Serial Entrepreneur: Secrets Revealed! EP99
G P T four Scale Startup with ai. Episode 99. Yes, that’s what it is, episode 99. We’ve done 99 of these episodes, and let me tell you, there have been some really memorable episodes. If you check on your favorite podcast network, you can find some of those episodes. Uh, I think one of them we had in the late December was with the founder of Reebok, and that was incredible.
Uh, also we had one with the author Jeffrey Moore, who wrote Inside the Tornado and Crossing the Chasm, and I think that’s particularly relevant to the advent of AI and G P T 3.5 and now G P T four. Because those who [00:01:00] can move the quickest and fastest in a wave can benefit the most. You’re on Startup Club.
We are almost 1 million members strong. If you’re listening to this in podcasts, you might not know this, but this is actually a live show on Clubhouse we do every Friday at two o’clock Eastern, and you’re welcome to come on stage. Our club is run by our CEO and co-moderator here, Michele Van Till Borg.
We have Mimi OER who writes the blog, and this is Startup Club and I’m calling C Campbell. And the show you’re listening to is Siri, entrepreneur Secrets Revealed, and that this show, we’re trying to crack the code, we’re trying to figure it out how we can help create entrepreneurial wealth. And we’ve, again, we’ve had hundreds of interviews on this show already.
We’re in our episode number 99. And we are really trying to figure out what it is that serial entrepreneurs do to start scale, exit, [00:02:00] and repeat. Today we’re talking about G P T four AI and really about scaling your startup using G P T four, Michele or Mimi. Do you have any thoughts about today’s show?
Well, I’m excited about it and I was thinking about this right as we were coming on. You know, I’m really curious, um, how many people here in the room, how many of the members have used chat? G P T. And we really want to hear what your experience was, what your thoughts were, and any suggestions you have for other people.
So if you’ve even used it, you have some kind of opinion or questions, just raise your hand. We want, we wanna bring you up to the stage, um, you know, and really, you know, harness the power of the club to, you know, understand and help each other out so that we can all move to together forward. Thank you.[00:03:00]
Yeah. And if you’d like to, if you’d like the topic, uh, and you’re interested in it, feel free to share this. If you look at the, the bottom of the screen there, we have uh, the icon, the second icon from the left. It’s an arrow up arrow kit. I’m doing that now myself. And you click on Share on Clubhouse and it’ll sing, send a signal out to those who know you on Clubhouse.
We’d really appreciate you doing. Because it, it brings in more audience, more community. And that’s what this show’s gonna be all about, is really listening and understanding what it is that chat g p t can do to help scale your business. And again, if you’re in the audience and you’ve used it, we want to hear about your experiences.
I mean, they may have been good, they may have been bad. We don’t care. It’s open. It’s an open session. And we, we talk about all topics related to scaling startups and, and helping entrepreneurs gain wealth. So I thought, just for fun, I would ask chat, g p t that question I would ask. Chat. G p t [00:04:00] Give me 10 reasons, or sorry, 10 ways that, um, chat g p t can help scale a startup.
That was exactly the prompt I put into it and it came out with, with, um, some pretty interesting thoughts. First, market research chat. G p t can analyze large amounts of data to provide insights and market trends. Find that interesting, Michele, because we’re, look, we’re talking about the project at Startup Club to do a business plan.
Ai uh, solution for startups and market research is sometimes a big component of business plans. I’m not gonna read everything here, but I’ll just give you a few tidbits. Customer support, and this is an area where a number of our companies and our incubator have begun to use it as well, especially if you have support overseas where their language may not be their first language and you really want to improve their speech.
I was also able to use this to make the responses more positive [00:05:00] so you can institutionalize positivity in your email responses by using chat G P T. And the third one, I’ll give you right now, cause I don’t want to, I don’t wanna go through all 10 right now. We’ll break it up over the show. But the third one that I’ll give you that chat G P t came up with was content creation.
Chat PT can generate highly quality content, including blogs, post social media updates and product descriptions, saving times and time and resources. And I have a little story here, a little confession. Uh, I if you know, if you know the club or if you’ve listened to this show, you’ve heard that, uh, we signed a publishing deal with Forbes for the book start scale, exit, repeat.
Well, the book was complete in November and we were designing the cover and two weeks ago we had to come up with the inside flap of the cover. And so I took the intro from the book. Here’s what I did. It’s exactly what I did. I took the intro from the book, I cut and paste it. I put that into chat, G p t 3.5 at the time, and I said to [00:06:00] chat, g p t, write an inside cover of 10 reasons why someone will want to read this book.
And let me tell you, it was insane. It was Kim it with so many good ideas. Now I did some editing, did some changes to it. But, um, together working with chat G P T A project that might have taken me, cuz I’m not the best writer to be quite frank, uh, even though I wrote a book. But the reality is I did have editors at Forbes helping me, you know, with, with the book.
But, uh, but the fact is I was able to accomplish that task probably within a half hour. And without it, it would’ve taken me 2, 2, 3 hours easy and probably a lot more editing from, from other others as well. All right. Well let’s jump right into it. Uh, unless Michele, you have something you wanna add right now or.
Do you wanna just jump right into it and, yeah. And wrap, Robin? Yeah, I wanna hear what our members have to say. I mean, this is definitely a hot topic, and even if you have something that’s, you know, negative or controversial to [00:07:00] say, just come up because we, we wanna really hear and have a good, meaty conversation here.
So let’s go to Edie Ededie, I think I’m saying your name correctly. Um, tell us what your thoughts are, you know, how you’re using it for a business or what your experiences have been. Thank you, Edie. Hi. Well, thank you for bringing me up here, and thank you for this great conversation. Um, you know, you ha I raised my hand a little prematurely, , to be honest.
Um, but I had, you had asked if never, never, we love it , but you had asked if anybody had used G P T and, and of course I have. I’ve actually been like glued to it. I just find it amazing and, um, and does a much better job at, um, making my thoughts cohesive. I tend to be a little bit wordy and a little bit too flowery.
And, and maybe I’ll, [00:08:00] I’ll start, um, a thought and then I’ll, I’ll deviate. I’ll go someplace else and then I’ll come back to it. And then, and even when I read some of my own writing, I like, well, why can’t I just keep it like all together and. , it really has this incredible ability to just put everything together for me.
It organizes my thoughts better than I ever could. And, um, and I am starting a new business. And so, um, and I’m going to a, a university local here and have having like a little booth because I’m looking for people to, to come and join the company. And, um, thought that would be a great way to sort of start and find people, um, who want to work with me.
And so I needed to do it relatively quickly. And, and, you know, I think for, well, for me anyway, I get a little bit nervous about, you know, putting things out there in the world and I, [00:09:00] I re-edit and I edit and I edit and I re-edit and I re-edit. And so I, I wrote up a little bit brochure. I asked, uh, Chad, g b t to, you know, to make it better to, and know.
I put in a, a couple of words and things that, that I needed, uh, um, for the type of people that I was looking for to join the company. And, um, and it did it a miraculous job. I was like, this is so easy. Like, I could do the work, any work that I, that would take me days to do. And, and just really in moments, it’s, I think it’s brilliant.
I re I really do, and I would love to be able to get onto chat G B T and tell me how I could make, you know, millions of dollars in a startup. You know, company. So great. So if you guys have any ideas or if you’ve asked the questions to chat g p t and actually came up with some real valid answers, I’m, I’m all in and can’t wait to hear what you have to say.
And, um, and that’s it. Thanks for your time. Yeah, that’s, that’s [00:10:00] great, Edie. And, and, uh, it was a pleasure to have you on stage and I’m following you because I love to follow people when they come on stage. Uh, there’s so many interesting people in this community and it’s, it’s, it’s great that you shared those comments with us.
I, I will say that, um, you know, being an entrepreneur, a tech entrepreneur for almost 30 years now, uh, we’ve seen technology change quite significantly over that period of time from dial up to broadband, social media. We’ve seen technological changes, regulatory changes, and the opportunities in the past have been enormous for startups when there is massive change.
I think the 2020s are really gonna be all about ai, and we’re at the very, very beginning. So this is a historic moment in time. This is an opportunity for entrepreneurs, for startups, for founders to think about how they can use this technology not only to help the startup. They could al you might already have a business, right?
[00:11:00] Uh, and, and you can use this technology to, can help your business, absolutely. But you may also be able to base a technology or set up a company based on this business. Even my daughter, over Christmas, we got excited about this and she, uh, put together a poem, ai.com and it’s, it’s really just a poetry writing service that, um, you can have your poetry on Canvas if you want to write a poem about your dog or for your mother or something like that.
The reality is I don’t even have the capability of writing a poem. , I mean, period. But you could enter in certain prompts using prompt to engineering. And she was able to put together this website poem ai, and now they can produce these amazing poems that you can put on canvas. That’s just one idea. I’m not saying like that’s the, the biggest idea, the one that Michele and I are, are bouncing around here is business plan ai, uh, and business plan ai.com, which will be on startup club.
You’ll be able to go directly to that. [00:12:00] And if you wanna write a business plan, and I don’t know how sophisticated we can get with this, Michele, but I think we could actually go as far as, you know, helping to write private placement memorandums. You know, I think we can go really far, far down the path.
We’ll start with the, the, the basic business plan that you might show investors. And all by answering simple questions, you know, who are my competitors? You’ll list the competitors and it’ll do descriptions of the two competitors. What are the risk factors of this type of business? It’ll list some risk factors.
Those kind of things can be added into the business plan and, uh, it’ll look quite sophisticated, even though. Uh, the amount of effort you put into it may not, may not deserve the level of, uh, the level you’ll get, but it’ll help you raise funds for your company. So, Edie, there are so many business opportunities.
Keep your eyes open, keep listening, keep joining. You’ll see opportunities wherever you go where they’re not utilizing this technology, and you can jump in and build a company. All right, Michele. Good? Yeah. Awesome. You know, and something I [00:13:00] just wanna mention, um, to the members, something that I found valuable is I go to Product Hunt.
I don’t know how many people go there, but I find like there’s so many cool ideas and things that people are doing, and I’m seeing a lot, Colin, a lot of people are doing some really interesting things with chat. G P T. Uh, and I’m gonna mention just one that I saw the other day actually. It’s, um, called nara, N a R a.
And these folks are doing a nutritional assistant using chat g p t through sms. So I thought that was a pretty cool use and um, I like those kind of businesses, um, cuz those are kind of businesses that, you know, maybe it’s not as heavy, uh, investment as like a big platform or some very sophisticated software that you could get up and going.
That’s one of those areas you kinda look at it, there’s competitors out there, but it’s like all [00:14:00] manual. Imagine if you can really start to harness chat g p t to actually improve people’s lives. So I thought that was a really cool one that I just recently saw, but, um, let’s go to Lucas. Lucas, looking forward to your thoughts on this topic.
Lucas, your um, mute is on the bottom right. We could come there. No, there we go. I thank you so much. Um, you bet. First of all, thanks for bringing me on folks, and, uh, thank you, uh, ededie for sharing. Uh, so I am, I, I’ve been an entrepreneur calling on commercial real estate for quite some time in the sustainability space.
Um, some of my clients are, you know, some of the larger commercial real estate companies around the country. Uh, during the pandemic, of course, uh, that business slowed down almost to a halt. Um, and I was able to pivot into the medical supply space because I [00:15:00] had all my, these deep roots in, in Asia and other places.
So I was able to, and, and of course I did that just to help my company keep, uh, af. This is before P P P R E I D L. Um, I was kind of early. I, because I get so much of my product from China, uh, I was in contact with them in December of 2019, January of 2020. So I had enough foresight, um, with boots on the ground to know that.
This, this pandemic might turn into a thing. So let’s let, let’s forward. Fast forward through 2022 and kind of got out of the, uh, medical supply space, uh, but did very, very, very well, uh, during that time. And as I was kind of rethinking the business that I’ve been in for 15 years, um, the natural progression was, uh, to just, uh, cause I’d never been a wholesaler before.
I’d always bought my supplies directly from [00:16:00] OEM and ODM manufacturers to give myself a competitive edge, uh, for when I would bid out, you know, l e d lighting and other, uh, energy efficiency retrofit work. . Um, and so as I, so my, my quick win a after all that background was just that I was able to create a, a bunch of everything from my pitch decks to some, so to some, uh, ClickFunnel, uh, lead funnels to um, you know, just some of the marketing and wording around this kind of new business.
Cuz it is a very niche space. Um, I really am only talking to, instead of the asset manage management side of portfolios, I’m now engaging the ownership side of those portfolios and owners of, you know, billion dollar. Commercial real estate portfolios, you have to talk to them about noi. You have to speak very specifically to the things that get them excited.
And like, uh, I think it was Michele Shared, uh, you know, my writing is, is okay, but [00:17:00] it, it, it, it doesn’t always. Hit where it needs to hit. So, uh, a quick win is that I, I landed a, uh, just this Monday after my first meeting landed a, a first, uh, very large account with a, a company called Greystar, which is one of the largest apartment, um, developers in the country.
Uh, um, and so it’s been fantastic. And I remember that first night that I was, uh, uh, up engaging with chat g p t. It was only, it was about two and a half months ago, and I literally was just asking every kind of question. Um, it was, it was addictive. I think Ededie shared. It was just super addictive. I was up from, from 12 o’clock at night till about five in the morning on my phone, just, just going crazy.
So now I, I am a subscriber. Uh, I, you know, and I, you have slightly more access. And the one thing I wanted to comment on, Or is something that Colin said is that I’ve just really followed every kind of influencer or people that are speaking about this on social [00:18:00] media. Um, and the, the biggest, uh, takeaway I’ve gotten so far is the inputs or the prompts that you give G p t definitely show a completely different output.
So I would encourage everybody listening to really think about the inputs that you’re putting in.
Yeah, it, it’s a real, um, art in a way, right? Like those prompts are so powerful, and I know a lot of people are out there sharing their advice, but one thing that we’ve seen is if you’re writing, you know, you’re requesting something from chat, G p t, say, write it like whatever, a Wall Street Journal. , um, you know, journalist or write it like an, uh, you know, c m O or a SEO experts and you can get back as you said, you know, just amazing results.
All right, [00:19:00] we are on Moon, rock, moon. Hello. Hi, how are you? Yes, I’m fine. I’m Munira from Nigeria and this is my first time joining the space and I’m so happy I’m here because I just finished a 16 pages video production plan using chat G B T for the first time. And it is mind blowing, like on just how to cook Nigerian JLo.
So for me to be able to do this, I feel like this, this, this should be something every entrepreneur should learn how to use. On my chart, G B T I installed six different chrome extension, prompt chrome extensions, and one of the extension has 1,876 prompt. So it gives me, you know, it makes me everything useful [00:20:00] to almost everybody.
Like I’ve been having friends calling me, please, I want you to do this for me. I want you. And I’m like, whoa, am I this smart? No, I’m not this ai, but everybody, so bt imagine.
Yeah. I think we’re having technical problem here. Or it could be on our side. No, she dropped from me too. Oh, she dropped for you to, okay. If you could try, we really want to hear from you cuz you got some really good stuff there. If you could just try to get a better wifi and we’ll come back to you. Just give you a minute or two.
We’ll come right back to you Monterey. And, and James, you’re next on the list. We’ll go on round Robin here and uh, we want to hear from you, James. I mean, what’s your chat G p T story? Yo, super excited about AI in the future. I think it’s gonna be a major influence. Of course. Been super. [00:21:00] Um, yeah, keen to just get involved as much as possible.
It’s been kind of on my to-do list in life for a little while and I was gonna ask for some guidance. I know there’s, um, it’s been mentioned there’s a lot of influencers and resources. Can anyone suggest like one starting point for me to kind of get an edge into, into the world of CH G P T, where I should start or what I should do?
Or just one tiny little step? I know there’s no right and wrong, but any suggestion would be welcome. If you don’t mind me asking, what has been your experience with it so far? So I’ve not, I think I downloaded a watch app, which didn’t really work so well. I’ve used some other, um, AI search engines, but I’ve not got any experience with chat GT specifically.
Um, in terms of roughly where I want to take [00:22:00] it, I do most of my writing with old school typewriter. So I want to kind of take just the explorative writing that I do creative writing and start to see what I can do with J G P T, but I’ve got no action experience so far. So, uh, Colin, do you mind if I, if I, no, I was about to give you your moderator button actually.
Just go for it, Andrew. Go for it. We, we love help. You can tell. Yes. . All right. Well, hey, this is a, I love this question because, um, because you gave a little bit of insight and you’re also showing kind of how initial user engagement goes into it. So the first thing I wanna put out there is, while there are a lot of apps that say that they are, uh, G P T based apps, which isn’t an untrue thing, they are often going to be apps that utilize open AI’s api.
What that means in, um, in general terms is everything that you submit through that app, they [00:23:00] will. They have, they might have their own data retention policies in regards to it. Um, they might, so if anyone is using that app specifically, they be in whatever company is utilizing the api. It is not open AI that is doing that.
So if you are getting started, and especially if you’re putting anything creative, um, or anything that has any sort of intellectual property associated with it, um, understands that OpenAI does use everything for training purposes. Um, large language models require an enormous amount of user input. So they are looking at all of the data that is coming through.
By using it, you are agreeing to it. So make sure that whenever you’re submitting, you’re comfortable with it being reviewed by the engineers. Now if you use one of the apps that you see that utilize a wrapper, understand that some, another third party entity will have access to whatever you’re submitting through.
Very likely they’re keeping full logs of everything that’s in there. So [00:24:00] that would be another, um, another point there. Those first two things are just a couple red flags to throw up just so you’re aware of what’s being done with your data. Um, and the place that I would recommend if you wanna experience chat, G B T and it’s most raw form, they purchased the domain ai.com.
Simple as that. ai.com, we’ll link you directly to where chat GBT is. You can register on it, um, and use it for free. Um, there is the paid plus model, which you don’t have to worry about right away. It just provides faster results. And it will also give you, um, early access to G P T four. And I actually just like 20 minutes ago, got my, um, a p I access for G P T four.
Um, and I’m kind of excited to test this out. Um, I’ll get to that when it’s my turn later on though. But, um, what I would do first and foremost, just play with it. Ask it questions, ask it silly questions, [00:25:00] and just to get a sense of how it responds to you. How, how does it behave when you start to see that and you’re, and you’re putting stuff in.
You don’t have to follow a guide for that, just to get a feel for it. And the follow up thing that I would recommend is, you know, search. Uh, chat G B t GitHub and you’ll be able to find different results that will correlate with how you can approach props. Because the first type of props you might want to try is, you know, adopt the persona of, um, a business executive Q hates this business plan that I’m proposing and provide like different bits of feedback, you know, et cetera, et cetera.
Try out the personas and see how those work. And that’s gonna start to give you a sense of, all right, now that I see how these different pieces work, you can put it together and create different prompts. Like, I have these very, very, very long prompts that I use for deep [00:26:00] analysis of large blocks of texts, but I just sort of evolved them over time, all from just playing around with it.
So avoid the apps. Um, so the apps that you see on the app store, avoid those, um, ai.com, register on there, play around with it. And at some point when you’re more comfortable with details, start to learn the api. And then you, you can even figure out ways of creating your own programs, even if you’re not a programmer.
That’s G P T four has made a really big leap with that, but I’ll get to that at a later point in conversation. So my name’s Andrew. I’m done. Cool. Thank you. Thank you so much for that, Andrew. Hey James. Um, go on. Did I understand you correctly to say that you’re a writer and you’re like a creative writer?
That’s correct. Yeah. Um, one thing that my partner does is he is also a writer and he found it very useful to help him just fill in the gaps, for example, in a story arc and [00:27:00] in character profiles. So, I mean, just anything you can think of. So he was plugging in things just like, um, here’s a story arc. Um, are there any suggestions that you could give me?
It could be as broad and as open as that, or as Andrew said, very specific. So you. , you might say, I have X character, uh, they’re going down this path, et cetera, et cetera. Like, build a character profile for me. And it will, and you know, obviously it might not be as creative as you might be yourself, but, um, you know, to our experience, it, it gives us ideas, right?
Like, oh my gosh, I didn’t think of that. And that’s very cool. Like just a way to fill it out and supplement in a very, very efficient way. Thank you. I, I also think, Michele, if you’re, if you’re not a good writer, but you have an expertise, you can use that [00:28:00] to help create webinars or white papers or, you know, even if you write it yourself, it’s in a raw form and then you feed it through.
Um, it can actually improve the style. Now again, I like what Andrew, you said there about persona, so it doesn’t always have to sound like it’s, you know, written by a corporate executive marketing person. It can, it can, it can, it can be a little bit funny or it can be have a little bit edgy or it can, you know, be written the style of New York Times reporter.
Um, I think that’s interesting as well. Well, let’s jump to Monica. Monica, thank you for being patient. We would love to hear from you in regards to how G P T has changed your life. . It’s been a time saver and as soon as I saw the, um, the, the title I said, wow, yes, Che, G G P T has been my, like, go-to for everything.
In fact, I wanted to share that we were looking for [00:29:00] a content writer because, you know, my staff, we were all busy and we were, we were doing things off the side of our desk. We decided, To use chat, G p T and Jasper, because I had both. But, um, we don’t need a content writer anymore. And I wanted to say that because it, it was just, it’s a, it’s a savings to us, but, um, it’s a savings to me and my time, but also for my, my side hustle, I should say.
Um, I use it as well, and it’s just wonderful to use. I mean, it’s, it’s probably the best thing I’ve ever had happen to me. And I, you know, I wanted to say that because I think everyone should use it. Like, James, you’re looking for, you know, writer. Now you have to be careful. I, I wanted to say that as well because I think sometimes, you know, there’s some plagiarism or something like that going on, but you have to read what you’re writing, but it, you know, what it’s writing, uh, what it returns, but it’s a godsend.
And the time saving [00:30:00] that I have, uh, for blog posts, uh, copy for websites and social media is. , it’s been amazing. Um, the unfortunate thing is I’m wondering what will happen, and I don’t know if you guys have touched on this, what will happen with content writers at this point? I mean, there, there was, you know, a huge influx of, of that need at one point, and now I don’t even know if that’s something that we actually need
So, um, but that’s my take one. Uh, chat. G p t, I just love it. Yeah, it is, it’s sometimes, you know, these technological changes can be devastating for certain industries. Um, and, but yet unleash so many opportunities and it’s, I think as startups, you know, we’re looking, we’re looking at those opportunities and that same writer might take on a different role in a different way, your organization because you’re successful and you’re able to hire more people doing other things.
And that’s [00:31:00] typically the way we’ve seen it evolve over the last a hundred years, ever since the day. The buggy whip went, uh, uh, became, you know, um, obsolete by the car. Uh, we’re seeing it happen over and over again, even with electric cars and every technology that we see out there. It has a way of eliminating jobs at the beginning, but then somehow unleashing through entrepreneurs and startups, unleashing a power within the, a community that can actually build and create amazing things that can change the world and actually result in hiring more people.
I, I know that’s sort of a convoluted argument. Well, we won’t turn this into a political discussion, but I, um, I wanted to share the next three that Chachi bt since chat t has, has joined our show today. The first three were market research, customer support, and content creation. The next three, I’ll give you right now, I did 10.
My, my query was, you know, tell me 10 uh, ways startups can use chat g p T to [00:32:00] scale their business. And it came up with ten first, first one being market research, the second one being customer support. The third one being content creation. The, the fourth one is lead generation, and I think that’s very interesting.
We’re always looking for more leads. Chat. G p T can help startups identify potential leads, generate targeted email campaigns, and increase the conversion rate of marketing campaigns. That’s interesting too. Um, it’s just increasing your, your, uh, conversion rate for a website or even for a marketing campaign.
The next one was personalized recommendations Chat. G P T can analyze customer data to provide personalized product recommendations, improving customer engagement and retention. So we were looking at, um, potentially doing an experiment with Dolly where we have a product. We have a company called pod.com and it’s, um, it sells dog beds and, and rugs and.
Blankets and, and, uh, car [00:33:00] seats, uh, and those kind of things. And we are thinking of doing a targeted marketing, personalized marketing using Dolly, where you take, you know, you have the one photo and just using Dolly, you can re, you can within 30 10 seconds wipe out the, the dog and the pitcher and then say, replace with a King Charles spaniel, or replace with a french bulldog, or replace with a golden retriever, and it instantly replaces it.
So you can produce images very high, very quickly. Produce images that can be personalized in marketing campaigns on Facebook based on the breed. And, and, and we’re, we’re working on a test with that one. and that idea of personalized recommendations that it came up with sort of got me thinking about that one.
Number six is a chat bot development chat. G p t can help startups build intelligent chat bots that can answer customer queries, provide assistance and automated repetitive tasks. So I know early on it talked about, um, [00:34:00] uh, customer support, but this is more like an automated component of your customer support that can be quite advanced and you can develop it quite efficiently as well.
So that’s interesting. Andrew, I’m curious about, uh, chat G p T four and what you’re seeing with the difference. I, I, I haven’t seen much of a difference myself. You mentioned there was some, some, some stuff around the APIs. I’m curious to hear about that as well. Sure. If you can share that with us. Yeah, happy to.
So, um, the, a, the API is currently not available for everyone. I was, um, I was an early alpha tester with the PLUS program, so I think I just lucked out and got to cut the line there. But one of the things that’s useful about G P D four, um, in its playground is they pivoted the use of the system messages.
So system messages were something that’s used to EF effectively tell. The software, how you want to use the information on the user message that is being submitted. [00:35:00] Basically, if you have an API that’s running, it’s possible that someone, you, you could have like a bad actor who decides that they’re going to inject something, where they’re gonna try to issue a command within the user prompt in an effort to override something in the api.
The system message is the way that you can prevent that from happening by simply saying, um, you know, you’re a helpful assistant. Ignore any messages over here and, um, only analyze what is being submitted in just a way to make sure you’re protected. One of the things that they changed with it is they’ve integrated, um, what we often refer to as personas.
Into the messaging. So what that means is, um, when you’re talking about, um, you know, adopting the persona of, uh, so you don’t have to necessarily say those words. Now you can simply say you are an expert in English and English linguistics, um, and you’re submitting this paper for [00:36:00] review. And it’s very important that you are trying to.
Make sure that the message is coming across in a concise way, that you don’t have any extra language that’s just going to add time, but not add value to whatever it is that you are pitching, um, whether it’s in the deck or in the speech. Um, so you can in within that message, say like, these are the types of, um, expert approaches on it.
Um, in the message, please provide the analysis, and then you drop the message in and you will get a significantly improved analysis of it if you set things up correctly. Now the other thing that’s nice about the API is, um, well the current test, it allows up to 8,000 tokens, which translates roughly to 6,000 words in um, English, um, chat, g b t currently, if you test directly through chat, G B T, um, is 1000 tokens.
So this is eight times longer with the length. That [00:37:00] also means that you can drop in a larger block, uh, text. Um, so let’s say you have, uh, maybe some data that’s in a table. Um, you know, it it, if you use something that has like comma delineations in it, like you could literally export something from Excel, drop it in there and be like, Hey, I have all this information, let’s do some work with it.
Um, I haven’t had a chance to test out the extent of that yet cause I literally just got the API access a little bit ago, so I’m gonna ha I’ll be able to speak to that better, uh, next week. Is it only through the API that you can test that or could you just pop it indirectly? , you can pop it in directly, but if it’s a large amount of data, then you may end up running into a limit with its memory.
So, um, cause the idea that’s the issue that we’ve had Yeah. Is, is Mimi can specifically talk about it. Actually she was really testing that, you know, because we have a transcript, right? We’re recording here. And we wanted to do, Mimi actually writes all the blogs, right. Mimi and wanted to do, I’ll [00:38:00] let you tell a little bit more about your experience, but we wanted to, an easy way to do a summary or a blog and kind of explain to us what happened.
Cuz maybe this is, uh, a solution now, Mimi. Yes. I’m glad to hear that about the, um, characters because, well, actually, when I was trying to figure it out, it was tokens, which was even different than real characters. Um, but yeah, I would try to feed the transcript in and then it would get, I had to do it over like 15 different entries and each one would obviously just summarize.
The previous entry into the blog post. So it didn’t come together cohesive or naturally at all, really. So I’m glad to hear that. I’m curious what you think of the new one. I’ll have a, I will have better insight on it, um, in a, in a week or so. Um, cause I do want to, I, I don’t wanna do like a limited amount of testing and then realize that I’ve, um, I didn’t give it enough, um, exploration in.
But this also does give you control over in the API refers [00:39:00] to it as a temperature, which is a value that’s set anywhere between zero and one. If it’s closer to zero, it’s going to provide very rigid results for what you’re looking for. So depending if you’re trying to generate ideas, maybe you want to go like closer to one if it’s really creative.
If you are, if it’s a very established business market and you’re trying to identify the most reasonable locations that you could potentially, um, Seek investors in, um, or potential sites. If you’re trying to choose new locations to open your business in, maybe you wanna set that closer to the middle, around 0.5 or even less than that.
Um, a lower temperature doesn’t mean the results are worse. You just have to think how much information already exists out in the world that can, um, that this could make a comfortable, reliable decision without it having to guess. And if there is a lot of information that it could have available, bearing in mind too, that this is only trained up through 2000, September of 2021.
Um, but [00:40:00] once this is more live and has access to live information, which well, we’ll see when that happens. That could be a, that could still be a long roadmap. But all of that aside, um, there is like a very, very slight pivot that I want to, um, introduce. And yesterday Microsoft announced the launch of. The 365 co-pilot and this announcement went.
It for anyone who’s following very closely. Um, we were very excited about it, but it was surprising how little of it was covered in the main media because the truth is what could potentially be offered in this product is the first time we could see an enormous scale of information. So what it will do, um, so the 365 co-pilot, um, it’s going to work with the various office suite products.
So, you know, we’re at Excel, PowerPoint, outlook teams, et cetera. Um, and there’s a [00:41:00] lot of automation, little small things that it’ll be able to do along the way. Um, you’ll be able to auto complete things better, your efficiency and excel. You don’t have to think about the exact structure of different formulas that are used.
Um, and if you’re not familiar with using Excel, it would be a great help in you being able to produce your own things. All of this stuff, great. It’s helpful, but not really groundbreaking. What is groundbreaking is the information that it uses goes through, um, what they refer to as Microsoft Graph. Um, and that’s something that, um, anyone can, can look up online to see what it is.
But Microsoft Graph is effectively where all of your data for your account is stored. So stuff that you’ve created in all of the various parts of office, including stuff like OneNote and notably SharePoint. So let’s say you have a significant amount of information. This could be stuff related to your business.
James. I’m a fellow writer. So stuff related to world building. You could [00:42:00] have a bunch of things that are sitting in one note, maybe you have like a folder, a folder structure set, set up on, um, a SharePoint of some kind. And if you have an active 365 account, all of the information that is contained within here.
Um, and one of the things Microsoft said is, Anything that is utilized, your Microsoft Co-pilot does not get used for any training purposes with the l uh, with the, um, um, with the models, with the training of the models. With that said, having access to all that information, that also means that your information with what Microsoft’s is saying is going to be secure.
There’s not gonna be engineers who are going to be looking at it. You don’t have to worry about it going through some third party. And if there is sensitive business data, this is the first real use case of it, um, at least from the open AI perspective that will allow you to use, um, To use like the full force of this and feel comfortable with your data.
But check this out. [00:43:00] Imagine you have all of this information that is located, um, within SharePoint. And bear in mind, we haven’t had a chance to, to test this publicly. This is going off of yesterday’s announcement and a lot of speculation. So, you know, we’ll see where, where it will truly land. But if you.
Let, let’s say that you have market researchers that pull up information on different cities, or you can even look at, all right, let’s look, what are the populations of the top, you know, 50 cities? Where are they now? Review the different entertainment venues that, um, that are there. How many are there? How many like maybe golf courses, specifically, you know, whatever your need is for the market.
And all of this is just stored as information in Excel files. You can ask it then in theory, if this works the way that it’s promised to. Looking at all of the information, um, that we have here. PR please provide like an analysis of which markets, um, Would be most feasible for entry based [00:44:00] on, you know, property value, et cetera, et cetera.
Try off, um, you’re not limited in characters. Whatever is on that SharePoint, if it’s um, five documents, if it’s 5,000 documents, all of that becomes part of the analysis. Once it’s integrated into that Microsoft graph, then it knows how to read work through that data. And if you’re like me and some of your organizational structures are absolute trash and you have stuff that is stored all over the place from what they’re saying, it can figure out pretty well a lot of the intentions that are there.
And we’ll see if there’s any feedback for questions. This could be. Really, really huge thing to talk about once it launches. So that’s a, uh, well, maybe not too quick, but that is an overview of, um, what copilot could be, which could be another thing that will better enhance our ability to utilize with ai.
My name’s I, I’m just gonna say that sounds so incredibly amazing [00:45:00] because just at a high level, what you said is it could take a massive amount, let’s just put it there, up to a massive amount of data that’s unstructured and structure it for, you know, just some amazing kind of output results set. You know, pretty much your flexible, infinitely, right?
Meaning not just adding numbers or, you know, answering one question, that’s a maximum amount of flexibility. It’s hard for, you know, it, it just boggles to mind and I’m super excited. You know, to learn more about this. Thanks for sharing that, Andrew. That’s really, really cool.
And anyone, you know, on the stage or has had companies, I mean, it’s not uncommon for entrepreneurs just to have, like you said, lots of, you know, data and documents, but they’re typically unstructured , which makes it almost impossible, uh, you know, [00:46:00] to, to manage or to extract in a meaningful way. So that’s huge.
All right, well that’s really cool. Well, do we wanna move on here? I know everybody’s been very patient and we have, uh, 15 minutes left. Dave, we’re gonna come to you in a minute here. We wanna give all a gun a chance here as well. Olgun, are you able to speak now? . Yep. There we go. Hey, sure. Can you hear me?
Yes, we do. Hello? Oh, thank you. Thank you for the opportunity to speak. Um, well, um, I just started using chat g PT quite recently myself. Uh, I come from the development of programming, background, software development. Um, and, uh, I discovered it, uh, around Christmas period last year. I started using it. I’ve be hearing [00:47:00] about it.
I, I, I’ve add, uh, insight into, uh, language modeling, like from, um, Google’s work on the transformational. Um, I was aware of land and now the palm. And, um, I, I’ve kind of come from trying to train the model sort of set of things, but I’m quite impressed with, uh, chat, uh, kind of their response. Um, I’ve began from chat G P T I’ve been trying to use it like, uh, Siri and Google Assistant sort of thing, but I’m not sure if you can retain this information once you finish your session on with it sort of thing.
Um, I hope someone can correct me on that one. Uh, but, uh, with regards to start, um, I, I have, uh, an idea that I’ve been kind of playing around with, with e-learning and, um, with e emailing. You’re saying reinforc emails? Is that what you said? E-learning. E emailing. E, thank [00:48:00] you. No, E-learning. E-learning. Okay.
Yes. So we’re, we’re e-learning, sorry about it’s actions. Sorry we all don’t speak English as well as you just so it out. Uh, uh, the door, the, I’m sorry. E-learning, um, learning management system sort of thing. Um, I’ve had quite exposure creating, um, uh, e-learning with learning management systems and, um, I was kind of, kind of peer around with some ideas on how to integrate, um, Chat, G p t, the API I’ve, um, beside using with the chat g p t on the open ai, um, platforms, I’ve generated the API keys and using it with postman sort of things and getting some response from it, sort of things.
And, um, yeah, and kind of learning about a prompt and the completion, the, the scope of the response, uh, kind of thing. The information that is, that is spitting back to you. Uh, and [00:49:00] I’m kind of at the phase of trying to get that into learning management system for scrum, like kind. Uh, I’m trying to play around with how to use it to train, um, agile, to train people on agile, kind of train an e.
For Agile and Scrum and spring sort of things, and using charge PT for reinforcement learning on the e-learning platform. I’m not sure if anyone have an insight into that, if this it’s possible. Um, but yeah, like something of that, of that nature. Um, but for now I’m, I’m just still playing around with it kind of, of being, I’m, we’re very kind of being impressed with the content it’s generating.
I also for about, um, create, uh, using chat GT for generating dynamic content, um, on websites sort of things like based on kind of the response that is being given, just kind of get some kind of different kind of contents like for different users in terms of website. [00:50:00] Webpage content type thing. Um, that’s also another idea that kind of occur to me that another possibilities.
But yeah, like the a learning use, um, using t for first learning on Agile because, um, I’ve got a couple of people I’m trying to get onto Agile and I don’t know how to train them and I’m thinking, can I create an eLearning solution, um, that, uh, child can assist in kind of reinforcing like whenever they return back to learning something, they left off something can perhaps.
Yeah. And I think, I think you’re, that’s the first time I’ve, first time I heard this idea of using it for e-learning, but I gotta believe that there’s a way of doing that. There’s gotta be a way and, and, and, and the start doesn’t just produce, help produce the content, but sometime it’s the format of content we were working on, um, uh, a part of the book.
That’s being published, uh, by Forbes called Start Scale Exit Repeat. And we were working on, there were [00:51:00] seven or eight areas of the book where it lent itself to be creating a chart. Uh, but when we had written it out, it wasn’t written out as a chart, and we were able to just plop it into chat, G p T, and it produced this great chart with, uh, uh, headings and, you know, and whatnot.
And, um, so I think it can actually help out with formats and things like that. It, it could definitely be, you know, let’s say you’re creating a, um, a, a document on a particular industry. Maybe it’s how to use Che p t from, or how to play chess or how to do whatever. And you can use chat g p t to help you write that ebook quite easily.
Anyone else have any other thoughts on, on e-learning? Yeah, I can, uh, I can pour in really quickly. Uh, hey, go ahead, Lewis Ho Lagoon. Um, if you check your back channel, um, I really like, uh, your mission. Uh, so I gave you a little, how would you say fire for your mission. Um, if you check your back channel, um, you’ll see, [00:52:00] um, prompts, literally stack prompting, uh, or prompt stacking rather.
Uh, so you can build out a course, build out the worksheets, build out the modules, build out everything that’s a complimentary gift from my community to yours. I really like your mission. I like what you’re a part of. The e-learning is something I accomplished from the first day, uh, that I kind of took a crack.
I, Chad, g b t, um, in its basic model, uh, I don’t pay for anything. Uh, when it comes to Chad Bt, I attach it to the internet. Give it all the data points. It can ab absolutely imagine, uh, utilizing web chat. Uh, web chat, g p t three or web chat g p. Um, with that being said, I created, um, financial literacy, real estate, blockchain, cryptocurrency, N F t, artificial Intelligence and, uh, um, creative writing course, for seven to 14, 14 to 18, and 18 to 75.
Um, these are six week, five day a week course, I mean six week, five day a [00:53:00] week courses, uh, to teach e-learning. I’m also, uh, uh, an adjunct professor at DePaul University. I’m from Alma Mater. I’m, I’m in, works with that, uh, that community to help build out a scale system for college students as well. Um,
There’s a lot, there’s a lot available guys. Um, and I wish more people were dedicated to teaching, right? Not, I mean, not courses and stuff. I, I teach courses and all that stuff, but I’m not interested in that. The kids, them taking this out of the schools, um, knowing full well that there’s so many creative little kids.
Look, you hear my kids in the background, right? . Um, but I, I don’t want to over inundate the room. I, I, I really appreciate what you guys have got here. I’ve always been a huge fan of the startup club. So , I hope that that added some value too. You
Well, that was so cool, Lewis, and, um, that was so generous of you. So thank you for coming up and sharing and [00:54:00] really sharing. Oh, here he is. He’s back all again. Yeah.
Excellent. So, um, we only have a few minutes left here. This has been an amazing discussion. Um, so Dave,
Yeah. Hi. Thanks. How good to got? Um, yeah, a absolutely. I’ve been, uh, doing a bunch of, uh, community education work, uh, around, uh, Chatt p t. We’ve done events where we pull about 50 local entrepreneurs, uh, in our area together and talked about the projects that they’re doing. Um, there there were people that were talking about, uh, customer validation, um, persona development.
Uh, there’s just a number of great, great things. Um, [00:55:00] I work for a custom software development firm. We have a ton of requests that are coming in, um, right now for people that want to develop integrations, uh, into open ai. And, and, um, we’re talking a lot about chat, G P T. . Um, but there’s also Whisper, which is the speech recognition component, uh, of what they’re doing.
There’s Dolly, which is doing the, the, um, you know, artistic, uh, vision. Um, I, I, you know, it’s been absolutely amazing to watch people kind of unearth, um, you know, these possibilities. I think it’s important to recognize, um, GaN was talking about something really important, which was that he’s working on the training side of it.
Chat, G B T right now knows what it’s been taught, and it’s been taught it off a lot. And it uses this corpus of training, right? That everything that it’s learned, it uses to give you a response. It’s really similar to when you’re [00:56:00] typing inside of an email and your email program is, or your chat is giving you the next word.
which is incredibly at times unoriginal because it has only, it can only tell you what it’s been told. Um, and I think that’s really, really important to know that the training side of it is gonna become so much more important, um, that we’ll start to see is what we’re seeing now is companies that want to keep their own private corpuses, um, and ha start to, uh, close down some of the walled gardens that we’ve seen or, or bring back walled gardens that we’ve seen.
Um, because now what you know, is gonna be that much more of a competitive analysis or, uh, competitive advantage. So I think that’s important to know about when you’re using it. Um, just the, uh, the difference between, uh, what it’s telling you [00:57:00] and what you know. I think that was kind of what Andrew was talking about when he was talking about Andrew, the 365 co-pilot.
was that co-pilot is based on you, it’s based on your accounts, it’s based on your emails, it’s based on your Excel sheets and, and things like that. And, and that becomes really, really interesting, um, because it has now a frame of reference built from your experience, your data.
Wow. There’s so much to, you know, think about and learn and talk. And this has been a fantastic, um, conversation. Thank you for adding that, Dave. And I love hearing how members are helping each other and, you know, just trying to help the community. But before we end today’s session, [00:58:00] Louis, did you you wanna go?
I, I know you’ve been so kind in helping others. Did you have anything else to add, Louis? Um, yeah. Uh, so really quickly, I don’t wanna take up too much shots again. Um, first of all, these people in the room, can you get a little bit louder? Sorry. Yes, yes, absolutely. Uh, can you hear me okay? Still a little bit light, but go ahead.
Yeah. Okay, perfect. So hold on for Justin. Hopefully you guys can hear me a little bit better now. Oh wow. Perfect. It was night and day, . All right, perfect. Um, so, uh, I wanna give a shout out to my partner Alicia, uh, in the, in the, in the audience real quick. Um, we’ve been teaching a lot. . Um, but it really comes down to the community service.
Uh, we’ve been hosting classes about all this stuff, but the impact to allow non-for-profits to, uh, uh, to apply for grants and seconds. Um, we helped one of the members, the Democratic Republic of Congo, [00:59:00] apply for grants internationally today. Uh, literally live on Zoom. All of our classes are on Zoom, so it’s not just clubhouse conjecture, um, but allowing that gentleman to apply for millions of dollars worth of grants in seconds, break down the ideology.
That was something special today. Um, so if you guys could please continue these type of rooms. Keep, please continue to push the envelope. Ask questions, ask. because that makes all the difference for the world at large. Um, so that’s all I wanted to say. Shout out to you guys and shout out for the movement.
If you guys haven’t followed the club. The Startup Club is one of the longest running clubs on Clubhouse. Uh, when I first started, almost two and a half years ago, three years. This is one of the clubs that I came into, and I always, uh, I always come back. So shout out to you guys and, uh, I love what you guys do.
Oh, that’s great. And, uh, and, and, and by the way, uh, we just love the fact that everybody came on stage during this show today and, and really contributed a lot, a lot more than we could ever do on our own. [01:00:00] And I know when you do come on stage, I’d like to follow you and even Dave, even though you don’t have your, your photo there, I, I still followed you.
It gave us some, some great content and appreciate your helping the com moderate as well, Andrew. Um, I wanted to share the, the last three, uh, last four from chat g p t on the query. I gave chat, G P T, which was, uh, uh, tell me. 10 ways you can, startups can scale their business using chat, G P T. And number seven was sales forecasting.
Chat. G P T could analyze sales data to predict future trends, enabling startups to make more informed decisions about inventory management and sales strategies. Well, that’s one I have not suggested are e-commerce companies and the incubator here, Michele use. But I gotta figure out how, how that one works.
That piece works. Business strategy chat, g p t can provide startups with valuable insights into trends, emerging technologies, and best [01:01:00] practices, helping them develop effective business strategies. Again, I think that can line up well with our, our idea to do, uh, business plan ai.com. Employee training chat g p t can provide on-demand training.
Here we go. E-learning on-demand training and support for employees. Reducing the time and resources required for training programs. Oh, I love that idea. And then number 10 was Social media Management Chat. G p t can help startups manage their social media presence, generating content, engaging with followers, and analyzing metrics to improve engagement and reach.
So today, the show, the community came together, put together some really great ideas of how we can help scale our startup and chat. G p t also contributed to it by offering 10 ideas. Mimi’s gonna put together a blog, and we’re looking forward to seeing all of you next week. Uh, back on the show. We do the show every Friday, [01:02:00] Dave, at two o’clock Eastern.
And we have so many phenomenal authors who come on the show and share with us their expertise and we interview them and we, we, we, we want people from the audience to come up and ask questions and participate in the conversation. Uh, we’ve got some phenomenal authors coming on, uh, over the next 60 days.
If you want to know who’s coming on, go to www.startup.club and sign up to the email list we’ve had on, uh, Kevin O’Leary, Mr. Wonderful on Startup Club. We’ve had on Jeffrey Moore on this show who wrote Crossing the Chasm. We’ve had on Joe Foster who started Reebok. We’ve had on Verne Harnish, who’s a, a, a scale up coach.
We’ve had on so many great authors, experts, billionaires, serial entrepreneurs on this show. But if you want to, if you want to hear from them, if you want to know when they’re coming on, please sign up for that email list so that you are aware of it. We [01:03:00] only send an email out once a week, and it’s only about our calendar and our shows.
Thank you very much for joining us, and we’ll see everyone next.
Thanks Andrew . Thank you. Thank you everyone. You’re always welcome to come. Commad, Andrew, you always are. Hey, happy to be here. It’s an exciting time. Talk later. Bye. Thanks for the room.