Speaker 1: 0:00

Welcome RVA to Inspire AI, where we spotlight companies and individuals in the region who are pioneering the development and use of artificial intelligence. I'm Jason McGinty from AI Ready RVA. At AI Ready RVA, our mission is to cultivate AI literacy in the greater Richmond region through awareness, community engagement, education and advocacy. Today's episode is made possible by Modern Ancients driving innovation with purpose. Modern Ancients uses AI and strategic insight to help businesses create lasting, positive change with their unique journey consulting practice. Find out more about how your business can grow at modernagentscom, and thanks to our listeners for tuning in today.

Speaker 1: 0:56

If you or your company would like to be featured in the Inspire AI Richmond episode, please drop us a message. Don't forget to like, share or follow our content and stay up to date on the latest events for AI Ready RVA. All right, Welcome back to Inspire AI, a production by AI Ready RVA, and this is a special event with Vic Rogers, board member of AI Ready RVA, alongside with me. Vic's been working with AI Ready on a number of educational opportunities and he has a ton of expertise that we'd like to dig into today. On prompt engineering, With the recent recording of prompt engineering, we thought that this would be a great opportunity to share some open insights with the community at large and just see where it goes. So, Vic, welcome.

Speaker 2: 1:53

Good to be here, appreciate you having me and I'm looking forward to the conversation Awesome.

Speaker 1: 1:58

Why don't you tell us a little bit about your experience with prompt engineering? Where did it start? What do you do with it these days, like, why is it so special to you?

Speaker 2: 2:07

yeah, I think that's really the catalyst to me, kind of learning how to leverage artificial intelligence effectively.

Speaker 2: 2:12

Um, it was like around february of 23 is when I really um, got into the weeds of understanding ai and initially one of the things I wanted to do was, hey, write a business plan for me, and and I and I told it, but then it gave me this long book and then the vernacular that it used was totally off base.

Speaker 2: 2:31

So I figured out eventually, like, oh, you have to, I guess, iterate more. So I guess that initial use would have been zero shot prompting and then few shot prompting over time, like that's the actual terminology I believe that is used. But I just started talking to it more. And once you started talking to the machine more effectively, I realized that, hey, if I say this same thing over and over again, I'll get here quicker. So I created a prompt library and that prompt library ended up growing and growing and it's been one of the things that I enjoy sharing with people and it helps them really get around that curve of prompt engineering. So, like, if you find a good prompt, copy and paste it, throw it in a Google Doc and just keep building over and over again, because what's going to happen over time is it's going to allow for you to have a better output and get to where you want to look quicker.

Speaker 1: 3:19

Okay, yeah, so help us understand. What is. What is it like to have a prompt library, is it? It's a document that stores these concepts of how to interact with AI, right? So can you break that down for us a little bit? Imagine no one's ever worked with this, with a prompt library, before, and you're trying to describe it to them. What do you, what are you telling them about?

Speaker 2: 3:45

and you're trying to describe it to them. What are you telling them about? Old school people will feel this it's like a Rolodex. So what it is is just a place where all of your prompts are stored. I'm pretty basic with it, so that's why I put it in a Google Doc that I can share with folks. But, notion, I believe you can also put one, utilize a Notion and have like a website where you can go to and you can do it that way. So I'm eventually gonna take all the prompts that I have here and put it there.

Speaker 2: 4:12

But pretty much what it is, it's just like another throwback example. Anybody who used to pay like Sega or like Super Nintendo, game Genie that's a throwback Used to have a whole bunch of cheat codes, right, but it'd be a book. So what a prompt library is? It's just that book that you go to. You find the cheat code, you put it in and it's like the Contra code or whatever you know. It's like your God mode or Doom or whatnot, right? I hope some of my gamers are feeling me with these old school references, like your God mode or Doom or whatnot, right? I hope some of my gamers are feeling me with these old school references, but it's just a way for you to have all the prompts in one place at one time. So I shared my prompt library with you and I'm open to getting your feedback, but, like one of my actual first prompts on here a DEI prompt, so it's called Eternal Ebony, celebrating Ageless Beauty and Resilience.

Speaker 2: 5:06

Why did I create that prompt? Well, if you go into chat GPT and you ask it to create an image of a person that you think is wealthy, a lot of times it might not pop out what you want. Or if you ask for it to give you an image of what you think I look like, if I describe myself, it might not look like me. This prompt in particular, though, it makes the machine or the LLM that you use come back and give you an image that looks like very, very good and the way I built it. I just just started iterating, so like just kind of to read a little bit of it, because it's kind of it's going to, it's going to, it's going to sound funny, but I create an image that vividly embodies the concept of ebony, along with the timeless adage black don't crack.

Speaker 2: 5:56

Now, why did I say black don't crack? In the black community. We understand what that means. If you look at me, I get a lot of times especially now, since my hair is, you know, wrapped up I look a lot younger than I am. But you know I'm pushing 40, you know, and it's one of those things where people are like oh, black, don't crack, right? So the image that you would get from a LLM if you put give me an image of a black person with black don't crack, put that in if you're listening to this, and then just give me it and then go to another LLM and say give me an image of a black person Two totally separate, different things.

Speaker 2: 6:28

But sometimes you might forget what the prompt was. If you have a prompt library, though, you keep it all in one place. You can go back and rinse, wash, repeat, use it over and over again. A couple of other things in here and of course it's like it's long. It talks about skin complexion In the black community. A lot of times, fair skin versus dark skin, hues can be depicted differently. So this one prompt in particular, it just gives you the ability to create an image, maybe of a woman or multiple women that you want to depict maybe a relative, or my mother or my wife or my nieces or anything like that.

Speaker 2: 7:07

It gives you diversity in how the image would come back to you versus just asking a zero shot. Create me an image of a black family, because a lot of times it doesn't give you what you want. You'd be surprised how often it comes back and you know it's not a positive image.

Speaker 1: 7:25

And it's not a positive image. Yeah, so you give it as much description as possible to depict what you're looking for. So you're building a prompt here in a paragraph format so that you can instruct the large language model to depict exactly what you're looking for. And this one particularly is for image development and you can imagine the user of this prompt library thinking about whatever image that they're looking for and you kind of lay out how specific terms you're looking for so that the experience of that large language model can determine exactly what it is that you're referring to, because in quotes you say black don't crack, and it knows exactly what that means and therefore it can leverage that knowledge to develop what you're looking for here. That sounds, that sounds really good yeah, 100 percent.

Speaker 2: 8:27

Um llms understand colloquial speaking, um I can't think about the exact terminology, but um, I always shout out um. There's a customized gpt on chat gpt called chat black gpt. Just use that for anything that you want um and that's. And if you want like a good understanding of like how I could be um more intentional with my communication with the black community or anything like that, that's a great, great GPT. I know the person who created it and she's a rock star. Her name is Erin in the AI space. But even if we want to talk about something that me and you both are really, really interested in, I have one on here called Teach Me Money and what it says is I want to assist uh, I want assistance um by a qualified financial service provider for 20 years, specializing in x, um and enabled with experience on understanding charts using technical analysis tools. Why? Interpreting macro economic environment prevailing across world. Consequently, assisting customer acquisition, long-term advantages. It's a lot of words in there, right, but I built that over time with the chat GPT itself. So, like, if you want a good prompt, just you could tell chat GPT to make a prompt for you and keep building it over and over again, but in particular this one, when I want to learn about something, um, I might like right now.

Speaker 2: 9:54

Um, nvidia's earnings came out last week. Uh, they did well, but the market didn't like it. So now nvidia's looking bad like this we're going into sunday. What's the date of second right? So, um, we're going into monday. I don't know how the market's gonna look, but if you want to understand how to better understand, like, have an idea, like, well, why did the market dip? Why? Why is this? Why is that? You can use this prompt and then in chat, gpt, hit the search button and it'll give you the most up-to-date information around how you want to be taught money, and then you can make decisions from it.

Speaker 1: 10:28

That's wild. Okay, so this, this particular one, leads, leads up with give me a qualified financial service provider that has 20 years experience specializing in. You said X. And if, if I were to look at NVIDIA, is the example you provided? Maybe I would say 20 years specializing in market analysis, or something like that, and that would put this prompt, or the AI with this prompt, into the mindset that they particularly have that kind of experience and leverage the exact or the precise type of information that a 20 year experienced market analyst would have when being asked specific questions.

Speaker 1: 11:29

You know and so like you could change that by saying teach me money, as if I were an eight year old. You could do that too, and then you would get a completely different answer. So one you're. If you're willing and ready to talk with a financial service provider that does have 20 years of experience, then this is the prompt for you. Yeah, I really like that. And and I heard you say this a couple of times but you started with you know, teach me something about this. And then, as you learned how the AI operated, you, you updated, you modified this one routinely went back to it and said I, you know, I want, I want a little bit more in terms of macroeconomic, environmental, environmental-like type descriptions here, or I want to know more about factors related to inflation, and then you would just plug in those types of elements and, based on whatever got you looking at this that day, if it's NVIDIA, you might change the prompts so that you're leveraging things about you know, tech, industry, you know, or hardware, or AI, even to get different, more specific results.

Speaker 2: 12:46

Yeah, 100%, and like what you're looking at again. What we're talking about is like what you're looking at again what we're talking about is like why a prompt library is effective.

Speaker 2: 12:55

So you're at prompt number 10. Scroll down to prompt number 58. This is called the ultimate stock analysis trading prompt. Now here I have technical analysis evaluating 200 day and 72 day moving averages, financial health, assets, year over year, market position, market sentiment, risk assessment, entry and exit strategy, growth potential, sector comparison. And then I want you to give me a clear, concise, two to three sentence output telling me when I should get into a stock. Is it a long-term play? Is it a short-term play, if I want to do options or things like that? But the reason why I say scroll down to number 58, because you'll see the evolution of how I've been able to enhance my prompting over time. That prompt that I created at the top I did that probably a year and a half ago this ultimate stop analysis one I did that probably about six months ago and I'm just getting stronger and stronger with it. So, like, the more you play with ChatGPT and the prompts and if you have a use case, you'll get better, and I'm always open to having these conversations with people.

Speaker 2: 14:04

Victorrogersai, r-o-g-e-r-s. Because we're not Aaron Rodgers fans, victorrogersai, come holler at me but, um, yeah, like it's one of those things where it's just fun, because I've worked with real estate people. I've worked with people trying to make milk preps. I worked with parents trying to make um, you know sleep structures for their children. Like you can do a lot of different things with ai, but you just gotta um practice with it. It's like playing a video game. Like you know, if you're uh, I've been on this gamer thing all day. But if you know, if you're playing Madden and you're getting beat on all Madden, move it down to maybe all pro and you'll get a little better. Then over time you can compete in Madden. You play online and when them kids from overseas are just whipping you and talking smack to you in another language, then you just hold your hand and shout at them.

Speaker 2: 14:47

It's a great opportunity to learn, and it's free too. You shouldn't necessarily have to spend a lot of money to learn, and it's free too.

Speaker 1: 14:51

Um you know, you shouldn't necessarily um have to spend a lot of money to learn how to prompt yeah, yeah, I kind of I try to emphasize that in the podcast, that it's you know it's never too too late to start, and the more you do this, the better you'll get at it as well as uh, the more ahead of of the change that you're going to be right Like. This sort of thing is going to continue to change over time. Getting the foundational understanding of how to leverage these tools is going to make things happen for you a lot faster. Yeah, I really love how like the depth that you go into these. I see you have a lot of interest in investing.

Speaker 1: 15:37

So I would love to look at this a little bit more, because that's definitely up on my list of hobbies, for sure. How successful would you say you've been with prompt engineering and investing specifically? Not that you're a financial advisor, so those listening to this shouldn't take it for that, but for what it is right. You're leveraging the intelligence of of all things Internet and AI here. So well, give it. Give it a score, one to ten, like how do you feel like this is done for you?

Speaker 2: 16:13

Yeah for a guy. You know I went, I went to school for marketing, you know, got a minor. I think, well, actually I got a minor. And I think, well, actually I got a, yeah, minor in management, right. So like coming into, like investing in things of that nature, like I didn't really know a lot, all I did was start listening to like financial podcasts and if you look on here, michael Burr, ray Dalio, I know- the Kramer one Of all of them.

Speaker 2: 16:45

there's a guy named Ian Dunlap, number 54. What I did was I listened to what they were talking about on their podcast. I took a transcript of what they would say and I said turn what they said into a prompt and then it a hundred percent allowed me to be a lot more better, like what I, what I say. If I had to scale myself on one to ten I say at eight because I'm not better than these people, right, I'm just able now to catch up to like a learning curve like that I would have. It took me a lot longer to, you know, learn like, I would say. I learned this stuff 10 times faster than most people and I take for granted how far along that I've come. A combination of using AI and just being a student of the news constantly. I've gotten really, really strong. I think I am bullish, though in the semiconductor space I need to probably be a little bit more diverse. Eli Lilly and healthcare is something that I think I want to diversify in and read a lot more into. But as it stands right now, look, I'm looking at an AMD and TSCM and ASML and you know I can read. I can rattle all those ticker symbols off. So, like I'm understanding that, that semiconductor sector and everything around it whether it's Amazon, whether it's, you know, meta, youa, all those companies that have tentacles in artificial intelligence I'm looking at those companies and finding out how to invest.

Speaker 2: 18:04

I'm mainly doing long-term investing. I'm scared to do options. A few options that I've done they're always call options. I'm never going to do a put. I just don't feel comfortable because you can lose everything. I had a buddy of mine lose his whole 401k because he didn't put a stop. Put, I think, is the term, and yeah, it just kept going down. So you know you gotta be safe with that, but you know it helped a lot. I'll say an eight, though, because I mean, like I can't say I'm not a professional like financial services provider, but if you want to talk finance and numbers and stuff, I'm, you know, pretty good, and if you look at how much I've been able to accumulate over a short amount of time, I think I'm doing pretty average, you know. I mean I'm confident that I'll do better over time, but humble about it because I know at any given time the market can dip.

Speaker 1: 18:54

Yes, yes, indeed. Yeah, I'm a long investor as well.

Speaker 2: 18:57

Do you have any long-term plays that you want to recommend? I always like having those conversations Long-term plays or like companies to invest in. Like when you do long-term investing, how are you doing it? Is it in like Roths, usually ETFs? I do like the semiconductor industry.

Speaker 1: 19:12

I do like the semiconductor industry. I like tech and healthcare. I don't do anything in the commodities and consumer goods. I think it feels like I'm chasing something there and I think that when I invest, I invest for growth. They're not all considered growth stocks or or anything like that, but um, I do look for the long-term wins, um not the.

Speaker 1: 19:43

I don't chase the short-term wins. Um, sometimes if someone says, hey, palantir is doing really well, uh, you need to get on this train, I'll take a look at it and and say, what, where does this fit into my long-term approach? And say, how much money would I be willing to kind of chase this Because it's a chase, right, it's a risk, a bet on some company that you don't really know much about unless you do your research. It sounds like you do a lot of research. That is me.

Speaker 1: 20:18

Personally, I try to try to stay away from too much of the the risk, risky type moves in the short term and and just stick with that long-term stuff. And and and I make bets on on companies that are, are responsible. I make bets on companies that that have good books in place and, uh, they, they aren't you know um disrespecting humanity in any way. Right, like I, I, I care about those things and uh, you, you vote with your wallet these days. So that's that's kind of how I do it.

Speaker 1: 20:53

I um try to max out my 401k, 401k and it's barely risk averse by putting money in 401k before you put money into direct equities, you know if you have any money left over and you can. You can put some in your direct equities. That's great, and you can play with that and and you can. You can do some other things. I could talk all day about this stuff, but this isn't a financial service podcast, right, it's an AI podcast, and I haven't really gotten the nerve to ask AI about my financial strategy in any way. But you've built my confidence here and I say to myself I'm going to take a deeper look at your prompts and and and think about how maybe I can, I can use that for my Lawrence long-term goals as well. But hey, man, this has been awesome. I really appreciate your time today and I think I think we got something good here. So I'd really like to work with you in the future on on maybe some additional co-hosting podcasts on various AI topics. So yeah, man, let's make it happen.

Speaker 2: 22:03

Oh yeah, I'd love to do it, man. I think it'd be, I think it'd be a good little conversation, man. All right, yeah.

Speaker 1: 22:08

And tell everybody again where they can find you.

Speaker 2: 22:11

Yeah, so Sustainable Growth Creative that's the name of my venture. You can look it up at sustainablegrowthcreativecom or org, but the easiest place is just victorrodgersai and that's V-I-C-T-O-R-R-O-G-E-R-Sai and the company name Sustainable Growth Creative. I'm going to just do a shorthand now. Sgc you can find that on LinkedIn. We got Instagram. Doesn't have a lot on there, but we made an AI agent on Instagram. That that's really what it's for. So if you are interested in oh, if you're a federal worker trying to figure out how to navigate these bumpy times go to our Instagram page. We made an AI agent that help people through the job market. Right now. It's kind of spooky and like that's the type of stuff that we're doing over here at Sustainable Growth Creative. We're trying to help folks feel comfortable with the technology and then feel comfortable to get out of that, get out of any kind of situation they're in. But again, victorrodgersai is the quickest place to find me. I love to work with new clients and hey, jason man, I appreciate you sharing your platform with me.

Speaker 1: 23:21

Um, I know you could be doing a lot of other things besides listening to me ramble about you know random stuff. But thanks a lot, man. It was a pleasure. All right, talk to you later. Later, and thanks to our listeners for tuning in today. If you or your company would like to be featured in the inspire ai rich episode, please drop us a message. Don't forget to like, share or follow our content and stay up to date on the latest events for AI Ready RVA. Thank you again and see you next time.