Pod Insights Michael Covel's Trend Following Ep. 1350: Steve Burns with Michael Covel on Trend Following Radio About this episode My guest today is Steve Burns, the founder of New Trader U in 2011. He started investing in 1993, and trading his own accounts in 1995. It was love at first trade. A natural teacher with a gift for cutting through the bull and making complex ideas simple. He offers an extensive blog resource with thousands of original articles, as well as online courses and best-selling books covering a variety of topics. The topic is AI. In this episode of Trend Following Radio we discuss: Rise of AI in content creation and personal branding Content monetization through AI-enhanced platforms Ethical and legal implications of AI usage Accuracy, reliability, and prompt engineering in AI tools Future of trading, jobs, and identity in the AI era Jump in! --- I’m MICHAEL COVEL, the host of TREND FOLLOWING RADIO, and I’m proud to have delivered 10+ million podcast listens since 2012. Investments, economics, psychology, politics, decision-making, human behavior, entrepreneurship and trend following are all passionately explored and debated on my show. To start? I’d like to give you a great piece of advice you can use in your life and trading journey… cut your losses! You will find much more about that philosophy here: https://www.trendfollowing.com/trend/ You can watch a free video here: https://www.trendfollowing.com/video/ Can’t get enough of this episode? You can choose from my thousand plus episodes here: https://www.trendfollowing.com/podcast My social media platforms: Twitter: @covel Facebook: @trendfollowing LinkedIn: @covel Instagram: @mikecovel Hope you enjoy my never-ending podcast conversation!
Transcript 0:00 Speaker 00 My guest today is a repeat guest, Steve Burns, author, trader, social media expert, artificial intelligence expert, which is the subject of our conversation today. I wanted to have an AI conversation and Steve was the first person that came to mind. Whether you don't know anything about AI, or if you just want to listen to two guys who are madly in love with AI, myself and Steve, I hope you enjoy.
0:43 Speaker 00 Even though you and I have got this interesting trading background, and maybe some people don't think it's interesting, whatever, I think by the numbers of book sales for both of us and the interest in our work, I would say, okay, people like what we do. Even though we're still in the same kind of headspace, you're probably a little more of a generalist and I'm a little more specific on one strategy, but hey, it's still all the same stuff. But here's where I wanted to go today because I thought I have since last fall, one assistant in particular was really pushing on AI usage. And I just looked at it as not discounting it, but I was, eh, whatever, what am I going to use it for? And then somebody went to an app, somebody that works for me, went to an app, fed it
1:26 Speaker 00 my podcast, and then she sent me just a 15 second excerpt of AI Mike. And I was like, what the blank? This is, I don't know where it's going. Now, from somebody that likes to produce content, it's great. It really surprised me. I really had no idea, because we're going to be the people that are most probably in tune with our own voice and our own mannerisms and whatever. And so I'm immediately thinking before I even click, ah, this is going to sound like shit. And it didn't. It did not. That surprised me. I assume you've had a similar experience?
2:02 Speaker 01 My wife Holly, that was her rabbit hole she ran into right after her metaverse adventures. She was all in on AI. Years ago, so she's so far ahead and she drug me into it because I'm the same way you were. It's like, what? I didn't get it. I just didn't get it. And you know, like you said, she uses it to create images, to clone my voice, which I'm like you, I thought that's ridiculous. And the minute she cloned my voice and could show me my own voice, which shows the danger of all the, what do you call it? The deep fakes and everything. She's gone on a rabbit hole in every possible facet, from image creation to voicing to comment, videos, audio. It's mind boggling. She's like got 10 business ideas for her own, for herself, for AI usage. She's much more in depth than I am, but I'm like you.
2:46 Speaker 01 I started understanding the power because like, say, for instance, you wanted to write a book about Warren Buffett's shareholder letters, like the biggest lessons. That would take a normal human, what, a year? Two years to go through all the letters and find his wisdom and do. You could create a Warren Buffett bot, train it on the Berkshire Hathaway shareholder letters and just generate a book in a day. If the proofreading and everything would be as hard to do, but from a research standpoint, you can take a two hour podcast, copy and paste the script of the YouTube video into AI and say, give me the five lessons in this video and it'll do it instantly. So I think it's the most powerful tools ever created. And even like perplexity now is like the new Google where it's like 100 times better than Google. It'll search the entire web and every search platform will give you exactly what you want.
3:31 Speaker 01 It's shocking the depth that it can go and the speed of anything you want to do, the speed of the research or the analytics that it can do for you is mind boggling.
3:41 Speaker 00 I think for me, the thing that I'm realizing, and I want to share this as a potential lesson for everyone. You gotta have the ideas. The ideas are going to be valuable. The ideas are essentially a prompt. And for me, for example, if I wanted to take a book of mine and make it a film script. The book, the whole book is essentially a prompt. And this is the thing that I think for people that are trying to understand AI, even if you don't understand it, what it is doing is forcing everyone not to create crappy content. Crappy content will get figured out, but to create good content, to create some kind of a personal brand.
4:26 Speaker 00 I know that sounds trite. People have heard that forever. But if you have a personal brand of some sort, I don't care what the topic is, the content, the subject matter, but a personal brand can, at least I'm going to argue at this moment in time, can be defended. But if you don't have any content, and look, okay, I've written a bunch of trading books. I have a podcast where there's a lot of great traders. And so you could say that's trend following content, but I'm just speaking generally. To find something that you love, something that you want to do and start to produce. Because if you don't, I don't know where you're going to make your money. Okay, professional trading. I want to do trading 24
5:08 Speaker 00 and 7. Okay, that might be tough for most people. You got to pay the bills and stuff, and you got to get a certain amount of wealth first. So it just strikes me that we're at this interesting crossroads where I feel completely contented and I feel excited about how many new things I can create. But philosophically, I worry about the masses.
5:29 Speaker 01 Going back to Karl Marx, the means of production, that was his whole communist manifesto gripe was the capitalists controlled the means of production. They created everything and the masses were succumbed to whatever they offered them. But I believe what AI has done that people aren't clicking, while it's going to destroy millions of entry level jobs and all the mundane stuff, everyone has the ability now and all the platforms are open now. I mean, if you want a radio, you had to have a radio station to produce content. You had to have a TV station to produce television content. But now you can use the means of production of AI to go on YouTube and create your own channel. I think if everybody, like you said, has to have their own piece of the puzzle. Like, it's been democratized now where they're having trouble like the movie studios and TV stations and all the monopolies are being crushed because of YouTube and because of podcasts
6:14 Speaker 01 and because of blogs and websites. And I think the key is you have to have your own, like you said, your own brand, your own production. You need to have your own website. You need to have your own investing methodology. You need to have your own trading system. You need to have your own YouTube channel. You need to have your own social media presence across all platforms. You need to have your own channel. I think the division of value is going to be each person, what they can create out of the AI revolution. When the Industrial Revolution or the Information Age. On a podcast we had a long time ago, I think I said like the future is going to be ever can aggregate the information because the information is overloaded now into the world. So whoever can aggregate it into their own platform and monetize it, that's where the future of income is. It's not going to be in these jobs, working nine to five for these corporations.
6:58 Speaker 01 Right now, all big corporations, their only job is to minimize expenses and maximize profits. And that's what AI is going to do for them. Only small companies had jobs and grow. So I think everybody better take their future into their own hands. Like you said, create their own channel of monetization for themselves and not look for entry level jobs going forward. Cause AI is wiping that out very quickly. It's already triggered hundreds of thousands of job losses already.
7:24 Speaker 00 Let me push back on something. When do you recall was the first moment that you had Steve Burns as a brand on the internet? Where people could say, okay, this guy Steve is involved with trading. Do you remember the month or the year?
7:41 Speaker 01 I think it was 2014. Cause I remember, you know, started writing different trading books and different topics. And I remember I had four books go number one in the best sellers in stock market investing, four in a row. And I was like, that's crazy. I didn't ever thought that like me, myself, but I would have a brand where I could launch a book and they would buy it because I wrote it.
7:59 Speaker 00 10 years.
8:00 Speaker 01 Yeah, 10 years ago.
8:02 Speaker 00 I think for me, probably, I think it's for the first handful of years. I had the Turtle Trader website, the trend following website. But it was more about the content and less about me. It didn't become about me as much until 2004 with my first book. But the reason I bring this up is for two guys that have had their name out there for a long time. It's an interesting advantage. It's an interesting distribution advantage. For example, I don't know the exact numbers and I'm sure there's somebody that I don't know about.
8:38 Speaker 00 But in the trading world, I don't know too many people that have got a head start on you when it comes to distribution. So you've not only creating content, you've got distribution. If I come back to your point about creating content, which I agree with completely, what other choice are you going to have? People are going to have to go that direction to produce some kind of service, some kind of content. But then how do they reach people? Because if everyone is forced to deal with AI cannibalizing so many people so quickly, how does a brand new person with no brand get started? How do they get started? Getting to your point, which is ultimately, you don't have a choice.
9:19 Speaker 00 But philosophically, if everyone's starting from a zero personal brand standpoint, getting off the ground when everybody's trying to get off the ground at the same time, because everybody's had their come to Jesus, there's a lot pitfalls ahead.
9:34 Speaker 01 The great advantage we have now is that the algorithms now rule social media and distribution. You can start, as long as you target on a sub-needs that people are interested in, that has high demand and low supply, the algorithms will share. You can start from zero now. And I've had to start over accounts that were already called, that were jacked and stolen. It was TikTok, and I could launch and get built back up just as quickly, because really it's more based on the content you create, the hashtags you use, and how the algorithm sends it through the network more than somebody following you. Because you could tell, like, YouTubers that have 4 million followers will get a quarter million views, but sometimes 300,000. Now, new YouTubers can launch and get up to tens of thousands very quickly based on the content they're creating.
10:15 Speaker 01 I think we've shifted from a follower, a social media world, to a niche media world, where everyone's looking for what they are interested in. So you have to find your lane of interest that you're interested, passionate about, and what there is an audience for. And the algorithms will match you to people even from zero. I started Threads from scratch, and I'm already at, I think it was 70,000 followers on Threads. I started that, what, a couple of years ago? I think it'd be a decade to get to 70,000 on formerly known as Twitter X. So I'm seeing in my new TikTok I had launched, I'm getting, even though with a little bit of followers, I'm still getting better distribution. So I think the shift now is you just find your lane. I think it's also had regated the podcasting audiences where podcasts may have peaked on big people. Okay, there's more and more podcasts and every time they do it, they fragment the audiences.
11:00 Speaker 01 People find their podcast that picks them the best. If you look at the numbers, especially even YouTube videos, look at the numbers of YouTubers peaked. Six, seven years ago, YouTubers peaked and they're now like they're getting way less than used to get. But all the new YouTubers are getting audiences. So I think that's a big shift now to more of a target algorithm will share you and will use you if you give a good content to share.
11:22 Speaker 00 Have you done any AI video yet?
11:24 Speaker 01 Holly has done some of that and we've tried some of it out. It's interesting, but she's still working on making it really good and engaging. But it's way better than anything I could create manually, but it's a work in progress for sure. So we just now started some of that on some of the social media pages. So it is a work in progress, but it's like that VO, Google VO, you can make the most amazing thing you ever saw. But it only is like for 10 seconds and it's very expensive to make. Have you seen those? Those are all gone viral and wild across social media.
11:53 Speaker 00 I think primarily Korean cheerleaders, not that discloses any of my watching habits, but at least that's pretty tame. But I think that's the entire world, Korean baseball cheerleaders. I've noticed that sometimes I'll see these little reels, these little shorts and then I realize it's AI. But if I look closely, it looks super realistic, but they don't have the backgrounds figured out yet. So you know it's AI, but it's amazingly realistic from the skin tones, the facial expressions. And I think to myself, gosh, if they've got that at that level now, it's only a matter of time. Look, it's almost like you, Steve, have to pick what year of your life do you want to be?
12:37 Speaker 00 The ongoing Steve going forward for the next thousand years. Assuming some of these social media apps are going to be there and successful for.
12:48 Speaker 00 Decades. Why would Steve ever want to pass on? Why would Steve ever pass on? Wouldn't Steve's family just keep Steve going on if Steve wasn't here?
12:58 Speaker 01 Have you seen the new Netflix documentaries that now use AI voice? I think it was Ron Perlman for the Hensink documentary and this new recent one with the one that always had Elvis and the Beatles on his show, Ed Sullivan. They use their AI voices to voice over their voice, what they wrote in their books for the documentaries. So like you said, intellectual property may now be their entire, everything, their voice, their writings, what they say, the documentaries recreate their, even their visually and audio.
13:26 Speaker 00 I hired a very nice guy for, I think, six of my books as audiobooks because there's a certain skill to reading an audiobook and it takes a lot of patience. And okay, if people are used to my voice and used to the way I do things on a podcast, they might not want to hear the book that same way. However, I am most likely releasing my next book as an audio mic. Audio AI mic because it sounds like me. It blows me away. I'm just completely shocked that you can do that. So back to my question, why would anybody in our family let you or I ever go away? Won't there be a podcast?
14:11 Speaker 00 Won't there be a video podcast with Mike and Steve? Who the AI will know how Mike and Steve talk? And Mike and Steve can talk 75 years from now and have a conversation. And how will the average person know that's not Mike and Steve? It will be Mike and Steve AI. There's no way you're going to know that it's AI.
14:31 Speaker 01 There's rumors that that's already happening where people are creating 100% AI visual and audios for podcasts and they're not real. There's already rumors that's going on, but they said it won't diverge who's doing it because it hurts their brand.
14:42 Speaker 00 How many people right now, well-known, are dead and are out there right now, quote, producing content, still participating? And this gets back to my point that I was trying to make earlier. Okay, not every person out there is going to want to create a lot of content, blah, blah, blah. However, even if it's for your own, I hope this doesn't sound too cocky or whatnot, even if it's for your own immortality, producing enough video and enough audio that at least there's enough training data for the AI in the future to create, recreate you. Just seems like everybody's gonna want that. They might not have it, but they're gonna want it.
15:19 Speaker 01 Yeah, sure like our library of books is could be a programming for a future AI. You could have a bot, they could go write new books, they sign everything. That's what's crazy about AI, the analytical skills of AI. Talk to somebody with a thousand IQ where you can put all your stuff in and go, what am I missing here? The AI is only as good as the prompts you use. So the better your prompts are, that it becomes magical, the better your prompts are. If you have an entire library of work you could program into an AI, it can find your missing points and it can even like reiterate based on the things that you're saying that you have been connected in your own books across the platform. That's the amazing thing is the questions you can ask about what am I missing here? What are my fundamental beliefs? That's what I've found is the questioning of my own content is incredible, the insights you can get.
16:01 Speaker 00 This gets into the creating of an AI agent as well. And then whether or not you keep that agent offline or online, or whether it's behind a wall or not. But one of my friends, quite a few steps ahead of me on the AI, has said, look, with all your content, and especially in my world, I've got so much digital content that I've never released, not necessarily all mine, just so many files saved over the eons. It's like a library that has never been released. And a lot of this was print material, this just stuff you're never gonna find again, stuff that's just not out there. But putting all this stuff in kind of a local AI, then you have something that's almost protected.
16:46 Speaker 00 Like you have your protected AI that you can produce content with. And like you said, you can then go ahead and say, okay, in my world, I've got this trend following direction. Then I could pick some subtopic to compare and contrast trend following to. And then that is then reaching out to the AI universe and bringing in content that perhaps is similar, but not thought of as normally being similar to trend following. And the next thing you're producing really interesting stuff that only furthers lessons that I've been trying to share with people for over 20 years.
17:24 Speaker 01 Of the speed, you could write a book with your own. If you have that big of a library of content, you could write a book and put all it into an AI and ask it, what would be the best book to write off this content that would be most interesting to readers and it will tell you. And then you can say, what would be the chapters in that book? And it'll tell you. And then you can say, expand on this and you have it prompted. It will explain this principle using a metaphor or whatever. It's the speed and it's only as good as how you use it. It's not magic. You can't just push a button and write the book, but you can prompt it into brainstorming and into your own, like you're the editor and you're using your own content to write a book. You still have to proofread it and all the other stuff, but the power of that is mind blowing. It'll tell you things you never thought of. It'll connect dots that are never connected.
18:04 Speaker 00 I'll give you an example. When I did my turtle trader book, which has sold pretty well over the years. When I did that book, I had interviews, I had my personal experiences with people, and then I had third-party content, quotations and whatnot. So all this information. And so, yes, I do think that the book, The Complete Turtle Trader, has everything in it. However, when I think about the database that I used to pull the information that made up the 80,000 words in that book, the amount of content that's not in the book, I've got
18:43 Speaker 00 multiple, I don't know how many, 10, interviews of turtles, several of which at least three or four have passed on. There's so much content. And getting back to your point, you could ask yourself what's the best way to do something, or you can ask AI to take a look at what might be the best way to do something. And then I could even think to myself, what happens if I take that turtle trader book and then say, just reimagine a prompt that looks at all that content in an entirely different way? I don't even know what that might be today. It's just really cool. And I guess that's what I wanted to chat with you today was just to get people into this headspace. Cause I don't talk to enough people that really get it.
19:26 Speaker 00 Right now, there's a couple of people that work for me that really get it. I got one good guy friend who's in the AI space and he really gets it. But in terms of people that are like really getting it, I know a lot of smart people. If this topic comes up, they might parrot something, they might, oh yeah, I know that. And they're smart guys, but they're not living it. They're not really seeing what's shifting under their feet.
19:49 Speaker 01 And it's always been that way, too. I remember people like come up in Twitter or now X, come up in conversations and people go, oh, that's nice. You're messing with Twitter. Like it's no big deal. It's the most powerful form of communication ever created in the history of mankind. And they're thinking it's something you share your practice on. Like it's so funny, like every iteration, it's step by step. It seems like they just miss it all every time. It's always miss the internet. I have no greater impact than the fax machine with Paul Krugman and the social media. And they saw social media as toys. I think you're right. Their minds are short circuiting and every time I hear somebody talking about AI on a television, like they do not have any concept of what's going on. They have no understanding of how powerful it is. The most powerful thing of AI you can use is to ask it to write the prompt.
20:32 Speaker 01 You ask it, what prompt should I ask you to tell me the best topic to use a content in a book? And it'll tell you.
20:39 Speaker 00 You've still got to give it some of that Steve original. You could give the AI and say, listen, I've got these ideas revolving around this. Can you help me with a prompt? You could go in cold turkey with nothing and ask for a prompt, but you're still giving it some kind of a prompt, even if you're trying to get to the final prompt.
20:58 Speaker 01 If you give it your information or your content, whatever you want it to do, you give it the information first. And then you say, this is the content. What have I missed in this? Or if I was going to expand this more chapters in this book, what would I write? It's only as good as the information you feed it. That's where the danger of these things are too, is where you let it go run around the internet and get all of the information from the internet because they're all connected now. But the danger of that is it gets false information. If there's fake news shared 10 times, it will think it's real. So, you know, it's not infallible. You put something in and say, only based on this content. Then you can filter it down to your content or whatever you're learning or a source of content, or only use this content. But that is the danger is it will believe falsehoods.
21:35 Speaker 00 Yeah, and I think it's really interesting when it comes to numbers and specific things like, let's say, a Wall Street number or a historical sports number. Many times I have done some type of a prompt, and I think I did something where it was like, "Share with me," just speaking generally here to Grok or something, I think it was Grok, "share with me your perspective, top 10 most successful college football teams for the last 25 years." And number one was Alabama, and I think number two was Clemson. And I was reading each of the little, and it was very interesting, and I was reading their statistics that they were citing for each of these write-ups. And it said Clemson won a national championship in a year that they did it.
22:19 Speaker 00 And I said, "Hey, Clemson did not win that year." "Oh, sorry, you're right." I don't know how long that's going to go on to the, "Oh, sorry, you're right." And then I'm saying, "How the hell were you wrong on something so basic like that?"
22:32 Speaker 01 I see that all the time. All I think about them is each one has its own skill set. Claude's great at writing, Perplexi's great at research, Jack UBT's great at number crunching. What is it? Grok is good at using Hex's data for stuff, but there's so much falsehood in there. I'll say, "Who are the top 10 traders of all time based on risk-adjusted returns over 30-year periods?" I treat very specific on the question. And then if something's not right, I'll take it and copy paste it into Perplexi because it's the new Google search. And I say, "Fact-check this and rewrite it to be accurate." So it'll go through and fact-check it, and Perplexi will give you the right thing. But like you said, it's wronged a lot about a lot of things. It's only as good as the garbage. Reddit can actually mess it up very quickly.
23:09 Speaker 00 That's a really interesting one that you just asked, though, because I personally would not have any confidence, ever, that what you just brought up about the risk-adjusted returns, that I would ever get accurate data. For me, when I first got started, it was like, okay, at least in the CTA trend-following world, if I have the disclosure document, what they put on file with the Commodity Futures Trading Commission, and I can see their audited numbers, then, okay, that's something. But so many traders will release a new fund or something, or they will change the name, or there's not Or somebody like Jim Simons, we're never going to know what he did. There's no one that's allowed to tell you, no one is going to ever know what Simons did.
23:54 Speaker 00 Period. It's not going to ever be knowledge. Not only that, you're not going to really know what his performance was. Maybe they shared it with somebody, but I don't know if they even had to. So it's a really difficult question that one of performance to try to make comparisons. It gets harder and harder, I think that every year that goes by because early on, let's say, pre-internet and pre-AI, traders did have to be more forthright. Today, then if you get online, people that maybe I've half heard of or never have heard of, and they purport to be blah, blah, blah, blah. Then you're just left back to my Clemson example with data. You throw your hands up a little bit.
24:37 Speaker 01 You're still a human filter to everything that you see. Some of your books that have all the performance data in the back, you can input all of your data that you have into the books and then say, quantify this by average annual return and max drawdown. It's only as good as the prompt you use and it's only as good as the data sample you use. And using the internet, and they're all connected to the internet now. So using the internet, it's going to carry all the falsehoods of the internet. Like you said, every day I see some kind of falsehood. So it's not foolproof, but it's learning how to use it in the best way. But you've seen how law firms are using it to write all kinds of case studies on a matter of hours. It used to take a paralegal a week to write. But there is software now, there are AI agents are locked into just contract, just the legal, the law books. They're just locking the law books and all the cases. So they're just in a enclosed system analyzing that data. That's where it does best.
25:22 Speaker 01 When you're just going on a cowboy on the internet, looking at everything, that's where you start having problems. And I bet in the future, they'll choose better data sources. Right now, it's just a wild west of data. That's the problem.
25:34 Speaker 00 It does seem like that one thing we're going to need in this world is verification. For example, how do we know that is Mike and that is Steve? And I don't think that company even exists yet. The company that will provide a service to say that this is actual content. This is actually Steve Byrne's content. This is Steve Byrne's. This is his video. This is his voice. Whether it's AI or not, that's also included. But it comes under Steve Byrne's control. And that doesn't exist yet. Now, a lot of privacy people might be thinking, gosh, that sounds like you're giving up privacy. I think it's about controlling yourself. If
26:19 Speaker 00 we don't have that in the future, how will you be able to stop somebody in Kazakhstan creating you all day long with hour-long videos that sound and look just like you? How will you be able to stop them? If there's not something where the audience, all of us, comes to believe that I will only trust something that has this verification seal? I don't even know what.
26:46 Speaker 01 And they're already doing that on TikTok with fake Joe Rogan advertisements. There's all kinds of fake AI, deep fake advertisements on TikTok using famous people. Yeah, but US-based, but once again, most of that's China. The main company's based in China, so how do you even begin to combat that? That's what Joe Rogan talks about. It happens all the time. That's another Wild West where you have to trust the sources. Like, you've got to follow their channel with their blue checkmark and only trust their sources. But like you said, once that cat's out of the bag, like right now, you have Disney, was it Disney suing one of the image creators and videos creators for copyright infringement? There are lawsuits going on because they're generating videos and images of copyrighted content. So when would that expand to individuals?
27:24 Speaker 00 What happens if some AI is just scanning one of my books and training on it? Okay, anybody can summarize any document. That's always been the way that it is. But AI summarizing, is that the same thing? It gets tricky fast.
27:39 Speaker 01 How you profit from somebody else's copyrighted material, copyright law and trademark law? How is that all going to play out in the AI world where they're not going to be allowed to use trademark and copyrighted content? That's probably the legal future of AI is going to be fascinating, too.
27:53 Speaker 00 Music on YouTube, they enforce that like crazy. I think on music, they will figure that out quick, but on writing so far, not much protections there. Let me throw another one at you. How far away are we from this podcast right now, me and you, we're talking on Zoom, there's video? How long, how far away are we from one button pressed after we're done and every language in our voice, you and I are having this conversation? So it immediately gets created into every language, me and you, our voice, we both become experts in speaking in Vietnamese and Chinese and French and you name it. How far away are we?
28:38 Speaker 01 YouTube has been working on that for a couple of years. I saw a YouTube guy on a podcast saying that was YouTube's vision is to, when you post a YouTube video, they can instantly translate into all of the languages and post it to all of the different Youtubes around the world. That was what the Google alphabet is working on for at least two years. So I think it's very close because Mr. Beast opened that up. Mr. Beast had the voice people do that and that's how he exponentially grew his channel.
29:05 Speaker 00 What's interesting about that too is it from a standpoint of back to the beginning of this conversation, a personal brand standpoint, do we really want the big Google controlling all these things? Or do we want to control these things? Or will we have the option, meaning it will become something where it's just advantageous to do it towards Google? And again, like you just pointed out, will you and I even have to worry about it? Will it just be a built-in service to where I host my podcast?
29:35 Speaker 01 Yeah, that's Alphabet's plan, which I guess it makes sense. The size of the population growth in the world is to that's how they can exponentially grow. And I guess once you do your YouTube video, you post it, you're in their terms of service and they can do what they would like, as they have always done. That's just why the YouTube has overtaken network, television, cable, even Netflix. Like YouTube is now the biggest broadcasting platform in the world for video. That's just crazy.
29:59 Speaker 00 It's the most interesting thing. If you think about it, and maybe a lot of people will resonate with what I'm about to say, but during COVID time, I think a lot of people saw a lot of content, whether it was movies or TV shows. And then everyone probably started to realize like I did, which is frankly, there's not a lot of good stuff out there. Going back 75 years, there's only a limited number of movies that are ever worth watching again. There are several TV shows that I could probably binge watch every couple of years, Breaking Bad, Sopranos, stuff like that. But relatively speaking, there's not a lot. But when it comes to YouTube, you can flip on YouTube and there's always new content.
30:36 Speaker 01 It's always interesting and you can drill down to exactly what you are interested in. And YouTube can show, the algorithm will show you things that you didn't know you were interested in related to what you were interested in. The algorithm is just amazing. Like there is no algorithm on television or cable. That's the power even. And Netflix does not have the best library to think YouTube, all the care is working for them to create all that content for free versus Netflix paying a billion for a noob junkie movie.
31:03 Speaker 00 Let me phrase it to you this way, though, in terms of, let me take it away from the podcast in languages. What about all of your books? And you know how this works. If you write a book, especially with publishers, eventually they'll come to you and they'll say, hey, we sold your rights to Japan, Korea, et cetera. Then you wait a couple more years, the book finally comes out. Hey, I think it's really interesting. My books are in like 16 languages. But with AI, what's interesting about that? With AI, how far away are you from, Steve, all of your books being available in all languages? And differently than, for example, the podcast stuff, books do get monetized at each language.
31:48 Speaker 01 I think there's already ones that my wife thinks can do it, but I would still think you need a proofreader and somebody who has a language. But she thinks there already are ones good enough where you just put it in there, it translates it, you're done. So in a couple of days, you could translate in 20 languages.
32:01 Speaker 00 Well, you know it's coming. That's what I'm getting at. It's like, to me, it feels, if I ask that kind of a big picture question, my gut says, there's no way it's longer than five years. No way.
32:12 Speaker 01 No, there's already services that say they provide that right now.
32:16 Speaker 00 So then the only issue is, can we do it now and get 80% accuracy or 99% accuracy? But I think the 100% accuracy, there's no way that's five years out. It's going to be a given. That's just unbelievable to potentially release a book and as an author, have it in every language. The reach is crazy.
32:37 Speaker 01 India alone has a billion people in India and half of them are Forex traders. So but think about the people the job is going to cost. If AI can translate, how many jobs of translators that will cost when AI, you push the button, it's translated. But we'll still need proofreaders, but that'll wipe out the translators, the virtual voices where they can clone your voice and use it to make an audiobook of your book version. You're getting rid of voiceover for books. The amount of job destruction here is mind-boggling. That's just two entire fields that are going to be gone within, like you said, a year or two, if not already.
33:08 Speaker 00 You said proofreading. Why won't it be that you can have an AI proofreader that maybe doesn't get everything perfect, but can at least point you to potential problem areas that a human could then check? The need for proofreaders will go away.
33:24 Speaker 01 Yeah, you're right. You could have an AI, you could have an AI agent that translated it. Or even the same AI agent could go back and proofread it in its language and give you all the possible things that are wrong, or even be, like you said, in two years, good enough to change it all. You don't need a proofreader or a translator.
33:39 Speaker 00 I have two people working for me that are, English as a second language, and they help to produce content in English, and it's perfect. I don't think English speakers, especially Americans, being a little bit cocky, what percentage of Americans understand that non-English speakers can produce content that English speakers don't know is AI? That's wild.
34:06 Speaker 01 Every day I see something like that. I can't tell if real influencers are AI or real people because they're using the strange camera filters. Where I'm thinking, is that AI? Is that even real?
34:15 Speaker 00 That's my point to the verification thing. It sure seems like there's an opportunity, if there's a buy-in, if people want to buy in, you might get to the point where it's like, maybe that's several years away, where the masses all of a sudden understand, oh my gosh, the opportunity in this AI world to be verified is quite valuable, that people actually know, it's me.
34:38 Speaker 01 You know, what, even a platform that doesn't allow any AI, a new social media platform that's all real.
34:44 Speaker 00 And then what happens if some social media platform says that content can't be posted because it's AI, but it's not AI? And then you have fights about people saying, what's real and what's not real? What's real? What's memorac?
34:58 Speaker 01 Oh, that's already going on, where I had people accuse my blog posts of being AI that I wrote years ago. Like, look at my old one. The date is in the Earl. I wrote that seven years ago. This is bizarre, because I sound like ChatGPT. I have that droning ChatGPT sound, and it's to the facts. And so I have been accused of blog posts that I wrote years ago. That's just funny. That didn't even exist back then. So that's what people are starting to question reality.
35:23 Speaker 00 For the audience listening, I cannot verify that Steve is actually Steve. I see him on the screen, but I have no idea if he's real anymore. I don't know. He doesn't know if I'm real anymore. Maybe people have not been aware or sure if I'm real for a long time, since I've been on the move outside of America. Hey, let me take it back to something you said earlier, perplexity, and Claude, and ChatGPT, and X. Those four, let me let you Give a little bit of insight from your perspective on using them, pros and cons. Let me let you start with Claude.
36:00 Speaker 01 Claude is the best writer. If you wanted to write, or if you prompted to write something like a social media post, you can say, "Write me a post about this," and then say, "Okay, clarify and make it better." Claude sounds the most like human writing. It's a writer. It sounds really good for a writing perspective. But the problem is Claude hallucinates a lot. It'll make up just crazy stuff. Like you said, it was like, no, that's not hacker at all. And then you'll say, and sometimes it'll crack itself, but sometimes there's a glitch where it will not. So you have to fact check Claude off of perplexity. But Claude's one of my favorites. He's just very clean, well spoken, well written. It's good for idea generation and for writing.
36:35 Speaker 00 But you don't see those benefits across perplexity, Grok, or ChatGPT? Claude is far and away above better?
36:44 Speaker 01 In my experience, absolutely. It's a good writer, but it is not the most accurate factually. And it's creative, like a human perspective, but it doesn't do the best deep research.
36:55 Speaker 00 Give me an example, or just explain in general perplexity on deep research. What you're seeing in perplexity that makes you believe in it more as a research platform?
37:07 Speaker 01 It shows you all the sources it draws from cleanly. You ask it a question, and it'll give you the answers, and it'll give you all the sources from it. Like I had some medical issues a year ago, and I had perplexity. I use it like a medical encyclopedia. And I said, "How does a medical encyclopedia and what do these symptoms lead to?" Because I was having no luck with the doctors in the emergency room. And it would cite the actual medical journals itself, and it would go through step by step, and I could cross-reference to the actual sources. So it was a good research tool. It's also great for just saying rewrite something and make this accurate. It'll do it, and it'll be accurate. Perplexity is rarely ever wrong unless it draws from a false source, but its accuracy rate's probably 90, 95, 96% in my experience.
37:50 Speaker 01 Maybe 98 on some topics that are simple and to the point, while Claude might be 80, 80, 80%, 85% accurate on Claude, just in my experience. So perplexity is like Google on steroids.
38:01 Speaker 00 I think we're subscribed to Claude. And Brock, I sometimes use chat GPT through the Microsoft browser just for a quick and dirty check on something. But I've noticed, and I'm pretty sure that's chat GPT running through the Microsoft browser. What I've noticed, though, with that, which infuriates me, is it's so woke. If you say one wrong word, it gives you a lecture, or it says, I will not talk about that.
38:30 Speaker 01 It's trained. This mental model is whatever the term with a woke-ism was. That is its mental model. Brock is not that way, for the most part.
38:39 Speaker 00 I'll give you an example on chat GPT. Also, I was doing some research on President Lyndon Johnson in the 1960s. And I was asking to see the most egregious things that LBJ said about certain topics, right? Controversial topics. For example, race, religion, sex. Everyone knows LBJ said all kinds of crazy stuff. And chat GPT said, that's too controversial. I'm paraphrasing. That's too controversial. I cannot discuss that. That actually shocked me because this was like the President of the United States, known things that he said, and we've got a situation where a major AI
39:24 Speaker 00 doesn't even want to participate on the basics of the American First Amendment, or the basics of just known documentation, or known things said. And it has been told you cannot play in that sandbox.
39:39 Speaker 01 Yeah, it's literally trained to the class of like ultra sensitive topics. So it's trained that way to avoid anything that it conceives would be the ultra sensitive topics, or even words, the way it'll use different words, the politically correct words instead of the actual original words.
39:54 Speaker 00 How can anybody, though, if you're trying to use an AI, and you and I have already given examples today how this can happen, but if you're trying to use an AI to get to some truth or to get to some understanding, and then that same AI all of a sudden censors itself, it seems like it's a really bad business model to censor your AI because ultimately, who's going to want to use an AI that you can't be for sure it's giving you accurate information? Look, if it messes up and you and I can catch it, fine, but when an AI just says, "Sorry, I can't play there", that strikes me as a really bad business model.
40:31 Speaker 01 That's the edge that Grok has, because that's what Elon's trying to make Grok, where it's just open and honest, which also has its problems at times, where people go down bad roads with it. That's nobody's business, but the people doing it. Like, how they're trained is how they act.
40:44 Speaker 00 What do you think about Grok?
40:45 Speaker 01 Grok is the one, like you said, the best to go down those rabbit holes. Like, it will take you down rabbit holes that the other AIs will not. You know, be more into reality and gritty than other ones will not. It's trained more on HX data, so it's about as smart as like the combined intelligence of HX as a platform. And it does really good at breaking stuff down. If you ask it questions, it not only gives you the answer, but it gives you the reasons for the answer. Like, it'll break down. It does a really good job of that. It's really good for a non-biased, non-censored, for the most part, analytical tool for understanding. It does a really good job of that. I'll put in the same things with Perplexity, Claude, Grok, and ChatGPT, get different answers. I find that fascinating.
41:27 Speaker 01 And Grok usually has, can be the most different of the other, of all four of them.
41:31 Speaker 00 From a trend following perspective, it's pretty easy to see how these tools can be used. Backtesting, run your system, okay. What I think is also interesting is that there's going to be a lot of data mining. A lot of young people are going to say to themselves, "How can I use all this information to data mine?" And they won't go to the basics that have been around for centuries. "Hey, price is king, let your profits run, cut your losses short." That will probably seem too simplistic. But I can see a lot of young people, perhaps, or even older people, a lot of people, going ahead and saying, "I can use all of this fundamental information to make predictions."
42:15 Speaker 00 This is going to be the biggest use of fundamental information ever.
42:21 Speaker 00 But I don't think, conceptually or philosophically, that's ever going to beat price. Price is going to be king. But I'm sure you can see that people are going to just fall in love with these AIs as a predictive tool.
42:35 Speaker 01 I've already seen it. There's all kinds of, shared on social media where people are using it to trade, analyze this data. What's the best buy point? What's the best sell point? And they're being successful. But what I see with them doing this and sharing it publicly, whether it's true or not, is the cognitive biases. They're already falling into cognitive biases because it worked one time or two times. It's going to work every time. And they're not understanding like, did you win because of what it did? Or did you win because we're in a raging bull market? Did you win because it was a V bottom? Because Trump backed off the tariffs? Did you actually have a trend following system where you bought the high and let the trail to stop and let it run? Which I find it fascinating is it does a good job. Most of the AIs will train you on the profitable principles of trading. If you put it in there, it will explain in the great depths. It'll give you, teach me about risk port ratios. Teach me about a positive expectancy model.
43:20 Speaker 01 It will train you on that. I think that's a better use of it than, or designing a trading system based on backtested data. It can do that. I think those are better paths to go down than trying to get it to make decisions. Probably the best use of it, people want to use it for a fundamental standpoint. Probably the best use would be a Warren Buffett bot where you trade it on the Benjamin Graham Dodd fundamentals. That's the other thing proven to work, which they still have to have the patience to understand. They're probably still going to lose 50 to 60% of the time even on that. But the huge winners will pay for it. So I think that would be the best use is to do proven models. And you could go down, you could have it do a Warren Buffett program it with just his newsletters and his books that he loved and use that. Let it do it, if anything.
44:02 Speaker 00 It's a wild time. I'm sure after witnessing what happened this spring with Trump, I'm sure many enterprising individuals said to themselves, let me create a trading system based off Trump talking. I don't think that's such a good idea, but I can see how the power of AI would get people looking to data mine, to pull together signals that like you said, hey, there might be a one-off where it Works, but it's not a full system. You're just data mining. It's going to be the mothership of all searching for what people think is some type of expectation, but really isn't.
44:42 Speaker 01 If we knew what Jim Simmons used for his Renaissance technologies, that would be a great models to put in there and figure out his principles of how he did what he did, but not just a person trying to make up stuff off the top of their head.
44:53 Speaker 00 I think the most interesting thing I know about him from that one book about him, I think Zuckerman wrote it, was that he got started or really got into it by seeing what Richard Dennis was doing. And that his early years were trend following years. And then he morphed to something bigger. I'm sure trend following was always an alpha component for him, a source of profit. But he probably, with technology and all that brain power, he probably was, there's a trend following component, there's a high frequency component. There's every component under the sun. And so for me, I never bothered to think to myself, let me know exactly what he's doing because A, we're never going to know.
45:36 Speaker 00 And B, if we did know, we couldn't execute everything he was doing because it's infinite strategies, most likely.
45:43 Speaker 01 From everything I've read of him is it was math. He had a mathematical model, but what I can understand, what he said and spoke, it was like a risk reward ratios, repeatable mathematical model, what Edward Thorpe did where he saw the entire market through the filter of math, which comes out to positive expectancy, bet size, small risk, letting winners run. That's what I see. I think that Jim Simmons broke the mathematical code, which is the same thing trend followers did, for the most part, just different timeframes for him.
46:10 Speaker 00 I have a feeling that he did it in a way with, like you said, timeframes and strategies and markets on a scale that no one could compete with because who has that, you know, how big was his trading floor? How many traders? How much technology? What is he producing? 50% a year after fees for 30 years? There's only one person doing this. We could all theorize about what he was possibly doing, but I don't know. When we finish this call, we can get off and see if AI can give us a decent answer about Renaissance technologies and what the hell they were doing. But I don't think we're going to get very far.
46:46 Speaker 01 Nah, he kept that very secret, which what you would expect from the closest to a holy grill anybody found in trading.
46:51 Speaker 00 Hey Steve, is there anything else about AI that I didn't bring up today? I mean, we could talk about all kinds of stuff, the philosophical negatives, the philosophical positives. We've talked about a lot of the positives today, some of the downsides, potential job losses. There's other downsides that I can't even imagine. I think privacy is going to be a beast because eventually all these cameras are going to catch us at every location where we are. Will governments be the only people that have access to the camera data? Because if governments are not the only people with access to the camera data, then our potential privacy loss is pretty huge. We would be tracked 24 and 7, essentially.
47:33 Speaker 01 We already are between phones and debit cards.
47:36 Speaker 00 Steve, I just wanted to catch up today. It's always good chatting. I didn't have any notes here or anything, but it's a fun conversation because I just keep getting farther and farther into it and I'm just shocked. It's exciting as hell to me and I don't know where everyone is that's listening, whether they've had their lightbulb moment yet, but all I can tell you, if you haven't had your lightbulb moment yet, have it and then you can make sure that everyone else, friends and family, have it because there's no other game in town.
48:03 Speaker 01 I'm finished saying that this is the greatest technology ever created and the winners, whatever platform wins, it could create the next trillionaire, whoever wins this one. So I would highly advise whether somebody wants to or not to start getting into it and playing with this because this is far bigger than social media or even the internet was. This is going to change everything. It'd be interesting to watch this interview 10 years from now and how the entire world has utterly changed.
48:28 Speaker 00 Yeah, in ways that you and I are not even theorizing probably today, we're probably not even pondering the way things are going to be changed. In fact, I paraphrase from Mark Cuban and I have mixed feelings about Mark Cuban, but he did say that we're reaching the point in time where the first trillionaire very well likely could be a dude in his basement in his underwear.
48:51 Speaker 01 I like that quote. If Elon doesn't beat him to it.
48:54 Speaker 00 With AI, if you catch that exponential rise, even Elon, who's been on an exponential rise, could be caught potentially. It's an amazing time. Steve, great stuff. I'm sure you've got a million new books coming. People can find you everywhere and I've said this before on the podcast with you. If they can't find the real Steve Burns associated with trading, Wall Street, etc., maybe your IQ is just not high enough. Maybe you're down around 50 and you have to be able to figure it out. But for Steve, for the people that might want to come check you out at one of your social media presences, and they are everywhere, where's the best place to go, Steve?
49:33 Speaker 01 I would say this X, @sJosephBurns, and I have a blue checkmark, and I think it's 666,000 followers on X to know it's the real one because there's plenty of imposters, but no blue checkmarks.
49:46 Speaker 00 And I do follow Steve on my X account. So if for some reason or another you are mentally challenged and you can't find Steve, without that, you can go to my X account, look at my followers, who I'm following, and I do follow only one Steve Burns, and it's the real one. I think it's the real one. I really don't know. I have no idea if the guy I talked to tonight is even real. He could be fake for all I know.
50:07 Speaker 01 I'm just a hologram.
50:08 Speaker 00 Steve, good stuff, man.
50:10 Speaker 01 Good time to you as always.
50:18 Speaker 02 I see a time when those awake will understand how to make money in up, down, and surprise markets. Whether a new trader or experienced, college student or financial advisor, protecting against a crash, or just trying to make a lot of money, trend following offers everyone an answer in uncertain times. To get started immediately, send me an email, [email protected] . I will send you the right trend following steps to take, along with my free video. But if you want to buy and hold, trust the government and trust Wall Street, this is absolutely not for you.
Previous Ep. 1349: Practice Makes Perfect with Michael Covel on Trend Following Radio Next Ep. 1351: Henrietta Moore and Arthur Kay with Michael Covel on Trend Following Radio Page updated November 7, 2026 · transcript and summary are machine-generated
