Camille Rapacz: How are you today, George?
George Drapeau: I'm doing very, very well. Thank you. How about you?
Camille Rapacz: I'm doing okay. I get a little bit sad at the end of summer.
George Drapeau: Yeah. Especially you. Oh, can I just say this thing? I sometimes I talk about my sister when I'm meeting somebody and we talk about, you know, I still have family in the Pacific Northwest and for a while when you had convertibles all the time, I'd say like, I think my sister and I really need to swap locations 'cause I don't mind the clouds and I love the rain.
And my sister is a sun goddess. I mean, she drives a convertible in Seattle and people are like, what? Why? There's like three days. It's like, well, it's more than three days. But look, just to give you an idea of how much she really loves the sun.
So it makes sense to me. Of course, you're a little bit sad when summer goes away. 'cause those three days you can drive your convertible.
Camille Rapacz: I know it is super sad though. We had an amazing summer, so I really can't complain like we definitely didn't just have three days of sun, we had miraculous, perfect weather. But whenever we get to this point, because this is where all the fall things like the pumpkin spice freaking latte starts showing up and I'm like, no, it's not even September yet. You just slow your roll and wait over there for fall. I like fall once we get way into it. Once we're in October, I'm good. But this transition August into September that's kind of hard. So I'm having to, having to perk myself up.
George Drapeau: All right. Well, this outta do it.
Camille: / Welcome to The Belief Shift. The show that explores. What you really need to know about building a successful business.
I'm your host, Camille Rapacz: business coach and consultant who spent too much of her career working in corporate business performance.
George: And I'm George Drapeau: your co-host and her brother. I'm a leader in the tech world bringing my corporate perspective, but mostly my curiosity.
Camille: Together, we're exploring beliefs about success and how to achieve it. But mostly we're bringing practical solutions so you and your business can thrive.
Camille Rapacz: /So today , I don't know if you've heard about this, George. The topic for today, I'm really curious, do you know about this thing called ChatGPT?
George Drapeau: Chris Christie mentioned it the other night on the Republican debate, so I've heard about it from that. Apparently it's a Republican scam. That's what it's, yes, of course I've heard of ChatGPT. It's all I hear about at work is ChatGPT. It's all I hear about my personal life is ChatGPT. Yes.
Camille Rapacz: Okay, so I'm doing an episode on this today. I was hesitating whether to do one because I thought everybody's talking about it, but that doesn't mean we shouldn't talk about it. Absolutely. Because it's been showing up a lot with my clients and I have a, a wide range of ways that I see it showing up. So I thought it would be worth talking about because I can see some people have figured out how to use it well in work, like how to use it as a useful tool.
I can see other people are just fumbling through and don't quite understand so they've made some mistakes and kind of fumbled through. Yeah. And then other people aren't using it at all, I think. 'cause you know, rightfully so they don't really know how. It is a tool that is just new and kind of sprung upon us out of nowhere, seemingly, for those of us who aren't in your world.
Yeah, so it just feels like a, it's not like, oh, there's just a new app to try out, or a new social media platform. We're sort of used to those by now. This is a really different type of tool, so I thought we should just talk about it for a little bit.
George Drapeau: Fantastic. Can't wait. Seriously. No joke.
Camille Rapacz: I know, I know.
I can't wait to hear what this is. Might be the episode where I can't wait to hear what you have to say about it more than the other way around.
George Drapeau: Yeah, we'll see about that. I'm gonna talk about Skynet. I'm not gonna talk about Skynet.
Camille Rapacz: Promise me you're not gonna scare our listeners.
George Drapeau: I won't scare our listeners. No, no, no. I promise
Camille Rapacz: into thinking ChatGPT is gonna take over the world 'cause we've had some of those scary articles and things. Yeah. Sigh. Humans. Why do we always do it?
George Drapeau: We like being scared.
Camille Rapacz: That's true. I do actually. I do like being scared. You do like being scared that that is the best part about fall for me is it's scary movie time. So I do,
George Drapeau: I did not see that coming.
Camille Rapacz: Really? I go on a whole bingey watching of all the old scary movies. Yeah. Yeah. I do.
George Drapeau: Wow. That's fantastic.
Camille Rapacz: That's, that's for another episode. Okay. Okay. So today let's talk about what ChatGPT is, right? I'm wondering how would you define it, George? How would you define it?
For the, for the lay person?
George Drapeau: Okay. The, for the lay person, I would say ChatGPT is the most amazing chat bot ever created by man. I hope that everybody understands what a chat bot is. So if not, I would start this way. If you are ever in customer service, which you've noticed in the last 10 years, maybe 15 years or so, is instead of talking to a human, you have the opportunity to talk, to chat with somebody online.
And sometimes it's an actual human, but sometimes it's clearly a computer. And you know, you'll, you'll get a prompt that says, hi, how can I help you? You could start your responses with one of these five things, I've lost my luggage. I need to make airline reservation or something like that. And it gives, because the, the computer has fairly limited in what, what it can parse from what you type.
So you type I need to change my reservation. And then the bot reacts in this really clumsy, awkward way, you know, it's a bot, but it's able to sort of understand some of your tests. That technology well or different technology has been applied to chat bots. ChatGPT is an amazing, amazing, amazing version of that with all kinds of interesting nuances and differences that are significant for this podcast.
But basically, if you think of ChatGPT as the most amazing chat bot you've ever met, that's what it is.
Camille Rapacz: That's a great definition. I like that. Thinking about it from a chat bot, but once I heard that as part of the, I was like, oh, now I get it. So yes, that makes a lot of sense.
George Drapeau: Yeah. There's all these terms that we hear about it these days. Maybe you'll get into that and I could explain some of those terms, but the terms are useful if you have some background and I don't think they're so useful If you don't have that like large language model, you have to explain large language model. So I just say it's an amazing super chat bot.
Camille Rapacz: Yeah, no, I think that's perfect. So I'm gonna give the Wikipedia definition 'cause I also felt like it was helpful.
George Drapeau: Yeah. And I wonder if ChatGPT created the definition.
Camille Rapacz: I know. I was thinking about that too, and then I was like, should I just ask ChatGPT to tell me what ChatGPT is?
But I didn't go there. I know I probably should have. Okay, so the Wikipedia definition so ChatGPT stands for chat generative pre-trained transformer. I know we're getting bored already. Wow. Yeah. But of course I heard Transformers and I thought like Bumblebee, like transformers. That would be awesome. That would be awesome. Yeah. It should show up like that. I am a decepticon or whatever. I don't know my transformers that well, but okay, so it says. It is a large language model based chat bot. So to your point, the chat bot makes sense. The language model maybe not so much developed by Open AI and launched in November, 2022. It's notable for enabling users to refine and steer a conversation towards a desired length format, style, level of detail, and language used.
George Drapeau: May I put that in practical terms by giving the audience a homework? Yes. So folks, here's the homework that will help you understand that Wikipedia generic link. Go to ChatGPT and ask it please define ChatGPT in the style of a limerick. Please define ChatGPT as if you were explaining it to a seven year old. Pick several different ways for it to do the same thing, and you'll get an idea of what it can do because it will give you explanations, but it's able to modify its style into whatever writing style you ask of it. Please define ChatGPT as if you are writing a verse in the Old Testament of the Bible.
Do that. And you get an idea of just what it can do. Same question, different styles. And that's what it's talking about when it says what desired length format, level of detail. Yes. Yeah. Does that help?
Camille Rapacz: Yes. Yeah. That's fantastic. And then while you were talking about that, I did one of 'em 'cause I wanted to know what ChatGPT would say to a seven year old.
So here's what it says to a seven year old, here's how it responded. Sure thing. Exclamation mark. This is one of the things I do love about it is it's like talking to another person. Right? Okay, so here's what it says.
Sure thing. Imagine you have a really smart robot friend. You can ask this robot friend any question you have and it will try. Its best to give you an answer that makes sense. It's like having a talking buddy who knows a lot of things and loves to help you learn and understand new stuff. Exclamation mark.
It tries to be very excited.
George Drapeau: That's totally. First grade, second grade, seven year old patter. Okay. Yes. Will you type another one? Yes. In the style of the Old Testament.
Camille Rapacz: Are you ready?
George Drapeau: I'm so ready.
Camille Rapacz: Okay, so I asked it, please define ChatGPT in the style of the Old Testament.
It says, in the days of knowledge as tapestry, there are rows of creation known as ChatGPT, woven from the threads of vast wisdom. This digital oracle fashioned by the hands of ingenious minds did possess the ability to engage in discourse with seekers of understanding like a wellspring of ancient knowledge it did respond to queries with words both profound and enlightening. As a modern day Oracle ChatGPT offered its counsel to those who sought to unravel the mysteries of the world, using the language of the present age to reveal the treasures of the past and the secrets of the future.
Very different response.
George Drapeau: Yeah, I mean, I personally prefer the seven year old one as more funnily excited. I do, but this is not bad.
Camille Rapacz: So yeah, I think that's a great demonstration of how when you interact with ChatGPT you can get it to give you all of these different layered and nuanced responses in these different voices, basically. So let's get into a little bit of Like that was fun, but is it actually useful? And spoiler alert, yes, it's useful.
That's why we're talking about it. Yes. If it wasn't, this would be a very different podcast, what not to do to improve your business. That's what we would be talking about.
George Drapeau: We should have a second podcast.
Camille Rapacz: We should have a second podcast. Yeah. The what not to do actually, that there could be lots of topics.
Alright, so let's talk about how it is useful. How would you use it in your work? Or how do your, how do your peeps think about using it?
George Drapeau: We don't that much actually, and it's there's this, I can tell you a longer story about why my team is not currently using ChatGPT, but as a manager, as a director, I have used it recently and instructed one of my directors to use it.
Here's a specific example that happened four weeks ago. I'm hiring, I'm hiring everybody and open some job reqs in areas that are kind of new for us and we were struggling a little bit about how to write the job description in a way that we thought would capture current talent in an area that's new for the company.
My engineering director showed me something and it was kind of an old standard template, frankly. I mean, he is good. He does a good job of writing the job descriptions, but it was he old school and I said, why don't you ask ChatGPT to write your job description? He did and he said, wow, thanks. It gave me some really good ideas.
Here's why, because large language model, what it means is that chat. The ChatGPT robot has this huge bin full of words and where you get the words or they're conversations and the conversations have been scoured all over the internet. These companies like Microsoft and Google have poured all of the texts they've ever gotten, ever.
They can get their hands on into this thing. So basically what we're seeing is ChatGPT has access to recent conversations, recent job descriptions written in this area. I haven't looked at a job description this area, but it has. So it's a right front job description for solutions manager from the cybersecurity area.
It's probably seen hundreds or thousands of them post and it's using that. So of course it's gonna give me more recent stuff than I had. And that's a, that was a great start. We did not use it word for word, but it was pretty good.
Camille Rapacz: That is such a great example of how to use it, and it's number one on my list of how is it useful, which is just generating ideas or brainstorming like it's a primer to get your brain working on a, on a problem or on a solution that you're trying to come to.
So, yeah, I wanna make this thing better, or I need just ideas about how to do this. It's a great source for that, and you start pinging it with the right questions and you keep going or you just ask it to give you some words. I've used it in that way a lot where then I get, oh, I like this phrasing, or, I like that.
And you can keep asking it to give it to you in different, like you said, voices, like what we just did, right. To just see what shows up. So, yeah, I think that's a fantastic way to use it. I think it's the way that, the number one way I think that you know, business leaders or even business owners can use it is if ever you feel like you are, and we all have this, we're just like, I, I, I don't know, I'm kind of stuck. I, I, yeah. I need something to help get my brain either just out of its stickiness or to just set me off in the right direction or a new direction. It is a great place to go for that.
You said something about that it does have all of this knowledge and information from, I mean, it's us, right? This is all of our words. Yeah. But it's all of it that's been on the internet ever.
Mm-hmm. So that has its pros and cons, but also I understood ChatGPT only has information from the internet up to 2021. Yeah.
George Drapeau: That sounds about right. Yeah. I mean, I don't know the exact dates of when they stopped absorbing data into three, 3.5 and four. But that, I mean, that wouldn't surprise me.
Camille Rapacz: So, yeah, so I think when it comes to like the kind of research that you're talking that you were just talking about, totally used for, there's so much information out there. Yeah. I mean, 2021 wasn't that long ago.
But keeping that in mind I think is important because you're not gonna get any factual information for anything that happened after that date.
Yeah. But when it comes to anything that we wanna do, so brainstorming, generating ideas, it's got all the stuff in there for you to, you know, tap into and learn. I think other ways it can be useful is just to get different perspectives on an idea. So even just asking it from the perspective of somebody that has a different role or that comes from a different part of the world, or you can ask it to generate that same idea from other perspectives, which can be really helpful.
So business owners will use this a lot for writing, like headlines and descriptions, social media posts. So if you're trying to create content for your business it's great for that. And then it can also do a good job of summarizing things. So if there's a book or a movie that you want it to write a summary of, it'll just boop spit that out for you. Yeah, that's really cool. It also says it can debug code.
George Drapeau: It could actually write code. I, I mean, it can debug, but even more importantly, And when you think about it code is a conversation of a different kind, a very structured kind. All kinds of code has been put in there and you could say, Hey, generate, see some code that does this kind of thing. There's probably logic that's done. It's a great acceleration tool, especially for people who are unlike me. I'm a software engineer by training, right?
For people who are not software engineers by training, but are familiar with the internet, and this is a great way to get some scaffolding code. It's amazing. It works. Wow. Pretty well.
Camille Rapacz: That's amazing. I also know you can have it help you write. So if you're an Excel user and you need help writing formulas, which I often, I know what I want it to do, but I, I don't know the exact formula to get it.
George Drapeau: That's an awesome idea.
Camille Rapacz: It's really great. So I found a good article about this. I'll put in the show notes too that helps you to get it accurate. Oh man. 'cause you have to be really precise when you're talking to ChatGPT to get it. But it is great use for, if you're trying to get a formula in Excel and you're just not exactly sure how to get to it it's a great resource for that because again, all of those things exist on the internet, so it's just looking for the right one for you.
Right. Yeah, absolutely. So that's another great use. And then the other thing that I love about it is it's, you know, even if we think outside of just using it for work, you can use it for life too. Mm-hmm. So you can generate ideas for, like, activities with your kids or meal planning or you know, ideas for making your next family Christmas more memorable and it'll just come up with all sorts of.
It's actually an, again, it kind of relates to the idea generator. Mm-hmm. Just helping you brainstorm new ideas, but it does it in a really cool way. And then you can start to give it more guidance on, I like these ideas, but give me more like this. Yeah. And it'll learn from, you can start having a conversation, which is, I think gets into the, how do you now, how do you get the most out of it?
So I wanna talk about tips for how people can get the most out of doing all of these things. Okay. Because it works differently than a Google search. Yeah, it's a conversation you create.
So you can sign up for ChatGPT 3.5. It's free so you can get your account and you can use that version for free.
And I haven't had to, I haven't felt the need to move to the paid version I think most people can use the free version the way that it is now and get tons out of it. So that's the other bonus is it's a free tool. And when you go in there, it'll show you, if you sign up for an account, then you start what they call a new chat.
And what you wanna do is every chat should be on a specific topic. You don't wanna have multiple topics in one chat because it's. It's like trying to have three conversations with the same person and have them track all three of 'em at the same time and bouncing around. Right. You don't wanna do that.
Yeah. So you wanna have one chat be one topic because you can conversationally improve on its responses within that same chat. Yeah. So some ways that you can do that is you can give it a role or persona, like we just started with that kind of idea. Give it to me in this voice, a seven year old, or act as a professor of literature or write something in the voice of Jerry Seinfeld or you can tell it how you want it to respond to you.
Right. Which is really cool. Yeah. And when you do that, so if you asked it to do something like what we did, You can then, in that same chat, have it improve on it by doing this, by either giving it a different persona or, but just by changing any of your, of what you're asking it to do.
You do wanna be specific and just give it one task at a time or it just, again, it gets confused about what it's trying to deliver to you.
And the other thing I love to do in it is to ask it to give me three versions of this. So I could say, I am doing a podcast that's about ChatGPT give me three versions of a headline to use for my podcast episode.
George Drapeau: How good has it been at doing that for you? What have you observed?
Camille Rapacz: What I have observed in doing that is that it's great at giving different versions, but what you have to be careful of is it can be a little wordy and I have to be careful that it doesn't not sound like my own brand voice as a business owner because it's just spitting words out, but it's not necessarily, there's no emotion in there.
And it doesn't know my way of speaking. I don't exist on the internet enough for it to know that, so I have to bring that to the table. But it will, sometimes what'll happen will be like, oh, I really like that phrase, and I might just pull like a couple of words out and then I'll use that for myself.
But get all these different versions and keep going until you get stuff that you like.
George Drapeau: Do you find that when you ask for different versions, they're different enough to be useful?
Camille Rapacz: It depends. It depends on what I've given it. The quality of the output depends on the quality of the input that I give it, the way that I prompt it.
So if I ask it too generic a question, I'm gonna get pretty generic answers back. Yeah. So sometimes I'll start with the generic question. It'll give me some, you know, responses and then I'll say, Re rewrite this in a more witty tone or rewrite this in a way that really speaks to a small business owner, or rewrite it in someone else's voice just to see what shows up.
Yeah. So doing all of that and getting all that variation, what I end up usually doing is I never find, like I'm gonna just copy paste this one. And I don't recommend people do that. I use a combination of all of those outputs to generate my own thing.
George Drapeau: Yeah, absolutely. Totally get that.
Camille Rapacz: And you do have to really provide context, right? So talk to it like you would a human. So talk to it like I'm giving a human an assignment. Hey, ChatGPT. I'm writing an outline for a podcast on how to use ChatGPT. My audience are business owners.
It's not just one command, it's not a Google search of, I just had to plop in a few words. It's conversational. You write a whole paragraph of what you're asking it to do and usually it starts with sure, it's very excited and then gives you an answer.
George Drapeau: May I elaborate on that?
Camille Rapacz: Yes. Go do more.
George Drapeau: So if you think about it again, about what the Super Bott is doing here, remember that it's got every conversation that's ever been on the internet. So if you're saying, I'm gonna give a talk about tiring personal leadership. Previous podcast, and I'm speaking to a college commencement audience.
If you're providing that context, what you're telling it to do is, okay, start looking at all the speeches you've ever seen, commencement speeches, all of them, and take this task and start generating text that looks like it fits into commencement speech style. That's what it's doing. It's going to that part of the corpus, and if you say, okay, I'm speaking to the Rotary Club. Knows what Rotary Club is. 'cause Rotary Club, there's a bunch of speeches that for business owners. So when you're narrowing, when you're giving it more context like that, you're telling it which of the trillions of conversations to ignore and which to start looking at. 'cause basically what it's doing is it's saying, well, statistically speaking, here's what the next sentence would be for an audience like that, I think how, based on millions of other sentences I've seen in for that audience, that's why the context is helpful.
You're telling it where to where to start copying from. Start ripping off.
Camille Rapacz: It's interesting 'cause that is basically what you're doing is you're just taking from what's already been created. It's not really inventing new No. As much as it is combining what's already out there into some answer for you.
George Drapeau: Yeah. It's not at all inventing . What's going on is it's using a statistical model to predict the next word and the sentence it thinks you want it to write based on the input that you've given and it'll keep going that way.
Yeah, absolutely.
Camille Rapacz: Do you have other tips for how to get the most outta ChatGPT?
The next thing I'm gonna talk about is what, like what to look out for, but tell us more.
George Drapeau: Yeah. So I'm gonna make a statement and I want the audience to prove me wrong 'cause I think I'm wrong. Ooh, don't ask it to argue with you. So ChatGPT is also unfailingly polite. It's excited to work with you and it's polite.
And if it actually gets something factually wrong, which will I think we're gonna talk about later, yeah, you can tell that, no, no, no, that's wrong. And it'll say, oh, I'm so sorry. And I don't think it'll justify why it was wrong, but it will apologize and try again. Try something, and it's like, well, that's nice.
So I was gonna say don't get it to argue with you because it doesn't really wanna do that. I think though, I think there probably is a way to get ChatGPT to argue with you. Maybe if you said things like, I'm gonna make this statement about what I think is the difference between leadership and management.
I'm just making an example. I want you to take the, the opposite view. Take the counter position, please and see what it does. And so if you could force it in a positive direction to argue with you, maybe it can do that. But otherwise, if you're looking, if you're just looking for it to disprove everything you say, I don't think it's really wanting to do that.
Camille Rapacz: I think that's a really good point. You can get a counter argument, but it's still pleasing you by doing what you asked it. It's not genuinely arguing with you.
George Drapeau: Yeah. Not like we would do with each other. Or you would do with a trusted buddy.
Like, no, no, no. Come on, let's, let's argue.
Camille Rapacz: Yeah. I think that's a really good point. But I do think that asking it to give you a counterpoint in that way sort of give it an argument against what I'm saying is a great way to get a different perspective.
Especially if you're really trying to write some sort of eye-catching headline or blog or topic or something like that. You're doing a presentation and you wanna introduce the counterpoint as an interesting way to come at that topic.
Great place to do that, right?
George Drapeau: Yeah. One other tip. Yes. Which we've kind of gone through here. Get goofy. I mean like, just try asking questions. Like I said, so one of my favorite examples I read about, I think in the New York Times, was somebody had said ChatGPT please describe to me how to get a peanut butter sandwich that's stuck in the VCR and do it in the style of the Old Testament.
And I mean, that's a weird problem to ask it, but it had an answer and it did it. And that was such a goofy way to ask. But it's fun. And I think if you just let yourself get goofy, that will help you kind of break your own structures about constraints, which we've been talking about this whole time. Just get goofy.
It's not gonna do you any harm.
Camille Rapacz: That's a cool example. I hadn't seen that one. So I think including what to look out for in our tips is also really important.
The number one is that it's not always right, so you gotta remember there's a lot of untruthful things on the internet. Mm-hmm. And sometimes ChatGPT is really trying to be a pleaser and it gets it wrong.
George Drapeau: Yeah. The thing that's dangerous about that to me, I think is chat PT is a really, really good conversationalist and like in real life we can be swayed by good speakers.
They're speaking so powerfully that you believe them, even though they're factually wrong. And that's exactly the problem here. You need to be aware of that a lot all the time, if you're asking, it is best used. I think for subjective stuff, gimme some writing ideas about copy like we've been talking about.
If you're asking for fact-based stuff, it is not the best tool for that yet.
Camille Rapacz: That's right. I have tested this out and when I did it, so you can, you can ask it for sources. So give me your source for, so when I went into fact-based stuff, I then asked it to give me sources and then I went and looked up that source to see if it existed and if it was accurate. So you can fact check it by doing that, saying What's your source for this? I haven't had this happen yet, but people have reported that they've done it and they even made up the source.
So you can't just say, give me the source and then say, okay, got it. And then assume that the source exists. You need to also search that that source actually exists and that that source is saying what ChatGPT is saying.
So there's a couple extra steps there, but it's still faster research than if you were trying to discover that source on your own on the internet. I think. Yeah, I think if, if I wanna get multiple sources on a specific topic that I wanna create some, you know, facts around, yeah, I'm gonna start by a narrative conversation in ChatGPT get it to focus, you know, gimme facts on these things. Give me the sources and then I know what to start researching for myself on the internet.
George Drapeau: That's awesome.
Camille Rapacz: The other one to look out for was just the idea that it doesn't have recent events, so it can't tell you anything. You know about the movies that came out this year, right?
For example. Mm-hmm. And it's learning from us, so it's also biased. Yeah, there are built-in biases in all of the conversations that we've had on the internet, and that's what ChatGPT is pulling from. So it has those two and it doesn't have any depth or insight again 'cause it's just giving it, I think your description was right.
It's just figuring out what parts of the internet it's gonna pull the information from to give to you. And then most of the time I find it's just too wordy.
George Drapeau: It's like me.
Camille Rapacz: Because lots of flowery language, I'll find. Oh yeah. And I'm like, oh, that, I mean, now I just sound like a, nobody's gonna even know what I'm saying. There's so many flowery words in here. Right. So usually when I'm editing something, I'm just taking a lot of descriptors out. Because it just gets very wordy. Mm-hmm. And you can tell it, simplify it. You can say, give this to me. And, and, you know, 50 words or less, you can say, do, do it in five words. You, I mean, you can be really specific about how long or short that, you know, response is gonna be. And that will help. That will help it get simpler.
You can also tell it, stop being so ver verbose, simplify this or say it like I'm talking to a, you know, 12 year old or something. So there's ways to adjust the wordiness, but most of the answers, like sometimes I have to look at, I'm like, Ugh, I don't even wanna read that. I gotta get it to give me another answer.
Any other what to look out for you can think of George?
No, I think these are the really, really good ones.
So the what not to do as you use ChatGPT I have three. You might have some more to add to this. Okay. Some, we, we've pretty much talked about these, but I, I wanna cover them again, do not copy, paste the answers.
Number one, we can tell when we read it like people are getting really smart about noticing what's a copy paste from ChatGPT and also that's just, It's just bad form and don't copy paste and use it for yourself.
George Drapeau: Yeah, great.
Camille Rapacz: Don't assume it's right, we already talked about that. Yeah. And then, and don't expect it to do all the work for you. Expect it to be a primer into your work. I know. I mean, it's a really happy, friendly, wanna be helpful robot, but you know, humans still needed.
George Drapeau: Yeah, as we were talking about this, I was thinking, I mean, this is one of the reasons why the writer's strike is happening because of this kind of model is very useful for the creative process and the creative process when you're stuck or when you're in a group of people and you, you're in this early stage where you want to generate ideas.
It's great at that. I mean, it's generating ideas that were all in existence, but there's so many of them out there in the world. You probably didn't know them all, but it's great at generating ideas. It's not great at kind of finishing everything off, but that's a significant part of the creative process.
Bouncing ideas off somebody. You have the whole writing room here, ChatGPT, but it's got the same caveats. Yeah, of course you're not gonna have it write the whole script. That would not work well for you.
Camille Rapacz: Yeah. Did you wanna say something?
George Drapeau: Yes. I have two more examples from the beginning of homework just to spur more. I really hope that you guys in the audience will do this homework.
Go and talk to ChatGPT and ask it to define ChatGPT in different voices, and I'm gonna give you two more. I'm gonna read the answers. You ready? Yes. Go. Question. Please describe ChatGPT in haiku form. Ooh. Five syllables, seven syllables, five syllables. That's a haiku.
Text flows like a stream. Ideas bloom within code. ChatGPT's soft dream. Not bad.
Camille Rapacz: Not bad chatGPT.
George Drapeau: Example number two. Please describe ChatGPT in Limerick form.
In the digital realm of thought, ChatGPT's responses are sought. With language so clever it chats on forever. A marvel of tech over rot.
It's amazing. It's just amazing.
Camille Rapacz: It really is amazing what it can do. This is a great assignment for people because I think it really will give you a sense of all the various types of responses you can get. It helps you think outside the box. 'cause I find that, you know, in the beginning when I started to use that, I definitely was boxed into only the way I thought about, you know, conversing.
Yeah. And if you expand in that way, it really, that's the beauty of it.
George Drapeau: So if we've poisoned you and you actually do this, you spend half a day doing it. Try to at least stop between ideas and see what it's giving you. Try to learn, like try to really absorb what it's doing for you.
We've asked it to do the same thing in different forms and look what it can do for you. Just don't just be in the rabbit hole. Try to absorb what it gave you. It truly is amazing technology.
Camille Rapacz: Yeah, it truly is. So if we haven't made it clear already, your reasons to use it are to help get the creative juices flowing.
When you're stuck, when you're trying to get out of your rutt and you're not sure what to say or how to say it or an approach to take, like your example of the writing a resume and wanting to make it better. Like we're kind of stuck in a rut and of how we do it.
That's just such an excellent reason to use it to help your brain. It's really about, I think about firing off new little ideas in your own brain. Yes. Because like you said, it, I mean, it's not inventing something new, but it's new to you. Yeah. And you can get the new to you ideas in your brain faster that will help you generate even better ideas.
Absolutely. That's what I think it's great for. And then also that broadening your perspective, getting perspectives from different angles. Looking at your problem or the scenario or the topic just from a different perspective, also a great reason to, to use it. Yeah. Okay. And number three is also, it's fun.
George Drapeau: It's fun. See,
Camille Rapacz: So last thing I really wanted to comment on about this is I've noticed there has been, of course, a huge explosion of just tools and apps using AI. And I don't know if I noticed this more so as a small business owner, I mean, everybody now wants me to, you know, use AI for my podcast, for my social media, for my, like every single thing that exists in my business.
There's an AI tool for, there's one for creating PowerPoint presentations, specifically a presentation. You plug in your idea and it spits out a slide deck for you. Yeah. With all the words on it. Converting, you know, transcripts into blog posts, social media posts, all that kind of stuff.
Just through an AI machine. Yeah. So while there are a ton of them and it seems like, why wouldn't I just use that? Mm-hmm. So far my experience, this is just my personal experience with this, is there's a, just a lot of bad ones. Like the output that I'm getting is just not great. It's, yeah, it's okay.
And if you wanna use it as a starting point, that's great, but I also think, well, I can just do that for free on ChatGPT I can tell it to write me three LinkedIn posts about this podcast episode, and it'll do that. I don't need to go to a special app for that. Mm-hmm. So that's just something I tell people to kind of watch out for.
That everybody's trying to benefit from this , everybody's into ai and hey, maybe if I create a tool that specifically solves the, the way that they wanna use it I can, you know, market that to them. I think some of these better tools are really gonna stand out and be great, but they're so, such an explosion of them now.
I just kind of say caution to people in which tools will be useful and which won't.
What are your thoughts about that?
George Drapeau: It's fantastic. I, I mean, I can say something different, but I can't really top that. I think that's a really, really cool way of putting it. You are right. A bunch of people are taking the technology and packaging it, packaging it to make it very easily consumable.
Love that. Yeah. Not great quality stuff yet. You're absolutely right. But, you know, I give people credit for doing it. And no, I think it's great.
Camille Rapacz: I mean, and some of those I think will end up going down a path of developing something really cool, right? Mm-hmm. But they can't make it really cool without putting it out in the world now and testing it out.
But the ones that I think are most useful for me so far are ones that were already working in that space to begin with. There's tools out there that we're already starting down this path that are just getting better faster, but all the new ones that are trying to jump in and take advantage of this, that weren't doing anything before, they're behind.
So they're not quite as, that's just not quite as good. The output just doesn't feel like it's quite as solid as I would want it to be. So just I see everybody keep that in mind. Yeah. Cool. Do, do you have other thoughts on the future of AI?
George Drapeau: Right now what's getting all the attention is this large language models stuff, the super bot technology, but AI is flourishing in, there's a bunch of different aspects of artificial intelligence right now.
For example, image recognition, which we know about is getting very good or image modification. Like if you go on Instagram and you see these influencers, If they're either using makeup to make themselves look better or they're using AI-based filters to make themselves look better, and it's the artificial intelligence technology to modify them is mind blowing.
It's different than this kind of, it's a different form of artificial intelligence, autonomous driving. So Tesla is the, probably the company you know about the best that has. Cars that can drive themselves. That's a different form of AI. But there's a bunch of stuff happening right now.
We are on the cusp of Indescribable amount of technology advancement. It's crazy. And I would just say, I guess I'm getting into philosophy here, the technology is advancing so quickly that people are worried about it overtaking our ability to absorb it societally, you know? But I, it is an amazing space to watch right now.
Camille Rapacz: Yeah. I, I think that was a really, that's such an excellent way to put it. The, the technology is outpacing our ability to adapt as a society to what it means to us and what we do with it. I just think that for, for us, as you know, working and trying to make our businesses better, whether we're business owners or business leaders, just back to the context of the podcast and even going back to the belief shifts, this is definitely a space to be curious and to learn, like go in with that learning mindset because again, this isn't going away and the businesses that are going to do well going forward are the ones who learn how to leverage these new tools.
Mm-hmm. They learn how to make them work for them. You can pretend that they don't matter to you and your business and you don't have that kind of business or whatever you wanna say. But the reality is it's just gonna become more and more prolific. It's probably existing in your world, in ways you don't even recognize, like mm-hmm.
Some of what you were just talking about. So it's already out there and you're gonna get left behind if you don't figure out what that means in your own world and understand it. I mean, you don't have to become a tech expert to understand how to use some of this stuff. A la, we did this podcast for you.
We sure did. So there are plenty of other ways to learn how to use this as a tool. Again, it can be used just like any tool. You can use it well, or you can use it poorly. And of course, all of my listeners are gonna learn how to use it well. So get out there and learn how to use it.
Try all the stuff that George said to try, absolutely ask ChatGPT to describe itself and all those voices. Super fun, but figure out where, for you in your professional work, it's gonna make a difference for you. It can help you be, ideally, it's gonna help you just be more efficient. That's what I think is really the benefit, is I can break through my own mental barriers faster.
I have an instant brainstorming thought partner person to go to, to start getting things moving. Whatever topic I'm stuck on, so that I find is just fantastic about it, and the more you use it, the better you'll get. There are tons of resources for this as well. Just Google, how do I do X, Y, Z on ChatGPT, man, and they're pretty good.
There's some great lists of tips and prompts and things like that. That's where I've just googled a bunch of questions to learn a ton of stuff, so I'll put some links in the show notes. Yes, sir.
George Drapeau: Yes. Actually I wanna make a, a shameless plug to get sponsored by Microsoft here by pointing out that I think Microsoft, you talked about Google.
I think Microsoft is kicking Google's butt in terms of incorporating ChatGPT if you go to the binging search engine, which has been fine. Oh, this is true. They have normal bing mode. They have chat mode and they're very clear. These are two different ways to search, which is brilliant, and it's the right way to teach the public about what this could do, not make them one thing.
So Microsoft, if you like that honest evaluation of what they've done and you feel like you wanna support the podcast, we would be very happy to receive your support.
Camille Rapacz: Absolutely. Great job, George. Yeah. You know?
George Drapeau: Yeah. Actually, what you should do is you should have some seminars for having Camille on campus and teaching them more leadership skills, but you know, also support the podcast. I mean, sure. Why not?
Camille Rapacz: That's so true. I am only right in their backyard, so, you know, you would think. But yeah, I think that's a great point. I, I forgot to even talk about that. I'm so glad you brought up Bing, because nobody even thinks about Bing. We just think about Googling things, but we should be binging now.
Like that's actually a great place to go now to, to do the search for how to use ChatGPT better and that you can access both just from that place. Yeah. Fantastic.
George Drapeau: They've done phenomenally well. Bravo, Microsoft. Yeah.
Camille Rapacz: Yeah. Bravo. Gotta say.
George Drapeau: Good for you, Michael Soft.
Camille Rapacz: Oh my gosh. Yeah, that is an old joke.
People listening. So he did say Michael Soft, and it's because long, long ago I had a coworker when Microsoft was just becoming a thing, I had a coworker that actually thought it was a guy and that his name was Michael Soft. So that's what we called him now privately. That's hilarious. It's hilarious.
Alright, any last thoughts on ChatGPT before we jump outta here? Nope. All right, well thanks for listening everybody. If you have thoughts on ChatGPT or anything else that we talk about on the podcast, leave us a voicemail. We really wanna, you know what? We're kind of lonely and we really wanna hear from you.
This is like a one-way street, and that's the only bummer about podcasting is we just put it out there and we don't hear back. So we'd love it if you left us a voicemail. Just go to the belief shift.com and you'll see a little widget on the side that says Send voicemail, and you can just do it right there.
Also, if you just wanna do some work with me and maybe some leadership developments, Microsoft, or maybe some, whatever you feel like you wanna work on. Or if you're just thinking about maybe hiring a coach or a consultant, but you don't really know what that means or what kind of budget you'd have to put together.
You just need to get on the phone with me in a free consultation, so you can go to camillerapacz.com/bookacall, that's all one word, book a call, and you can book a free consult with me. Chat for an hour. Awesome. Not ChatGPT chat. It won't be a bot. It'll be the real me. The real me talking to you. That's all I got for this week.
Awesome. All right. We'll be back in your ears next week everybody./
