In this episode of Digital Marketing Masters, host Matt Rouse welcomes science fiction author Tom Easton to discuss the intersection of AI and science fiction. Tom shares his background as a retired science professor and his journey into writing science fiction since the 1970s. The conversation delves into the upcoming anthology "Tales of Galactic Pest Control," featuring Matt's own story, and explores the intriguing theme of AI's role in society, touching on topics like AI hallucinations, the impact of AI on jobs, and the potential need for universal basic income.
Matt and Tom engage in a lively discussion about the evolution of AI technology, its implications for the future, and the ethical considerations surrounding AI training on existing works. They also reflect on the potential societal changes AI could bring, such as shifts in identity and the role of education. The episode concludes with insights into the creative process behind the anthology and the Kickstarter campaign supporting it, highlighting the collaborative spirit of the science fiction community.
Tales of Galactic Pest Control: https://www.kickstarter.com/projects/galactic-pests/tales-of-galactic-pest-controla-sci-fi-and-fantasy-anthology?ref=mattrouse
Amazin Stories: https://amazingstories.com/
matthewrouse.com
HookDM
SMBAutopilot
https://serve.podhome.fm/digital-marketing-masters
(00:19) Introduction to the Episode
(01:00) Meet Tom Easton: Science Fiction Writer
(02:09) AI in Science Fiction and Reality
(05:05) Challenges and Opportunities of AI Research
(09:18) Copyright and AI: A Legal Perspective
(11:45) The Future of Jobs and Universal Basic Income
(18:09) AI's Impact on Vocational Training and Education
(26:10) Galactic Pest Control Anthology
(31:09) AI's Potential for Greater Good
(37:31) Exponential Growth of AI Technology
(43:02) Closing Remarks and Future Prospects
[00:00:19]
AI Narrator #1:
Today on the digital marketing masters podcast, science fiction legend,
[00:00:55] Matt Rouse:
And welcome back to Digital Marketing Masters, everyone. I'm your host, Matt Raus. And today, I'm super excited. My guest is Tom Easton. How How are you doing, Tom? I'm doing fine. You wanted a quick introduction? Well, I'll just say I'm a retired science professor who's been writing science fiction since the nineteen seventies.
[00:01:14] AI Narrator #1:
And if you take a quick look at the t shirt I'm wearing, Amazing Select run by Steve Davidson and Kermit Whittle want to celebrate the hundredth anniversary of magazine science fiction next year at Raven Con in Virginia. I don't know if I'll be able to make it, but I did buy their t shirt so I can why get in front of audiences like this. That's right. So Amazing Tales, a hundred years now. Right? 1926 was when Amazing published.
[00:01:44] Matt Rouse:
Nice. The reason that we have connected is because Amazing Stories has an anthology coming out. Actually, you may not know this, Tom. This was my submission of a science fiction story that I ever did, and all the books I've written in the past were self published. So this is the time I ever asked anybody if I could get my story in their book. Sure enough, I got invited to be in this one.
[00:02:09] AI Narrator #1:
Well, with your abiding interest in AI, you have you touched some very interesting bases in that story. Way back when, people started talking about where science fiction might go and what apply where AI might go and whether we should be scared of it. And there were people who were saying, we shouldn't be scared about it because they will regard us as gods. Their creators. You may remember reading comments like that. Personally, I think that was an awful lot of wishful thinking because after all, we don't do a very good job of listening to our gods or agreeing on just which gods to pay attention to. That's right. But in your story, you don't go into that very much, but you do mention that the robots should pay attention to the humans because the humans created them. Right. That just touches on that very, very old notion.
I don't did you have to resist temptation to dig further into that?
[00:03:09] Matt Rouse:
Oh, a little bit, but I also didn't want to kind of extend the length of the story too long. I really wanted to try and keep it down to, like, the essentials.
[00:03:19] AI Narrator #1:
This is usually a good idea.
[00:03:20] Matt Rouse:
I I wanted to try and do a little bit of world building in that without getting too extreme. I thought that was a safe bet. You know? And I guess it worked because, because my story made the book. And I should mention that the the book is being published by Amazing Stories. Right? And, Galactic Pest Control was the theme of the stories for the book, which is clever in itself. Like, that kind of story writes itself, I think. But
[00:03:50] AI Narrator #1:
Steve has been posting things today, making graphics and jokes about the Galactic Pest Control Agency. Right. You know, when things are crawling up your pipe, this is who you have to call. And then the link, of course, is a Kickstarter.
[00:04:05] Matt Rouse:
It's almost that like a spin off of, like, pest control was ghostbusters. Right? Yep. I remember when the movie came out, and and I I had a conversation with somebody about it. And they were just like it never occurred to them that I was like, they're making the ghosts out to be like the rats of of the sewers in New York City. Right? Because they're coming up into the hotels and stuff from the sewers, and and they were just like, ah, it never even dawned on me that that was how the story came about. Right? Yep. I don't know if that's originally how they made the story, but that was just my take. I think that even when they did it, they were riffing on something from the past,
[00:04:39] AI Narrator #1:
but I can't remember where the previous incarnation of who you're gonna call was.
[00:04:48] Matt Rouse:
Yeah. That might be worth digging a rabbit hole on the Internet to see if I can find out where that had came from. Well, I think that if you do a search on who you're gonna call, you're gonna get deluged by Ghostbusters references. That's right. We'll do, like, exclusionary search or something, see if we can find out. You know, there's a very clever thing that you can do with the deep research models of AI right now, which is you can say, I want you to research this thing, but I don't want you to include anything about, you know, Ghostbusters. Right? And then it'll go through and try and exclude anything that had Ghostbusters in it and just research the rest. The trouble with using AI for research, though, as some lawyers are discovering, when the judge notices
[00:05:33] AI Narrator #1:
the fictitious references in their brief, they're in trouble.
[00:05:38] Matt Rouse:
That's right. Well and you know what? This brings up an excellent topic, and this is something that I've been working on a lot in other kind of science fiction stories. And the novel that I'm working on is the idea that the AI internally, like, if you if you look at how AI systems are developed now, essentially get a reward, almost like a dopamine response when they fulfill the user's query. Right? Is is the idea is how do I how does the AI solve the problem that you're giving it? How does it find the information? Whatever. There's a reward mechanism in there. And part of the reason they hallucinate is that they're saying, well, I know I don't have the right answer, but I want the user to be happy with the response I'm giving, so I'll make something up that sounds plausible.
[00:06:27] AI Narrator #1:
Do you remember the old search engine Metacrawler? I do. Most people, I think, would say, but that was back when the Internet was new, and its basic gimmick was that it checked multiple other search engines
[00:06:42] Matt Rouse:
to give you results. That's an idea that could actually bear repeating. It is. I think, actually, one one of the ways now is is they call it mixture of experts. And Mhmm. Mixture of experts basically is multiple AI systems that fact check each other before it comes up with the result. And another way that it's done is sending a whole bunch a lot of times they call it agentic now, which is just the we wanna sound fancy word for an AI agent. Do the shopping for you. That's right. Well, another thing that they do is they could use them internally so they can say, okay. We have we have a project for this AI to do. Go out, and we're gonna send a 100 AIs to go do the same task. And And then when they come back, they go, okay. Well, 60 of them got the same answer. So we're gonna say that is likely the best answer. Oh, democracy.
Right. It's it's like Woah. Woah. It's like a right. A democratic answer, but then you use the mixture of experts model to kind of double check and fact check it to make sure what their sources existed and all these things. So there's a lot of ways that they're trying to get that hallucination out of the system. Yep. But it's only on the newest models. So if you're using the free chat GPT for your law firm instead of paying the $20,
[00:07:59] AI Narrator #1:
you know, just just pay for it. Get the good one. You know? Well, you you mentioned sending out multiple AIs to do the work for you. It makes me think that one AI is very, very expensive in terms of energy to the point where people are beginning to worry about this very, very seriously. And if you're sending out multiple independent AIs to do things for you, oh my.
[00:08:22] Matt Rouse:
Well, so there's another thing get the energy cost down. And and you and I had talked a little bit before we started recording about about exponential technologies. One of the things that they found is that there is an exponential decrease in the energy use because of a number of factors, including research and and fixing things and being able to determine how much compute they need to do an inference task, which is after it's trained and you ask it something. It's called inference. So the amount of energy needed to do the inference is dropping. And by improving the data they're using to training, they're also reducing the amount of energy it takes to train the model.
So be able to train a model that's 10 times bigger than the last model, you know, a year from now, but it still uses the same amount of energy used to train the smaller one. Yep. Now you mentioned the training. That was one of the things that I think you wanted to talk about when we were initially talking about this evening.
[00:09:27] AI Narrator #1:
Have you done done the lit gen search on your own work to see how many times your work has been used for training?
[00:09:34] Matt Rouse:
I have not. I have.
[00:09:37] AI Narrator #1:
And, basically, everything I have ever written, fiction and nonfiction, for fifty years, since long before the Internet even K. Has been processed without any attribution or permission or payment.
[00:09:51] Matt Rouse:
Well, I do know that my books were in the llama one for that training, but I didn't do the full at Genentes.
[00:09:58] AI Narrator #1:
It's an insidious kind of a thing to do to check your own work there. One of the things that I find is not really being discussed an awful lot. People talk about copyright infringement. It's not copy. Not really. But copyright law has something else that's also supposed to be protected, and that is the production of derivative works.
[00:10:21] Matt Rouse:
Right.
[00:10:22] AI Narrator #1:
If you train an AI on my fifty years of writing, is that AI a derivative work in the meaning of the law? I don't think anybody has asked that question yet. I mean, that is a good question. I just had an IP attorney on the show, like, two episodes ago. I probably
[00:10:39] Matt Rouse:
I should go back and ask her that question. We did cover a few things that are interesting, though, and I think one of the most interesting things is that the law is different by country. So if your work has been trained in been used for training in one country, it might be illegal or it might be a future court battle. But in another country, it might be completely legal for them to trade on your work. Well, international
[00:11:03] AI Narrator #1:
copyright law is a whole another thing. Right. National is complicated enough. Actually, on my upcoming
[00:11:11] Matt Rouse:
book, the will AI take my job to, I put a passage in the copyright section to say that this is allowed for training purposes for AI models. And the reason I put that in is because that book specifically is gonna go out of date really fast for one thing. But, also, I want my view on it to be part of the training data. And so I'm, like, saying, look. I want my opinions on how this works and stuff to be in the future of the training data, and I also want, you know, if people are looking me up to be able to find that information in the future. You're saying
[00:11:47] AI Narrator #1:
before this recording session that you're talking a lot about the impact on jobs and the eventual need for universal basic income. Yes. Or it has a couple of other names as well. But among the set you know, that one, I think, is going to be essential just to keep the economy working. People don't have jobs. They don't have income. They don't can't spend money. The economy dies, to put it simply. You'll find my discussion of that in the Destinies book. Right. Another complication of the rise of the AI and its impact on jobs is identity crises.
People tend, at least in this country I understand other countries, it's not so bad. But people in this country tend to say who answer when asked, who are you? I'm a librarian. I'm a teacher. I'm a I'm a I'm a whatever. Right. I you identify in terms of your job. If you can't get a job, what happens to your identity or to your sense of identity? And among the suggestions that I make in my own book, my own essay there, is that you're going to see people looking for various other things from churches to political parties to causes, increased tribalism, and we are already seeing that in the last few years of politics.
[00:13:12] Matt Rouse:
Yeah. I think that's a 100% true. How bad is it going to get? A lot worse. Sorry. I don't want to speculate, but I am writing a book that speculates about the future of AI. I'm a restriction writer. I speculate all the time. I a 100% agree with that. Does AI create civil war?
[00:13:32] AI Narrator #1:
Well, I think the AI might stop civil war. Another impact that we're seeing, which is more subtle perhaps, is if the AI takes so many jobs. Higher education now is extraordinarily vocational. When you and I went to college, liberal arts was the standard mode. Right. Since then, even though liberal arts college has added criminal justice programs, child rearing, teaching, counseling, many other things, comp well, just a computer science program for that matter. These are aimed at jobs. If jobs are no longer available, why should people pay for vocational training? What happens to the schools? We already have many small colleges that have already closed, and there are other factors feeding into that as well. Families have gotten smaller, so the catchment area for the students has been shrinking.
But right now, what we're seeing with the Trump administration is they've got an out and out attack on higher education. And you combine it with the big push for AI that they're also going in for, we are going to see some very dramatic changes there. Yeah. As well as the, population decline in most countries in the world. And when I was writing in Destinies, that was that book came out five years ago. And looking at what's happening right now, a science fiction writer is not supposed to predict the future. In fact, a futurologist really shouldn't be thinking that they can predict the future either.
But we can be thinking of, well, yes, we can see certain things trending.
[00:15:09] Matt Rouse:
Right.
[00:15:10] AI Narrator #1:
If this goes on and if that isn't a classic science fiction line, if this goes on
[00:15:18] Matt Rouse:
Right.
[00:15:19] AI Narrator #1:
Actually, that's, that kind of connects to another anthology I would like to do someday if I live long enough and get a publisher interested. Did you ever read Larry Niven's store essay more than story? I don't think so. Anne of Steel, Woman of Kleenex. But I'm gonna write it down. You can look to it up. It is a remarkable reductio ad absurdum, and I would love to do an anthology where writers like you, for instance, were invited to take a science fiction trope and beat it into the ground with a big stick. And the title would be Right. Oh, the absurdity. Does that sound like fun?
[00:15:58] Matt Rouse:
I know I know the absolutely. I'm in. But I know I I'd looked that book up, and then, of course, I got a summary of it because they have AI summary of everything on search now. And I know of the story. It's it's actually a very common, like, schoolboy age story where they say, well, if On the web if anybody Googles. If Superman was gonna, you know, go out with a woman, what would happen? Like, a human woman, you know, that could Well, he'd be punching.
[00:16:28] AI Narrator #1:
His sperm would be poking holes through everything. He would die of peritonitis.
[00:16:33] Matt Rouse:
There's a that it actually kind of reminds me of an old joke where these guys are like they go up to, like, the top of this building with a couple drunk friends or whatever, and the one guy keeps jumping off and and flying back up onto the roof. And he says, if you jump in this specific spot Superman being drunk. Right. Superman, you're being drunk. That's right. So they have the guy jump off the building and execute himself basically because it's All you have to do is believe. All you gotta do is believe. Right? So, anyway, I think I think, I do have some contrary opinions about AI that I think are kind of uncommon. I 100% believe there's gonna be, like, job displacement for sure, if not a lot of job loss. I do think there's gonna be quite a bit of job creation, but I don't think it's gonna be nearly enough to cover the job loss. And, also, you can't just reskill somebody into being, like, a robotics engineer.
Right? Right. It's not the same as when they went from using, you know, going from You have to remember the bell curve. Right. Right. I mean, there's just they can't even find enough people to do the AI work that they're doing now. So as we train people in very hard, high level mathematics and computer science, I mean, there's a pretty slim picking of graduates to get out of those programs compared to the amount of people
[00:18:06] AI Narrator #1:
who work in, like, knowledge worker jobs. Right? Did you see the recent news about this company that actually announces startup actually announces that it is developing AI to do all jobs.
[00:18:23] Matt Rouse:
Oh, I'm sure. I mean, that's isn't that what all AI is eventually? Company that says it is explicitly
[00:18:30] AI Narrator #1:
aiming at creating the software that will replace humans in all jobs. Now there are jobs that they can't replace.
[00:18:39] Matt Rouse:
Carpenter is hard. Plumber is hard. You know? I think that those in the kind of let's call it medium term. Right? Like, the five to ten year term. I think those jobs will be heavily augmented by robotics, but not like a robot who comes to your house as a plumber. It'll be a plumber who has a tiny robot. He throws down a pipe, and it goes and fixes the thing. Or, you know They already have robots that will explore the pipes for you and send back to you. So why can't I have one that just goes and fixes the hole in the pipe? Right? And, you know, like, there's there's a lot of these robotics and stuff that are already being built. Right? Okay.
Maybe your construction guy has a team of 10 autonomous robots that work with them, but the carpenter is still the one who needs to go, okay. I gotta do the fiddly work or, you know, I gotta be the one who makes this decision when there's an edge case of something.
[00:19:35] AI Narrator #1:
Well, here's another one. How about hairdressers?
[00:19:37] Matt Rouse:
Oh, hairdresser's a tough one. And you know what? The thing about hairdressers is if you're a good hairdresser, you're a good hairdresser because of the relationship to your clients, not good because of how well you cut hair. But the work is also very, very far from routine It is. Which makes it harder for a robot. That's right. Because there's not enough examples, and you everybody's hair is different. I think that there's this thing that happens at AI laboratories where they say, okay. We've got thousand low hanging fruit things that we could work with the AI to try and get done. Well, we've only got enough people to do 10 of them. Right?
So I don't have enough time to work on hairdresser because hairdresser, even if you replace thousands of hairdressers, it's still not an economically valuable task compared to doing cheap. Well, it's it's in comparison to, like, a pharmaceutical research company. Right? Right. So if if a pharmaceutical company is worth $3,000,000,000 for the new drug or $30,000,000,000, you gotta replace a lot of hairdressers to make up $30,000,000,000. Right? I'm not saying they work for cheap. I'm saying compared to that industry.
[00:20:51] AI Narrator #1:
Right? Okay. I put it more crudely than you would. Right. But so the point, I think, is valid. The jobs, which would not generate so much income for the AI owner, are the last to be replaced. Right. Your point, in principle, eventually, even those jobs can be done. I mean, eventually we get into the whole UBI thing.
[00:21:16] Matt Rouse:
Yeah. Another I've I've actually heard it called recently. A lot of people call it big, which is basic income guarantee, which is more of an economist term. But same thing. It's paying people money not to work, you know, or work if you want to. Just give everybody some money so they can live. Well, actually, the
[00:21:35] AI Narrator #1:
the real UBI, universal basic income, doesn't say we're giving you money not to work. Right. We're giving you money because the jobs aren't available, and we're covering your basic needs. If you can find a job
[00:21:50] Matt Rouse:
that will pay you in addition, that's cool. I don't I don't think there's anything wrong with the society where I can decide if I wanna work or not. Like, that sounds pretty good to me. I think that should sound good to most people. But, you know, there's always people who were like, well, where's the money gonna come from? Well, where do you think it's gonna come from? You tax the AI companies who took all the jobs of everyone else. Right? And whether that happens or not, that's gonna be another story. Yep. You know, I'm he wasn't the person, I think, or the last person to come up with the idea, but my dad used to teach a case in business school that if a company replaces a worker or a shift of workers with an automation system like a robot, they should have to pay the same amount of income tax as those employees would have generated.
[00:22:38] AI Narrator #1:
Fat chance. But you're Of course. That never happened. You're looking all the way back to CNC. That's right. So For the, for the novices out there, we should probably say what CNC means. Computer numerically controlled. Right. And the CNC machines actually would be trained by a human worker guiding the machine's working parts through the task over and over and over again until the machine could do it. And, of course, then bye bye,
[00:23:12] Matt Rouse:
Right. And now nobody's out there picketing saying get rid of all the CNC machines. Right? Or, like, we brought out a train, so now a 100 people don't have to drive cars or horses across the country to deliver packages, and nobody's out there picketing trains. Right? And what happens is with each kind of technological push, there's pushback at the time, but each kind of advancement does still have displacement. But I think the biggest difference here is the displacement hasn't been between taking a worker and retraining them for a different trade or skill that is, you know, reasonably easy to retrain for. It's In my in my teaching career It might remove their job completely.
[00:24:03] AI Narrator #1:
A college actually would take money from the state to run retraining programs for people that were being left unemployed when the mills in Maine closed. Right. In Waterville, Maine, there was a big stock paper plant, for instance. You could tell you were in Waterville by the smell of sulfur dioxide.
[00:24:23] Matt Rouse:
But that closed down, and the state said, well, you know, okay. We have to retrain people. Let's put some money into this. And I think there's a lot less of that available now. There is. And I think that also depends where you are. Right? Like, I think there's quite a bit of that still in some countries. I know Canada has quite a bit. I when I lived in The United States, I noticed that it wasn't as as much of a thing. Like, it was more of a you we'll give you a loan so you could go back to school versus, like, we'll pay you to be retrained for a new job.
[00:24:54] AI Narrator #1:
Well, in some countries, the the whole retraining or even the initial training thing, college is free. Right. That's like public education.
[00:25:04] Matt Rouse:
Yeah. They're doing that some of that here now with medical staff because we're completely understaffed in the medical system. And they're like, okay. Well, we'll pay for your school, but you have to sign a contract that you're gonna work with us for so many years after you get out of school. Yep. Which I think is
[00:25:19] AI Narrator #1:
fine. You know? Well, that's one of the things that Trump has been killing. There were programs like that with the federal government. We will fund your education, and you will work with, let's say, on an Indian reservation or in a park or someplace for ten years to work it off.
[00:25:37] Matt Rouse:
Right. No more. Yeah. Well, there's definitely a lot of stuff going on in The United States politically right now that I'm not a huge fan of. I mean, I'm not even going to visit, like, the conferences I usually go to in The United States just because I don't wanna get hassled when I go to the border, you know, because I'm not a US citizen. Right? Bad things
[00:25:58] AI Narrator #1:
on something that's going to be posted in public, we could both get in trouble, I suppose.
[00:26:03] Matt Rouse:
That's right, Tom. Stop getting me in trouble.
[00:26:06] AI Narrator #1:
I'm too old to worry about the trouble. I won't let you get into it.
[00:26:10] Matt Rouse:
With some of the time that we have left, I do want to ask you a couple questions. So the the anthology is Tales of Galactic Pest Control. It's on Kickstarter, so people can go I think it's, like, $12 to get a digital copy, $30 if you want a paperback.
[00:26:29] AI Narrator #1:
You you you can back the Kickstarter for basically what you would spend preordering a book on Amazon.
[00:26:36] Matt Rouse:
That's right. So why not? But there is a lot of stories in this book. I think there's what? Is there Well, I just read and four by Steven Silver,
[00:26:45] AI Narrator #1:
and that takes us up to over 30 stories, which is gonna be expensive for the publisher. They've they really got their hopes riding on this Kickstarter.
[00:26:53] Matt Rouse:
That's right. Does you mean the publisher?
[00:26:56] AI Narrator #1:
Even though I don't have to fit the bill.
[00:26:59] Matt Rouse:
Right. So besides my story, which is the obvious choice, what do you think your favorite story in the book is that you've read to you? I'm not going to even try.
[00:27:10] AI Narrator #1:
Come on. You don't wanna be like, all my kids, I love them all equally. Well, I have to talk about yours because there you are right there in front of me, and we've already done that. There are others well, let me say it this way. Every story is one that I liked. Right. Many, many more stories were rejected. And some of them, somebody else might have liked. But I said, nope. Doesn't do it for me. Right now, when somebody says, can I send you a story? I'm saying the book is just about full. You can send it. I'll look at it, but you have to knock my socks off. And when it comes time to reject one, I say, sorry. My socks are still on.
[00:27:51] Matt Rouse:
What happens? I mean, you've been editing for, like, science fiction books and magazines for decades at this point. I haven't edited magazines, but I've edited I've done
[00:28:01] AI Narrator #1:
what, this is the or anthology dating back I think the one was dated 1997.
[00:28:08] Matt Rouse:
And you guys are paying the authors of the stories, which I know sounds to maybe to most people sounds like, well, why wouldn't you pay them? But if you look at the publishing world as it sits right now, especially for short stories, there's not a lot of people paying anything for
[00:28:25] AI Narrator #1:
anything. Well, I I've done three anthologies for I and Randall Strock's fantastic books. The one was science fiction for the throne, one sitting reads. And the idea was to get stories under 2,000 words, preferably reprints, and put about 40 of them in the book. Mhmm. Obviously, it's a bathroom reader. When that was launched, people were picking up three and four copies at the con and, obviously, planning to give copies to their friends. Then we did fantasy for the throne, and then we did horror for the throne with the inevitable joke about it will scare the something out of you. Right.
These things write their own jokes. They do. They write themselves sometimes. And now Ion has actually given us a story for this anthology. So I think turn about there is Fair Play. He's been my editor. Now I'm his. But, I've also known him a lot longer than that. Even before Fantastic Books, He used to work for Analog at one point. Anthologies give me some editing experience. I've had coeditors. That's usually a way to help cut the work. I've also learned that one of the easiest ways to do an anthology, and that's the way David Hartwell used to do it, was to go to a large con and walk through the dealer room and glad hand people and say, I have a sandbox. You wanna
[00:29:55] Matt Rouse:
play? Right. And with
[00:29:57] AI Narrator #1:
the galactic pest control, here's a sandbox. You wanna play? And we have had a lot of good people say, yeah. Sure.
[00:30:05] Matt Rouse:
And I think that when the book comes out, it's gonna be a very a very good one. Yeah. I'm excited, actually. I'm excited to read everybody else's story and see what their take on the on the same idea was. I I think that's also one of the exciting parts about having, like, a theme with so many different people is everybody has their own. Like, when they get when they hear something, they have their own kind of twist and take on it. And,
[00:30:29] AI Narrator #1:
This is one of the things that makes it difficult for people to squeeze in at the end because they might be too similar to something you've already accepted.
[00:30:39] Matt Rouse:
Right. Well, hopefully, nobody's too similar to mine. Well, like I say, at the end, you have to blow my socks off. On the AI front, I know that we we did talk a bit about, you know, AI and and the potential of job loss and things like that. Do you see a possible future where there is some kind of, like, greater good that comes out of AI rather than just everybody's gonna lose the job and nobody has any money?
[00:31:09] AI Narrator #1:
Well, we we tend to focus on those impacts that so pose problems. But I did a a story that explored some of this. It appeared in Black Cat Weekly a year ago. And, basically, it what it said was that once you start having to face the identity crisis thing that I mentioned Mhmm. You need to find something else to do. So we have a future in which people have battles of the bands in the streets, saxophones against Right. Bagpipes. You know? You could really get into that. Right. Well, if you got time You don't have anything else to do, so why not learn an instrument?
[00:31:57] Matt Rouse:
Well, you know, I
[00:31:58] AI Narrator #1:
And you can also do the whole, what what what is it? The the anachronism people who do jousting and so
[00:32:08] Matt Rouse:
Right. I was gonna say what if what would happen to the world if the starving artists didn't have to be starving? Suddenly, they can actually where you can go. Do their work. Right? And Well, some I mean, I live in an artist community. So Well, some artists
[00:32:23] AI Narrator #1:
would say that's cool. I don't have to worry. I can just concentrate on building this one massive masterwork. And some people manage to do that anyway. I mean, you think about, what is it, Barcelona with that cathedral?
[00:32:36] Matt Rouse:
Mhmm. And then you have a whole lot of other people saying, oh, I don't have to be so goddamn obsessive, and they take a nap. Right. They're like, I don't have to make. I can focus on on what I wanna do for work instead of making 17 pottery cups a day to to put food on the table. You know? Yeah. Where how many paintings do I gotta sell this month to keep my kids in school? My brother was an artist. He died over twenty years ago, but
[00:33:04] AI Narrator #1:
he was a sculptor, painter. Mhmm. His name was Walter Easton. You can look him up. Oh, perhaps the best thing that you can find is, a movie that he made with the head of the art department at Colby College up in Maine, and it was up here. I think that's it. Deep Trout is the name of the trilogy. I have it here on DVD. It it can be found online. And he had the kind of artist ego that said he didn't need to sign his work because everybody should know it was his. Right. That's deep in your mouth. In my mind, a little extreme. I'm not a credit hog.
I'm willing to share, but I do want my name on it. Right. I think you do too. Deep Trout was an obvious, play on words At the time, Deep Throat was in recent memory.
[00:34:12] Matt Rouse:
Right.
[00:34:13] AI Narrator #1:
And he did some very, very interesting sculpture. A lot of it is showcased there, but he never made any money. One problem was that he had some things that could easily have been not exactly mass produced, but at least cranked out with, you know, variations. Small sculptures that might have qualified as motel art. He could have sold a ton of them, but that was beneath him. This is the kind of thing AI specializes in. Oops. I shouldn't say things like that. Right.
[00:34:48] Matt Rouse:
Yeah. I mean, if it's something that is repeatable and has some sort of kind of guiding guideline structure to it that could be, you know, even even with modifications, but still repeatable a while. That's definitely something in the AI, you know, frontier. I think people are gonna really quickly start to look at some of the stuff that's come out really recently. Like, cloud four just came out from Anthropic, and it can work autonomously on things. Honestly, I was I was looking through something today, and it didn't know the answer, like, in its own training data. It it, like, admitted. It was like, well, I don't know the answer to that, so I'm gonna go look it up. And then it went and looked some stuff, and then it told me what it looked up. And then it gave me the sources of where it looked it up from. And, you know, it did all that without me having to ask it. Well, go look it up or go do that. You know? And then there's just, like, v o three just came out from Google, which you could type in a text prompt, and it will make HD video. It keeps the characters, and it makes the audio as well. So it lip syncs to the people's voices and things like that. It is essentially indistinguishable from, you know, well shot video.
I think this this generation of stuff that's out right now is is indistinguishable from human made content for people. And the next generation is better than human content, which better is is, you know, and how do you measure better, right, is is the thing like you can you can tell if it's better at math than it was before, but you can't really tell if it's, like, more of a creative writer than it used to be, You know? So some things are a bit squishy. If we can get the AI to the point where it doesn't hallucinate
[00:36:35] AI Narrator #1:
I'm not sure I like that particular term, but you know what I mean. Right. If if we can get it to the point where it doesn't do that, and if it's doing a search and it comes up dry and it says, I will look that up instead of manufacturing it Then we're pretty much there. That's very useful. Yeah. We're there now. It's just
[00:36:53] Matt Rouse:
only on the cutting edge models.
[00:36:55] AI Narrator #1:
And if we have an AI that says, oh, yes. I can work with your drafting machine to generate a file through your three d printer, that will be pretty cool.
[00:37:05] Matt Rouse:
Yes. Actually, the program called might be cooler than the real thing. There's a program called Mesh. It's an AI three d model creator. Yep. So, I mean, here's the thing. It's it's like there's so much out there now, and it's being developed so quickly by so many different people, and it has so much money behind it and so much momentum behind it that it's gonna get better at a pace that people can't comprehend.
[00:37:31] AI Narrator #1:
And That's what I was saying about wetting the bottom of Lake Michigan. That's right. Exponential growth can fill the lake up in a hurry.
[00:37:39] Matt Rouse:
Yeah. My dad used to have this this thing that when he was teaching, he would say if if the amount of lily pads on the pond doubles every three months
[00:37:50] AI Narrator #1:
Yeah. And Or the the classic chess puzzle from India. How should I pay you? Oh, just put one grain of rice on one square, two grains on the next, four on the next.
[00:38:01] Matt Rouse:
Double it on every square.
[00:38:05] AI Narrator #1:
Bankrupt pick up the raja.
[00:38:07] Matt Rouse:
I used that analogy in my book of will AI take my job. And on the last square, it would be pile of rice the size of Mount Everest if each grain of rice was the equivalent of every grain of sand on the Earth? Every equivalent of what? Every grain of sand on the Earth. Yeah. So it would be so big, it would be bigger than the planet, basically.
[00:38:32] AI Narrator #1:
That's the whole point. Right. Through, the whole thing of exponential growth, this is one of the classic examples. And, it was a sucker punch for the Raja who made that stupid offer, and it has been proposed in various other forms as well, though I think mostly as illustrations rather than actual history. I don't see anybody actually buying into it. And how long would it take the Raja after all to say off with his head? That's right.
[00:39:03] Matt Rouse:
Well, everybody out there, start counting your rice because, current the current model of, of AI is is doubling currently every seven months. Every seven months? Every seven months, it's doubling down. Like Michigan's bottom. That's right. Yep. I was talking to a guy who had a company. I said, how long do you think it's gonna be before AI can do what your company does? And he said, oh, it's at least three generations. I said, well, the next generation's coming out next month, so that means you got you got a year and a half until it could do what your company does. And he's like, ah, shit. And I'm like, this is why we need to think. Right? We need to think about stuff in advance. We gotta go look.
If this got better that quickly at this thing, what would it mean for me and my job or my family or, you know, my business or whatever it is. Right? Well, but this this has all come in my lifetime.
[00:40:03] AI Narrator #1:
I told you earlier before we did this, my AI exposure goes all the way back to the McCullough PIX artificial neuron. That was dated to the nineteen fifties. In the nineteen sixties, there was the perceptron. Mhmm. This is when a mini computer was the size of a refrigerator, the PDP eight, PDP 10. They didn't have microcomputers yet. But the Perceptron was a small neural net, and they could already do some very interesting things, like checking credit card fraud. And it was forgotten because it was so small and so limited. Right. And then twenty years later, people started talking about this hot new thing, neural nets.
Right. I think it was 20. It might have been 30. But it was abandoned for a long time to the point where most people don't even recognize the term perceptron anymore. Do you recall it? I don't. The only reason I know of it is because I did the research for the book of where did kind of AI get started. Right? It's out there. But things have come a long, long way in one man's lifetime. What happens in the next lifetime? It's true. Woah.
[00:41:15] Matt Rouse:
Yeah. Well and you talk about abandoned technologies, like, the LLM was essentially the the trans the pre trained transformer was an abandoned technology. Right? Google invented the transformer. They built it, and they said this is a dead end. There's nothing we could do with this. And then, you know, OpenAI's researchers kind of went and said, look. You know, Ilya Sutzkever was like, I'm gonna pump a ton of data into this thing of this dataset that I have. Everybody was on reinforcement learning. And as soon as the large language model came out and went crazy, everybody started building them. But now they're going back and using the reinforcement learning on top of the large language model,
[00:41:55] AI Narrator #1:
And now they're getting these results that are just incredible, like, you know, multi layered brains and circuit design and synapses. Term small language model yet? I don't know how. Meaning, basically, not so extensive on the training. That's right. It's gonna have to get there. And then we have the whole issue of embodiment.
[00:42:15] Matt Rouse:
Mhmm. Embodied robots and
[00:42:18] AI Narrator #1:
a humanoid AI and raise it like a child. There's been a certain amount of science fiction that has dealt with that. You know what's funny?
[00:42:26] Matt Rouse:
My story, my novel that I'm writing is called AGI Next. I'm I'm currently on chapter 23, so I'm I'm working my way. I don't know how long it's gonna be, though. But, the the idea in my book is that there is an AGI system who also has an embodied robot of itself who is raising a human child. So it's kind of the reverse.
[00:42:52] AI Narrator #1:
Do you think we've given people enough keywords now?
[00:42:57] Matt Rouse:
No. I haven't said edge case more than twice, I think. So that's my my buzzword of the day. I did talk in,
[00:43:04] AI Narrator #1:
Destinies about the impact on education. The you know, as education becomes more vocational and the jobs become fewer, boom, there goes higher education.
[00:43:15] Matt Rouse:
Yeah. Well, I I mean, I know a lot of people who taught. My my uncle and my parents all taught in higher ed. It's quite a thing. I mean, I talk about education in the book, but I mostly talk about, you know, what are you gonna do in the next few years, not what are you gonna do ten years from now. Okay. You know? But say, you know what? I don't wanna cut it too short, but I also I don't want people to have to spend three hours in their car to listen to our whole talk here. So I think maybe we should over your time limit. That's alright. It's okay. It's it's artificial time limit until the AI imposes it on you. Waiting. I'm having fun. But
[00:43:49] AI Narrator #1:
if people expect a half hour or forty five minutes, and here it is, quarter past five almost. Right. We're almost almost an hour in. The plug for the Galactic Pest Control Anthology, and there's the the link for the Kickstarter there in the private chat. It is I'll put it in the I'll put it in the show notes for the show, and we'll share it out. While I'm at it, I should unplug the five year old book, Destiny's Issues to Shape Our Future, which has obviously dealt with some of the things we've been talking about.
[00:44:21] Matt Rouse:
Absolutely. And, Tom, thank you so much for coming on the show. It's so exciting to talk to another science fiction writer, especially somebody who's accomplished as yourself, and I really appreciate it. And, again, it's Tales of Galactic Pest Control, a sci fi anthology. It is on Kickstarter right now. Go back it back real authors doing real work and some of us talking about AI, but not everyone. And,
[00:44:48] AI Narrator #1:
Math will get paid double for a story.
[00:44:53] Matt Rouse:
I'll get paid I'll be getting paid twice of what I put into the Kickstarter. Yeah. It's not about the money, honestly. I I just I love writing, and I was so excited when I got Kermit sent me the acceptance email, you know, saying that we think it's gonna make it. And then I finally did, and and it was great. So I I appreciate you guys. Okay. It's been a pleasure. Thanks, Tom. And, man, we'll have to talk again soon when, when the next book comes out. Where are you located? I'm in Nova Scotia. Nova Scotia.
[00:45:27] AI Narrator #1:
Do you get to cons in The US?
[00:45:29] Matt Rouse:
I have sometimes. The guy used to live in Oregon, but I have not recently. Okay. Because we moved up here at COVID. My next one will be ReaderCon
[00:45:39] AI Narrator #1:
in Burlington, Massachusetts, which is not too far from you That's right. From Nova Scotia. You can actually take the ferry and drive. A few years back, my wife and I took that ferry from Bar Harbor and drove all the way out to Halifax. And it was it was interesting. The idea in Nova Scotia of an overpass is particularly interesting. It's a stop sign.
[00:46:07] Matt Rouse:
That's right. Well, the Bar Harbor ferry goes to Yarmouth, which is about a hour drive from my house. But, yeah, it's I I live kind of in the middle of nowhere, but they call it the the cradle of civilization in in Canada because I live on the oldest road in North America. So I live on Granville Road that's something like 400 years old. And, Well, I have ancestors
[00:46:30] AI Narrator #1:
that went to Nova Scotia as Tories at the time of the revolution.
[00:46:35] Matt Rouse:
Oh, yeah. They probably live in Bear River. After everything settled down, they came back. Right.
[00:46:42] AI Narrator #1:
But That's good to say all the time. They went to Lunenburg.
[00:46:45] Matt Rouse:
Oh, yeah. I know Lunenburg. It's just across the across the province. I gotta drive across, but it only takes an hour. It's it's kinda like a peninsula here. So I appreciate it, and we'll talk again soon, and let's hope, Kickstarter goes well. And then We'll drop to a quick start. And, yeah. And then we'll have to get on to the next one.
[00:47:07] Unknown:
Thank you for listening to the Digital Marketing Masters podcast with your host, Matt Rouse. If you enjoyed the show, please leave us an honest review on your favorite podcast app. Be sure to check out Matt's book, Will AI Take My Job? And have a wonderful day.
Today on the digital marketing masters podcast, science fiction legend,
[00:00:55] Matt Rouse:
And welcome back to Digital Marketing Masters, everyone. I'm your host, Matt Raus. And today, I'm super excited. My guest is Tom Easton. How How are you doing, Tom? I'm doing fine. You wanted a quick introduction? Well, I'll just say I'm a retired science professor who's been writing science fiction since the nineteen seventies.
[00:01:14] AI Narrator #1:
And if you take a quick look at the t shirt I'm wearing, Amazing Select run by Steve Davidson and Kermit Whittle want to celebrate the hundredth anniversary of magazine science fiction next year at Raven Con in Virginia. I don't know if I'll be able to make it, but I did buy their t shirt so I can why get in front of audiences like this. That's right. So Amazing Tales, a hundred years now. Right? 1926 was when Amazing published.
[00:01:44] Matt Rouse:
Nice. The reason that we have connected is because Amazing Stories has an anthology coming out. Actually, you may not know this, Tom. This was my submission of a science fiction story that I ever did, and all the books I've written in the past were self published. So this is the time I ever asked anybody if I could get my story in their book. Sure enough, I got invited to be in this one.
[00:02:09] AI Narrator #1:
Well, with your abiding interest in AI, you have you touched some very interesting bases in that story. Way back when, people started talking about where science fiction might go and what apply where AI might go and whether we should be scared of it. And there were people who were saying, we shouldn't be scared about it because they will regard us as gods. Their creators. You may remember reading comments like that. Personally, I think that was an awful lot of wishful thinking because after all, we don't do a very good job of listening to our gods or agreeing on just which gods to pay attention to. That's right. But in your story, you don't go into that very much, but you do mention that the robots should pay attention to the humans because the humans created them. Right. That just touches on that very, very old notion.
I don't did you have to resist temptation to dig further into that?
[00:03:09] Matt Rouse:
Oh, a little bit, but I also didn't want to kind of extend the length of the story too long. I really wanted to try and keep it down to, like, the essentials.
[00:03:19] AI Narrator #1:
This is usually a good idea.
[00:03:20] Matt Rouse:
I I wanted to try and do a little bit of world building in that without getting too extreme. I thought that was a safe bet. You know? And I guess it worked because, because my story made the book. And I should mention that the the book is being published by Amazing Stories. Right? And, Galactic Pest Control was the theme of the stories for the book, which is clever in itself. Like, that kind of story writes itself, I think. But
[00:03:50] AI Narrator #1:
Steve has been posting things today, making graphics and jokes about the Galactic Pest Control Agency. Right. You know, when things are crawling up your pipe, this is who you have to call. And then the link, of course, is a Kickstarter.
[00:04:05] Matt Rouse:
It's almost that like a spin off of, like, pest control was ghostbusters. Right? Yep. I remember when the movie came out, and and I I had a conversation with somebody about it. And they were just like it never occurred to them that I was like, they're making the ghosts out to be like the rats of of the sewers in New York City. Right? Because they're coming up into the hotels and stuff from the sewers, and and they were just like, ah, it never even dawned on me that that was how the story came about. Right? Yep. I don't know if that's originally how they made the story, but that was just my take. I think that even when they did it, they were riffing on something from the past,
[00:04:39] AI Narrator #1:
but I can't remember where the previous incarnation of who you're gonna call was.
[00:04:48] Matt Rouse:
Yeah. That might be worth digging a rabbit hole on the Internet to see if I can find out where that had came from. Well, I think that if you do a search on who you're gonna call, you're gonna get deluged by Ghostbusters references. That's right. We'll do, like, exclusionary search or something, see if we can find out. You know, there's a very clever thing that you can do with the deep research models of AI right now, which is you can say, I want you to research this thing, but I don't want you to include anything about, you know, Ghostbusters. Right? And then it'll go through and try and exclude anything that had Ghostbusters in it and just research the rest. The trouble with using AI for research, though, as some lawyers are discovering, when the judge notices
[00:05:33] AI Narrator #1:
the fictitious references in their brief, they're in trouble.
[00:05:38] Matt Rouse:
That's right. Well and you know what? This brings up an excellent topic, and this is something that I've been working on a lot in other kind of science fiction stories. And the novel that I'm working on is the idea that the AI internally, like, if you if you look at how AI systems are developed now, essentially get a reward, almost like a dopamine response when they fulfill the user's query. Right? Is is the idea is how do I how does the AI solve the problem that you're giving it? How does it find the information? Whatever. There's a reward mechanism in there. And part of the reason they hallucinate is that they're saying, well, I know I don't have the right answer, but I want the user to be happy with the response I'm giving, so I'll make something up that sounds plausible.
[00:06:27] AI Narrator #1:
Do you remember the old search engine Metacrawler? I do. Most people, I think, would say, but that was back when the Internet was new, and its basic gimmick was that it checked multiple other search engines
[00:06:42] Matt Rouse:
to give you results. That's an idea that could actually bear repeating. It is. I think, actually, one one of the ways now is is they call it mixture of experts. And Mhmm. Mixture of experts basically is multiple AI systems that fact check each other before it comes up with the result. And another way that it's done is sending a whole bunch a lot of times they call it agentic now, which is just the we wanna sound fancy word for an AI agent. Do the shopping for you. That's right. Well, another thing that they do is they could use them internally so they can say, okay. We have we have a project for this AI to do. Go out, and we're gonna send a 100 AIs to go do the same task. And And then when they come back, they go, okay. Well, 60 of them got the same answer. So we're gonna say that is likely the best answer. Oh, democracy.
Right. It's it's like Woah. Woah. It's like a right. A democratic answer, but then you use the mixture of experts model to kind of double check and fact check it to make sure what their sources existed and all these things. So there's a lot of ways that they're trying to get that hallucination out of the system. Yep. But it's only on the newest models. So if you're using the free chat GPT for your law firm instead of paying the $20,
[00:07:59] AI Narrator #1:
you know, just just pay for it. Get the good one. You know? Well, you you mentioned sending out multiple AIs to do the work for you. It makes me think that one AI is very, very expensive in terms of energy to the point where people are beginning to worry about this very, very seriously. And if you're sending out multiple independent AIs to do things for you, oh my.
[00:08:22] Matt Rouse:
Well, so there's another thing get the energy cost down. And and you and I had talked a little bit before we started recording about about exponential technologies. One of the things that they found is that there is an exponential decrease in the energy use because of a number of factors, including research and and fixing things and being able to determine how much compute they need to do an inference task, which is after it's trained and you ask it something. It's called inference. So the amount of energy needed to do the inference is dropping. And by improving the data they're using to training, they're also reducing the amount of energy it takes to train the model.
So be able to train a model that's 10 times bigger than the last model, you know, a year from now, but it still uses the same amount of energy used to train the smaller one. Yep. Now you mentioned the training. That was one of the things that I think you wanted to talk about when we were initially talking about this evening.
[00:09:27] AI Narrator #1:
Have you done done the lit gen search on your own work to see how many times your work has been used for training?
[00:09:34] Matt Rouse:
I have not. I have.
[00:09:37] AI Narrator #1:
And, basically, everything I have ever written, fiction and nonfiction, for fifty years, since long before the Internet even K. Has been processed without any attribution or permission or payment.
[00:09:51] Matt Rouse:
Well, I do know that my books were in the llama one for that training, but I didn't do the full at Genentes.
[00:09:58] AI Narrator #1:
It's an insidious kind of a thing to do to check your own work there. One of the things that I find is not really being discussed an awful lot. People talk about copyright infringement. It's not copy. Not really. But copyright law has something else that's also supposed to be protected, and that is the production of derivative works.
[00:10:21] Matt Rouse:
Right.
[00:10:22] AI Narrator #1:
If you train an AI on my fifty years of writing, is that AI a derivative work in the meaning of the law? I don't think anybody has asked that question yet. I mean, that is a good question. I just had an IP attorney on the show, like, two episodes ago. I probably
[00:10:39] Matt Rouse:
I should go back and ask her that question. We did cover a few things that are interesting, though, and I think one of the most interesting things is that the law is different by country. So if your work has been trained in been used for training in one country, it might be illegal or it might be a future court battle. But in another country, it might be completely legal for them to trade on your work. Well, international
[00:11:03] AI Narrator #1:
copyright law is a whole another thing. Right. National is complicated enough. Actually, on my upcoming
[00:11:11] Matt Rouse:
book, the will AI take my job to, I put a passage in the copyright section to say that this is allowed for training purposes for AI models. And the reason I put that in is because that book specifically is gonna go out of date really fast for one thing. But, also, I want my view on it to be part of the training data. And so I'm, like, saying, look. I want my opinions on how this works and stuff to be in the future of the training data, and I also want, you know, if people are looking me up to be able to find that information in the future. You're saying
[00:11:47] AI Narrator #1:
before this recording session that you're talking a lot about the impact on jobs and the eventual need for universal basic income. Yes. Or it has a couple of other names as well. But among the set you know, that one, I think, is going to be essential just to keep the economy working. People don't have jobs. They don't have income. They don't can't spend money. The economy dies, to put it simply. You'll find my discussion of that in the Destinies book. Right. Another complication of the rise of the AI and its impact on jobs is identity crises.
People tend, at least in this country I understand other countries, it's not so bad. But people in this country tend to say who answer when asked, who are you? I'm a librarian. I'm a teacher. I'm a I'm a I'm a whatever. Right. I you identify in terms of your job. If you can't get a job, what happens to your identity or to your sense of identity? And among the suggestions that I make in my own book, my own essay there, is that you're going to see people looking for various other things from churches to political parties to causes, increased tribalism, and we are already seeing that in the last few years of politics.
[00:13:12] Matt Rouse:
Yeah. I think that's a 100% true. How bad is it going to get? A lot worse. Sorry. I don't want to speculate, but I am writing a book that speculates about the future of AI. I'm a restriction writer. I speculate all the time. I a 100% agree with that. Does AI create civil war?
[00:13:32] AI Narrator #1:
Well, I think the AI might stop civil war. Another impact that we're seeing, which is more subtle perhaps, is if the AI takes so many jobs. Higher education now is extraordinarily vocational. When you and I went to college, liberal arts was the standard mode. Right. Since then, even though liberal arts college has added criminal justice programs, child rearing, teaching, counseling, many other things, comp well, just a computer science program for that matter. These are aimed at jobs. If jobs are no longer available, why should people pay for vocational training? What happens to the schools? We already have many small colleges that have already closed, and there are other factors feeding into that as well. Families have gotten smaller, so the catchment area for the students has been shrinking.
But right now, what we're seeing with the Trump administration is they've got an out and out attack on higher education. And you combine it with the big push for AI that they're also going in for, we are going to see some very dramatic changes there. Yeah. As well as the, population decline in most countries in the world. And when I was writing in Destinies, that was that book came out five years ago. And looking at what's happening right now, a science fiction writer is not supposed to predict the future. In fact, a futurologist really shouldn't be thinking that they can predict the future either.
But we can be thinking of, well, yes, we can see certain things trending.
[00:15:09] Matt Rouse:
Right.
[00:15:10] AI Narrator #1:
If this goes on and if that isn't a classic science fiction line, if this goes on
[00:15:18] Matt Rouse:
Right.
[00:15:19] AI Narrator #1:
Actually, that's, that kind of connects to another anthology I would like to do someday if I live long enough and get a publisher interested. Did you ever read Larry Niven's store essay more than story? I don't think so. Anne of Steel, Woman of Kleenex. But I'm gonna write it down. You can look to it up. It is a remarkable reductio ad absurdum, and I would love to do an anthology where writers like you, for instance, were invited to take a science fiction trope and beat it into the ground with a big stick. And the title would be Right. Oh, the absurdity. Does that sound like fun?
[00:15:58] Matt Rouse:
I know I know the absolutely. I'm in. But I know I I'd looked that book up, and then, of course, I got a summary of it because they have AI summary of everything on search now. And I know of the story. It's it's actually a very common, like, schoolboy age story where they say, well, if On the web if anybody Googles. If Superman was gonna, you know, go out with a woman, what would happen? Like, a human woman, you know, that could Well, he'd be punching.
[00:16:28] AI Narrator #1:
His sperm would be poking holes through everything. He would die of peritonitis.
[00:16:33] Matt Rouse:
There's a that it actually kind of reminds me of an old joke where these guys are like they go up to, like, the top of this building with a couple drunk friends or whatever, and the one guy keeps jumping off and and flying back up onto the roof. And he says, if you jump in this specific spot Superman being drunk. Right. Superman, you're being drunk. That's right. So they have the guy jump off the building and execute himself basically because it's All you have to do is believe. All you gotta do is believe. Right? So, anyway, I think I think, I do have some contrary opinions about AI that I think are kind of uncommon. I 100% believe there's gonna be, like, job displacement for sure, if not a lot of job loss. I do think there's gonna be quite a bit of job creation, but I don't think it's gonna be nearly enough to cover the job loss. And, also, you can't just reskill somebody into being, like, a robotics engineer.
Right? Right. It's not the same as when they went from using, you know, going from You have to remember the bell curve. Right. Right. I mean, there's just they can't even find enough people to do the AI work that they're doing now. So as we train people in very hard, high level mathematics and computer science, I mean, there's a pretty slim picking of graduates to get out of those programs compared to the amount of people
[00:18:06] AI Narrator #1:
who work in, like, knowledge worker jobs. Right? Did you see the recent news about this company that actually announces startup actually announces that it is developing AI to do all jobs.
[00:18:23] Matt Rouse:
Oh, I'm sure. I mean, that's isn't that what all AI is eventually? Company that says it is explicitly
[00:18:30] AI Narrator #1:
aiming at creating the software that will replace humans in all jobs. Now there are jobs that they can't replace.
[00:18:39] Matt Rouse:
Carpenter is hard. Plumber is hard. You know? I think that those in the kind of let's call it medium term. Right? Like, the five to ten year term. I think those jobs will be heavily augmented by robotics, but not like a robot who comes to your house as a plumber. It'll be a plumber who has a tiny robot. He throws down a pipe, and it goes and fixes the thing. Or, you know They already have robots that will explore the pipes for you and send back to you. So why can't I have one that just goes and fixes the hole in the pipe? Right? And, you know, like, there's there's a lot of these robotics and stuff that are already being built. Right? Okay.
Maybe your construction guy has a team of 10 autonomous robots that work with them, but the carpenter is still the one who needs to go, okay. I gotta do the fiddly work or, you know, I gotta be the one who makes this decision when there's an edge case of something.
[00:19:35] AI Narrator #1:
Well, here's another one. How about hairdressers?
[00:19:37] Matt Rouse:
Oh, hairdresser's a tough one. And you know what? The thing about hairdressers is if you're a good hairdresser, you're a good hairdresser because of the relationship to your clients, not good because of how well you cut hair. But the work is also very, very far from routine It is. Which makes it harder for a robot. That's right. Because there's not enough examples, and you everybody's hair is different. I think that there's this thing that happens at AI laboratories where they say, okay. We've got thousand low hanging fruit things that we could work with the AI to try and get done. Well, we've only got enough people to do 10 of them. Right?
So I don't have enough time to work on hairdresser because hairdresser, even if you replace thousands of hairdressers, it's still not an economically valuable task compared to doing cheap. Well, it's it's in comparison to, like, a pharmaceutical research company. Right? Right. So if if a pharmaceutical company is worth $3,000,000,000 for the new drug or $30,000,000,000, you gotta replace a lot of hairdressers to make up $30,000,000,000. Right? I'm not saying they work for cheap. I'm saying compared to that industry.
[00:20:51] AI Narrator #1:
Right? Okay. I put it more crudely than you would. Right. But so the point, I think, is valid. The jobs, which would not generate so much income for the AI owner, are the last to be replaced. Right. Your point, in principle, eventually, even those jobs can be done. I mean, eventually we get into the whole UBI thing.
[00:21:16] Matt Rouse:
Yeah. Another I've I've actually heard it called recently. A lot of people call it big, which is basic income guarantee, which is more of an economist term. But same thing. It's paying people money not to work, you know, or work if you want to. Just give everybody some money so they can live. Well, actually, the
[00:21:35] AI Narrator #1:
the real UBI, universal basic income, doesn't say we're giving you money not to work. Right. We're giving you money because the jobs aren't available, and we're covering your basic needs. If you can find a job
[00:21:50] Matt Rouse:
that will pay you in addition, that's cool. I don't I don't think there's anything wrong with the society where I can decide if I wanna work or not. Like, that sounds pretty good to me. I think that should sound good to most people. But, you know, there's always people who were like, well, where's the money gonna come from? Well, where do you think it's gonna come from? You tax the AI companies who took all the jobs of everyone else. Right? And whether that happens or not, that's gonna be another story. Yep. You know, I'm he wasn't the person, I think, or the last person to come up with the idea, but my dad used to teach a case in business school that if a company replaces a worker or a shift of workers with an automation system like a robot, they should have to pay the same amount of income tax as those employees would have generated.
[00:22:38] AI Narrator #1:
Fat chance. But you're Of course. That never happened. You're looking all the way back to CNC. That's right. So For the, for the novices out there, we should probably say what CNC means. Computer numerically controlled. Right. And the CNC machines actually would be trained by a human worker guiding the machine's working parts through the task over and over and over again until the machine could do it. And, of course, then bye bye,
[00:23:12] Matt Rouse:
Right. And now nobody's out there picketing saying get rid of all the CNC machines. Right? Or, like, we brought out a train, so now a 100 people don't have to drive cars or horses across the country to deliver packages, and nobody's out there picketing trains. Right? And what happens is with each kind of technological push, there's pushback at the time, but each kind of advancement does still have displacement. But I think the biggest difference here is the displacement hasn't been between taking a worker and retraining them for a different trade or skill that is, you know, reasonably easy to retrain for. It's In my in my teaching career It might remove their job completely.
[00:24:03] AI Narrator #1:
A college actually would take money from the state to run retraining programs for people that were being left unemployed when the mills in Maine closed. Right. In Waterville, Maine, there was a big stock paper plant, for instance. You could tell you were in Waterville by the smell of sulfur dioxide.
[00:24:23] Matt Rouse:
But that closed down, and the state said, well, you know, okay. We have to retrain people. Let's put some money into this. And I think there's a lot less of that available now. There is. And I think that also depends where you are. Right? Like, I think there's quite a bit of that still in some countries. I know Canada has quite a bit. I when I lived in The United States, I noticed that it wasn't as as much of a thing. Like, it was more of a you we'll give you a loan so you could go back to school versus, like, we'll pay you to be retrained for a new job.
[00:24:54] AI Narrator #1:
Well, in some countries, the the whole retraining or even the initial training thing, college is free. Right. That's like public education.
[00:25:04] Matt Rouse:
Yeah. They're doing that some of that here now with medical staff because we're completely understaffed in the medical system. And they're like, okay. Well, we'll pay for your school, but you have to sign a contract that you're gonna work with us for so many years after you get out of school. Yep. Which I think is
[00:25:19] AI Narrator #1:
fine. You know? Well, that's one of the things that Trump has been killing. There were programs like that with the federal government. We will fund your education, and you will work with, let's say, on an Indian reservation or in a park or someplace for ten years to work it off.
[00:25:37] Matt Rouse:
Right. No more. Yeah. Well, there's definitely a lot of stuff going on in The United States politically right now that I'm not a huge fan of. I mean, I'm not even going to visit, like, the conferences I usually go to in The United States just because I don't wanna get hassled when I go to the border, you know, because I'm not a US citizen. Right? Bad things
[00:25:58] AI Narrator #1:
on something that's going to be posted in public, we could both get in trouble, I suppose.
[00:26:03] Matt Rouse:
That's right, Tom. Stop getting me in trouble.
[00:26:06] AI Narrator #1:
I'm too old to worry about the trouble. I won't let you get into it.
[00:26:10] Matt Rouse:
With some of the time that we have left, I do want to ask you a couple questions. So the the anthology is Tales of Galactic Pest Control. It's on Kickstarter, so people can go I think it's, like, $12 to get a digital copy, $30 if you want a paperback.
[00:26:29] AI Narrator #1:
You you you can back the Kickstarter for basically what you would spend preordering a book on Amazon.
[00:26:36] Matt Rouse:
That's right. So why not? But there is a lot of stories in this book. I think there's what? Is there Well, I just read and four by Steven Silver,
[00:26:45] AI Narrator #1:
and that takes us up to over 30 stories, which is gonna be expensive for the publisher. They've they really got their hopes riding on this Kickstarter.
[00:26:53] Matt Rouse:
That's right. Does you mean the publisher?
[00:26:56] AI Narrator #1:
Even though I don't have to fit the bill.
[00:26:59] Matt Rouse:
Right. So besides my story, which is the obvious choice, what do you think your favorite story in the book is that you've read to you? I'm not going to even try.
[00:27:10] AI Narrator #1:
Come on. You don't wanna be like, all my kids, I love them all equally. Well, I have to talk about yours because there you are right there in front of me, and we've already done that. There are others well, let me say it this way. Every story is one that I liked. Right. Many, many more stories were rejected. And some of them, somebody else might have liked. But I said, nope. Doesn't do it for me. Right now, when somebody says, can I send you a story? I'm saying the book is just about full. You can send it. I'll look at it, but you have to knock my socks off. And when it comes time to reject one, I say, sorry. My socks are still on.
[00:27:51] Matt Rouse:
What happens? I mean, you've been editing for, like, science fiction books and magazines for decades at this point. I haven't edited magazines, but I've edited I've done
[00:28:01] AI Narrator #1:
what, this is the or anthology dating back I think the one was dated 1997.
[00:28:08] Matt Rouse:
And you guys are paying the authors of the stories, which I know sounds to maybe to most people sounds like, well, why wouldn't you pay them? But if you look at the publishing world as it sits right now, especially for short stories, there's not a lot of people paying anything for
[00:28:25] AI Narrator #1:
anything. Well, I I've done three anthologies for I and Randall Strock's fantastic books. The one was science fiction for the throne, one sitting reads. And the idea was to get stories under 2,000 words, preferably reprints, and put about 40 of them in the book. Mhmm. Obviously, it's a bathroom reader. When that was launched, people were picking up three and four copies at the con and, obviously, planning to give copies to their friends. Then we did fantasy for the throne, and then we did horror for the throne with the inevitable joke about it will scare the something out of you. Right.
These things write their own jokes. They do. They write themselves sometimes. And now Ion has actually given us a story for this anthology. So I think turn about there is Fair Play. He's been my editor. Now I'm his. But, I've also known him a lot longer than that. Even before Fantastic Books, He used to work for Analog at one point. Anthologies give me some editing experience. I've had coeditors. That's usually a way to help cut the work. I've also learned that one of the easiest ways to do an anthology, and that's the way David Hartwell used to do it, was to go to a large con and walk through the dealer room and glad hand people and say, I have a sandbox. You wanna
[00:29:55] Matt Rouse:
play? Right. And with
[00:29:57] AI Narrator #1:
the galactic pest control, here's a sandbox. You wanna play? And we have had a lot of good people say, yeah. Sure.
[00:30:05] Matt Rouse:
And I think that when the book comes out, it's gonna be a very a very good one. Yeah. I'm excited, actually. I'm excited to read everybody else's story and see what their take on the on the same idea was. I I think that's also one of the exciting parts about having, like, a theme with so many different people is everybody has their own. Like, when they get when they hear something, they have their own kind of twist and take on it. And,
[00:30:29] AI Narrator #1:
This is one of the things that makes it difficult for people to squeeze in at the end because they might be too similar to something you've already accepted.
[00:30:39] Matt Rouse:
Right. Well, hopefully, nobody's too similar to mine. Well, like I say, at the end, you have to blow my socks off. On the AI front, I know that we we did talk a bit about, you know, AI and and the potential of job loss and things like that. Do you see a possible future where there is some kind of, like, greater good that comes out of AI rather than just everybody's gonna lose the job and nobody has any money?
[00:31:09] AI Narrator #1:
Well, we we tend to focus on those impacts that so pose problems. But I did a a story that explored some of this. It appeared in Black Cat Weekly a year ago. And, basically, it what it said was that once you start having to face the identity crisis thing that I mentioned Mhmm. You need to find something else to do. So we have a future in which people have battles of the bands in the streets, saxophones against Right. Bagpipes. You know? You could really get into that. Right. Well, if you got time You don't have anything else to do, so why not learn an instrument?
[00:31:57] Matt Rouse:
Well, you know, I
[00:31:58] AI Narrator #1:
And you can also do the whole, what what what is it? The the anachronism people who do jousting and so
[00:32:08] Matt Rouse:
Right. I was gonna say what if what would happen to the world if the starving artists didn't have to be starving? Suddenly, they can actually where you can go. Do their work. Right? And Well, some I mean, I live in an artist community. So Well, some artists
[00:32:23] AI Narrator #1:
would say that's cool. I don't have to worry. I can just concentrate on building this one massive masterwork. And some people manage to do that anyway. I mean, you think about, what is it, Barcelona with that cathedral?
[00:32:36] Matt Rouse:
Mhmm. And then you have a whole lot of other people saying, oh, I don't have to be so goddamn obsessive, and they take a nap. Right. They're like, I don't have to make. I can focus on on what I wanna do for work instead of making 17 pottery cups a day to to put food on the table. You know? Yeah. Where how many paintings do I gotta sell this month to keep my kids in school? My brother was an artist. He died over twenty years ago, but
[00:33:04] AI Narrator #1:
he was a sculptor, painter. Mhmm. His name was Walter Easton. You can look him up. Oh, perhaps the best thing that you can find is, a movie that he made with the head of the art department at Colby College up in Maine, and it was up here. I think that's it. Deep Trout is the name of the trilogy. I have it here on DVD. It it can be found online. And he had the kind of artist ego that said he didn't need to sign his work because everybody should know it was his. Right. That's deep in your mouth. In my mind, a little extreme. I'm not a credit hog.
I'm willing to share, but I do want my name on it. Right. I think you do too. Deep Trout was an obvious, play on words At the time, Deep Throat was in recent memory.
[00:34:12] Matt Rouse:
Right.
[00:34:13] AI Narrator #1:
And he did some very, very interesting sculpture. A lot of it is showcased there, but he never made any money. One problem was that he had some things that could easily have been not exactly mass produced, but at least cranked out with, you know, variations. Small sculptures that might have qualified as motel art. He could have sold a ton of them, but that was beneath him. This is the kind of thing AI specializes in. Oops. I shouldn't say things like that. Right.
[00:34:48] Matt Rouse:
Yeah. I mean, if it's something that is repeatable and has some sort of kind of guiding guideline structure to it that could be, you know, even even with modifications, but still repeatable a while. That's definitely something in the AI, you know, frontier. I think people are gonna really quickly start to look at some of the stuff that's come out really recently. Like, cloud four just came out from Anthropic, and it can work autonomously on things. Honestly, I was I was looking through something today, and it didn't know the answer, like, in its own training data. It it, like, admitted. It was like, well, I don't know the answer to that, so I'm gonna go look it up. And then it went and looked some stuff, and then it told me what it looked up. And then it gave me the sources of where it looked it up from. And, you know, it did all that without me having to ask it. Well, go look it up or go do that. You know? And then there's just, like, v o three just came out from Google, which you could type in a text prompt, and it will make HD video. It keeps the characters, and it makes the audio as well. So it lip syncs to the people's voices and things like that. It is essentially indistinguishable from, you know, well shot video.
I think this this generation of stuff that's out right now is is indistinguishable from human made content for people. And the next generation is better than human content, which better is is, you know, and how do you measure better, right, is is the thing like you can you can tell if it's better at math than it was before, but you can't really tell if it's, like, more of a creative writer than it used to be, You know? So some things are a bit squishy. If we can get the AI to the point where it doesn't hallucinate
[00:36:35] AI Narrator #1:
I'm not sure I like that particular term, but you know what I mean. Right. If if we can get it to the point where it doesn't do that, and if it's doing a search and it comes up dry and it says, I will look that up instead of manufacturing it Then we're pretty much there. That's very useful. Yeah. We're there now. It's just
[00:36:53] Matt Rouse:
only on the cutting edge models.
[00:36:55] AI Narrator #1:
And if we have an AI that says, oh, yes. I can work with your drafting machine to generate a file through your three d printer, that will be pretty cool.
[00:37:05] Matt Rouse:
Yes. Actually, the program called might be cooler than the real thing. There's a program called Mesh. It's an AI three d model creator. Yep. So, I mean, here's the thing. It's it's like there's so much out there now, and it's being developed so quickly by so many different people, and it has so much money behind it and so much momentum behind it that it's gonna get better at a pace that people can't comprehend.
[00:37:31] AI Narrator #1:
And That's what I was saying about wetting the bottom of Lake Michigan. That's right. Exponential growth can fill the lake up in a hurry.
[00:37:39] Matt Rouse:
Yeah. My dad used to have this this thing that when he was teaching, he would say if if the amount of lily pads on the pond doubles every three months
[00:37:50] AI Narrator #1:
Yeah. And Or the the classic chess puzzle from India. How should I pay you? Oh, just put one grain of rice on one square, two grains on the next, four on the next.
[00:38:01] Matt Rouse:
Double it on every square.
[00:38:05] AI Narrator #1:
Bankrupt pick up the raja.
[00:38:07] Matt Rouse:
I used that analogy in my book of will AI take my job. And on the last square, it would be pile of rice the size of Mount Everest if each grain of rice was the equivalent of every grain of sand on the Earth? Every equivalent of what? Every grain of sand on the Earth. Yeah. So it would be so big, it would be bigger than the planet, basically.
[00:38:32] AI Narrator #1:
That's the whole point. Right. Through, the whole thing of exponential growth, this is one of the classic examples. And, it was a sucker punch for the Raja who made that stupid offer, and it has been proposed in various other forms as well, though I think mostly as illustrations rather than actual history. I don't see anybody actually buying into it. And how long would it take the Raja after all to say off with his head? That's right.
[00:39:03] Matt Rouse:
Well, everybody out there, start counting your rice because, current the current model of, of AI is is doubling currently every seven months. Every seven months? Every seven months, it's doubling down. Like Michigan's bottom. That's right. Yep. I was talking to a guy who had a company. I said, how long do you think it's gonna be before AI can do what your company does? And he said, oh, it's at least three generations. I said, well, the next generation's coming out next month, so that means you got you got a year and a half until it could do what your company does. And he's like, ah, shit. And I'm like, this is why we need to think. Right? We need to think about stuff in advance. We gotta go look.
If this got better that quickly at this thing, what would it mean for me and my job or my family or, you know, my business or whatever it is. Right? Well, but this this has all come in my lifetime.
[00:40:03] AI Narrator #1:
I told you earlier before we did this, my AI exposure goes all the way back to the McCullough PIX artificial neuron. That was dated to the nineteen fifties. In the nineteen sixties, there was the perceptron. Mhmm. This is when a mini computer was the size of a refrigerator, the PDP eight, PDP 10. They didn't have microcomputers yet. But the Perceptron was a small neural net, and they could already do some very interesting things, like checking credit card fraud. And it was forgotten because it was so small and so limited. Right. And then twenty years later, people started talking about this hot new thing, neural nets.
Right. I think it was 20. It might have been 30. But it was abandoned for a long time to the point where most people don't even recognize the term perceptron anymore. Do you recall it? I don't. The only reason I know of it is because I did the research for the book of where did kind of AI get started. Right? It's out there. But things have come a long, long way in one man's lifetime. What happens in the next lifetime? It's true. Woah.
[00:41:15] Matt Rouse:
Yeah. Well and you talk about abandoned technologies, like, the LLM was essentially the the trans the pre trained transformer was an abandoned technology. Right? Google invented the transformer. They built it, and they said this is a dead end. There's nothing we could do with this. And then, you know, OpenAI's researchers kind of went and said, look. You know, Ilya Sutzkever was like, I'm gonna pump a ton of data into this thing of this dataset that I have. Everybody was on reinforcement learning. And as soon as the large language model came out and went crazy, everybody started building them. But now they're going back and using the reinforcement learning on top of the large language model,
[00:41:55] AI Narrator #1:
And now they're getting these results that are just incredible, like, you know, multi layered brains and circuit design and synapses. Term small language model yet? I don't know how. Meaning, basically, not so extensive on the training. That's right. It's gonna have to get there. And then we have the whole issue of embodiment.
[00:42:15] Matt Rouse:
Mhmm. Embodied robots and
[00:42:18] AI Narrator #1:
a humanoid AI and raise it like a child. There's been a certain amount of science fiction that has dealt with that. You know what's funny?
[00:42:26] Matt Rouse:
My story, my novel that I'm writing is called AGI Next. I'm I'm currently on chapter 23, so I'm I'm working my way. I don't know how long it's gonna be, though. But, the the idea in my book is that there is an AGI system who also has an embodied robot of itself who is raising a human child. So it's kind of the reverse.
[00:42:52] AI Narrator #1:
Do you think we've given people enough keywords now?
[00:42:57] Matt Rouse:
No. I haven't said edge case more than twice, I think. So that's my my buzzword of the day. I did talk in,
[00:43:04] AI Narrator #1:
Destinies about the impact on education. The you know, as education becomes more vocational and the jobs become fewer, boom, there goes higher education.
[00:43:15] Matt Rouse:
Yeah. Well, I I mean, I know a lot of people who taught. My my uncle and my parents all taught in higher ed. It's quite a thing. I mean, I talk about education in the book, but I mostly talk about, you know, what are you gonna do in the next few years, not what are you gonna do ten years from now. Okay. You know? But say, you know what? I don't wanna cut it too short, but I also I don't want people to have to spend three hours in their car to listen to our whole talk here. So I think maybe we should over your time limit. That's alright. It's okay. It's it's artificial time limit until the AI imposes it on you. Waiting. I'm having fun. But
[00:43:49] AI Narrator #1:
if people expect a half hour or forty five minutes, and here it is, quarter past five almost. Right. We're almost almost an hour in. The plug for the Galactic Pest Control Anthology, and there's the the link for the Kickstarter there in the private chat. It is I'll put it in the I'll put it in the show notes for the show, and we'll share it out. While I'm at it, I should unplug the five year old book, Destiny's Issues to Shape Our Future, which has obviously dealt with some of the things we've been talking about.
[00:44:21] Matt Rouse:
Absolutely. And, Tom, thank you so much for coming on the show. It's so exciting to talk to another science fiction writer, especially somebody who's accomplished as yourself, and I really appreciate it. And, again, it's Tales of Galactic Pest Control, a sci fi anthology. It is on Kickstarter right now. Go back it back real authors doing real work and some of us talking about AI, but not everyone. And,
[00:44:48] AI Narrator #1:
Math will get paid double for a story.
[00:44:53] Matt Rouse:
I'll get paid I'll be getting paid twice of what I put into the Kickstarter. Yeah. It's not about the money, honestly. I I just I love writing, and I was so excited when I got Kermit sent me the acceptance email, you know, saying that we think it's gonna make it. And then I finally did, and and it was great. So I I appreciate you guys. Okay. It's been a pleasure. Thanks, Tom. And, man, we'll have to talk again soon when, when the next book comes out. Where are you located? I'm in Nova Scotia. Nova Scotia.
[00:45:27] AI Narrator #1:
Do you get to cons in The US?
[00:45:29] Matt Rouse:
I have sometimes. The guy used to live in Oregon, but I have not recently. Okay. Because we moved up here at COVID. My next one will be ReaderCon
[00:45:39] AI Narrator #1:
in Burlington, Massachusetts, which is not too far from you That's right. From Nova Scotia. You can actually take the ferry and drive. A few years back, my wife and I took that ferry from Bar Harbor and drove all the way out to Halifax. And it was it was interesting. The idea in Nova Scotia of an overpass is particularly interesting. It's a stop sign.
[00:46:07] Matt Rouse:
That's right. Well, the Bar Harbor ferry goes to Yarmouth, which is about a hour drive from my house. But, yeah, it's I I live kind of in the middle of nowhere, but they call it the the cradle of civilization in in Canada because I live on the oldest road in North America. So I live on Granville Road that's something like 400 years old. And, Well, I have ancestors
[00:46:30] AI Narrator #1:
that went to Nova Scotia as Tories at the time of the revolution.
[00:46:35] Matt Rouse:
Oh, yeah. They probably live in Bear River. After everything settled down, they came back. Right.
[00:46:42] AI Narrator #1:
But That's good to say all the time. They went to Lunenburg.
[00:46:45] Matt Rouse:
Oh, yeah. I know Lunenburg. It's just across the across the province. I gotta drive across, but it only takes an hour. It's it's kinda like a peninsula here. So I appreciate it, and we'll talk again soon, and let's hope, Kickstarter goes well. And then We'll drop to a quick start. And, yeah. And then we'll have to get on to the next one.
[00:47:07] Unknown:
Thank you for listening to the Digital Marketing Masters podcast with your host, Matt Rouse. If you enjoyed the show, please leave us an honest review on your favorite podcast app. Be sure to check out Matt's book, Will AI Take My Job? And have a wonderful day.
Introduction to the Episode
Meet Tom Easton: Science Fiction Writer
AI in Science Fiction and Reality
Challenges and Opportunities of AI Research
Copyright and AI: A Legal Perspective
The Future of Jobs and Universal Basic Income
AI's Impact on Vocational Training and Education
Galactic Pest Control Anthology
AI's Potential for Greater Good
Exponential Growth of AI Technology
Closing Remarks and Future Prospects
