy.

S: We'll see. We'll see when that's official. We'll let everybody know.

E: Right. She explained to us there's a specific point or criteria, a line specifically across it.

S: Yeah, we'll let you know when we cross that line. This episode is coming. We're recording a little bit early because we're going to DragonCon this weekend. This episode will be coming out the week after DragonCon, so we will have already been there in podcast time.

E: Oh my gosh. We had a blast, I think.

S: I'm assuming. Yeah.

J: Right?

S: Hopefully the hurricane didn't interfere with our trip, hurricane barreling up the Atlantic Ocean. Right now, it's supposed to miss us. We will see. We will have seen by the time this episode comes out.

J: What could it potentially, what's the worst case scenario?

S: It grounds planes. We can't get out.

E: Yes.

J: Okay.

E: Well, we can't get in.

S: Can't either.

J: Didn't we have something like this happen when we were in Tennessee and we drove back?

S: Yes.

E: Oh. It was called, I think, what was it? Sandy? You may have heard of it.

B: It's like a 10-hour drive home, too.

S: Yeah, it wasn't bad.

E: And we rode the edge of that storm the whole way home, remember?

B: Yeah.

E: Remember the ominous skies that were chasing us?

S: We were riders on a storm.

B: I remember the really cool Halloween store we ran by and went in.

J: I remember by the time we got home, I was delirious.

E: Well it wasn't just the drive, if you remember. It was like the 12 hours before we had to drive when we kind of realized, all right, we have to start putting a plan into action because our flights had been, what, canceled I think by that point. So we were effectively stranded 800, 900 miles from home and with no plan whatsoever. So having to scramble to get the van and change and just rally it all together, it was really more of like a 24-hour ordeal.

J: Yeah, for sure.

S: Now, I know you guys have heard this because we talked about it, and I'm going to play it right now in the episode. Listen to this and tell us what you think this is. [plays audio] So that was Pink Floyd, The Wall.

E: Yes.

S: Right? Very recognizable.

E: On a crummy broken tape player or something.

S: Yeah, it sounds a little distorted in a weird way. That is an AI-generated clip based upon basically the brain activity of somebody listening to the song.

B: That's pretty amazing.

E: How? How?

S: Yeah, I mean, so we talked about this, and I think it was a science or fiction a couple months ago, that when you add the AI to interpret the brain signals, it really makes a huge leap in its ability to go from brain activity to audio, to what the person is thinking or hearing or trying to say. It really is amazing technology.

E: And with the clarity of the lyric as well, it's one thing to sort of get a musical representation just with no lyric, right? You can come close, and I think your brain can have a lot of latitude there to say, yeah, this is pretty close. But then you start hearing a lyric kick in, and it's recognizable to the sounding of the word. That sort of brings it to another level.

S: And I was reading a news item today. I was going to use it for science or fiction, and I realized it's too close to what we talked about a couple months ago, where just another study looking at this technology, AI-assisted interpretation of brain waves to figure out what words somebody's thinking of, and they claimed a 92 to 100% accuracy. I'd have to look deeper into exactly, sometimes there was a limited choice of things that they were choosing from. But even still, this is years ahead of where I thought we would be, because of that introduced element of, oh yeah, you just have AI do the analysis and figure out how to. There's got to be a learning curve. It's like a trained AI on going from words to brain activity, brain activity to words. You have to have a feedback loop where it's trained. Then it gets really good at interpreting an individual's brain activity. It does not carry over to another individual. You've got to be trained on one specific person.

E: Clinical application here?

S: Oh yeah, you're locked in. You can't talk. You know what I mean? You can't communicate. Strap an EEG on your head, hook it up to one of these AIs, train it for a little bit, and then you're communicating.

E: That's such an incredible thing for people who find themselves in that super unfortunate situation. I mean, that is just beyond a world of difference.

S: I always think of Pike on Star Trek. It's like, really? 200 years in the future? They're saying yes, no? Even then, even when it was first coming out in the early 70s, late 60s, I thought, really? We're not going to be more advanced than that? We're already way past that.

B: Yeah, there's big, big internet arguments going on about how ridiculous that was, and people try to retcon it and make sense of it, and other people are like, that's just what they decided to do.

S: It was kind of important for the narrative, because you couldn't have Pike just communicating with them and giving away the whole storyline.

B: That basically shaped the whole thing. That happens quite often.

E: And the Futurama parody of that one, look it up. It's one of the funniest episodes you've ever seen.

S: Are you enjoying the reboot of Futurama, Evan?

E: Oh, damn, man. I got to start watching. Steve, shame on me. I have not yet watched.

S: It's good. I've watched like five or six episodes. It's really good.

E: I'm so far behind. I got to catch up.

News Items

J: I'm at the stage now where I am fully intimidated by artificial intelligence, because--

S: --Well, let's talk about that, Jay, since that's my news item. Let's just slide right into that.

B: Bam!

ChatGPT Performs at University Level (6:31)

S: We've talked about the latest crop of artificial intelligence, basically the generative pre-trained transformers. Transformers is a type of AI technology. Generative means that it's generating content, and pre-trained means that it's already been trained on a lot of data. Another way to refer to these applications like ChatGPT, again, the GPT stands for generative pre-trained transformer, is these are large language models. They are modeling language, so the AI itself doesn't necessarily have to understand anything about the words that it's generating. It's just predicting the next word chunk, and remarkably, remarkably effective at generating natural sounding, real sounding text, poetry, prose, answers to questions

B: It's striking, yeah.

S: Yeah, it is pretty striking.

B: Really blowing away deterring tests at this point, right?

S: Totally. There's been a number of studies looking at how ChatGPT specifically does when given certain standardized tests, or how ChatGPT does on academic performance, right? For example, recently, one study showed that ChatGPT was able to pass the medical boards, there you go, with a just passing score, but still, just passing is passing. All right, so there's one recent study, which I thought was worth updating this, and then we could discuss. This is looking at how ChatGPT performs at a university level, right? So now, essentially what they did was they took essay type test questions in 32 different fields, and they gave students the questions, and they gave ChatGPT the questions. And they had blinded evaluators, people who took that course.

E: Sure, they didn't know what they were.

S: Yeah, just grade them, like how did they do? And the results were that ChatGPT did pretty well. So in 9 of the 32 courses, it did as well or better than the students.

B: This is ChatGPT 4.0, right?

S: Yeah, this is the latest ChatGPT.

E: Wow.

S: And most of the rest, it was in the range. It did pretty well. Yeah, it wasn't quite statistically as good as the students, but it was up there. There was a few courses where it did poorly.

E: Shame.

B: Poorly, flat out poorly?

S: Flat out poorly, like significantly less than the students. One was mathematics. So we already know ChatGPT, not so good at math. The second one was economics, because it's very math-based. And the third one was coding, also because of the math-based component of that. So basically, that's just telling us ChatGPT isn't good at math. But other than that, as long as there wasn't a huge math component to the course material, it did as well or better than the university students at basically doing a test or homework type essay-type assignments. The obvious concern here is that students will be using this to do their assignments, right? And they did a couple other elements of the test. They put the ChatGPT's answers through two standard bits of software whose job is to detect ChatGPT-generated answers.

B: Oh, give me a break. Does that really...

S: So the two tools, the two ChatGPT detecting tools, failed to detect the AI's work in 32% and 49% of the time. So in one of them, it was a coin flip, right?

B: Yeah, to roll the dice at this point.

S: The other one was only a third of the time they didn't detect it. So that's obviously not good.

B: Yeah, at this point in time, I think it's pretty, people generally agree that those kind of tools to detect it are just not even good. Don't even use them at this point.

S: There's a lot of false positives too, claiming that a student's work is an AI's work. So yeah.

E: Yeah, that's true too.

S: All right. So a little bit of nuance here, and this is similar to other research, because I've been following this type of research pretty closely. And so you could start to ask, what kind of content is ChatGPT good at and what kind is it not so good at? And we could break it down into a few categories. So one is factual knowledge, another is cognitive analysis, and a third is creativity. So in general, as you would probably guess this, ChatGPT does best at factual knowledge.

E: Sure.

S: It's just knowing stuff. If the test is just, do you know facts, ChatGPT does really well.

B: Well, is it so much facts as do you know your training data?

S: Well, yeah, that's the same thing.

B: Yeah, it is. Just throwing that out there.

S: It obviously doesn't understand anything, but it can reproduce factually correct answers to fact-based questions. Creativity and cognitive analysis, it did not do as well as the students. It did better in the fact-based questions. But this is interesting. If you evaluate the questions based upon the difficulty, as the questions got more difficult, the gap narrowed.

E: I see.

S: Meaning that the students got harder, quicker for the students than it did for ChatGPT. I don't know if that pattern was going to hold up and be reducible, but that's interesting to think about that. You would think that, oh, for the really challenging questions, ChatGPT faltered more than the ChatGPT did. And in fact, it started to become harder to separate them the harder the questions became, the more challenging they were in terms of abstract thought and creativity.

E: Creativity is a little more subjective, though, so harder to measure.

S: Yeah, maybe that's an artifact of that, you're right. The authors do mention that, the more subjective measures, it's harder to make firm conclusions about. There's a couple of things we could chat about here. One is, this is a snapshot today. This is going to be different in six months, in a year, in five years.

E: Six months, yeah.

S: Back to 10 years from now, forget about it.

J: It's unpredictable.

E: Oh, we're going to look back and laugh at this news item.

S: I would say we shouldn't assume that we're going to keep the same pace of improvement as we've been seeing over the last few years.

E: No Moore's law here?

S: No. And there are some experts who are saying, and I've read their essays where they argue specifically that we're going to plateau. We're already getting diminishing returns. And so you have to think, why are they so much better than at doing this kind of task than previous AIs? One is that it's being trained on massive amounts of data, but there's only so much headroom there. There's only so much more data you could train it on. And again, you get diminishing returns. Once you're training on billions of samples of data, do you really get better if you do another billion?

B: Right. You would need orders of magnitude more, I would think, and that's not going to happen.

J: Steve, I was under the assumption that we were going to be moving into expert systems where it's trained very, very specifically, like to be a lawyer or to be able to solve medical issues.

S: Yeah. So that's a question. How much room for improvement is there on specially trained GPTs? And also, how much more room for improvement is there just on fixing things like the hallucinations, where it sort of makes up stuff? So you might have to add some subroutines in there that reduces the errors in the hallucinations. You could have specially trained versions of ChatGPT that's not just trained on the internet, but that's trained on all the medical stuff or all legal knowledge, legal textbooks and cases and everything.

B: I've heard ideas, Steve, about taking those, making many, many of those like little expert systems and then linking them together, which sounded intriguing. Who knows if it's a viable path forward, but it does sound intriguing. But yeah, clearly the low-hanging fruit has been picked, but I still think we may be a little bit surprised as to all the different ways we can apply it and use it and tweak it. Like typically, right? It's the classic sci-fi scenario where it's like no one's really seeing how it's going to be used and how it's going to be applied, so we may see more of that. But yeah, we definitely have gotten all this low-hanging fruit and who knows, we may need a paradigm change to really have that next leap.

S: But I think we could say, it's safe to say, like according to this and other studies, ChatGPT is up to the task of doing any student's homework, right? At this point, and it's only going to get better. It's already at the point where it's so hard to detect there's no reliable detection method. Plus the other thing is, students can easily subvert any detection method by just lightly editing the output of the ChatGPT, you know what I mean? It would be kind of silly to just use it completely unaltered, but you could have it do 95% of your work for you and turn a several-hour task into a 10- or 20-minute task just by, it might take you a few tries to get the good prompt, and then you get the result that you're happy with, and then you sort of lightly edit it so it sounds like your prose. And that's basically undetectable at that point. No one could detect who did the actual work. So what are teachers or professors to do, right? That's the question.

B: I think it boils down to the testing, because the kind of testing where you can't use AI to help, and at some point it may come to the point where you have to be in a specific room where technology won't even function. Like you have some super slick kind of ear insert, inner ear insert, and computerized glasses, whatever. You would have to even try to make sure that wasn't in there, and to just be like, it's all on you. It's just you and a pen and a paper and no technology to help you. That's how they may really assess how much are you actually learning, because homework is going to be worthless in terms of assessing.

S: Or oral exams.

B: Exactly. Same idea.

E: Yeah. But isn't that – that presents a whole other range of difficulties, right? Because some people just would do much better in a written test rather than an oral test on the same material.

B: Well, write it in the class where you can't have any help. I mean–

E: Yeah, I mean what I suppose you would have is a database where you would log in to do your assignment, which the school or the institution controls, and it's impervious to being able to be manipulated by the ChatGPT. I mean–

S: So there's another approach. So first let me back up a little bit and say that if you read commentary about this, I've three basic approaches to this problem. One is to say if kids want to cheat, screw them. They're cheating themselves out of an education. That doesn't work, I think, because teachers need a way to assess the effectiveness of their own teaching. And if that is taken away from them, they may not realize that what they're doing isn't working. And also grade sometimes – you know, some classes, some schools grade on a curve. There's a certain competitiveness in terms of going to the next level of your education or whatever. So there has to be a level playing field, and you don't want to punish the students who are not cheating, basically. So I don't think that works. The other one is really aggressive detection methods. I think we agree that's going to fail. That is failing. So we're left with basically methods of evaluating the effectiveness of the teaching and how much the students have learned that are ChatGPT proof. This may require a significant reorganization of the entire process. This may not be a quick fix. So for example – and I think the good news is I think we're already moving in this direction. And I think I've said on the show at my medical school, we don't lecture anymore. We don't give lectures. Students consume the material on their own time via video, podcasts, audio, and written material. And then when we have class time, we're doing workshops. So I think that's what we need to – we need to go to that model where it's like, okay, on Tuesday's class, we're going to be talking about chapter three. Here's all your material. Here's the chapter. Here's a video about it. Here's lectures about it, whatever. By Tuesday, you should have consumed all this material. And then during that classroom time, you're talking about what do people think about it and analysing the content or solving problems with the skills that they were supposed to have learned. And so that not only becomes a superior way of teaching the material, but also a way of evaluating the students. And I think any testing, as Bob said, needs to be live in the classroom, free of technology.

E: And the student would know that if they are kind of more limited than the other students in their participation, that their grade may suffer as a result of that, right?

S: Yeah. I mean, this exists. I mean, I had classes where even in college 30 years ago, it was like classroom participation will be part of your grade. That was just baked into some of the classes because they were so – like we had one class on the Canterbury Tales. And the discussions, you had to have read it. And if you didn't read it, you weren't going to be able to participate in the classroom discussion. And that would have negatively affected your grade. That was part of your evaluation. So that just needs to come forward as the primary way of evaluating students, not like here write an essay because that's just useless now as a way of evaluating students. And so the final thought I'll leave you with is that I think it's pretty clear that this needs to happen or something like this. We need to brainstorm other ways to sort of make ChatGPT-proof classwork. But universities and colleges and schools can't just put this on the teachers. They can't just say, like, do all this additional work to sort of reimagine education. They need support and infrastructure. And this is a disruptive technology. It has successfully disrupted institutions of learning. And there needs to be, I think, a significant change to how we evaluate students. And that's going to require resources. You can't just dump that on teachers.

E: Should there be punishment if students are caught cheating?

S: Well, that's the second approach is like trying to detect and punish cheating. I think because there's so many false positives, false negatives, it's problematic. You end up punishing students who didn't cheat.

E: Yeah, it would have to go through-

S: I would say no, I don't think we're not effective enough to do that.

E: Let's ask ChatGPT what's the best way to get around this problem. It has some ideas.

Geothermal Energy (21:59)

S: Right. All right, Jay, tell us about geothermal energy.

J: Well, Steve, you and I are fans of geothermal. We've talked about this in the past. We've discussed briefly where you could get a geothermal heating and air conditioner put into your house, right? And this is a similar concept, except on a huge, huge, like an order of magnitude, larger scale or even more. So the fact is the United States, basically every country is sitting on a significant and largely untapped source of clean renewable energy. And this is geothermal energy. So this energy source is capable of potentially powering the entire world, since heat energy is everywhere. If you drill deep enough it gets warmer and warmer.

S: That's the key. That's the key. Drilling deep enough.Of course.

J: As a renewable energy, it could potentially be a game changer if things go well now compared to wind and solar geothermal would be consistent. There would always be that heat there that we could use to produce steam, to turn turbines, to produce energy. Wind and solar as you guys know, it's unpredictable. We have external factors, weather conditions. Is the wind blowing? Yes or no. Is the sun shining? Yes or no. Sometimes it's yes. And sometimes it's no. Well, geothermal energy, it's always there. The heat is always there. So this makes it a consistent and reliable source of clean power. Now, historically, it's been difficult to collect geothermal energy efficiently because we've had to find unique geological conditions that are required for traditional geothermal plants to take advantage of it. So today's geothermal power plants tap into naturally occurring underground hot water reservoirs and things like that that can drive turbines that generate electricity. But there's only a certain number of these out there that would work by putting a fact putting a an energy generator on top of. So only a small fraction of America's electricity is produced through this method. And it's zero point four percent. It's not a lot. It's relatively insignificant today how much we're getting from it. Existing drilling and fracking technologies, though, have been steadily getting better as the years go by. There's been a significant technological boom for drilling and fracking since the early 2000s. And these advancements now make geothermal energy collection feasible. Like we are we are at that point where the technology has progressed far enough, where we have dozens of new energy companies popping up and are all trying to take advantage of this new these new fracking and drilling machinery and techniques that we have. And they want to get in on the geothermal potential here because it's likely to happen. It's just a matter of time. We have companies out of out of all the companies I pick out a couple here like Fervo Energy. They're using fracking techniques to try and crack open hard rock formations and then create artificial geothermal reservoirs. Basically they drill deep enough where there's a constant store of heat and then they inject water into the fractured rock. Then the heat would create steam from the water. And that again, that's all you need, right? You need a way to collect that steam energy. And they have different techniques to do that. I can get into a little detail. Like in this instance, they would create they would drill two holes very, very deep vertical hole and then a horizontal hole and then do it and do it again like a kind of like a mirror image. And then you have basically a way in and a way out, right? And then when the water is injected into it, the steam would come out like the other hole not the hole that's injecting the water. And then there's another approach that this company, Evor, is doing. It's called a closed loop approach. And this is essentially exactly like what geothermal is in the home on the home level, except it's much deeper and much, much, much more dangerous and complicated. But this closed loop approach is it involves like drilling sealed pipes into the hot rocks, right? So the water is not touching the rocks, the water is inside the pipes, but you pump the water down into the pipes and then it becomes it turns into steam it picks up an incredible amount of energy from the heat. And then on the other side you have the heat energy coming out as steam and then again you have a generator there, but it's not not just putting water out into the rockets all stays inside the pipes. So the main factor that's driving interest in expanding geothermal energy is this constant need, of course, for more clean energy solutions than we have today. And we need them to operate consistently meaning 24 hours a day now, particularly during times when wind and solar energy production fluctuates. So when wind and solar have made huge technological strides, they're doing really, really good. They're they're kicking ass. Every time you look at the charts and read what they're capable of doing, how much energy they collect and everything, the numbers just keep going up, the costs keep going down. But they still and always will need backup energy sources or at least we need to have this amazingly aggressive batteries connected to our grid to store the energy. And a lot of experts are saying it's very unlikely that we're going to get to that, especially in the short term. So geothermal energy could potentially fill this gap that we need to fill when we're using solar and wind. And look, it's a reliable carbon free alternative. There's a lot of untapped potential here. There's challenges and risks, of course, that come with this. Fracking could cause earthquakes. And they have to be very careful about how they're handling that. Allowing these kinds of projects takes a lot of time. And it has a complex process, especially if it's happening on federal property. There's a lot of regulatory reviews that take years and years. You know how slow governments can move on stuff like this. And again the government, the U.S. government in particular, isn't funding geothermal anywhere near its funding wind, solar and hydrogen. So I personally have a lot of hope for this technology. It seems almost like a no brainer. It's clean. We're using technologies that we already have. We just need them to get more refined and we need them to basically get to the finish line now. There's these companies are trying to do it. They're looking for funding and who knows what the future will bring. But I am guessing I think that that this is very likely to happen. It's not going to happen quickly again, because the money isn't isn't there yet. But I do think that in the within the next 10 to 20 years, like geothermal could be a huge, huge supply of energy as long as things go well.

S: Yeah, I think there's no question that there's going to be increased geothermal going forward. The only question is how much and how much really depends on how deep you have to dig. So like all all forms of energy, we need to pick the low hanging fruit first. There are locations where you don't have to drill that deep to get basically to to build a geothermal plant so we could do all of that. Then there are ones where you have to drill deeper, but within our current drilling technology. And then there's the deep geothermal where we're basically like, yeah, we've got to like, we've got to develop the drilling technology itself to be able to drill deeper and economically. And that's the big question mark. So it's really just a matter of how much are we going to have based on how deep we need to dig. And I've read estimates that, yeah, 10 to 15 percent is kind of the low hanging fruit in terms of like an existing energy demand in the U.S. Other countries will have a lot more. This is very region specific. It just depends on your country. And I agree that also, like the main advantage here you glossed over kind of a lot of debate and controversy about like how much wind and solar can we really get to? And it really, it really depends on grid storage. With enough grid storage, we can have 100 percent wind and solar. There's no question about that. But that but enough grid storage is a lot of freaking grid storage. And it's still questionable how quickly and how much we're going to be able to develop things like closed loop hydro. If that turns out to be something that we can do in the next 20 years, then that would give a huge boon to wind and solar. But the big advantage to like hydro geothermal and nuclear right now is that they just require a single connection to the to the grid and it could be an existing connection. And they can produce a lot of electricity and you just plug into an existing grid infrastructure. The biggest the biggest hurdle with wind and solar is that it's more than the intermittency is that it's widely distributed. We have a backlog of years of new connections to the grid and we need lots of grid upgrade if we're going to be able to move like solar power from one place to another or wind power from one place to another. And so that may be a limiting factor. So things that can just like swap out coal or natural gas for another large base load source, there's a certain there's a huge advantage to that in terms of how quickly we could decarbonize our energy infrastructure. If it didn't matter how long it took, then we could get go we could say, OK, we'll just do all wind and solar however long it takes. But we're not we don't have that luxury. We got to do it as quickly as possible. Which means we've got to pick the low hanging fruit of every type of energy and every type of low carbon energy. And then the final thing about geothermal, I think, is that it's one technological breakthrough away from being an order of magnitude, more important than these more modest projections.

Failed Star (31:55)

S: All right, Bob, tell us about this failed star. Why are astronomers excited about that?

B: Radiation belts, radiation belts like Earth and Jupiter's have for the first time been found outside our solar system and they're around a brown dwarf. Why should we even care about this? We shouldn't. Back to you, Steve. Oh, wait, I got a few more things to say.

E: It does not impact my immediate.

B: I totally lied.

E: The leads on this paper are Melodie Kao, fellow at the University of California, Santa Cruz and Professor Evgenya Shkolnik of ASU’s School of Earth and Space Exploration, published in Nature on May 15th. Now, it's easy not to think about radiation belts, even though one begins, look straight up a thousand kilometers straight up. There's the beginning of a radiation belt. I personally, I don't see a lot of news items about them. I haven't come across them in a while and I'm always kind of looking around for something. But they are fascinating regardless. And it looks like they just got way more fascinating. Radiation belts are donut shaped magnetic structures found around the planets and moons in our solar system that have essentially large scale magnetic fields around them. So Earth, Jupiter, other gas giants, some moons as well. These are and all of these radiation belts are filled with charged particles, high energy electrons and protons. That's what you find in them. Sometimes you find more protons in one and more electrons in the other. But it's essentially high energy electrons and protons. Now, it's believed that the electrons and protons come from our sun's solar wind and they just get permanently trapped in our own magnetosphere. You may be thinking about auroras right now. And if you are, I send you a polite but earnest golf clap right now. The aurora borealis and the aurora australis, otherwise known as the northern and southern lights, are probably caused by those same electrons and protons trapped in the radiation belts. But some of them get channeled by the magnetosphere down to the Earth's polar regions when the solar activity warrants it. And then there they interact with the atmosphere and they produce the beautiful glow that we know as the northern and southern lights. Now, our own inner and outer radiation belts were discovered in 1958 by the first American satellite called, come on, people, the first American satellite.

S: Telstar?

B: What? Closed. Not bad, Steve, but it was Explorer 1.

E: Oh, that's good trivia right there.

B: It is. It is. And I didn't think of it either.

E: Everyone knows. It's weird. Everyone knows Sputnik.

B: Oh, my God. Right. Nobody knows Explorer. Right. Explorer 1. But this was very an important thing, an important satellite for many reasons. Space scientist James Van Allen and his team were behind this discovery. And the belts are still called the Van Allen belts to this day. You may come across that term, but they are dangerous areas for sure. They interfere with electronics and satellites. They can harm or even kill astronauts. But limited exposure is safe. It's safe enough. You can go. You can fly through it without really worrying about it. I mean, if you if you went outside of your ship, you wouldn't even notice it if you were just even floating right in the middle of one. But your instruments would certainly notice it. Now, Jupiter, as you might imagine, has a monster radiation belts, many, many times more powerful than the Earth's. They're actually recognized as one of the most hazardous areas of the solar system, besides Jay's bathroom, of course. So charged particles from the sun and interestingly, vulcanism on Jupiter's moon Io are what we believe cause a good chunk of these radiation belts in Jupiter. These charged particles are accelerated within the belt, like accelerators on Earth, like the LHC, Large Hadron Collider. So if you landed on Jupiter's moon Europa within one of these belts, some estimates say that you would get a lethal dose in hours, hours. Like you got a lethal dose. It's a deadly place. And of course, nearby, if you had electronics nearby you on Europa it would meet a similar fate. So yeah, when the aliens in Arthur C. Clarke's sequel to 2001 is based on Odyssey called 2010, when those aliens said all these worlds are yours, except Europa, attempt no landing there, maybe they had maybe they had our safety in mind. Dudes, don't go to Europa. That's a deadly place. Stay away. I don't think that's what they had in mind though, at all, Steve. So don't worry. Okay. So we have identified these radiation belts in our solar system. Okay. That's fine. How common are they? Well, we don't know, but we now have one new massive new data point. And it doesn't evolve a planet or a moon, but a brown dwarf. Now, brown dwarfs are, these are oddball astronomical objects. They are really are bizarre. They're at a huge 13 to 80 Jupiter masses. They're heavy. They could be up to 80 times the mass of Jupiter, but they're too big. So they're too big to be a planet. And they're not big enough to be to sustain hydrogen fusion in their cores. So they're not really considered true stars either. You'll often hear them referred to as a failed star or a substellar objects, which is fair enough. But they do fuse deuterium and even lithium if you're over 65 Jupiter masses. I say, if you fuse new elements in your core, you're a star. I call you a star. That's just me. All right. So the nearest brown dwarf to us is called LSR J1A35, blah, blah, blah, whatever. I'm going to call it LSR. It's only 18.5 years away. To be fair, this is a brown dwarf binary. It's a binary pair, but we're only talking about one of them. LSR, I'll call it. It's only 18.5 light years away.It's right in our backyard. In fact, the term backyard in this case is inadequate. That's so close. It's basically touching the deck that's attached to the back of our house. It's like right there. It's the third closest system to us after Alpha Centauri and Bernard's star. I mean, it's just stupid close. 18.5 light years is nothing, relatively speaking.

E: Let's send a probe there with a solar sail. Let's get there.

B: Yeah. Because LSR is so close though, that's why we are able to see its radio emissions in such great detail and make this breakthrough that they have made. In fact, the image resolution they got would surprise the hell out of me. They got, it was 50 times better than even the James Webb Space Telescope, 50 times better than JWST. So how could that even be? They did this by using very long baseline interferometry, VLBI. So they coordinated 39 radio dishes all over the Earth from Hawaii to Germany to create in a real sense a telescope the size of the Earth. Not in a full sense, but in a realistic sense, the telescope for some purposes is the size of the Earth when you spread it out over many, many radio dishes like that. So this is how they imaged the black hole recently that was all over the news was it last year, a year or two ago. They use this very similar technique. And without this technique, that brown dwarf would have just looked like a point of light, just like pretty much any other star. It really wouldn't have been able to infer really much at all. So this very long baseline interferometry technique was critical. So they were able to actually resolve the dynamic magnetic field around the brown dwarf, its magnetosphere. This is the first time this has been done and its shape and nature was recognizably similar to Earth's. And even more so, it was more similar to Jupiter's, in fact. Now, from what they can infer, as you might imagine, the radiation belts around this brown dwarf belied its name. Some estimates put it at 10 million times more intense than Jupiter's. 10 million times more intense of a radiation belt than Jupiter. So how long would an astronaut last near this? Regardless of gravity or anything else, just the radiation belt. So it's hard to say how long an astronaut would last. But if you extrapolate from Europa, if a lethal dose in Europa, you would get in a few hours. So what's one 10 millionth of three hours? I figured a thousandth of a second you would last and you would have a deadly dose in just a sub-second. That's going to be a very, very nasty place to get close to. All right, so that's the radiation belt. But what does the aurora look like on this brown dwarf? Now, imagine right now a classic aurora on the Earth. We've all seen the pictures, right?

E: That nice greenish glow.

B: Amazing, among the most beautiful images, the most beautiful things you could see in the night sky, arguably.

E: Everyone but Bob can't see.

B: I saw one decades ago. It went way down into Connecticut. It was light. It was not a dramatic one, but you could still get the sense of the beauty of an aurora borealis. So imagine that. Okay, now multiply that by about a million. And that's the aurora you would probably see on this brown dwarf. Maybe you might need radio eyes to see it. I'm not sure what the actual wavelength would be, but that's a million times the intensity of a classic aurora on the Earth. Amazing. But it gets even better. That's not even the really good stuff. The researchers speculate that there may be an exoplanet in this system, an exoplanet. Now, they think it perhaps if it exists, there's no real hard evidence, but if it exists, it would exist perhaps like Jupiter in the sense that there's a body in a close orbit that's supplying many of the charged particles that compose its radiation belts and auroras. So there may be some volcanism on this exoplanet that is spewing out the charged particles that the brown dwarf then accelerates and causes such a nasty environment in creating the radiation belt and creating the occasional auroras. Now, I love the lead scientist on this Melodie Kao's analogy of a radiation belt. She compares it to a yard, but she said the yard doesn't have flowers, it has glowing energetic particles. And then I'll quote her directly here to finish this. She said, "The particular properties of each radiation belt tells us something about that planet's energetic, magnetic, and particle resources. How quickly it's spinning? How strong is the magnetic field? How close it lives to the sun? Does it have moons that can supply more particles? Or does it have rings like Saturn that will absorb them and more? For the first time, we're able to see what sorts of yards, brown dwarfs, and low mass stars have. I'm excited for the day that we can learn about the magnetospheric yards inhabited by exoplanets." So really interesting the fact that we could use these radio signals to infer what the radiation belt and auroras look like on these distant brown dwarfs that eventually, hopefully in the future, we could then fine tune this technique so that we could look at them for exoplanets. And then by observing this radiation belt on the exoplanet, we can determine all of these details about the exoplanet, about its spin, its magnetic field, how close to the sun, all these things we can infer just by interpreting these radio signals that were emitted by the radiation belt. Fascinating stuff. Thank you for your time. Back to you, Steve.

S: So that's why astronomers are excited.

B: Yes.

S: All right.

Doctor and NDEs (43:55)

S: All right. Evan, tell us about this doctor who thinks that near-death experiences are real.

B: Oh, God. (Evan laughs) Real what, though?

E: Here you go, Bob. Put aside the Van Allen belt for now. Pick this one up. I read this at a website called inside.com. It was featured on yahoo.com's website. And seeing as how I have a Yahoo-based email, I get all kinds of updates and news flashed at me. Comes at me a million miles an hour. So this was one of them. An article about a doctor named Jeffrey Long. Jeffrey Long is a medical doctor specializing in the practice of radiation oncology. That's, yep, radiation to treat cancer. Dr. Long is also the founder of something called the Near-Death Experience Research Foundation. And I went to their website and I looked at their mission statement, and here's what it says. "To research and study consciousness experiences and to spread the message of love, unity, and peace around the world." The banner on their homepage reads, "Where science and spirituality come together." Okay, all right. As a scientist, Jeff founded this organization in 1998. He wanted to know if near-death experiences, NDEs, I'm going to call them going forward, if they were real by directly asking the people who have experienced these things. And the answer is a resounding yes, as a result of his research he was able to author a New York Times bestseller book called Evidence of the Afterlife, the Science of Near-Death Experiences. And he's written some other stuff. He's been on all the major networks all the TV shows, Oprah, Dr. Oz. The only place I didn't see him was Goop. But he's pretty much been everywhere talking about this stuff for a long time. He says that studying near-death experiences, sorry, NDEs, has made him a better cancer doctor. Do I need to explain what a near-death experience is? Do you want to hear it?

J: Migh as well.

S: Give a bullet.

E: Yeah, well, here's his definition. I mean, these are his words. "A near death experience is someone who is either comatose or clinically dead without a heartbeat, having a lucid experience where they see, hear, feel emotions, and interact with other beings." That definition, I think, needs some clarity, right, Steve? I mean, clinically dead?

S: Yeah, that's how you define it. In most studies of NDEs, it's just defined as the heart is stopped. A period of time when there is no spontaneous heartbeat. Often, people are undergoing CPR, cardiopulmonary resuscitation, during that time. That's considered "clinically dead".

E: Right, clinically dead. OK, all right. So yeah, so there's the hard definition of it. Now, he came across, so he was in his practice. He was a resident at the time, the late 1980s. And he was in the Journal of the American Medical Association. He came across an article describing NDEs. And here's what he said, it stopped me in my tracks. All my medical training told me you were either dead or alive. There was no in-between. But suddenly-

S: Really? What course are you on?

E: Right. All his medical training told him this, Steve. But suddenly, he was reading about a cardiologist describing patients who had died and then came back to life, reporting very distinct and almost unbelievable experiences. And from that moment, he was fascinated with NDEs. So yeah, his organization collects stories from people who have these experiences. And then he, these are his words. "Evaluates them in the mind of a scientist and a doctor. I've made opinions based on evidence and came to this as a skeptic. But in the face of overwhelming evidence, I've come to believe there's certainly an afterlife."

S: Bullshit.

E: I'm a medical doctor. Steve, he's...

S: He's already thrown that authority away with the hyperbole that he's been using. You're not either alive or dead. There's a pretty big continuum between those two things, right? Your heart could be stopped, but your brain can still be functioning, for example. That's a pretty big point if we're talking about experiences that are remembered after somebody revives. He's clearly not coming at this from a skeptic scientific point of view, medical background point of view. He wants to believe this.

E: Oh yeah, absolutely. It's amazing that he's able to say that and has been saying this for decades with basically a straight face.

S: Or he's been bedazzled by the idea of it and is just engaging in motivated reasoning. But what's inexcusable is that he's promoting basically a religious belief by citing some kind of skeptical scientific authority, and that's BS.

E: That is exactly right. He's an oncologist, as I said before. He says, here, "I'm a medical doctor, but I've read brain research and considered every possible explanation for NDEs. The bottom line is none of them hold water. There isn't even a remotely plausible physical explanation for this phenomenon." Steve, those are his words exactly I pulled that.

S: Yeah, which also, I mean, right there, again, he's totally full of it because there are absolutely plausible explanations. He's not even addressing the most plausible explanation, which is that the memories that people are recalling didn't happen when they were "clinically dead". And research to prove that it's happening during a period of time with no brain activity have failed. They have failed to demonstrate that anything is happening at a time when you can document no brain activity. So wrong. That statement is a straight up wrong. Either he is just ignorant of the topic he professes expertise on, or he is being intellectually-

E: This is argument from authority, Steve, is it not? I mean, this oncologist basically-

S: Partly, I mean, partly, but he's also making specific factual claims, which are just not true.

B: He's a punk.

E: And he goes on. I mean, there's parts in here in which he talks about, what it boils down to is that he says his research, and he has studied apparently more than 5,400 cases, okay, through his organization for whatever that's worth. And he said, here's the thing. He said, "My day job still revolves around helping patients fight cancer. I don't tell my patients about my NDE research, and yet my work with NDE has made me a more compassionate and loving doctor. I'm able to help my patients face life-threatening diseases with increased courage and passion. My goal is to help them have more healthy days here on Earth, but I firmly believe that if and when they pass, they will be at peace."

S: Yeah, this is all subjective.

E: Steve, will he be writing for Science-Based Medicine anytime soon? Submitting an article?

S: Yeah, I doubt it. I'd be happy to engage in a written debate with him on SBM or elsewhere.

E: You recall, Steve, it was, oh my gosh, it's coming up on 10 years in which you had your debate, you and Sean Carroll.

S: That's right.

E: Against proponents of near-death experiences.

S: Yeah, where we kicked ass.

E: You did. You moved the needle, right?

S: By the rules of the debate, we won.

E: Absolutely. Yeah, no, it was significant, too. It wasn't just a small blip. You convinced a good chunk of that audience to think differently about this after it was all said and done.

S: Yeah, but the bottom line is, just as a scientific question, they haven't proven that NDEs are a phenomenon, that they are anything other than just brain activity occurring sometime around the entire experience, right? Because they haven't been able to demonstrate any mental activity when the brain is not functional. That's it. Until they do that, they have nothing, nothing. All right, let's move on.

Who's That Noisy? (51:54)

New Noisy (55:29)

[Eerie mechanical hums, groans, and screeches]

what this week's noisy is

Announcements (56:04)

[top]                        

Interview with Anna Blakney (58:26)

  • SciComm and Vaccines
[top]                        

Science or Fiction (1:29:34)

Theme: Negative results

Item #1: A study of the use of augmented reality glasses in social interactions found no significant impact.[5]
Item #2: A recent Finnish study found that the type and amount of fungus (molds and yeast) found in the home of young children do not correlate with the risk of developing asthma.[6]
Item #3: A study following children from age 10 to 16 did not find any correlation between social media use and symptoms of anxiety and depression.[7]

Answer Item
Fiction Use of AR glasses
Science Fungus & asthma risk
Science
Social media & depression
Host Result
Steve clever
Rogue Guess
Bob
Social media & depression
Jay
Use of AR glasses
Evan
Fungus & asthma risk

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

Bob's Response

Jay's Response

Evan's Response

Steve Explains Item #3

Steve Explains Item #2

Steve Explains Item #1

Skeptical Quote of the Week (1:42:22)


Not bound by facts, the hoax is free to fabricate feelings and the genres associated with them- it is this artfulness and ambiguity that help explain [its] popularity.

 – Kevin Young (1970-present), American poet and the director of the Smithsonian Institution National Museum of African American History and Culture, from from Bunk: The Rise of Hoaxes, Humbug, Plagiarists, Phonies, Post-Facts, and Fake News (2017)


Signoff

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 theskepticsguide.org. Send your questions to info@theskepticsguide.org. And, if you would like to support the show and all the work that we do, go to patreon.com/SkepticsGuide and consider becoming a patron and becoming part of the SGU community. Our listeners and supporters are what make SGU possible.

[top]                        

Today I Learned

  • Fact/Description, possibly with an article reference[8]
  • Fact/Description
  • Fact/Description

References

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