 This 10th year of Daily Tech News show is made possible by you, the listeners. Thanks to every single one of you, including Paul Teeson, Ali Sanjabi, Audrey Stoll Edler's spot and our brand new patron, Michael. On this episode of DTNS is AI coming for math. Is that a good thing? Also, AI is affecting celebrity in every which way. And why self-checkout tech works for nobody, nobody. This is the Daily Tech News for Friday, October 6th, 2023 in Los Angeles. I'm Tom Merritt. And from Studio Sardine, I'm Sarah Lane. I'm the show's producer, Roger Chang. And joining us, data scientist, comedian, circus performer and host of majoring in everything podcast, Andrea Jones-Roy. Welcome back. Thank you so much. I'm glad to be here. I'm not in any of the exciting places you all are. I'm in Drury, New York City. So you're not at Studio Sardine, New York City. No, no, alas. It would be OK if you were. Yeah. I just thought I might eat some sardines later. I had some sardines for lunch. You and I like sardines. Sardines get a bad rap. Free the sardines. This is the most effusive, positive sardine review I've ever heard in my entire life, I think. Yeah, we're Omega fans. Nice. I can see the glowing skin and fur. That's right. That's why our brands are so smart. Nice. What do we do next? All right, the quick hits. Microsoft is heading towards finalizing its sixty eight point seven billion dollar acquisition of the Activision Blizzard on Friday, October 13th. Oh, spooky day. The UK's competition and markets authorities deadline expires today. However, October 6th, which is meant to gather opinions on moving the deal forward. The CMA provisionally approved the deal last month. And unless something pretty big changes, this should let Microsoft close the deal. Epic CEO Tim Sweeney told Unreal Fest that his company will also introduce a per seat or subscription model for Unreal Engine in 2024. Remember, that's thing got unity in trouble. However, only non game commercial developers would be charged. So car makers, video producers, other non game uses. Sweeney clarified in posts on X that they would not be charged and that there would even be a minute or they would be charged. Sorry, but there would be a minimum revenue threshold for commercial products, including independent filmmakers. So if you're making a game, you don't get charged. If you're not making enough revenue, you won't get charged. Basically doing what Unity did, but only for non game uses that make a lot of money. Reuters sources say open A.I. is considering alternatives to Nvidia for the chips that open A.I. uses to run its A.I. models. Options include developing a chip in-house, acquiring a chip company or even just obtaining chips from companies besides Nvidia. Apple, Google, Amazon now all have their own in-chip designed in-house design chips, rather. And Meta is working on a design of its own. So open A.I. might go that route as well. Most Google Pixel users can now download the Magnifier app that Google showed off earlier this week at Made by Google. Magnifier makes small, dim or distant text easier to read, like a magnified glass, like you'd expect. So for example, a menu behind a restaurant counter. You could point your phone, use the Magnifier, be able to read it. A far off road sign. Magnifier is supported on Pixel 5 and newer, so it goes back a ways with the exception of the Pixel Fold and the Pixel tablet. Apple's iOS also offers a similar app. All right, let's talk data breach. It's news time for breaches. Thursday, MGM Resorts said in a regulatory filing with the SEC that the cyber attack last month that took US-based MGM property reservation systems and slot machines offline will cost the company more than $100 million in the third quarter. The company disclosed that some personal information, including names, contacts, gender, dates of birth, driver's license numbers of MGM Resorts customers prior to March of 2019 was stolen and that Social Security numbers and passport details were accessed for a limited number of consumers. All right, that's not it. There's more. Sony Interactive Entertainment notified around 6,800 current and former US employees that an unauthorized party exploited a zero-day vulnerability in the Move It transfer platform, which exposed personal and family member information. The data breach notification notes that the breach occurred on May 28th three days before Sony learned from Progress Software, which is the Move It vendor, about that flaw. But wait, there's more. Oh, no. 23andMe confirmed that due to a credential stuffing attack, some of its customer data has shown up on forums with intent to sell. This includes full names, user names, profile photos, sex, date of birth, genetic ancestry results and geographical location. That is a stuffing attack. So really, one might say, hey, people should have changed their passwords beforehand. But when we're talking about that kind of information, something to know. Yeah, don't use the same password at multiple places or at that kind of thing is going to happen. I use that constantly. Should I say that on this show? I have two passwords. I'm very proud of myself. Which password? Andria. Yeah. Yes. So they are password 123 and then my name. I will say I kind of, you know, I feel like I'm kind of, you know, I go like nine, eight, seven, six, five. Oh, no one's ever going to think of that. Yeah, yeah, people don't think that way. That's a good one. Yeah. Well, just today, I did one of those. It was like, answer the security question thing. And they asked my option was, what was your mother's maiden name? And I have a hyphenated last name because it's my mother's maiden name, which is my, my like security answer is in my name. So it's a terrible. Yeah, that's not a good one to ask. Not a good one. Yeah. Well, let's talk about chat bots. Shall we? The information reports that meta is paying millions of dollars to celebrities for the rights to recreate their personalities as chat bots. The bots are currently text only, but reportedly celebrities spent a few hours recording themselves so that their voices could be added later this year. And right in line with that news, the BBC published a feature article about all the companies trying to create celebrities from scratch using AI. So you don't have to pay them millions of dollars to get them to record things. The fine tradition of virtual stars isn't new. It stretches back from Hatsune Miku to gorillas are a kind of virtual celebrity all the way to Alvin and the chipmunks. Yeah, but all of those had humans doing the vocals. That may not be necessary now. And they have some partial examples in that BBC story. Andrea, as a performer, as not just a data scientist, but you're also a performer. Do you think there's a place for generative models in entertainment? Unfortunately, yes. As you were describing it, I was like, oh, yeah, they don't need voice actors anymore. Or I mean, maybe if you could imagine some version where it's like, this is the real thing, like just like you want a real Louis Vuitton as opposed to a knockoff Louis Vuitton. But knockoff Louis Vuitton is doing great. So I think synthetic voices that sound like Taylor Swift or whoever is on your mind. Hypothetically, I'm thinking about Taylor Swift all the time. Yeah, I think it's a real problem. I think, you know, and that's what we're seeing a lot of the striking about is replacing humans for these sorts of things. And what about the idea of just creating a celebrity that isn't even an imitation, right? That's just like, oh, we're going to we're going to create a singing star or a host of a show. Lil Micaela, that's a good example, you know, of something that I mean, I think a lot of people were very aware that this was all A.I. generated, but has a backstory, personality, you know, unique songs. I mean, I mean, we fall for fictional characters all the time, right? Like we get obsessed with people who are, you know, it's a real actor, but it's otherwise a story and it's the character that we're obsessed with more than the actor. And so I think it's completely feasible that we would be obsessed with made up celebrity. It kind of makes me think of like a celebrity version of what was that movie with Scarlett Johansson? Her. Yeah. Yeah. But it's like just a famous person, as opposed to someone you have like a weird phone relationship with. Not to be confused with the singer, her, who I saw, and is a real person. OK, well, now we're going to have to start someone in the in the comments that I'm not a robot, like in the login things. We're going to have to do that in real life. Like, hi, I'm on stage. I'm not like a hologram. I don't know how to prove it to you that this is real. That's a good point. I was sitting pretty far up at the Rose Bowl. Maybe she was totally virtual. Yeah. I mean, when when Alvin and the chipmunks was when you mentioned that, Tom, I mean, when I was a little kid, that was kind of, you know, they had like their like Christmas albums. And, you know, my parents thought that the whole thing was kind of lame. But I thought it was great. Yeah, I wouldn't. I mean, that was just a cartoon, I would say, that's voiced by a person. We see all sorts of, you know, that sort of thing. You know, anything that isn't live action can be voiced as a person. And that's one thing. I feel like what we're talking about here is something a little bit different, though. Yeah, we're talking about more like Hatsune Miku, who can appear on stage and sing and is the voice is synthesized. It's not an AI making it, but the voice is synthesized with now a generative models. I know there's a lot of Korean companies trying to make a K-pop group that's totally virtual catch on. There have been several attempts at it. None of them have exactly burned up the charts yet, but, you know, keep taking enough swings, eventually get a hit. It is kind of fun to watch from a if I were not a human who cared about the arts and the arts and I'm in the arts. It is fun to watch literally the last year. People like we went from like, well, the arts are the only area that will never be touched by AI to like, well, OK, like I've really it would be more fun if I weren't part of this planet, you know, and it's an interesting thing to watch. Yeah, and I think there's room for both. I think I think we overblow how algorithms and technology can replace people. We tend to catastrophize it a little. So even if we do get full blown virtual celebrities, I don't think it drives humans out of the marketplace. Case in point is our next story. Insider reports that some companies are cooling on the introduction of self-checkout lanes or at least adapting how they're run to include plenty of human help. Costco has humans checking membership cards to stop people from sharing them. Walmart has pulled self-checkout from three of its stores in Albuquerque, New Mexico. It says it has no widespread plans to do that elsewhere. But most Walmart stores are pursuing a hybrid approach. So they haven't reduced staff. You can choose to have as little or as much help from a person as you want as you come up to these little corrals. You can check yourself out or you can ask somebody to help you out. But there's lots of people there looking over, making sure you're not stealing stuff and and being ready to help you with the situation. On the other end of the scale, you've got Kroger, which has all self-checkout lanes, at least one of its stores, but also did not reduce its staff at that store either. Because of what I mentioned before about theft, ECR retail loss, which monitors specifically retail loss, says its research indicates that self-checkout results in a higher percentage of inventory loss. Its numbers are for 2018, but still, it's probable that that has continued. I'd be curious what, including more staff near self-checkout lanes, has done for that, though. Walmart and Kroger rely on video monitoring to track merchandise through the store and into the self-checkout lanes to try to prevent loss. Which makes me wonder, Andrea, should we just skip from self-checkout, which nobody likes, it's like 25 percent of the time I have to get somebody to come over and help because it does something weird. It sounds like it's not actually saving any money or any labor. Maybe we should go right to just check walkout technologies like Amazon wants to do. Well, the walkout technologies or the palm technology, which I guess is the next step before walkout, I will offer a contradictory view, which is I love a self-checkout. I was thinking, as you were describing this story, I was in CVS this morning and there was a human at a regular checkout and there was an open self-checkout. And the human was like, hello, right here. And I was like, I don't think so. And I actively went. I don't like talking to people. I get stressed out. You all are nice, but like in person. And then I get stressed out that they're going to do it wrong. I was buying like one thing. What are they going to do wrong? I get I actually prefer I like to press the buttons. That's probably bad for COVID purposes, but I really prefer it. And I'm glad to hear that they're not losing staff because you're right. Twenty-five percent of the time, if not more, depending on what sorts of things you tend to buy, like if you buy produce, heaven help you, right? With self-checkout. But I actually I prefer it. And I like the people floating as opposed to having to rely on someone because I have strong views. But just walk out would be even better for you. And I should be replaced by a nicer AI. I think that's just walk out sounds like what you're actually I did the other day, watch someone get caught stealing. It was very exciting. That was not just walk out. There wasn't just walk out technology. That wasn't just technology. Just walking out. It was just walking out. The self-checkout, you know, and I I'm all about the palm scan when I go to Whole Foods because, you know, Amazon. And and that's one of the grocery stores that I go to often. If I have, I don't know, like a bottle of wine, I can't do self-checkout because we're not there yet type thing. That isn't always the case. But I have I have been in self-checkout where I'm like, this will just be easier. And you do. There's a weird thing where I'm like, I am not stealing anything. But I can see the person watching me enter things, looking at my stuff, being like, is she going to say that she got that from the salad bar? Is she going to say it's organic broccoli rather than not? And it is like a weird thing where I'm like, I am stressed. I was bought once. I bought. I was OK. I also go to Whole Foods and I live in Lower Manhattan. Obviously, I go to Whole Foods, but I go to Whole Foods and I was at a machine that didn't have the beeping sound like every time you go through. And then that ruined, first of all, it ruined the experience. I really enjoy the boop sound. That's like half the reason I do it. And it wasn't working. And so it was really hard to tell if something was registering. So I already needed a bunch of help because sometimes I would scan something twice. But then I had raspberries and I I thought I scanned it and it didn't go through. And then I continued. And within seconds, there were two people like, you missed those raspberries. So you're right to be nervous. Like they're watching closely. Right. And it's like, listen, I, you know, believe me or not, I'm not trying to steal food from Whole Foods or any grocery store where there's self-checkout options. But I know that it's kind of easy to do so if the, you know, if they're not kind of vulturing over you. And it makes for an experience that I find stressful. But just walk out is better for everybody. That's what I'm here. Totally. Which is like, they don't have to watch you. The computer is watching you. Yeah. You don't have to think about, did I scan it right? Because you just walk out with it. Like, yeah, no. But I think the problem is Amazon hasn't quite perfected it well enough to convince other companies to adopt it. Because it isn't perfect. I know that the times I've used it. It feels shady, even though I don't quite know why. It's also really limited. I mean, even if someone has heard about this, it's going to be, you know, a little bit of a learning curve to get used to it. The times I've used it to, you don't get your receipt for like hours. Like you should get it immediately, right? If it's working really well. That delay is odd. Like, yeah. I mean, I'm the person who's like when they're like, do you want your receipt? I'm like, yes. And then I stand there and look at it and just make sure it was not overcharged because I'm 95 years old. But if you do self-checkout, any overcharges your own fault, whereas I'll do it. I'm more of an eagle eye when someone else is doing it. Like they didn't they put they rung up the wrong type of apple. That apple was on sale. This is this, my prime, blah, blah, blah. I'm I'm I micromanage. Everyone prefers me doing my own checkout. Well, doing your own checkout is probably wise. Deciding who the top five YouTubers on the Internet are is not a smart move, but I did it anyway. So go watch me go down in flames in the latest episode of Tom's Top Five, where I asked everyone on the Internet who they thought their favorite YouTuber was. And then I present what I found. The top five YouTubers, if you're curious, who's number one or two, three, four or five, go find it at YouTube.com slash Daily Tech News. That's just a magnet for all opinions ever had in the world. It just you've just created a black hole. Tom is a very brave person. Man, taking one for the Internet team right there. I am not that brave. I would write one that's like five people I've seen on YouTube. Like I'm just so afraid of like people just descending. Five people you might like on YouTube. Yeah, I did start the episode by blaming everyone who sent me the YouTubers like, these are mine. Yeah, that's the way out. Yeah, hive mind. Well, listen, we talk a lot about what AI does well, what it doesn't do well and how it will affect business going forward based on both of those parameters. What about math though? The New York Times, Siobhan Robert wrote a story this past summer called AI is coming for math to Andrea, your data scientist. High level math is kind of your lane. What do you make of how AI might help or hinder math? Well, my first reaction to I like that you think that math is my lane. I am a data scientist because I like computers to do the math for me, which is part of why I think this article is exciting. Manipulative math is your lane. Like I couldn't like I'm like an ideas person. Like, what if you multiply it by that? OK, go do it. Minions, AKA computers. But this article when it first came out, AI is coming from mathematics to I didn't read it for like a week because it was too much for me to handle. It was like all the articles coming out about how it's going to replace everything. And I was seeing it in research, you're seeing it in the arts. And it was really a lot to handle. And I was like math math feels like the most sacred spot where you just most untouchable, right? Because you picture like a mathematics professor and you have to say mathematics instead of math, you they're writing on chalkboards. They're scribbling proofs. It just feels like the most analog field ever. But the article itself actually opened my eyes in two ways. One, apparently, and maybe people watching are more aware of this than I am, apparently mathematicians have been using computers to solve proofs or evaluate the possibility of a proof, whether or not it would be possible. Even if you can't get through the reasoning, you can say, oh, there actually is a solution here. Since like the 1960s, mathematicians have been using computers. So one, computers have been helpful in math way beyond calculators for a long time. But two, it is very much emblematic of the difference, the limits of AI, which is you can use computers to crunch numbers and say, OK, there is a prime number with this many digits or oh, there is a seven dimensional space that you can manipulate. But there's no way to get to that answer. And so we still need humans for the reasoning and the intuition. And, you know, the actors are going to go. The poets are going to go. The K-pop stars are going to go. But someone somewhere is still going to have to sit with a pencil and paper to reason through a mathematical proof. But you might be asking, why do you need to reason through mathematical proof if you can just prove that it exists through brute computational force? And so far, the answer to that is, A, it's kind of fun. And B, it takes a lot of computing power to do some of these proofs. So think of the computing power and the chips that we're thinking about for generative AI and open AI and all these things and multiply that times infinity for the amount of computational power to prove that there's some nth dimensional thing out there that you could rotate in some particular way. But even the mathematicians in the last few years have had to reckon with the fact that AI might be coming for them in a way that the rest of us. So in data science, we think about coding. It's long been the case that that computers were going to start to be able to do more of the coding than we are. So that's sort of we we had it coming, whereas I think the mathematicians have to catch up with the rest of us who've been reckoning with this for a long time. Yeah, I get the sense from reading this article that it's the difference between showing your work and just having an answer. Right. The the the models are great at giving you an answer. I know that's seven. Oh, yeah. Because but because they are black boxes in large part. And we haven't really. That was another part of this article I thought is they're like, mathematicians should be trying to figure out how these things come up with their answers. That'd be a great thing for people to study. But because we don't really know how they come up with their answers, we still need to show the work. And so there's where the mathematicians can come in and say, oh, let me let me prove how it got there. Let me understand how it got there because that not only shows you, like you said, it's fun. It shows you understanding, but it also can be applied to other things in a way that just having the answer doesn't. Right. And that's exactly the the core of the problem that we've seen with machine learning and algorithms since way before generative AI, which is, you know, we have these machines that are really good at predicting, wow, this person might become sick. This person is a great candidate for whatever. Also, there are tons of biases and it's not perfect predictions, but but they're there, right? Big, complicated models to predict the weather, all those sorts of things. But if you don't know the why behind it, you're sort of limited in what you can do with that information, right? Just like if you're like told that, oh, a year from now, you're going to be in Spain. You're like, is that helpful to me? Like I'd rather know why I'm going to be in Spain. Like did something go wrong and something going right or whatever? I'm how did I get here? I'm crowdfunding to go to Spain, if that's not obvious from my appearance here. But the other thing that's interesting about this, though, to me is that there's another way in with the mathematical proofs. And I've started toying with this piece of it a little bit more in my own life, not nearly in a sophisticated of form as like Terence Tao, the world famous mathematician, but is a pedagogical way of working with generative AI. So there's an algorithm. I want to say it's from Microsoft. It's I might be wrong about that, but it's called Lean. And it's it's generative AI for mathematicians. And the way that it works is you describe your proof to it. And it tells you whether or not your reasoning is sound. So you're still doing the work. It's just going to say, yeah, this makes sense. Or, oh, you're missing something. A lot of times, apparently, Lean is very bad at understanding. You almost have to write your own kind of mathematics pseudocode to convey it. So humans are still doing a coding of translating the mathematical proof into something that computer can understand. But lots of times the computer won't understand it. So you have to like re-explain your own idea four different ways, kind of like you're you're being a tutor. And mathematicians are saying that actually going through that process of explaining it in a variety of different ways to the computer, because it's doing what's called proof whining, makes you better at your own proof. You see it in a different and anyone who's done any teaching knows. I got much better at data science when I started teaching data science, because the student says, wait, why do you do it that way? And you're like, why do we do it that way? And so there's this sort of iterative interactive component to using generative AI and mathematics that's intriguing. And of course, we could use that in other domains as well. I did a practice the other day after thinking about this. I did a practice to prepare for coding. I was like, all right, I just use chat GPTI. I said, all right, pretend you're a student of data science who knows no coding, no math. I'm going to describe to you how computer programming works. And I want you to just keep asking me questions. And it asked me some really interesting questions that helped me improve my lecture that I then gave. So that's kind of cool, I feel like. And it feels less terrifying than the brute force oracles that we're working on, on the other hand. Yeah, to my way of thinking, this is a great example of how these models are best going to be used, not by replacing people, but replacing the parts of people that are tedious or we're not very good at leaving us to do the part we are good at and actually helping us get better. That system lean, which was, you're right, developed at Microsoft, it's open source now, that is a great example of the kind of tools I want. Something that helps me along, catches my mistakes. So like, hey, it seems like you missed a step. Makes me explain things because the easier and simpler you can explain something the more you understand it. I love all of that, yeah. Yeah, no, and ultimately this article, you read the headline, you think, oh God, it's coming from math, now humanity is over. And ultimately I walked away thinking, oh, it really boils down a lot of what we're talking about with the, can we replace artists? Can we replace this? Can we replace that? And sort of in the most optimistic way, there's still this kernel of human, metacognition, reason, intuition that is, and ultimately we still, at least for now, are still telling the computer what to do. It's not like the computer is like turning on and saying, I did a proof that I thought of myself, right? Like we're still steering the ship. We're still saying I need a K-pop star that does this and I need a K-pop star that does that and maybe the career is less about the K-pop stars and more about the computer engineers who generate the K-pop stars. I mean, they become in the way that like a Pixar animator could be a celebrity in a way, I don't know. But it ultimately gave me hope that there's still this kind of kernel of human glow that's keeping us all going. Yeah, I have that hope as well. So keep hope alive. Yes, let's keep it alive. The other fun fact, though, is, I don't know if you saw this in the article, is that back in the 60s, 70s, when they started using computers to do the brute force piece, the finding that you can color a map with, you need at least four colors for a map so that none of the countries have the same color touching. That was the first mathematical proof that was solved through computational brute force. I imagine that that's one that since people have reasoned through sort of backwards induction or reverse engineering, but that finding, which is something that I think a lot of us learn when we start drawing maps or labeling maps, at least in middle and high school, is back in the 60s, they used a computer to figure that out. So Big Maps has been using AI this whole time. Well, Big Maps might be needed for big ships. Andrea. Yeah. Let me tell you a little bit about the industry of ship breaking. Germany's Leviathan and the Netherlands elegant exits to examples of companies that are trying to change the ship breaking industry for the better. Now, currently when a big cargo ship, maybe is taking cars to and fro or oil or that sort of thing, big, big ships, when they retire from their seafaring lives, they often end up on a beach somewhere, often in Southeast Asia, polluting the environment, harming the workers that are dismantling them for parts to be sold later. Like so many retirees. I was literally just thinking that. I was like, yeah, we just kind of wandered to Southeast Asia and ruined the local client. Yes. In this case, you're like a very large steel ship. Cargo ship, yeah. Leviathan has a huge robotic arm, kind of interesting in this BBC article, which we'll have in our show notes, that can slice through steel with something called a water jet, which is a combination of water and sand. It's so strong, it just slices through that thing, breaking the ship up into parts, letting the steel be recycled, keeping people from being harmed by trying to do that in other ways that are somewhat archaic at this point. The technology is precise, can also be used by bomb disposal experts to cut the fuses out of bombs. Currently in use in Ukraine, so says the company. Elegant Exit, a Netherlands company, uses gas-fueled cutters at its facility in Bahrain. It also says it removes hazardous materials before processing ships into large pieces of steel, up to 25 tons for transportation to be used elsewhere. Yeah, I know. That's impressive. I learned a little something today. A little water and sand, and you can slice up steel ships. Weekend plans are set, all right? I have everything set up, let's go. That was my absolute first thought, it's like, can you just give me like a hand first? And I'll just start slicing up stuff. Not people. I got some stuff to slice up. But, you know, just things that should be really hard to slice. Stuff you don't need anymore, you know? Just wanna fit it in the recycling bin and then you can. I mean, that's where we're getting to the, we're using AI to augment human cognition. We'll use it to augment our slicing powers, right? Like I'm all about the sci-fi modular enhancements to my hands and so on, I'll take two. Yeah, yeah. 100%. Well, Andrea Jones, thank you so much for being with us and telling us a little bit about the future of math, talking about slicing up steel cargo ships and everything in between. Let folks know where they can keep up with the rest of your work. Always a joy to be here. Thank you so much for having me back. Loved having this discussion. I can be found on the various social medias, Jonesroy, J-O-N-E-S, no-hyphen, R-O-O-Y. So Jonesroy1word and Jonesroy.com. There it is. Patrons, guess, go, do it now. Patrons, stick around for the extended show, Good Day Internet, where we will be talking some more. It's a Friday. So we do some lighthearted fun stuff on Fridays and it's time again for the great GDI debates. Join us as we tackle some of the most hotly debated questions of the internet. And in fact, we let the chat pick one of the topics this time, which is perfect for October. Well done, chat room. Why is it perfect for October? You'll see why. Stick around. Yeah, I'm really, crowd sourcing is an October thing. That's good, all right. Andrea, I'm historically bad at these games. So you've got the upper hand already. But just a reminder, everybody, you can catch DTNS Live Monday through Friday at 4 p.m. Eastern, 2100 UTC. Find out more at dailytechnewshow.com slash live. We'll be off Monday for the U.S. holiday. We'll see you back on Tuesday. Have a great weekend. This week's episode of Daily Tech News Show were created by the following people, host producer and writer, Tom Merritt, host producer and writer, Sarah Lane, executive producer and booker, Roger Chang, producer, writer and co-host, Rob Dunwood, video producer and Twitch producer, Joe Coontz, technical producer, Anthony Lemos, Spanish language host, writer and producer, Dan Campos. Science correspondent, Dr. Nicky Ackermans, social media producer and moderator, Zoe Detterding. Our mods, Beatmaster, W. Scottus1, Biocow, Captain Kipper, Steve Guadarrama, Paul Rees, Matthew J. Stevens, aka Gadget Virtuoso and J.D. Galloway. Modern video hosting by Dan Christensen. Music and art provided by Martin Bell, Dan Looters, Mustafa A, Acast and Len Peralta. Acast adds support from Tatiana Matias, Patreon support from Tom McNeil. Contributors for this week's show included Justin Robert Young. Our guests this week were Michelle Raman and Andrea Jones Roy. And thanks to all the patrons who make the show possible. This show is part of the Frog Pants Network. Get more at frogpants.com. The Diamond Club hopes you have enjoyed this program.