← The AI Podcast
The AI Podcast
Exploring Nvidia's Stand on Export Controls: Turning export controls into an opportunity
00:30:48
Loading summary
Transcript41 lines
- [00:00]
A
Today on the podcast we have the pleasure of being joined by the legendary Rajeev Kapoor. He's the best selling author and an AI strategist. He's the CEO of 1105 Media. He has about 10 years experience in leadership spanning between a whole bunch of different Fortune 500s. His latest book is called Prompting Made Simple. We're going to talk a little bit about that. And basically has a whole bunch of really interesting frameworks for helping professionals get smart, smarter, faster results from things like ChatGPT that we use all the time. He's a finalist for the Ernst and Young Entrepreneurship of the Year award and he is on a mission to make AI a little bit more accessible, actionable and results driven for business leaders and creative people. Like all of you listening. So welcome so much welcome today to the show, man.
- [00:45]
B
I appreciate it. Jaden, thanks for having me on board.
- [00:47]
A
Really excited to get you on board. Before we get everything kicked off, I have so many questions for you that I think the audience would love to hear about. Basically tips for becoming better at prompting and getting more out of AI models. Before we get into that, all of that, I'd love to know like, I guess a little bit about your background, what got you started with this, started writing your book, I guess a little bit about your process. Give us a bit about your background.
- [01:08]
B
Yeah, so you know, I was a tech executive for a long time. I worked, I worked at Dell. I was an executive at Dell for a long time. I worked directly for Michael back in the late 90s, mid to late 90s. It was a great experience and just kind of worked my way up at Dell. Eventually ran kind of part of the western region for Dell and then went to China and Hong Kong for Dell in 2000 and Singapore and helped build out the South Asia market and then was recruited to be the president of an E commerce company out here in Southern California in Orange county called Smart Home. We did that for a few years and then that actually led me to my next gig, which is I actually became then CEO of an audio technology company headquartered out of Switzerland of all places. Okay, amazing company started by a couple of amazing guys and they actually were building AI audio algorithms back then. So this was like 14 years ago. And so a lot of people don't realize that there's two kinds of AI. So there's the ChatGPT, generative AI stuff that, you know, prompting is basically, basically all about. And then you have machine learning AI. And so these guys were very much in the machine learning space all around audio technology. And so I was CEO of that company. That company eventually sold to Sennheiser. I was really enamored with AI. I wanted to learn more about AI, so I went to MIT and I got dual certifications at MIT on AI. And then when I was asked to be the CEO of 1105, almost 11 years now, basically, we were doing a lot. So we're a B2B marketing and media company, meaning we do a lot of stuff in the technology space, but all B2B. So our job is like buyers with sellers. So if you're a brand in the technology space and you want to find more buyers of your product, you come to us and we help you. So that's the basic way to put it. We do events, we do Legion, we do all that fun stuff. Right. But the areas we cover were data center, was cloud infrastructure, was cybersecurity, was big data, was business intelligence. So we were covering that already. And we started covering AI a long time ago, again, on the machine learning side, when ChatGPT came out, we obviously added that to our repertoire. But I remember the day ChatGPT came out, I literally fell out of my bed going, what the heck is this? And I immediately said, hey, everybody on the planet's gonna have to learn AI. It's gonna be really important. And at the time, nobody knew just how impactful it was gonna be. I knew it was gonna be pretty big, not as big as it had become and. But it's crazy to see where the future holds. But again, the science behind this really is going to all end up for the general consumer is going to be all around prompting. Right? I mean, we don't need to know how the Sacha is made. ChatGPT 5 will come out. We don't care the mechanics behind it, but what we are going to care about is how to get the most value out of it. And the way to get the most value of that is going to be things like prompting. So that's where we are.
- [03:43]
A
Okay. Amazing. And you've written this phenomenal book. I want to get into it, but I guess getting kicked off and asking you a little bit about how you're approaching prompting and some of your frameworks and what you're using, I guess for you, when you're actually going and developing a prompt, how do you actually work on refining it? Like, let's say you ask ChatGPT or any model, like, question, it gives you a response, that's okay, but it's not amazing. I feel like half the time I'm like, I feel discouraged. I'm like, oh my gosh, I just, I'll just go do this from scratch. Like, how would you work on refining that prompt? I think for a lot of business professionals, they don't want to just go start from scratch because they're, you know, maybe it's a prompt that's going to help them do something on a more repetitive basis, I guess. How would you look at that?
- [04:22]
B
Well, look, first of all, the good News is that ChatGPT has a memory function, right? So whatever you start doing, whatever you start prompting, whatever you've done in the past, if you're a business professional and you want to come back to that same thing, it's going to remember. So, so, so just remember that, right? But in terms of how I refine my prompts, what I start with is I always try to look at what I'm trying to accomplish. Well, what is that goal that I want to try to accomplish? And I actually kind of write that down on a piece of paper. I write that down. And so I kind of start with what is my goal for this prompt? And just kind of write down like a little quick sentence. What is my ultimate goal? I then go into the prompt window and I start thinking about, okay, do I want to be a Persona based prompt? So, for example, and I literally just did this today. So I have one of my business units is struggling with one of their events that they're putting on. So I asked ChatGPT to be, I asked to be an amazing event coordinator, an event planner, you know, that has a strong technological background, who puts on great events, you know, and imagine Steve Jobs is going to be the person on stage speaking, right? So I immediately gave it a Persona. I, you know, I gave it a multiple Persona. It, I asked it to kind of also think of it from a Steve Jobs perspective. And so that's my starting point, right? Where I just really want to clarify what I want that Persona to help me with. So I do a lot of Persona type based prompting. I'll do a lot of chain prompting as well. We'll talk about that in a second. But I start with that Persona kind of establishing kind of what type of support I want. Brainstorming, partner, consultant, you know, a data analyst, help me maybe analyze a set of data. So I asked, I asked you to actually take on a specific role, right? So once I have that Persona in place, then I, then I go out and I proceed to tell it the task. What is the task I wanted to accomplish? Like, well, what tasks do I Need help with, I need help with this. Here's the task at hand and I'll even use the word task. Here's the task we want to do. From there I then give a context. So what is that context that I want to provide? You know we did this event last year, it did really well, but this is the time of year we did it. Now we're doing it this time of year. Last year we had this speaker, this year we don't have that speaker or whatever context I want to provide. Right. So for example, like it doesn't have to be something around an event. It could be like if you are just doing something personally and you're looking to do a vacation plan, you could say hey, take on the role of an event planner. The task is to create a 10 day itinerary. Context is we want to go to Africa, blah blah, oh by the way, my wife is allergic to pineapple.
- [06:50]
A
Right?
- [06:51]
B
So that's context. You want to put context around the prompt and then you make the final ask, give me a bullet point framework, give me an outcome, give me a paragraph, give me a summary, give me a, give me a brainstorming session. And then, and then you know, in this case I gave, I gave it the role, the task, the contact, the ask. But in the ask I said hey, and if you need more information from me to clarify, ask me questions. And a lot of people don't do that last part. And that last part is huge. When you start realizing that you can get ChatGPT or Gemini, whatever, Grok, whatever you want to use to start asking you questions, that becomes a game changer because now you feel like you're having a full on conversation and that's when the, that's when you take that role task contact as type prompting and you combine it with chain prompting. So now ChatGPT is asking you a question, you're answering that, you're answering the questions that it might have. It might cause you to think for a few minutes, you might have to go back and pull up some data, some information, give us a chatgpt and it'll give you an answer and then you can just go back and forth with it a little bit. But once you do that, you're going to start to see the clearer you are, the more specific you can get, you know, the better your responses are going to be. And then once you, once I have that, it might give me, it might give me an A, B, C, D, E outline of whatever it is. Uhuh, the output was, I might say okay, Great. I love it. I want to go deeper and let's peel back the onion on a. Go deeper on a one. Da, da, da. Right. So then I just started going deeper, deeper, deeper, deeper, deeper. And you just keep going back, deeper, deeper, deeper. So that's really the goal. That's really how you prompt from a business perspective. And that's really the. And it's. And again, it's, you know, I have this talk where it's called A Taste of AI. Jaden and I talk about the six key takeaways for AI success. And one of the six is storytelling. So the more you, the more you become a storyteller, the more you're using imagination in the prompt. So the more you're defining what you want that prompt to be, the way, way better your results going to be.
- [08:39]
A
Okay, I love that and I love this concept of chaining prompts together. I'm currently building a software that basically lets you link together different AI models and you put your prompts in there. And we find we're getting way better results doing that than just having like one huge document that's a prompt. But I guess for people that are doing that from your perspective, what are you seeing in, you know, someone that would say, like, you know, I don't want to ask it questions, I don't want to have it go back and forth. I just want it to give me the thing one time, like, what are you seeing in how. How good a prompt or an output can be if you just give it like a huge, you know, blurb of text and instructions versus if you have kind of a conversation back and forth.
- [09:18]
B
And evolve the, you know, that's a good question. So I've seen both work, you know, I've seen like these massively look, the average prompt from what I've seen and from my studies and talking to others, you know, the average prompt for you to get a really good detailed response for what you're probably looking for is probably in the 300 to 400 word range. You know, it's not like just two sentences, right? Yeah, A tell me about Jaden or whatever. Right. Or, you know what I mean? It's got search capability, but it's really not a search engine, it's an answer engine. So it's really more about the more information you get it, the better your response going to be. So if you can figure out how to get to that level, it doesn't have to be 400 words. Like I said, that's the average. Some people put 800,000 words, whatever it might be, again, the more you put in there, again, the better you're going to be. Now, I also understand that I've seen it where folks, and I've done it myself, where you might say, hey, you know what? I'm thinking about opening an office in London. I don't need a big long prompt for that, but I could just say, hey, got HR GPT. I'm thinking about opening an office in London. What do you think I should consider? And that's, that might just be the prompt. And it's just like if I was just talking to somebody on the phone, so then it's just me naturally just having a conversation. Sometimes I'll just do that through the app because it's multimodal and I'll just have a conversation with ChatGPT. So, you know, so do this, this and this. Oh, that's a really good idea. What about this? And what about this idea? What about this? Do I need to have a bank account in London or whatever? Right? I mean, so I just go back and forth with it and I'll say, well, what are the pitfalls? What are the opportunities? What are the concerns? What are the challenges? So I just, I like having that conversation because sometimes I can get my flowing. I could say, I'll ask you, I'll even say the chat GPT and that chain prompting. I'll say, well, what do you think? It'll be interesting, the response I get when I say, what do you think? Right. And so for me, I think that's kind of a win. And look, here's the little life hack when it comes to prompting. If you get stuck prompting, Ask ChatGPT to help you write the prompt.
- [11:11]
A
I love that.
- [11:13]
B
So again, it's like having a conversation. It's no different than saying, you know, going to a peer, a boss or friend saying, hey, I really need help with this. What do you think? And have them help you. Right? So it's the same thing. If you get stuck with a prompt, if you don't think, if you don't think you're getting the right. If it's not, if it's not being cohesive enough or you're not feeling like you're getting to the right area with the prompt, then just stop, say ChatGPT, I think we're on the wrong page. I'm really struggling with this. Maybe I'm not explaining the prompt properly. How would you, can you help guide me how to write this prompt better for you? Right. And then go down that path and then Say, oh, okay, great. With that being said, do you mind writing the prompt for me chatgpt and let me see. And then I can modify it based on what you give me. Right. So that's, so that, that's another way to do it. And with things like Sora and these other things like Operator and all these other things that are coming out in VO2, that's a good way to start also playing with those.
- [12:03]
A
Okay.
- [12:03]
B
Getting the, is getting the AI to help you.
- [12:05]
A
Okay, I love that. I think that's a fantastic advice. One thing I would love to ask you about is so you've talked a lot about like having these conversations or being specific in your experience. Maybe for some people, they're, you know, they, they just don't feel like they can, they can get a great output on a specific thing because they don't feel like maybe they're like an expert on that area. I, I think when people are making prompts, where do you think people are kind of over engineering their prompts and where do you think they're being too vague on their prompts? Like what are thing ways? Maybe people are kind of overthinking how much information they need to add and what are places where maybe they need to add more information or be more specific?
- [12:39]
B
Yeah, I don't, I think if you go back to that role task context, ask, portion, I don't think people are putting enough context behind the prompt. You got to be able to give it the context of what, what you're trying to do. So if you're doing a brainstorming session for new, for, for 2026 business planning, but you're not asking for new revenue ideas and you're not giving it the context of what happened before, the concerns with issues or the challenges. Or hey, maybe you're in the hospitality industry and it does understand that tourism is down or whatever. Right. If you're not giving it the right context, it's going to be a struggle. Okay, so that's kind of what. So make sure that whenever you are doing your prompting, make sure it does have the right context to understand the position of where you are, what you're trying to do. Especially from a business perspective. Because ChatGPT doesn't know unless you've updated your business information, unless you've uploaded all of your financials or whatever, your history, whatever. Unless you've done all that stuff, it's not going to know. Right. So you have to give it the context of what's happening now. You can also give it links to articles about your industry. So it can understand that. Right. You can say, here's an article on my industry about tourism being down. Take that into account when giving me your answer. Whatever. Right. You know what I mean? So it's all about that context. So the deeper you can go in that context, the better off your response is going to be. And so that's what I think people are not doing enough of. I think what they might be doing too much of. And, you know, it's interesting, you know, if people, I don't know if we're at a point where people are doing too much. I just think, you know, you know, I think people are probably doing a little bit too much on the pleasantries, you know.
- [14:06]
A
Sure.
- [14:06]
B
On the platitudes and those kinds of things. I do some of that just because, you know, in case the Terminator robots come for us, they're going to remember I was nice. Right. But, you know, I do some of that. But I think, you know, people sometimes I think, you know, I think people think too much. Right. I think they overthink the thing. I think they overthink the prompt and just keep it simple. Stupid. Right? And just start. Just start by writing it.
- [14:29]
A
Okay, I have a question. So based off your experience, I have no idea what your opinion on. There was a story, oh, my gosh, maybe like a year, maybe a year and a half ago, where basically they're saying ChatGPT was going to give you a better response if at the end of your prompt you said, if you do a good job, I'm going to give you like a $500 tip? And then they looked at the response, they're saying it did better. This was older versions of ChatGPT. People had all these kind of funny hacks and things like that, the thank yous and all this kind of stuff. Do you feel like in today's models, the. Basically the frontier models we're seeing today out of Gemini and Anthropic and chatgpt, do those things matter as much? Or have we kind of moved past some of those tricks?
- [15:07]
B
You know, they still matter a little bit. They still matter a little bit. And I think they still matter a little bit. So, for example, I was reading an article yesterday about how a guy convinced ChatGPT to give him Microsoft key codes by pretending, I don't know if you saw this article or not, but by pretending, asking chatgpt to be his grandmother and. And to sing lullabies in the form of Microsoft keycaps. And ChatGPT produced Lullaby, a lullaby based on, you know what I mean? So, yeah, so some of that's still there, right? But you have to be really. You mean you gotta be really messed up up here to do those kinds of things, right? So, I mean, but that's possible. There are people out there that are gonna do that. I. You know, but you know, it's. Again, it's. To me, I want people to do whatever's gonna make them comfortable, to just start doing it. Right? So here's the thing. You and I live in a bubble, in an AI bubble. And the people most. Some of the people listening on this, to this podcast, and I hope they're getting great value, are also probably within our. Within that bubble with us.
- [16:04]
A
Totally.
- [16:04]
B
But here's the thing. 3% of the population around the world really gets ChatGPT or Gemini or Claude or whatever, right? But there's 97% that don't. And that 97% is really what I'm targeting my books at. That's what I'd be. That's what I'm focused on, is that 97%. And they're just afraid to even start using it or even if they are using it. They don't even know. To your point earlier, it's like, what do I even put in this thing? Right? And the problem is that you got OpenAI releasing. You go to OpenAI, you go to ChatGPT and you've got seven different choose from. It's stupid. It's like, what are you doing, Sam? I mean, come on, just. Okay, fine, two or three max, whatever. Beginner, intermediate, advanced, simple, right? Study mode here, whatever, right? You know, I mean, it's just simple and study was pretty cool. But, you know, it just. Just to start just. I just want people to take the first step and don't be afraid and just start talking to it and just. And forget about massive prompts. Just start talking to it like you're talking to a friend in the Philippines or you're talking to a colleague. I mean, you do the same thing on slack at work, right? If you're on slack or teams, you know, you have an office in Paris and you're in Los Angeles. Hey, what are you guys doing? What are you guys up to? Blah, blah, blah. Can you help me with this thing? Right? It's the same thing, right? So just. Just pretend you're talking to another human. And if you start doing that, you start thinking with that process, it's going to be a win for you.
- [17:21]
A
Okay, I love that. I think that's honestly great advice for Everybody. One thing that I would be curious, your take on is basically what do you have a preference or have you seen any difference in what you've seen with basically using embedded instructions? So let's say you want to give it some context. You copy and paste a huge article, paste in versus uploading a PDF document, like so basically the text versus just uploading documents. Have you seen any difference in results there?
- [17:50]
B
No, I haven't yet. I tend to lean towards trying to copy what's in the PDF and pasting it just because sometimes I've noticed that ChatGPT struggles sometimes with maybe sometimes converting and understanding and reading the PDF. It doesn't happen often, it's very rare. Yeah, but I find that typically if I can take the PDF and just kind of highlight it, or export the PDF to Word and just highlight it, copy and paste it, that tends to be me, for me personally tends to be a little bit of a better, you know, better experience, results wise. It's, you know, they're not that far off, maybe off. You know, they're about the same off by maybe, I mean give or take 5%, you know what I mean? So, but it just, to me, it's just easier to just copy and paste it because then it's also easy for me to go look at that PDF and understand what's written there. And if I need to edit something that I can't edit the PDF stuff, but I can edit the prompt. Yeah, I mean, I can add a text prompt.
- [18:40]
A
Oh, that's actually a great point. Yeah, no, that's a great point. Okay, just from your own usage, I'd be curious what's one basically prompt or automation that you use maybe in your business, an example of a great use case for ChatGPT that you use that other people might not be familiar with.
- [18:57]
B
Yeah, no. So I'll give you two, and I'll give you one personal one. And the personal one I literally just did this morning. So I broke this wrist about eight weeks ago. So my brace is off now, but I broke this wrist and it was a little freak accident. And I'm starting to go back to the gym now because my brace is off and you know, I'm still having a gripping challenge a little bit. And so, you know, I put into ChatGPT saying, hey, ChatGPT, I broke, you know, I broke my wrist eight weeks ago. And it said. And do you remember? And it said, yes, because what I had done was I actually took a picture of my X ray.
- [19:27]
A
Okay.
- [19:28]
B
Back when I. Back when I Did the X ray eight weeks ago and I put it to the ChatGPT, what did I do? And ChatGPT told me exactly what I did. So my buddy's an orthopedic doctor and I went to go see him the next day, I said, what did I do? And he told me what I did. And it's exactly what ChatGPT said. So now my point is that that image is now in the memory function of ChatGPT, right? So it's in my account, on my memory function, whatever, right? So now I could say, as you know, I broke my wrist eight weeks ago. Do you remember that? And it says, yes, I remember how. Whatever. It gives me a response. I said, okay, I'm starting to go back to the gym. Can you please give me an upper body workout so I can start to exercise again and don't have to hurt my wrist. That won't put a lot of load on my wrist. That can also help me improve my grip strength, et cetera, blah, blah, blah. So I wrote the prompt whatever and hit enter and it gave me a nice detailed prompt of what to do, what exercises to do, what banded, it was all kind of banded workouts and so what to do, how to do it, how to structure it, rest. And it literally gave me a nice full response. It was perfect. I mean, oh, to me it looked perfect, but I'm not a fitness instructor, so to me it looked great. It looked great. It was awesome. I could just literally just take that, it's on my phone now. I go to the gym and I could do what I'm going to do and whatever, right? It was awesome. So that's just one example of something that I did today. And I know people now who take their blood work and you want to anonymize all your information. I don't put your name on all that stuff. But people will take their old blood work but the new blood work and have it compared to H igpt. I want you to act as a world class doctor. What do you think? How does blood work compare? And then they take that answer, then they go sit with their doctors and the beauty there is that they're better armed now to go be able to go back to the doctor and have a full conversation, right? Oh my A1C level is elevated. Oh, you're pre diabetic. So now they can walk into the doctor and say, I think I'm pre diabetic. Now let's have a conversation. Instead of spending 20 minutes talking about being pre diabetic, I can have a 20 minute conversation. How do I, how do I fix the pre diabetes, for example? Right. So that's, so that's, so that's one thing on a personal basis, you know, from a business basis, I'll tell you, the one that is probably in my company, the one that is the most used is working with ChatGPT to come up with new business ideas, new growth ideas, brainstorming new programs, brainstorming new ideas, act like one of our customers. How would you think they would respond to this? What are, what are some new things that customers are going to want and see what are the case may be and really kind of building up those case studies. So I, I have a little small, little advisory business that I do. I advise, I advise companies and organizations. I'm a member of this group called ypo and so I advise a lot of YPO members. I'm sure you've heard of it. But, and, and so like I'll give you an example. So like I recently worked with a YPO member and she sold her company. So she had a big giant shipping packaging company, right? And her company, and she's out of Houston, and her company would pack goods and crates and wooden crates. They could ship everything from small vases to Lamborghinis. Okay? And what was happening was, so she sold the company to private equity and they did it. And about a year later, the private guys called her back, hey, can you come back? We're having a lot of problems. So she went back and she saw me speak at an event in Barcelona earlier this year and she called me up afterwards and she goes, I really need some help. Can ChatGPT help me? I said, well, what do you need? And she goes, well, here's a problem, here's the situation. What was happening was the wooden crates were sitting longer on the container ships and because of climate change or whatever the case might be, the water was a little bit more acidic and it was causing the, the wooden crates to deteriorate a lot faster.
- [22:41]
A
Oh, okay.
- [22:43]
B
And so what's happening now? The salt water was eating away at the crate faster was now getting into the merchandise and the goods. And so what happened was, and this is a TR story, so there was a Lamborghini in a crate, the water got into the wood, it started spraying the Lamborghini, it started destroying the paint of the Lamborghini.
- [22:57]
A
Oh no.
- [22:58]
B
So now the company's on the hook for the lamp for a brand new Lamborghini. Okay, so that's a lot of money. So what do we do. And so I work with her. We basically did a, we did a business prompt, we did a brainstorming prompt on here's the issue, here's a challenge, here's a concern, here's the opportunity. So we gave it the whole road to the road, task, contact, task thing. We asked it for business ideas, solutions or how do we fix this problem? And spit out a bunch of answers. But one of the answers that spit out was look at Honeycomb product. And I'm looking at her like, honeycomb product. And I could see this light go off in her head first. In my mind, I'm like, it's cereal, right? And then she goes, no, no, no. She was interested. She had heard that there are people who would take honey and coat wood in honey because the honey would keep the, would absorb the acidic from seawater, whatever, right? And they'd never done that. So they're now in the process of testing, putting honey on some of their larger items and even the center, some of the smaller items. So they have, so they have like, so they have like a smaller. About this big and they actually take it and they, they're, they're now dipping. They've taken one and covered it in honey and honeycomb painted honeycomb, a thin layer and they like literally dip it in, in the seawater to see how long it'll last. And then they blasting it with seawater to see how long it'll last. And so far it seems, it seems the results seem pretty positive. But if that works, it's a game changer for them. And it took 15 minutes to write the prompt. So it's those kinds of things that in the business side that I get really excited about.
- [24:25]
A
That is super cool. I mean, that's such a fascinating story. Great use case. And of course that's exactly what you want to use ChatGPT for solving real problems inside of your business. Something that you mentioned there that I'm curious on. You know, you talked about inside of your organization. This is something that you guys use a lot for a bunch of different purposes. I think a lot of people that are listening, maybe they have co workers or maybe they're a leader in an organization and they're trying to get more people in their organization on board with using some of these AI tools. What I. What do you think the best approach for business leaders is to, to get the people around them using these tools?
- [24:59]
B
Right.
- [24:59]
A
We've heard companies say, like, oh, like I think in Japan there's like this AI mandate that some companies roll out where Everyone has to use AI at a certain level. Or like, how do you best encourage people to use it? How do you get them on their feet? I know your book is an amazing resource that everyone should give to everyone in the organization. But beyond that, how do you approach it?
- [25:17]
B
Prompting made simple. There it is right there.
- [25:19]
A
Promptly made simple. We'll leave a link in the description of the show, notes to the to the book.
- [25:24]
B
So that's a great question. So here's the thing. To me, there's six key takeaways for AI success at any organization. Number one is you want to. It has to start at the top. It has to be the CEO. He or she has to own this. If they don't own it, it won't go anywhere. So they have to own it up here. They have to own it. They have to make it an initiative. They have to make it the core part of what they do. They don't be experts in AI, but they have to be an expert in terms of getting it implemented into the organization. So step one is, is the CEO. Step two is a C suite. If somebody in the C suite is not buying into doing this, you need to get rid of them, period. There's zero room, and I'll tell you why in a minute. So they have to get on board. They got to get on this bus. Number two is be curious, not judgmental. I won't tell you the story of why. Why. That's number two, because I don't think we have the time. But you want to be curious. Don't be judgmental about this. Be curious. Okay. Number three, Data, data, data. Data is a new oil. Look at your data. Understand your data. Understand, understand. My challenge to people listening and my question for you, not for you, but for the people listening, do you have a good command of your first party data? If the answer is you don't, you have to spend a little bit of money to get that data under control. Because the more you do that, the more successful you're going to be an AI, because AI is nothing without great data. You need to have great data to be able to have AI do something with what it is you want it to do. And number four, it's not about artificial intelligence. It's about augmenting intelligence. So that's an education process. So you want to make sure people understand. It's about augmenting what you do. It's about helping you what you do. So for example, what do we have to do before Microsoft PowerPoint? What do we do before Microsoft Word, Excel, These are all tools that augmented what we did. Like, I grew up at a time in the 90s when we didn't have PowerPoint and we had to use overhead projectors, right? And, you know, you know, I mean, so. And we had to draw on a. On a. On a plastic piece of paper or on a plastic paper, but on a plastic sheet. And, you know, you had to go to a library. You had to go to the library, get a microfiche or whatever. So, but. But all these things, all these new technologies dying to augment what you do, help you get better, you know, at what you do, right? And if you can do that, you'll do the one thing money can't do, which is buy time. It'll buy time. It'll make you more efficient, and at the end of the day, become a storyteller and use your imagination. If you can do those, if you understand those six rules, you're going to be great. Now in terms of how to get people to use it, in my organization, we first did mandatory training. Everybody's got to attend mandatory training. And we've done probably close to now, probably a thousand hours worth of mandatory AI training over the last year and a half, two years for all the employees. Number two is we start having contests. You know, whoever had the best prompt, whoever had the best idea. We feature those in our. In our newsletter or in our monthly newsletter recognition event, whatever the case might be, like, hey, Jaden came up with this prompt to help him with this issue, and he did it. And this. But say he solved the problem, whatever, right? So we do that, and, you know, we start doing a little bit, a little bit of recognition around it, right? We found little contests around it. We have little hackathons around it, little things like that. So the more you kind of make it fun, the more you start doing it. But, however, the most important thing is you, the CEO, you have to lead, be the leader, you do it. Make sure every staff meeting you have. Are you asking your chief marketing officer, how are we using it in a. How are we using AI? Ask your cto. How are we using AI? Ask your Steve, whatever, supply chain officer, what, whatever you have. Are we using AI to help put our supply chain right? So the more you start asking it, the more you start pushing it, the more you become a leader around it, an evangelist, the more people will use an organization.
- [28:44]
A
I love that. Yeah, that is some fantastic advice. Rajeev, it's been amazing having you on the show today. You've shared so much incredible advice for anyone Listening. Make sure you go and check out his book link is in the description for that. As we're wrapping up the show today, I guess I would love to get maybe one final piece of advice for you for people using AI. Anything you haven't mentioned yet on the show that you think people should be thinking about?
- [29:09]
B
Yeah, look, I mean, I think when you start to realize that we are headed towards a future where things like, let me back up. We are the last generation that's going to be managing humans exclusively in the future, we're going to be managing humans, AI agents and robots. So, you know, all I can do, all I can express upon you is this is really important. Don't be afraid. It's not going to bite. Don't, you know, make sure, you know, if there's two types of chatgpt, there's the regular version and then there's the team version. The team version is $5 more a month. Make sure you get that one because that doesn't use your data to train, you know, but just start to use it and don't be afraid. It's, it's not going to bite. It's not going to do anything to you. It's, it's going to help you. It's going to enhance you. And for those of you that are listening are worried about job loss and those kinds of things, look, anybody who talks about, oh, there could be a ton of job creation or people say, oh, there's going to be a ton of job loss, the fact of the matter is nobody has any fricking clue what's going to happen. The only thing I can pretty much guarantee is that if you don't learn AI, you're going to get replaced by somebody that knows how to use AI. So just start the process and listen to your YouTube channel, Jaden, or your podcast or others and just start to run, play, try new things and it's going to be great. So it's just like riding a bike. Just get on a bike and start riding and who cares if you fall off a couple times, get back on, it's fine. Nothing's going to happen. I love that.
- [30:36]
A
That is some amazing advice. Thank you so much again for coming on the show. To everyone that's listening, thanks so much for tuning into the podcast. Sure to leave us a rating and review wherever you get your shows and I will catch you in the next episode.