This podcast episode is sponsored by ChiefAIOfficer.com, offering training and certification through the International Association of Chief AI Officers. Interested in a new career or leveling up your value in the marketplace? ChiefAIOfficer.com can help.
SPEAKER_00Welcome to this episode of Using AI at Work, where we explore the forefront of AI and its impact on our work and lives. I'm your host, Chris Daigle, and today I'm honored to speak with Segovia Smith, the co-founder at Chief AI Officer and the visionary behind Prompt Vault Studio, our proprietary software that takes your prompt engineering learning curve from months to minutes. Together, we'll uncover how intelligent prompt management can transform your interactions with AI, enhancing both efficiency and creativity. So whether you're an AI enthusiast or a professional looking to optimize your workflow, this conversation offers invaluable insights to elevate your AI mastery. Let's get started.
SPEAKER_01Welcome to Using AI at Work. I'm your host, Chris Daigle. Each week we'll be learning how today's business owners, entrepreneurs, and ambitious professionals are getting more done with smart use of tomorrow's tech. Let's get started. Okay, so welcome to this episode. This is something that we've never done before, which is kind of weird. Uh Segovia is a co-founder of the uh company Chief AI Officer that's a sponsor of the Using AI at Work podcast, but we've never had him on as a guest, despite all the wisdom that he has to share. So um, Segovia, before we get started, I don't know, do you want to kind of share um a little bit about your background and how you found us?
SPEAKER_02Yeah, um, so I come from a long line the last decade, spending time in the uh what we call ed tech, you know, education online learning. Um, as a product manager, I've focused a lot on, you know, developing out curriculum and not only the curriculum side for um what we I used to sell in a previous startup into the B2B space. So we were essentially creating learning and development products, leadership courses, emotional intelligence training, team training. Um, we had one for DEI. There was just a ton of different courses that we had developed out. And uh that startup kind of got gobbled up by COVID in terms of, you know, everybody lost their firing, laying off tens of thousands of people. All the L and D budgets were drying up, and so we had to kind of put it on pause and look for other opportunities. And that was right, you know, around the time when AI started to take off in that uh early Chat GPT 3.5, it sort of broke out on the scene. Uh so I went into a year of, you know, sharpen my skills, you know, re-up on product management 101, just kind of go back to basics, first principles, and really look in and lean hard into what's going on in the AI space. You know, I'd been a part of some of the stuff with chatbots, and you know, those were cool, but you know they were a bit gimmicky and you had to have the right database to be, you know, not sounding like it was completely wrong answers. Um, so after spending uh a full year of going just deep into AI upskilling, taking every course I could get, spending at least $10,000 or $15,000 on a number of different programs from Udacity, Udemy, even some Ivy League schools, I kind of ran into chief officer, and I think I was actually uh impressed by what I'd seen, you know, in terms of how you guys were positioning the course and the curriculum. It wasn't just about, hey, this is another AI skill set, but it was more like how do we think about, you know, the strategic aspect of a business from that 30,000-foot view and then move forward across the org chart, bringing those high-level uh game plans down into um the movement of how do we deploy AI, right? It could be at the ML level for machine learning, it could be at the surface level, which is uh the gen AI. And then you and I got to conversing. Um, you were just kind of going through uh a round of kind of re-org uh within the organization and didn't have a product manager, so I was like, oh, let me get in there. So we started, you know, and uh the rest is kind of history. So anyway, it's been a fun journey, and I'm totally excited for where we're gonna take it in 2025.
SPEAKER_01Nice. I'll be honest with you, I didn't know what a product manager did before we hired you, but as soon as you told me what a project manager manager did, I was like, we need that now. So it was a great addition to the team. So when you joined us about a year ago, obviously the certification program had been built, it was ready for release. Uh however, we had a few projects that that were more technical in nature that were in development, some customization, custom software, that sort of thing. Now one of the tools that we had in mind was this AI tool for uh chief AI officers uh that was the prompt vault studio. What if you're not familiar with prompting, you're listening to the wrong podcast probably. But we knew that the shortcut for success for a lot of individuals wasn't necessarily them being starting with mastery, it was them starting with knowing how to communicate to the the LLMs, to chat GPT, in order to get the results that it was looking for or that that the individual was looking for. And that's based on intelligent prompt creation. The prompt vault studio that we kind of had gotten a lot of traction on but hadn't launched yet, was one of the first projects that you stepped into. Um can you kind of share what the goal like uh that we've got now and the direction and the vision behind the prompt vault studio for somebody who maybe isn't familiar with some of our software?
SPEAKER_02Sure. Yeah, it was interesting because you guys had a team, I think you know, you had a Scrum Master, you had, you know, uh an I uh somebody in kind of IT, a couple programmers, and it was interesting because you guys were using no code low uh low code, right? So it was using a tool that I'd never seen before. Um I traditionally had more of the the full stack developers working for me in the past. Uh so I had to learn a little bit about what we were using, and it was bubble that it was built on, um, bubble.io. And when I first kind of got my fingers into there, it was like, okay, what exactly are we doing? And there was more than just you know a prompt library. I think at the time, I don't think we actually had even come up with a branding around Prompt Vault Studio. It was like we had a prompt library. Um, we had another tool uh that was kind of like almost like a meeting assistant, like an otter or a fathom, you know, that would like download and stay it uh connected to all your meetings, which was pretty cool. Um what else was there in that mix? There was at least another two or three that were kind of in development. It was like a whole network, a suite, if you will, that you guys had envisioned. Yeah. Um, but none of them had actually gone to like shipping production. They were all, you know, still in their ideation, you know, UIUX was still getting worked out. And I felt like what we needed to do is look at what is our core business model. Um, and you you obviously had just finished developing the certificate training program to becoming a chief AI officer. I saw before I even really had any long-term conversations with you around where the product was going, when I kind of approached you to say, hey, I think I could be a good fit with the organization. This is what I bring to the table. I already had like a vision in my mind for where I thought, you know, chief A officer and being certified as a kind of plant your flag at the top of Everest, this is where we are in the org chart. And there will eventually be a trickle-down effect of all kinds of skills that'll be downstream from a chief A officer. So I wanted to make sure that the tools that we were actually putting together filled that uh as the end game and served that audience. But at the same time, I want to make sure that we didn't have the need for getting Sherlocked because, you know, at the time, you know, we were both, you and I were looking across the AI landscape. There's tools coming out almost like hundreds of them every single quarter. You go to Matt Wolfe's YouTube channel and some of these other guys are just covering, you know, tens of thousands of tools. And it's like we want to make sure that we're not in a position to create and put all this energy into something that all of a sudden, you know, uh Anthropic or OpenAI comes out, adds one feature, and boom, like lights out. You know, nobody needs what we're doing anymore. So although a prompt vault or what at the time was like one of five spokes on a wheel as you know, a prompt library, it didn't seem like the sexiest thing out there. It was the most functional thing that would serve our audience. And at the same time, it was scratching an itch we had internally. And I believe that, you know, even on my own year of upscaling in AI, once you get past the point in your own kind of, I'm on the command line, I'm typing in some stuff, you know, we all graduate from I'll just keep sneezing one-liners into this, to I actually need measurable outputs that are reliable every single time. And when you spend you know, a couple of hours going back and forth on a few sessions, you realize that maybe I should just copy that prompt and put it somewhere. And so the journey of many people, especially if you're positioning yourself as a certified, whether you're certified, you're an AI expert, AI consultant, et cetera, the chances are you probably have a few tricks up your sleeve and you store them, whether it's a Word doc, Google Doc, Note, Evernote. You know, some people had been developing Notion databases. Um, and so when I looked out at the landscape of what were people doing, the thing was there wasn't really anybody that was excelling in the space of being like a place you could store all your prompts. But this is prior to AI, like Anthropic in the last, say, six months has come out with a directory. Um, you know, this was even prior to Chat GPT releasing their custom GPTs, which is a little bit different from just a static level, you know, directory of prompts. Most of what we found in the marketplace at that time was places that were uh either just glorified notion databases, in which case, you know, you can you can taxonomize and organize them, but you can't do anything that's totally uninteractive. Um, and then the other thing we were finding was uh what would be coined as like a prompt marketplace itself. And they were kind of a little, they were mostly focused on image-related prompts, all the you know, dressing of anime characters and everything else. You could buy a prompt for between $3 and maybe $10. And it didn't seem like it was a good place for us to be fitting in to say we're you know supporting people at the highest level of the org chart, but yet we're gonna be charging them $3 at $5. It's kind of a nickel and diny, you know, system. So I I kind of honed in on that one product out of five and said, let's go just ship this, let's go take this to market, let's work with the UI guys, let's work with some design uh team and get this thing actually out the door. And as a PM, a product manager, first thing I did was go start talking to some customers, which meant interviewing people, like going out there, beating the pavement, if you will, going onto AI events, workshops, and just dropping a mention that says, hey, we got a software in beta. If you would like a free access, talk to me. And so through a series of dozens of these interviews where I would just watch them over their screen, do kind of an over-the-shoulder uh ride along and say, You should you use the tool. Tell me what you think it can do. What would you want it to do? We started to identify different features. And those features, you know, began with obviously organizing them. That was a given, being able to categorize them based on business use cases or departments, um searchable, indexable, et cetera. And then it moved from, all right, so now you can put all your prompts here, you can quickly access them. Uh, what about if you were to be able to run the models? And that's where I think things you know went from just a static, you know, better UI than a Notion database to, okay, you you let's talk to our development team and start to connect the dots. We'll go to all the APIs and we'll just hook this thing up to the latest and greatest of um OpenAI, uh Claude, you know, when when Lama 3 dropped, you know, we were in there right away. Within days, we were launching it. So inside the kind of UX user can take a prompt. So imagine this is just the user use case. You sign up for prompt fall. The first thing is if you're new, as you know, whether you're a chief officer or uh soon-to-be certified chief officer, um, or just an AI expert that's on the path of discovery, we always say, you know, taking your learning curve from months to minutes. And how, how is that really accomplished? It starts with being able to look over the shoulders of other people who have gone before you and really say, ah, okay, not just, hey, I think I'm just gonna get better through braille my way blindly in the dark, but let me see what other people who actually know how to do prompting. I don't think that any person who becomes a certified chief officer needs to be a prompt engineer. My qualification for a prompt engineer is more like if you are a Python coder and you're like in the deep weeds or you're like a master linguist and you're gonna be hammering away on prompts all day long just to find out like how these models are being tuned. I think that career-wise, there's an opportunity for people to still kind of go in the direction of being a prompt engineer. But on the masses on the whole, I think the same way we all ubiquitously know how to just get results out of Google by typing in directions to a restaurant or anything, prompting will become just a part of our fiber. Everybody knows how to do it, just like they now unlock the screen on their iPhone. But between here and there, there's gonna be a lot of upskilling required for people to be able to get the maximum value, not just the sneeze uh a parlor trick type of thing. So by us being able to uh run those models, it gave up some of our more advanced users and people who are like, all right, I'm gonna use this prompt, you know, uh ChatGPT 4.0 just dropped. Hey, I'm gonna try, you know, the difference, and I'm gonna take a prompt, run it in Claude, one click, okay, see results, run it in ChatGPT 3.5, see results, run it in 4.0, run it in Llama. And within three to four clicks, you're already actually seeing these and saying, uh, okay, this is the better result. Like for a long time in early 2024, it was very obvious. And most people that were talking about Claude were saying, if you need long-form written content that really emulates, I mean, the writer's circles, people that were um in uh the writing communities, they're all like, it's gotta be Claude or nothing. If you want, you know, programming and you know, coding, it's gonna be over in ChatGPT. And then you'll get different uses out of Llama or even Gemini, et cetera. So that was kind of the next evolution feature in beyond just an organization tool. And then, you know, again, through round and round of product feedback working backwards from the customer, we realized that, you know, rather than just having a couple hundred prompts in here, and again, all of the prompts that we have, when you look at a standard prompt library that comes when Claude released theirs and when ChatGPT released, OpenAI released theirs, you know, at least 80%, well, maybe 50% of them are business related. And then there's like, get me, you know, uh, how can I make a recipe out of the stuff in my refrigerator, or, you know, give me a nutrition plan for my, you know, 30-day workout challenge. And those are great. It's not that they don't have a use case, but in terms of where we're focused, it's not in the business mindset. We we want to work backwards from the business use case at the perspective of a chief AI officer or somebody who's an AI architect and claiming to make a living, you know, consulting on AI. So we realized that we could expand all of those product or all of those prompts that we had already been developing by parameterizing them. So very quickly in Q1, we moved from, okay, now we got a couple hundred prompts to let's add parameters to every single one of these in a specific way that allows, instead of being, I'm a business in niche X with a product in niche Y or target at this, by adding a parameter around that instantly all of our hundreds of prompts became use usable to anybody across any industry in any niche. And rather than having to do the input of that directly in line with the paragraph, we kind of created this color coding syntax, um, much to the way if you're using an IDE or your actual programmer, you would see as you're highlighting an HTML or to PHP or whatever, you'd see like certain keywords highlighted with brackets and parameters. So this allowed somebody to actually create those parameters and then it creates for the end user, again, if I'm the architect or the AI expert, I can create four or five, a dozen prompts, and then I can downstream them at because I know how to write the best prompts, and I can say, give them off to the marketing team and say, okay, all you got to do is fill in this little web form. It's like four to five keywords in a field, and it's going to allow them to generate it. So whether they're generating it on the spot inside of the prompt vault studio or they're copying it and going, you know, ChatGPT native and having a full-on discussion, you know, it allows us to be able to like give those things. So a user could come in, create a couple prompts, share them with a link, drop it in a Slack or a Microsoft Teams, and send you know somebody else to go do their bidding on this. So those were some of the early features that really, you know, we saw traction on and a lot of uh user feedback around. Um so there's lots of you know new things that we've been doing.
SPEAKER_01But one of the things that I was looking for everywhere, so anybody listening, if you've uh especially on on Twitter or X, if you find the AI expert, they'd say, you know, like, comment, and reply blank, and I'll send you my thousand best prompts. Wow, that's fantastic, right? Like for free. So I did it a few times, and what would I get? I'd get one line prompt. So build me a marketing plan, write an article for LinkedIn. I was like, man, I know enough about the use of these models to know that that's gonna get me who knows. That's really it's it's just not enough context for the models to be able to give me ideal output. But I thought, man, there's all these prompts out there, redoing them and and making them robust would be extremely time consuming, valuable, but very time consuming. Boy, it would be great if we had some way to just type in a basic prompt and have an intelligent algorithm to make it a world-class prompt to improve the prompt. And one of the things that you guys worked on for a while and now it's out in the wild in the promptable, was the prompt improver. And I think that that's a huge feature. And if somebody is confining themselves to where they have an idea of what they'd like to, you know, I want to use ChatGPT to do this, but they're not quite comfortable crafting a full prompt, you know, based on certain prompting patterns. The prompt ball has the ability for them to take that kernel and with one click of a button have essentially best practice applied to expand that into a full-fledged, like a professional quality level prompt. What was that like what was the process of figuring out the logic or the intelligence behind that?
SPEAKER_02Yeah, it was certainly um it was no small feat to kind of getting that thing dialed in. Because and and that becomes like the difference where we can all get good results out of the model, but when you've got to get measure like measure twice, cut once results where, like, for example, when somebody fills so so just to redescribe what you're doing, um, you walk into Prompt Vault, you've organized, you you've gone through and you said, here's a bunch of prompts, I'm gonna favorite some of these because they're categorically relevant, whether I'm in HR or marketing or what have you. Um, and then now you're like, I'm gonna input some of my prompts. So you click on new prompt, the pink button up at the top, and then you'd say, uh, all right, I'm gonna enter my one-liner. And whether you've got a one-liner or you've got a full-on like multi-paragraph prompt, it's really, you know, it'll work with either or. Um, but generally people tend at the early onset of their prompting experience to type in one sentence. It's just the nature of the beast. So we would be able to then take that and not only uh at the I think the earliest version of this, we were like, okay, we were having all kinds of problems getting it consistent. So we were like, we'll, we'll just we'll just make it easier so that when you go to the next form field to say, well, what is the title of the prompt and what is the description of the prompt? We were detecting what you were actually trying to do and saying, we'll come up with a three-word, you know, keyword title for this. If you want to change it, you're welcome to do it. But like, that's one less thing as a burden AI is helping you with, and it's creating an organized system that other users can use. Because although we contribute, you know, hundreds of prompts per month into the vault, it's a also there's UGC, user-generated content, where other users can make their prompts publicly available or they can keep them private. So, you know, this was a useful feature, but then we we knew we had to kind of crack the nut on how do we get it consistent to you know, dial in those results that we were looking for. And it wasn't just come up with a better paragraph because we had already introduced this other feature of parameterization, where instead of like if a user types out, my business is in um, you know, let's say the I don't know, pick a niche. The realistic dog grooming. Okay, I dog groom, my business is in the dog grooming space, blah, blah, blah. And you know, we featured product, you know, uh special magic watcher soap, I don't know, whatever that is, then they're gonna obviously type out for their own use case. But we had to make sure that the the actual engineering behind the prompt would evaluate what is the business use case, where or every opportunity within that sentence or paragraph or entire, you know, full enchilada prompt, and look for how do we parameterize this so that the user who's actually inputting the data can use it more effectively in the second, third, and fifth use if it's somebody downstream in the org chart. But also, if they share this open source as a public prompt, how would another user who has a car washing, not a dog washing, or a hairstylist or a whatever, you know, uh laundromat business, they could use the same prompt. And so it was one part was like identification of what could be added as a parameter. Then there's oftentimes users. That will put stuff in where they don't actually give us any indicator of what the business is, the niche is. They're just, it's so vanilla. So then we had to figure out how does a model actually not just come up with the stupidest thing like the balloon twisting business for kids, but keep it in the realm of like business-oriented topics and themes. So that comes down to like the data set that we're using, almost like a rag model of like, here's all the other documentation behind the scenes. If you think about how a typical like custom GPT would be written, you've got your uploaded assets and then you've got your prompt and instructions. So each one of those, based on the title, based on the description, and then based on the prompt itself, has to go through a few gyrations of output. And it's no good when you go, oh, it comes because again, if you've ever tried to do that measure twice, cut once and you get it working, you're like, oh man, it's working. And then you try it again. It's like, oh, it's now giving me four answers, A, B, C, D, which one would you like? Or it's giving me that little stupid sentence at the top that says, you know, some flavor of, you know, let me just be your friend here and you know, conclude with a sentence like, would you like to continue the dialogue? We don't want any of that. We just want only the title, three to four keywords. So that was the part where it was probably the most tricky and kind of layering it in to get consistent results. And I know, you know, we were trying to ship that, you know, as fast as we could, but at the same time, I kept pushing like the dates out to go, we got to get this thing nailed, you know, before we officially, you know, release it to the public. Um, but I'm I I've, you know, for the better part of the second half of the year, I've never seen it fail. We've done hundreds of demos. We do it live on the challenge all the time, and it's just it's magnificent. So you could bring in your one-sentence prompt and it'll turn it into a prompt that's designed. You know, it'll take the concept of what is the user role? Like it'll it'll use different frameworks and models that are commonly taught out there in the space and give you a full-flushed thing where it's it's actually identifying your blind spots. We have some spots, you know, different uh we kind of cycle through almost like round robin. So not every prompt is using the same formula, but it's like evaluate this prompt and look for the blind spots that are being missed by the user. And you may have thought to parameterize this one or two things, but it's gonna go, oh, okay, we know what the context is where you're trying to go. And you're not asking this question or these other five questions, boom, boom, boom, boom, it lays them all out. And maybe like 10% of them are going, that's where the human in the loop comes in. And you're like, you read it and you go, Yeah, I don't really need a date field. Like it doesn't really matter. You know, we'll just you know X that out and you can kind of massage it a little bit. Um, but then if you go and you give it like a full, like let's say you're really good at prompting, you've got this whole stricted thing or scripted out thing. I mean, even to the fact, like I taught a class inside the community recently on pseudocoding, okay, where you're almost, it's almost like prompting in the language and the syntax of programming, but pseudocode is more, it's it's not using actual code, it's just kind of like nesting things to say, I'm gonna use the the concept of a for loop or I'm gonna use an if-then-else statement in plain text with some you know square brackets around it. So you can run the you you know the user through something like do this as many, especially if you're getting into this thought, you know, leadership style prompting where you're saying, I want you to interview me and continuously ask me questions on a subject until I, as if you're running a software, type the word stop or exit stop, you know, and until you can kind of program that in using some pseudocode. So if you're an advanced, you know, prompt person, engineer, whatever, you can throw higher grade stuff in there and just say, look, I don't want the AI enhancer involved. I know what I'm doing. Or you can go to the second stage and say, yeah, let's just go ahead and give this a try and see what it comes up with. If it butchers it, okay, fine, you can just click on do. But if it still gives you some enhancements that were in that blind spot, um, you can go ahead and use them and tweak them. So that I think has been really exciting. Um, you know, I was just looking before we got on this interview to talk. It was like, how many prompts, you know, do we have? Like, what is the measured result of like, you know, looking at the the bar chart of like month over month? And we are now, you know, I think like maybe last week we crossed 2024. We have 2024 prompts and counting. Um, you know, that doesn't mean that everybody logs in. Uh, and then I was like, okay, sweet, I'm gonna be the 2025 and ring us into the new year, make that next prompt. But like when somebody comes in, rather than rather than dumping thousands of prompts on them, we give them a good sort of baseline starter uh uh set. And then through the chief officer community, whether you're an AI insider or you're you know in the alumni program or you're you know getting certified, you get access to these prompt packs. So every month the team is you know fine-tuning and tweaking on various departments. Like this month we'll be totally focused on accounting and you know, we'll make three or four different prompts for like advanced accounting and tax preparation. And you know, it's not to say that everybody needs that, but there are some times where either you that's your niche, right, and you're moving from a CFO role into a C AIO. And so that's like, oh, this is my this is my sweet stuff. I'm gonna take this all back to my, you know, my cohorts and my team. And there's times where you're an AI consultant, you're like, okay, we've identified and we went in on the beach head of marketing, and now we've tackled HR. What's next? Oh, let's go talk to accounting. Hey, we're gonna roll up with you know, three or four of these premium prompt packs of 30 or 40 very well scripted and optimized prompts, and we're gonna release them and get them all super excited and how to use them. So um we just continuously release these you know premium prompt packs that are available. Usually there's about 10 to 12 new packs, and they might range from like 10 or 15 or 20 prompts in every pack that are themed. We've done business negotiation, we've done customer service, we have a ton of them, obviously, in marketing and social media and content creation, um, strategic planning and all those different areas. So that's that's been another uh interesting space.
SPEAKER_01So uh just to give a recap, because my next question is gonna be what's on the roadmap for 2025, but just to give a re recap of the functionality. There's obviously over 2,000 highly calibrated prompts specific for business application. So any department, any role in the company can find some usable prompts in there that aren't just those one-liners like they get from uh X or Twitter or LinkedIn or whatever. We have the ability to take those same prompts and use the parameters or the variables specific to my company. The prompt may have been built for one industry or one activity or product, but because of the way that we've structured the prompts and the use of it in promptable, we have the ability to say I like to truly customize it for your specific use case, which is fabulous. I'm not saying that anywhere else. Additionally, that's two, the ability to take that same prompt and without leaving promptable, test it in chat GPT, chat GPT 3.5, chat GPT 4.0, Claude, Gemini, Malama. And we also have so for those listening, you're able to run it. What does ChatGPT say? What run it again? What does Claude say? So it's not like you're bouncing back and forth about amongst a bunch of screens, you're able to actually uh do all of that evaluation, analysis, and review right there within that same session within Prompt Vault.
SPEAKER_02And those are all premium accounts. So like even if you don't have a paid version of ChatGPT or Claude, like a lot of people will go, oh, I got $20 a month, I'm only putting it on, you know, table number one. Like if they're a betting man, I'm gonna go with ChatGPT. But I can't, you know, foot the bill for our Claude, so I just only use the free version of that or whatever. So we have access to all of the paid models that are kind of built in with our API condition. So, you know, that's another aspect, but carry on.
SPEAKER_01If you're enjoying this episode and want to learn more about how to start using AI at work, we've made it easy for you. For just one dollar, you can have full access to the Chief AI officer community, which will give you additional training, custom software, daily training calls on AI tools, using AI automations, getting more from your chat GBT sessions, and the business of being an AI consultant. Simply go to chiefaiofficer.com forward slash insiders to accelerate your AI journey. Now, back to the episode. The prompt improver, which was I I looked. When I was first getting started with this stuff, I was looking because I I had all those lists of one-liners and I thought, this is crap, but I I'd like I like the concept or the kernel of this prompt. How can I I I looked, I searched, and there were a couple out there that were that would show up in search results when it came to prompt improvers. I'd submit the prompt, hit it, disappointment. If you hit it, time out. And if you hit it, that's not even accurate. Never could find one. So that I think that might be a um uh kind of a hidden gem within uh the prompt vault. But one of the things that I particularly like was when you introduced the idea of these prompt packs, and I think that what started that was we had some opportunities for kind of collaborations and education when it came to certain frameworks of doing business consulting or certain uh styles of doing things, and we built these series of prompts that were specific to that framework or whatever, and then we would bundle them essentially for the users of those, like scaling up was one of those people. I think that that's very cool. And at this point, any idea how many um prompt packs we actually have in there?
SPEAKER_02Yeah, we got more than 200 at least, you know, probably creeping up on 250. I mean, some of those are probably still going through QA, like I can see a giant list in the database of you know, phase one, you know, creation, and then they go through review, then they go through testing, and then eventually we release them into the wild, or we'll put them out in these, you know, prompt packs in the community. Um, but that's interesting because that idea actually originated. So back in probably the first 10 or 15 interviews, like before we even launched from alpha to beta, right? I was talking to a guy, he was at probably you know one of the mega AI online summits or whatever, and I happened to, you know, be dropping some mentions to people that were talking about prompting, and he like my ears to the ground. Like I remember when I first got into web development, like it at the beginning, this is like 1996, and people you'd be out there and you'd hear the word website, and he was like, Oh, somebody said website. There's an opportunity, and I like you know, hustle over to wherever that conversation was because it was not a term you heard very often. So, like if anybody mentioning prompting or how do I like relate it to that, I would drop a DM in the chat and say, Hey, I've got this tool. Are you interested? And I do these, you know, interviews. So it was a guy who was a copywriter, and he was like, Oh, I I've got this, you know, course I teach on, you know, I can see obviously Chat GBD is coming for my job, like me and whatever, a million other copywriters. Uh, he'd been doing copywriting for like 30 years. I mean, he was a total vet, amazing copywriter, and he had a four-module course that he was teaching. And I think it was inside of perhaps you know, the end of module one or module two, he actually had a series of prompts that he would, you know, give people to say, like, here's how you become a better copywriter. Let's, you know, stand on the shoulders of giants. Uh, you know, let's let's use um some of the best copywriting. Uh, I can't think of top copywriter off the top of my head, but uh John Benson. There you go. John Benson, of course. Uh there would be one. And there's many others. Uh so so you know, you you would whether you're doing copywriting for long form or direct response or you know, just short stuff for even uh Amazon descriptions. I mean, all of it is is different kinds of copy styles. And so he had a set of prompts and he would give some everybody access to a PDF. And it was like, okay, here's a 10-page PDF that just has, again, you know, the equivalent of these one-liners, and maybe uh I didn't I never actually saw his original PDF, but I saw what he ended up putting in the prompt vault, and I'm like, okay, cool. So you've got these prompts that you're giving them in a single-page PDF download at the end of module one. How's about we create this feature where you could put all your prompts into Prompt Vault, and then we'll make a special organization structure, which was a folder in essence, but we ended up calling it a premium prompt pack just based on kind of product you know labeling. And he could basically give that custom link to his audience. He can make sure all those prompts are private. Nobody can see them unless they are a paid user of his course. And when they you know click on it, they'll get their invitation and they'll be able to register to the back end of prompt vault. But then it'll inherit all of the organized prompts, whether that's 10 or 100, that he has structured for his students. And so that was kind of like that first seed of, oh, okay, how would we use this, you know, times multiple different use case scenarios? And then I think shortly after you were speaking at the World AI Summit, and we're like, oh man, here's an opportunity. We we need to like get out there in front of the audience and say, look, we're probably not gonna be able to do like a pitch, if you will, but let's make a mention of some goodies that they can gobble up and they'll definitely want. So anybody that's a warm body that's paying attention to the front of the stage is gonna want to write this down. So we put together like a massive prompt pack of like 50 prompts. They were all targeted around, you know, AI and marketing and social media and the whole nine yards. And then you were able to stand up there and like put a QR code on your slide and just go, okay, scan this and you'll get access to this premium prompt pack with 50 prompts. So those are different ways in which, you know, and and since then now we've got like so many of our graduates, they're taking their version of customized prompts that they've put in and they're doing that same thing. They're strategically going out there, like Terry, she's out there talking in front of audiences every week, whether it's 12, 15, 75 people, and she can, you know, throw a backslide on there and go, if you want access to these prompts, you know, I know we've had other people that have even built custom landing pages with an opt-in that says, Harry, if you want access to prompt vault, go ahead and do this. So um, it's become a little bit of a lead magnet for other folks, you know, who are building their own list in their audience. Um, but yeah, that's been tremendous. The other thing that's kind of interesting and worth pointing out is, you know, in addition, in terms of like where I was bringing the vision of both the coupling of Prompt Vault and what prompting meant was this uh smart cuts. So the idea of people sharing strategies on how to use AI uh was not uncommon. There were a lot of people out there, you know, internet marketers and all different kinds of niches where they're teaching things. The term like AI playbook was often being used and a couple other, you know, uh AI strategy thing, whatever. There was a few, a few different common labels. So we decided to, you know, go and register the domain name Smart Cuts and trademark that. So we've got this idea of what is an AI-powered shortcut. It's a smart cut, right? So it allows you to, and a smart cut is essentially like a process map or like an SOP standard operating procedure for how somebody in the role of X would use any tool across the landscape that's you know in the Gen AI or otherwise space, but not just like, oh, here's you know, 100 tools, let's go make a little like glorified tutorial. But not, you know, we wanted to make sure we were coupling that because we already had developed Prompt Vault, that every single one of those smart cuts was married between an AI tool and prompting. So you would say, all right, let's take a tool like 11 Labs. Okay, what is it? It's categorically an audio tool, right? It's voice to tech or it's text-to-voice first. Then they have, you know, developed voice-to-voice and voice to text and all kinds of other little spin-out features. But how, you know, how do you create a prompt that will couple with that service? Well, you could generate, you know, a script for a podcast ad, for example, or you know, any one of a number of different things that would come in there. And then you take a different tool like Fathom or Otter or, you know, any of the other note takers and say, okay, well, how do you marry prompting with that? All right. Well, you've just had a meeting and now you want to do uh a meeting analysis or a sentiment analysis, or you're doing a sales call. You know, was there you know, uh buyer uh intent based on the dialogue that was happening between their sales rep, you know, and the customer, right? So we'd see we'd create a series of prompts for that. And so the the smart cuts are a way to kind of give people, as a again, going back to our customer, you know, ICP um ideal customer profile, it's somebody who's an AI expert, who's an authority on AI, who resonates with that consulting, whether they're fractional or chief AI officer, and say, I don't necessarily need to have read 200 of these smart cuts. But what I do know is that I have a whole arsenal available to me anytime I need to, that's categoric by department, by marketing, business, image use case, video use case, et cetera. And when I go in to do a consultation with a client, when opportunity starts shining like a beacon over there as a pain point, I can go into the back trunk here and say, ah, we've got three or four smart cuts that are specifically about HR-related AI tools with the relevant prompts. I can drop these into the team, have them begin flushing out and using these tools. Um, so that's been a real interesting thing. So inside of actually the prompt vault, we have kind of the smart cut logo as an identifier that's just kind of a watermark in the background of you know dozens of these different um prompts. So at the back end of every single smart cut, you'll actually be able to click on a resource that says, you know, I've just finished reading this whole SOP on how to do X. I can follow along, paint by numbers, but I can go click on this, instantly add that into my vault, and now I've got another tool in my arsenal. So that that's kind of a cool spin that we married, you know, into um another educational function of how do we level up the AI and the prompt engineering skill set.
SPEAKER_01Now, this is great. Somebody's listening to this, I'm sure and they're like, I want to get a hold of that. Uh you can't afford it, trust me. No, I'm kidding. Um it's free. So if you're in our community. But one of the things is that's great, I can do it, but I work with a team. I've got lots of people that I want to make sure that that they, if that we can collaborate on prompts, that hey, what was that prompt you used to create that that report or that piece of content? And one of the things that you guys launched with that product was an enterprise grade version of that that's usually only made available to chief AI officers, but talk about that a little bit.
SPEAKER_02Yeah, so that's actually one of the things that you know we did launch this year and are putting a lot more effort into, you know, going into 2024 in terms of the product roadmap. Um, the concept generally being, okay, I'm a single user, I get my you know, prompt vault. And for those who are listening before you know the episode hits stop recording, we'll definitely give you uh a special URL to get your invitation uh through the side door. Uh, but essentially the the idea was, okay, now it's more than just me. I'm going into either, let's say I work as a paid, you know, full-time or fractional C AIO within a single organization. Um, I've you know tuned up a bunch of prompts. Maybe I, you know, in addition to I can input and create my own, I can optimize some of the stuff that I work on. I could optimize things for other teams and departments because maybe they're not as skilled at that, you know, fine-tuning as somebody who's playing around in AI all day. And then I want to be able to share that. So on the one hand, you know, as a single user, you can still go in and say, here's a you know, uh prompt that I've optimized. Let me go ahead and grab the little share link and slack it over to somebody. I mean, that's obvious. Anybody can do that. You can obviously make it open to say, I'm gonna make it public, or even we've got a level that says like private, public, open source. The difference with open source is any other user could take yours and make a cloned copy and then they can modify and tweak it to their nature. Um, but the the enterprise was like, how do we make sure that teams can have they can still have that private layer where this is their, you know, their junk, their secret sauce? They don't want to like make it public just to be able to share from one department to the other. Um and you get, or you'll get the C AIO, somebody who's like, look, I can look across my arsenal of like 500 or a thousand prompts that I've inherited through all these prompt packs, and I'm gonna cherry pick the ones that are gonna be the best for you know this particular user or this department, and I'll just go favorite a bunch of them, throw them into a folder. So the whole structured folder organization is there at the enterprise level, and say, now I can invite in my team and say, here's a folder for marketing, here's a folder for customer service, here's a folder for legal HR, whatever, right? So each of those become an organized structure that allows you, including we played around with some stuff like tandem work where two people can be logged in on the same prompt and actually seeing the output of you know what's happening while one person's massaging them. That's some experimentation that we've kind of been working on. It's not necessarily uh shipped into production yet. But that folder permission is at one area, um, that sort of advanced level organization, the peer-to-peer sharing. Um, one of probably the most interesting, and even when you get into like how do you scale across teams, there's a lot of things that you don't think about. But as a product manager, you know, we have to look at those layers and layers of different permissions to say, if I invite this person in and they leave the company, how do we revoke access? And they don't just take like all the prompts with them, or how do we shut off access to these prompts? Um, you know, what happens if, you know, we we've got a model that says anybody with you know domain at Coca Cola that can just come in. We can give like, as opposed to individual invitations, like you could hardwire a custom link to any individual, invite them through the back end with an email. Or you could say, I just want to create one user evergreen link that I can throw on a Slack channel, pin it. And everybody, even if they come into the organization three months from now, can go click on that as part of the HR onboarding and they're inside the vault and they've got access to everything. So those are kind of the things that we've been working on. But one feature that's uh, as many people know, we've not moved the free line. You know, we've we've kept it as you're either inside, you're an AI insider in the community and you have access to this, um, or you're you know an alumni or graduate or in the program, then you'll obviously get access to all of the things that we have with that. But we're uh in development right now for the idea of probably one of the most requested features of the year is this threaded conversation. So if you're a user and you know and you're listening to this, you're okay, yeah, I always wish I could just go and continue the dialogue beyond, you know, run Llama 3, run you know Chat GPT, just keep running the model. But I want to like bust that into a longer threaded conversation where I can follow up. And the answer to that has always been all right, we we've seen that we're not trying to recreate the entire chat user interface to create like prompt vault is now you know the competitor of OpenAI. So we we've you know excluded that, but I've I've seen enough indicators for where that actually becomes useful for folks, especially because they can tap into all the premium models with one single subscription. So the idea in the future, uh, whether we get that out in Q1 or Q2, will be that you'll be able to take uh one of your sessions and say, okay, turn this into like into an actual chat, pull up a sidebar, and now it's that starter result, and you can just go and have like a half an hour conversation that's you know, saving the session, saving the dialogue, saving that thread. Um so that's gonna be really exciting. Another thing uh that's kind of slated on the roadmap, but we didn't really talk in this you know context about images, although you know business prompting tends to lean a little bit further away from images, but there are some use cases where images make sense. So we we natively are integrated right now with uh um with Dolly. Um you can store prompts with Midjourney and some of the other models, um, but we're gonna be bringing more image models online, you know, coming forward. Um so there's just a lot of you know little fine-tuning and UI enhancements and things like that that we're working on as well.
SPEAKER_01It's an incredible tool. And obviously, I'm biased. It's uh one of the tools that we created at Chief AI Officer, but I don't see anything else like that. I see prompt libraries and I see lists of prompts and I see notion databases and stuff like that. But this full kind of like prompt intelligence environment where I can go in there and I don't have to be a prompt engineer to be able to do some really cool stuff with my prompts. I haven't seen that out there, and it's one of the accomplishments, certainly on our technical side, that I'm most proud of. That that and the fact that we make it available for free is something that um we might want to have our heads examined, but so you know, we're we're um we've done a deep dive into the functionality of this, and and your work in this has resulted in you actually being the uh the prompt education lead within the community. You host a weekly call uh on Tuesdays where we cover um all kinds of topics related to prompts. So uh I'd like you to maybe talk a little bit about what how have you grown as far as your prompt capabilities by because I've always heard that you learn by teaching, right? You may you may know something, but when you have to explain it to somebody else, it forces you to really like, oh, how like think about it in a different way. And I know that I mean you've done uh probably 30 calls at this point um with our community. What what are some of the lessons that that you've learned that that you thought, oh, everybody knows this, but when you reveal it, people are like, oh my gosh, that was incredible.
SPEAKER_02Yeah, I mean, it it always comes back to first principle, right? Like the the thing is, and and I teach on a number of different topics from different prompt uh frameworks and uh you know, working backwards, having prompt, uh using kind of that thought leadership, you know, having interview you. I'll typically go through like inside the community, like a six-week kind of revolving series because we get people in the course and they're like, there's some people that are brand new. And if I'm talking about pseudocoding and prompting, it's like, whoa, that's way heady. You know, so every once in a while I got to dial it back based on the cohorts and the the live groups that we're bringing in. But yeah, I've certainly become much more. I I don't know that I'd qualify myself as a I'm a prompt engineer based on my definition would be I gotta be a skilled Python programmer and a linguist like master. Um, but I'd say I'm I'm probably you know as good or better than most people at you know, getting absolutely you know, expected results. Um and it's those those fine-tuned problems that we see not on the front end, but on the back end of trying to develop software that creates prompting, that's probably been the hardest thing for me to solve and sharpening the axe. But in terms of lesson planning, you know, I'm always out there anytime there's a podcast or any kind of you know episode um and mention of new models, like I gotta stay, you know, up to up to date on that stuff. Um, if there's new ideas that people are coming around with, um, you know, the the idea of mega prompting was a big thing for a while. And then it's like you know, thought, uh, tree of thought, and then there's like chain of thought, and there's there's all these different kinds of frameworks that you can leverage um to get use out of. And sometimes it's just little tweaky things about the right word and um using using like you know, that kind of uh blackmail, you know, the like my job depends on this, you must give me better results. I mean, there's all kinds of gimmicky things, but I try and bring the best things that I learn, you know, throughout my you know, weekly and monthly journeys and bring those into the curriculum that I kind of teach within that uh weekly series. So uh I'm constantly, you know, it's it's a never-ending thing. And as any chief A officer, like we always kind of go, oh my God, you know, I've been spending 40 hours a week while not employed, like drinking from three fire hoses with AI. What happens once I'm actually employed as a chief officer? How do I stay in this, right? And and that's I think the biggest probably aspect of what the community does for us is you are it's it's a hive mind mentality. You don't have to be focused on 18 different things at once. You can stay as an expert in your lane and you know, plug into the community and know that you're never gonna miss a beat. If something amazing drops, you know, somebody's gonna be talking about it. And it's just a question of whether you want to pinch and zoom.
SPEAKER_01You know, and and for those of you listening who are looking to get more education on that, obviously you can join the the chief AI officer free community. Um, and then at that point you'll have access to some of these tools. Um but we do have uh a level of it called the AI Insiders, where you get access to six um live training calls per week on all kinds of different topics, including Segovia's call on prompting. But one of the other things that we've done this year is I think that we at at least nine times we've hosted a monthly AI challenge that we call the Using AI at work challenge. And it's intended to take somebody who's AI curious or maybe struggling with the the path forward on the AI journey for career advancement, job security, whatever. And we always end up teaching um some portion of that on prompting. But one of the things that you're known as, uh, I think at this point is the candyman, because anytime we do a challenge, one of the things we do is a reward for individuals investing in themselves and sticking around is we give them something every day of the challenge or every time that we do the challenge. And not to put you on the spot here, but uh since you are already known as the candy man for giving away plenty of smart cuts, prompt packs, all kinds of stuff that we do uh in that challenge for free.
SPEAKER_02Um any ideas or thoughts on something maybe you could Yeah, I would I would not be uh showing up live on any kind of call uh if I didn't have something up my sleeve. So uh one I I will just mention because I don't know that this is as obvious. And if you are listening to this in the first episode you've ever heard, you know, welcome to the tribe. Um as an AI expert, we welcome you into our kind of inner circle. Uh, but at the same time, every single episode, so these smart cuts that I was talking about, the vision was always if you're if we are talking and you are interviewing, and I I mean you're one of the top podcasts I listen to, and I listened to like 30 or 40 hours of podcasts, you know, covering product management, covering AI, covering like, you know, the end of the world AI scenarios and all of that mix, all of it together, just to keep myself abreast. And what's interesting is I wanted to make sure that every episode would have like super value. So we actually have our team, and this is not as obvious um as I'd like it to be, but if you record an episode with somebody and they start talking about whatever it is, XYZ tool, they're like the last one we did, I think last week was talking about um uh Light Trick, you know, uh studio, right? So while you you'll record that episode in advance because you know we don't want to leave anything last minute, and we'll the team will have anywhere from seven to 14 days to go, all right, we we just dropped, we'll be dropping an episode in 14 days from now. Let's go and make sure that we have aligned and created a smart cut specifically for that episode. Um and the guests, they typically don't know that. Like they're just coming on and they're just talking whatever they get into. Yeah. And somewhere along the way, without fail, there's always a mention of like this tool or these two or three tools. So we actually have our team, we developed a custom GPT that says we can upload an episode transcript and say, you know, create a smart cut out of this, or rather, give us the idea. It understands all of what a smart cut is. Give us the best ideas in a list of 10 that we could create, and it'll come up with some that are hallucinations that like have nothing to do with anything that you or the guest spoke about. But you know, that's where the human in the loop comes in. And then the team will kind of narrow it down and say, okay, this is one of the tools you guys went, you know, pinched and zoomed on. Let's make a smart cut related to that. So you can essentially go to every single one of the, this is episode 49. If you went to like episode seven or 13 or 18, you could go uh smartcuts.io slash, you know, episode number and be able to get access to the relevant related kind of smartcut for that episode. So in that, you know, vein, since you know, many people, if you're not watching on YouTube and you didn't see that in the show notes or on iTunes or whatever, you may have missed that. So I want to kind of draw attention there. And in the same way, if you were to head over to smartcuts.io forward slash 49, you'll actually get access to a particular smartcut that the team just finished going through QA like in the last two or three days. Um, and it's titled, it's titled AI Thought Partner for Business Negotiation. So this is again, in the month of December, we're kind of focusing more on you know strategic thinking around business planning as you kind of look backward in the rearview mirror across what you've been doing through the the last, you know, 2024 and into 2025, around what is what are the big rocks that we could be working on, quarterly goals, KPIs, OKRs, all that stuff. So the concept around the smart cut kind of walks people through an idea and a concept that is, I don't know why it's so uncommon. I mean, for people inside of our little universe, if you ask any of them, they'd be like, oh yeah, I do that all the time. But outside of our bubble, you know, in in the kind of uh AI elite, if you will, uh, a lot of people have never tried this. It's they they think about prompting and what they can get as either parlor tricks or just content generation. Write me a LinkedIn post of 300 words, you know, blah, blah, blah, you know, uh, clean up my email, proofread my X or Y, you know, look for typos, just silly things that are everyday use cases, but to turn the model around and say, act as a thought partner, you know, specializing in X. So your role is this. I want you to help me solve something. And in this case, it's it's a business negotiation, because of course, any, you know, the use case of our audience is I'm a fractional CIO and I got to negotiate my salary, or, you know, I'm an AI consultant negotiating with a client and I'm gonna, you know, not just use the traditional hours for dollars model, but I'm gonna work my way into a structure of I'm gonna get points based on percentage of growth or reduction of cost. Yeah. So all these models we kind of explore in a program, but this allows you to like turn it into an interview where ChatGPT is now asking you the hard questions or whatever model you want to use this for. And it's getting inside your head, and you're like, sometimes you'll go, oh, damn, I know the answer. I wouldn't have thought of that question. Yeah. And you just bang out a paragraph. But sometimes you get into this like funky space where you're six questions in, you're going, shit, I don't know. Like I'm I hit a wall, like, what do I do? So there's like a little some tips in there that kind of walk you through like how do you navigate around that and continue on to get the maximum value, and then summarize the entire conversation into something that's like the executive summary of what you need to go, whether you're building a pitch deck, presenting to a board, fine-tuning that you know, contract or whatever. So that's one uh kind of freebie that we can throw out there. The other one I had prepared is let's see if I can find so that would be smartcuts.io forward slash 49. And then I have, it wouldn't be fair if I didn't give you guys a prompt pack. So because this month, um, we're, you know, not only can you get like if you roll up to prompt vault.studio and it's not .com, it's just prompt vault.studio, you would either get to log in or sign up for the wait list. Uh you can go to prompt vault.studio forward slash begin. So it's just dot studio forward slash begin, and that'll get you an invitation to register for an account, which if you are on one of our AI challenges, we typically would give you that link uh in the chat. But I've created a special link. It's actually C-A-I-O.Cx slash strategy. And that's a link that will take you in and it'll actually allow you to inherit a prompt pack, a premium prompt pack of about a dozen prompts that are in that kind of AI strategic thinking and you know, different whether it's strategy on marketing or strategy on business or strategy on X or Y. So that's a compilation of some prompts that you can kind of roll up and get in the mix with. So hopefully that you know gives somebody, you know, who's listening something to uh to play with.
SPEAKER_01And we're gonna be putting those links for sure in the show notes. So, Sagobi, this has been fantastic. Prompting, but we didn't cover the whole topic of prompting. So I have a feeling I'll be inviting you back on to now that people understand that this tool exists out there, and this is kind of a shortcut to uh again from months to minutes, taking that learning curve, you know, you can use these immediately to uh become somebody who is getting ideal outcome from your prompting sessions. But for those individuals who are gonna be maybe building their own prompts in Prompt Vault, I'd love for us to explore some of the uh the frameworks that you lean on at some point in the future on a future episode. So we're at the top of the hour here, and I guess next steps for individuals who obviously go sign up for an account at Prompt Vault. We'd love to have you in the community for sure. But um, how would what would you suggest for people as far as next steps when it comes to becoming somebody who is prompt literate?
SPEAKER_02Yeah, I I think you know, to understand prompts better, you just need to go in and look at what other good prompts look like, and you'll start to go, oh, okay, you know, uh so so the first thing is you create an account, go, you know, read through if if you're going into this particular prompt pack, read through all 10 prompts, you know, spin the dice and get a couple of them, you know, going. You can you don't have to, so there's we didn't mention this, but inside a prompt vault with a free account, you get a thousand credits per month. Okay. And every time you want to like click, you know, use a model, you're basically using up a couple tokens. And depending on, you know, um the number of inputs and outputs, how long the prompt is and how long the output is, like a hundred-word prompt output is much less than write me a 5,000-word essay. Um so you kind of get these credits. But uh a thousand uh credits is enough for any average user in a month. If you're a power user, you're an AI you know expert, this is your you know, job role, or function. Um, you know, those who are insider, uh either AI insiders or chief AI officers, they get like 25,000 credits a month. But you can you know accrue more credits by just referring other people. There's a referral program, you can you know go into the settings and just say, turn that on and um share it with your team and things like that. So you're welcome to you know go beyond just your own usage. Um but I think you know, just reading through whether you're in the featured prompts or you're inside of a prompt pack, you can just click on the bookmark and go, okay, this one's good, okay, this one's good, and read you know, a few of them and then start bookmarking the ones that are gonna really work for you. You know, definitely, I would say within your first or second session, get that magical experience of let me sneeze out a one sentence and you know, try the prompt improver. Like don't wait, oh yeah, that sounded cool, but I'll wait three months before I ever do that. Because unless you plan on not using any prompting at all for the next three months and you're gonna hide under a rock, then the chances are you're gonna get mileage out of that. And just seeing how it's rewriting your stuff will wake you up to, oh, okay, I could probably have written that whole thing myself. And what will it do? You know, and and there's another tip. Once you get past that first layer of it's improved, there's a little icon on that kind of you know, create prompt interface. It's a little pink magic, you know, uh AIFI this, you know, further. And you can click that a second time or even a third time, and it'll just continue to expand and enhance. And you know, there's a diminishing point of no return where you like I I uh have my 17-year-old son in there, and he's just like click, click, click, click, click, click. What happens if I click this 40 times? I'm like, well, you know, it's just burning up all your credits. Uh but yeah, it's it there's a lot that you can learn just by observing what other people are doing.
SPEAKER_01Awesome. So everybody who's uh been listening to this, I'm sure at this point you're probably intrigued. Most of you, if you're listening and you're in our community, you already have an account with Prompt Vault. You're actively using it and getting the value of it. But for I don't care what level of experience you have right now with AI, this is a tool, and again, all the stuff that we mentioned, nothing besides the chief AI officer certification costs you a penny. So if you're not using these tools and and heck, sharing them with your uh, you know, everybody else who's on the journey with you and your your team, your company, whatever, uh you're missing out. So what I would suggest is at the end of this episode, go download an account. Like Segobia said, you'll be gifted a thousand free credits every month to use this tool. There is no cost. Uh, and you're you the benefit that you gain from the use of Chat GPT, Claude, whatever your model of choice is, is gonna increase significantly with just from listening to this one podcast and using that one free tool.
SPEAKER_02So that's one more tip. Let me let me throw in one more thing. So yes, I will just say we're recording this on a Thursday. So two days ago, inside of my Tuesday call where I'm teaching, I was just thinking about there was a use case where somebody we like to share a lot of the wins, like, you know, different calls across the week. People will kind of ask for like, does anybody have any wins? And I usually try and zone those into, you know, not just wins like I want a new thing, but how does it relate to prompting? And I won't mention the name, but somebody was in there and they're saying, I just had a win. I was, you know, talking to an individual, sat down with them, showed them a couple things, and shared over my shoulder just a screen share in a kind of semi-casual environment, prompt vault, show them around a couple of things, clicked on, you know, the results out of one or two of them, and they were like, Oh my God, mind blown. And there's all these other ones, like I could go categorically into this, that, and they're like, okay, sign me up. I need, I am contracting with you. I need more training on this, get me an account. So immediately right there, and then she's like, let me give this uh access to you. I'll share a special link. You can sign up as soon as we're off the phone and we'll book a session for a paid consult, you know, uh like in a couple days, and then we're gonna bring in a few other team members. So there's a lot of creative ways in which like we keep seeing our you know certified AI experts using it in different contexts, whether it's speaking from a stage, you know, teaching courses and delivering prompts to those people, you know, sidecarring with executives on how to actually you know do prompting, because many executives they're like, yeah, our company does AI, but uh, I don't know about me, like not so much. Like there's so much that can be done. And this is just it's a good opener, you know, uh to a conversation.
SPEAKER_01It's a great tip. Awesome. This has been, we I'm sure we could keep going on this as well. And like I said, there's going to be a part two of this uh at some point in the near future. So for everybody listening to this, thank you so much for joining us. And um, we will see you on the next episode. Thanks for tuning in to using AI at work. Don't forget to subscribe for more conversations about how to use AI at work. And a special thank you to our sponsor, Chief AI Officer, for empowering businesses with AI education and training.
SPEAKER_00Visit their website for a free AI readiness assessment and AI strategy guide to help you get started using AI at work.
SPEAKER_01That's www.chiefaiofficer.com. Well, thanks to our producer, Evan Desalnier, for making this episode possible. Follow us on Twitter at the handle using AI at work and visit www.usingai at work.com for free resources to help you harness AI in your role.