← The Amazing Intern - Turning Your A.I. Curiosity Into Confidence
The Amazing Intern - Turning Your A.I. Curiosity Into Confidence
Ep 47: Just Have a Conversation: A Beginner's Guide to ChatGPT with Jim Sterne
00:45:53
In this episode, Cary sits down with Jim Sterne, a seasoned marketing consultant, speaker, and author known for his work in web analytics, marketing technology, and the impacts of generative AI on business. About Jim Sterne: Jim Sterne...
Loading summary
Transcript100 lines
- [00:00]
T-Mobile Advertiser
It's better over here at&T customers switching to T Mobile has never been easier. We'll pay off your existing phone and give you a new one free, all on America's largest 5G network. Visit t mobile.com carrierfreedom to switch today.
- [00:16]
Pay off up to $650 via virtual prepaid MasterCard in 15 days. Free phone up to $830 via 24 monthly bill credits plus tax qualifying port in trade in service and go 5G next and credit required. Contact us before canceling entire account to continue. You build credits, your credit stop and balance and required finance agreement is due.
- [00:32]
Jim Stern
And it will delight you and shock you and surprise you and scare you. And you got to go through all of that before you can say, okay, I am going to treat this like my best friend slash mentor professor who I'm talking to in a bar at closing time.
- [00:54]
Kerry Weston
Hey there. Welcome to the ChatGPT experiment, a podcast where we take all the confusing ChatGPT stuff and break it down so even Grandma could understand. If you're a curious beginner, well, you're in the right place as we'll help you understand what ChatGPT is, how others are using it, and practical takeaways that you can use right away. It's gonna be fun, it's gonna be informative and maybe surprising when you find out how helpful ChatGPT can be. Alright, let's get this thing started. Here's your host, Kerry Weston.
- [01:28]
Hey gang, how you doing? Hope you're doing well. Got a good conversation to share with you today. Welcoming Jim Stern to the show. First met Jim out in Denver at a conference called Building a Better Agency Summit and saw Jim on stage talking about artificial intelligence and marketing and data and I said this is a guy that I love to have on the show. Got to talk to him at dinner that night and brought him on. I'm very pleased and excited to share it with you. A lot of good stuff coming out of Jim's head. Let me tell you a little bit about Jim in the background here and a couple things you can look at for the show. Jim is the president of Target Marketing in Santa Barbara and in 2017 wrote a book called Artificial Intelligence Marketing. And since about 2000, I want to say 2002, he's had the Marketing Analytics Summit so they're more than 20 years in. It's out in Phoenix. I'll put links to the show or to the summit in the show notes so you can see it. But really interesting conversation, really interesting insights. I think you're going to get a lot out of this. One thing that came up very interesting during our conversation at dinner is Jim's got his degree in Shakespeare of all things. So from the world of Shakespeare to data analytics. And he builds the bridge here during the podcast so you can connect the dots. But we all do talk in stories, huh? And so whatever we do has to be delivered in a good story. So Jim certainly brings that through. A few things to think about or to at least listen for during this conversation. You know, Jim's going to share a lot about the trepidation that businesses still have about how to use AI tools like CHAT GPT. He's got some good insights on that. He emphasizes that folks struggle, and I see this too, folks struggle to really adopt and know how to use this because we're still using, using it like, like searching for facts. Yeah, he's got some good tips and insights on what's happening and how to really make the most of it. And he's going to share some practical ways that he uses ChatGPT for things like organizing ideas, improving document flow, using as a creative thinking partner. Just an all around good conversation. Going to be eye opening for a few, I think really valuable here to have on the show. So let's get to it.
- [03:56]
Jim Stern
Right?
- [03:56]
Kerry Weston
Here's my conversation with Jim. I hope you enjoy it. If you got any feedback or questions, shoot it my way. But otherwise, until we talk again, do stay curious. Okay, here's my conversation with Jim Stern.
- [04:15]
Jim Stern
Well, hey Jim, welcome to the show.
- [04:17]
Thanks very much, Carrie. I'm very pleased to be here. It is an honor to show up.
- [04:22]
Well, I appreciate you, I appreciate you saying that. Thanks for showing up. It's a much better interview when there's two people. So we met in Denver a few weeks ago at the BABA Summit for Agency Management Institute and I was fascinated by your presentation and thank you. Wanted to share what you where you've been, what you've done with the listeners. I know there's going to be tremendous amount of value. I want to start though, if you would, just to give context, would you just share with people your own version of who you are, what you're doing and who you're doing it for. These days.
- [04:53]
I am the guy who looks over the horizon at technology to see how it's going to impact marketing specifically and business in general. So I first job out of school was selling Apple 2es out of retail store and explaining computers to people. And then I sold business computers to companies that had never owned one before. And then I sold software development tools, so programs that help programmers write programs. And then in 1993 I tripped over the Internet and went, oh wait a minute, this changes all of marketing. And I ran around and asked a bunch of people what they were doing and they didn't have any answers. And they all asked, what are the other people doing? Oh, that's what it means to be a consultant. You ask a lot of impertinent questions and then blab to everybody else. And started writing books and public speaking and what, what should you do on the Internet? And how do you make a good website? And then in about 2000 I discovered analytics and went, oh, wait a minute, it's not my opinion that your website sucks. We can measure this stuff. And so in my, what, ninth, tenth book was Web metrics, How do you measure the success of your website? And I started a conference in 2002, the Marketing Analytics Summit, which I just finished up last week. So that's 23rd year.
- [06:20]
Congratulations.
- [06:21]
Thank you. Of, of like what, what can you measure? How should you measure? What will it tell you? And then that led to artificial intelligence. If you're doing analytics, you need to do machine learning. So my book Artificial Intelligence in Marketing Practical Applications is seven years old now. It's about machine learning and of course since November of 2022, all in on generative AI.
- [06:51]
In reference to the book, you had future predictions towards the end of the book and one of them was this far out theory that AI was going to actually start creating content for you seemed to come to fruition. Huh. Seemed to come to fruition. I thought that was, I thought that was interesting. I also liked your voice and visual search. So I'm going to combine those two. When you talk about what did you see then when you were talking about voice and visual that led you to believe that that was on the horizon?
- [07:20]
I was watching the fact that we were moving from deterministic programming, do what I tell you to probabilistic, which is machine learning and statistics. And if it could learn things, then it could learn by seeing and learn by doing and learn by example. And just the very basics of that were starting out. But it seemed obvious that that would be, if not technically feasible as quickly as it turned out. It was clearly desirable that smart people would be working very hard and spending a lot of money to make it happen. And oh my goodness, it's happening.
- [08:08]
Interesting too to see the, you know, the, the five thousand dollar gadgets hooked to a battery pack on your belt and you know, this is, this is the Jetsons coming to life. And someday it'll just be a pair of glasses, if not something embedded into our screen. But Those, those things 20 years ago that seemed futuristic are now. Oh, yeah, that's probably being done somewhere near where you are at the moment, actually.
- [08:27]
Right. Yeah. Back in, in 2018, I did a presentation where I imitated having a conversation with my personal bot. And it's, it's part of, it's one of the featured posts on my LinkedIn page of my, my butler telling me that I needed to get my car serviced. And that is the, the technology is there to do that today.
- [08:53]
And yet I can't start my defrost in my truck from my phone. And that irks me.
- [09:02]
You will.
- [09:04]
I know I will. But right now I can't. And I'm in Bangor, Maine, and the winters are long and the summers are short. All I want to do is start the defrost from my kitchen. That's all I want to do. So.
- [09:13]
And, and before, before you are able to start the defrost from your kitchen, the system will be in place to know when you want it. It will observe your behavior. You will be on your third cup of coffee, and it's going to know what your shopping list and errands are, and it will start defrosting 20 minutes ahead of time.
- [09:32]
Yeah, it's fascinating, isn't it? It's fascinating. So let's, let's go to what you see today, because the whole point of having this conversation for the audience that we do is to kind of shed some light on what's happening specifically around, you know, language models. And ChatGPT is where I focus, but helping beginners and curious folks kind of better understand where do they lie in the matrix of normal? Like, what is it that I know and don't know? And how does that compare to what the average Joe knows and don't? And are they far behind because they haven't touched this yet, or is everyone smarter than me? All that kind of stuff. You're spending an awful lot of time working with businesses, talking about generative AI and using tools like ChatGPT. Can you just give us a broad brush? Landscape of. If you walk into a room of a hundred people these days in business, what does the landscape look like as far as comfort and confidence in using these tools? You know, as a general percentage of.
- [10:31]
That ring, 3% have not touched it, 75% have touched it and thought it was broken because they thought it was Google and it gave them weird answers and they didn't understand it. And three people are automating things, building their own GPTs and, and just running full speed ahead like crazy. And one guy in the corner is actually a programmer and is not just building his own GPT, but is building his own large language model or customizing an open source. And that's a data scientist. I, I don't have a lot in common with him.
- [11:20]
So I hope the listeners just heard that on average 75, 80% of the people that you're going to run into at any normal event, conversation, presentation or whatnot are still in this. I don't really know how this works. I might have touched it and I'm just not familiar with it and who knows if it has any value for me.
- [11:40]
So there are, there are four stages. There are four stages. We, we start with, well, what I know is Google. And so I ask it a question and it gives me weird output. I don't get it. And then the next is I kind of lean into its creativity and ask it for its opinion. And that's really interesting. And I have a conversation with it instead of Q and A. And then the next step is automation where I get it to do things for me like clean up this list of email addresses. It's, it's, you know, there's all kinds of punctuation in there and emojis and stuff. It's like it can clean that up for me. And then the fourth step is really engaging with it in a creative way and understanding how different it is from computing as we know it. So what I mean by that, there is this new alien intelligence that has arrived on Earth and it's called a computer. And we have to learn to speak computer. We learn assembly language and we learn binary. Like it's either on or off. It's, it, it adds, subtracts, multiplies, divide. Now we're not used to that. We're used to having a conversation and there being ambiguity in words. No, no, no, this is all math, all the time. It's black or it's white. That's it. And if it doesn't do that, it's an error message. And if it's really confused, blue screen of death. And we have to learn how to get along with this alien. Now we've taught computers to speak English and we have to change what we believe about computers. The, that they're actually more humanistic and less deterministic. And that's, that's changes are the shape of our thinking about how to use them.
- [13:40]
And that's difficult For a lot of people, a lot of people to get their head around.
- [13:44]
It's a, It's a paradigm shift. It's. It's awkward.
- [13:47]
It is awkward. And there's. It's no surprise to you how much anxiety is created when we actually introduce technology. Not even AI, just technology and computers. Right. There's a frozen. I don't know how to do that. I don't know where I'm going to go. There's a comfort level. And in a conversation I had yesterday about training and where the future is going as far as learning and workforce and job development and whatnot, there's a. And thought about this way. But there's a sentimental value that we apply to things. Right. We apply it to cars and houses and possessions. So it's worth more we're going to sell it than maybe market value. But to us, it's worth more. We also do that to our own skill set and our own value in the work. Yes. Right.
- [14:25]
Well, there's also just fear of change. I mean, I got where I am because I know what I know and I'm doing it the way I'm doing it.
- [14:30]
And you want me to, what, protectionism there too? Yeah, it's like I need to. I need to hug it a little harder.
- [14:37]
Right.
- [14:37]
The cement life jacket. Right. I remember the cement life jacket.
- [14:40]
That's good.
- [14:41]
When, when the ISO 9000 standards came out, I remember walking through a manufacturing plant and asking what the ISO 9000 metric meant, what that sticker meant on the window. And he said, we've been certified by the ISO 9000 gods to say that we are consistent in our process, we are standard, we have done everything the same way over and over again. Then he said, you know, we could be making a cement life jacket and still have ISO 9000. There's no logic as to what we're building as long as we do it over and over again at the same time. That's right.
- [15:09]
The probably the most crucial thing for people to understand and the hardest thing to get around in this, like, it's a different technology perspective. It's not a database, it's not a search engine. It's not a logic reasoning system. It's not a calculator, it's not an encyclopedia. It doesn't know thing, but it acts like it does. All it is is a mathematical model of the English language. How far apart are words usually? So the cat in the hat, we know. We, I mean, we all, we all grew up reading Dr. Seuss. Well, actually, the cat could be in the tree or the store or the window or the chair. But statistically, the next most likely thing is hat that is all it does. And yet what it generates is actually mind boggling. We, we anthropomorphize it. We think it has reason and, and compassion and smarts, but it doesn't. It just fools us. So the, the trick here is that humans, we have, we have meaning and feelings and we try to find the right words. A large language model has all the words, but has no concept of feeling or meaning. So that's why it elucidates. Sometimes it just says weird stuff. If you treat it like a search engine, you will be disappointed. And if you try to use prompt engineering like a programming language, if you are actually fine tuning a model and doing embeddings and retrieval, augmented all the things data scientists do to try to make it deterministic, good if you're a data scientist. If you're not, prompt engineering is a bit of a waste of time. Instead, have a conversation like, I, I need to buy a new washing machine. What, what features should I consider? Is the question not. You are an expert at appliances for the kitchen and I need to know exactly the specifications of it. You should research. No, no, no. What should I consider? And then have a conversation with it, turn it around and get it to ask you questions.
- [17:28]
And that is exactly what makes people's eyes get big right there. And I can share with you that. And leading up to the university, there was a workshop today. I had 38 or 40 people on call with me and we went through this whole conversation prompt about finding value about the company and customers and all the stuff that we do in marketing. And then somebody asked me, do you think that it could make an elevator speech for me? Right. And I, and I said no. The question was, how long do you think it would take for them to make an elevator speech for me? Okay, well, I was sharing my screen, I said, can you make an elevator speech for me? And says, sure.
- [18:06]
Here's, you know.
- [18:08]
And he's like, oh, I didn't know it could do that. I was like, yeah, well that's, you know, that's what we're doing. We're just having conversation. If you don't know, ask. And what was really fascinating is when we started showing prompts about, I want you to interview me one question at a time so we can get to a conclusion. And Right. It's that it's, it's a computer. I'm not supposed to be able to do that. That you know, nobody says that out loud, but that's the reaction I get is that this is a computer. It's supposed to be me forcing stuff into it to a predictable outcome. But this is really two way street.
- [18:37]
And that's, that's we, we learned that computers are this binary thing, this weird alien tech. If we had this first, we'd know exactly what to do. You'd have a conversation. I mean, what I like to say is it's a conversation with a really brilliant complete stranger in a bar around closing time. Fascinating conversation, full of great stuff. But like I'm going to go home and sleep on it before I copy and paste or trust any of it. So don't ask for facts, ask for opinions. So, example of a fact. The conference where you and I met, I took three different pictures of the audience and I told chat GPT. Here are three different angles, different parts of the room. How many people are in the room? It said, well, in this picture there's 10 rows with about 10 people. So that's 150 people. No. And the second picture, same thing in the second picture, same thing. 150. 150. Therefore there are 150 people in the room. So you don't do multiplication and you don't do addition. I asked for a fact, silly me. Ask for an opinion. What can you. I then I went around the front and took a picture of faces. What can you tell me about the people in this room? Well, they have an average age of and they're professional and they're dressed nicely and they seem to be at a conference. Blah, blah, blah, blah, blah. It's like, great, let's have a conversation.
- [20:05]
No, it's fascinating. So you opened up by talking about your definition of what a consultant is. So I know that's what you're spending your time doing these days and you're talking to a bunch of people and organizations and whatnot. Could you share with us some of the normal conversations, the repetitive conversations that come up with businesses that are looking to. I don't know if adopt is the right word because that seems a little too formal, but utilize right training. Begin implementing what kind of conversations are coming from business owners and leaders when they are looking to engage in this stuff.
- [20:38]
The three conversations are fear first. So we are. Everybody's worried about privacy. We've been told not to upload any of our information. It's like, well, that was true up until a couple of months ago. If on an enterprise scale, bring in a large language model inside your firewall, fine. Or create a secure instance up in the cloud with one of the cloud providers. Fine, but now with ChatGPT, there's also a button, a setting that it will not learn from your stuff. Do you trust them? Well, I don't know. How much do you trust? Fill in the blank. Anybody else? So, yes. So the misconceptions about, about the IT side of it have to be overcome. Then how do we get the most value out of it? Right, actually that's the third one. Let me come back to that one. Because the second one is, is there a process by which we can bring this in and adopt these capabilities? And yes, I've, I've published the seven point plan of do a survey. How are you using it now? What, what value and what mistakes are people getting already? Executive alignment? How risk averse is your company? Or move fast and break things. Does everybody at the upper level agree on what the vision should be? Form an AI council so that the excited people get together and work together and keep track of what's going on week by week. It's just crazy how fast it changes. Number four, policies. Working with it and the legal department. What can you do? What you'd be allowed to do. What are you regulated against doing then? Now that I have policies, I can put together a training program and this is what you should do. This is what you may not do, this is what you should try. Here's how you should share with everybody. This is how we should go about using it. Then we get to the fun part of how do you add generative AI to your services and your products and your processes? Like we can automate stuff, we can do more things with the same people. We can figure out workflow, workarounds that will speed up everything. That'd be great. And then finally, oh, and adding it to your products because by golly, I want to talk to my car and say please defrost for me, thank you very much. And then finally, how do you measure whether or not it's valuable? And that one I actually published a paper on of measuring the success. The measuring the value of generative AI in a corporate environment is classic business metrics. You know, profit, loss, revenue, expense and, and then how do you measure creativity? Because that's the big, that's the third conversation is this is about a boost to your creativity to help you become a more strategic thinker, a more innovative thinker, a more critical thinker. I mean asking, asking questions that we don't know that we could ask. How, how I have this problem. How Can I deconstruct it into smaller bits? How, what are some alternative perspectives? What am I not thinking about? How can you expand on this idea? What are, what are the unintended consequences? How do I, how do I improve? Give me, give me 25 hypotheses about why this might have gone wrong. And you, and you share it. You don't ask for facts, you ask for opinions. And you think about how that can help you. You, you use it as a brain boost. I mean, when, if we really want to get up there. One of my favorite things is.
- [24:27]
Here.
- [24:27]
Are three people I'm going to, I'm going to, let's say, pitch a new product. I'm going after venture funding. And these three people are going to be in the room and here are their LinkedIn profiles and here's some blog posts and here's some emails that I've shared. So I want you to instantiate those three people. I want you to create a Persona for each one. Here is my PowerPoint deck, here's a video of me presenting it. And I want you to simulate their questions and answers so we can do role playing. And then I want you, them to, I want you to have them discuss this whole thing when I am out of the room and tell me how I can improve my presentation and what vocabulary would be most persuasive for each of the individuals. Go and like, it just, it just generates. Is it, is it solid, is it factual? I don't care. It's useful, it's valuable, it's creative.
- [25:22]
It's a five minute Internet research tool, right? I mean, I'm smiling as you say that because you know what I did 10 minutes before we jumped onto a cause, I went to chat and I said, here's who I'm interviewing for the show. Here's his LinkedIn profile. Search the web, give me a summary, give me some questions, give me relevance. Here's the part, right? I spend more time creating the question than I did reading. You know, it just spits it out instantly. Yeah. And so I've got a whole primer next to me of stuff that, you know, should I need it, and I'm not going to, but should I need it. I've got emergency questions and all kinds of background stuff here. I mean, it's just imagine using that to go into a sales pitch or just go into a job interview or just to have a normal conversation with someone that you're meeting for the first time. You know, it's, it's, there's a wealth of stuff out there.
- [26:08]
I was at the LinkedIn Learning campus yesterday filming a generative AI for marketing. And there's another session that was being recorded. Generative AI for salespeople. Like yes, of course, this is exactly that ideal. But you know, everybody's a salesperson. Everybody's job is to persuade somebody of something and oh, I need to improve my fill in the blank skills. Can you help me? Why, yes, I'd be happy to help you delve into the luscious landscape of the. Okay.
- [26:41]
Be succinct in an ever changing world. Yeah. So I've got my answer to this that is only 24 hours old because of an interview I did yesterday. And before yesterday, my answer would not have been near what it is now. But I'm going to ask you the question and see what your answer is as we take a look at. We looked at you. Obviously you went through the future of marketing and you spent a lot of time on how this application applies to traditional marketing practices and all the things that we need to be doing. What's the most, what would be the most creative, eye popping way a business you have seen is using AI in a business model? Now that you've run into or heard of, like, what would be a story that you would share that would be like, oh my goodness, I didn't know that was happening. I did catch off guard with that one. I know, yeah.
- [27:35]
Do I get 24 hours?
- [27:36]
So you get 24 hours.
- [27:38]
The reason that I'm stumbling is that this is this. Let me, let me go back to my history. Here's an Apple Iie. Hey, cool. What can I use it for? Anything. You can do anything with it. Well, that's not helpful. Look, here's the World wide Web. Really? What can I do with it? Anything. You could. You'll be able to sell books on this eventually. Like, I don't believe you. Oh, what can I do with generative AI? Well, what's your problem? What?
- [28:04]
Right?
- [28:05]
I, I'm stymied because every time I have a conversation there's a new opportunity, there's a new solution, there's a new business model. It, it, there's, it's so limitless. I do not have an answer.
- [28:18]
I only ask you this. Well, yeah, yes, exactly. And I only ask this because I never would. Well, I wouldn't have thought it possible right now. Let me say that out loud. Never is a strong word. I just, I didn't think we were there, but I interviewed somebody yesterday who's in the creative field. A creative director in the advertising world and has had 20 years of creative directors for doing creative work for cosmetics consumer brands. You know, the. The brand that spend $200,000 to send a crew of 10 to a tropical island to come back with 12 photographs. Right. That they're going to use in their Vogue magazine ad. You know what I mean?
- [28:57]
Yeah, yeah.
- [28:58]
His entire business now is AI. So he has prompt. He has created custom GPTs that link to Photoshop, that links to motion, that links to graphics, that links to polishing, that links to all these things, and is working with some of the world's biggest brands and is creating motion, is creating imagery, is creating photography, is creating graphic and print ads at a scale. He said 10 times scale. But when he got through, I'm like, that's not 10, that's a multiplier of a hundred. I mean, the ability for him to create. To take a cosmetic brand, 3D model the bottle, create futuristic worlds and output print ads, glossy banner ads, video motion in a matter of a day.
- [29:49]
Right.
- [29:50]
Is just. To me, I told him as I'm watching this is. I feel like I need to be eating popcorn and just watching this happen, because this is a movie for me. This is not real. Yeah, but that stuff's happening.
- [30:04]
The. The important thing here is this person you talk to has how many years of experience?
- [30:11]
That's exactly right. 20 years of doing the creative side.
- [30:15]
Right. So can I do exactly the same thing he does with exactly the same tools in exactly the same amount of time? Absolutely. My output would be absurd and unusable. Ridiculous. I can. I have written a dozen books. I can get ChatGPT to write me a book, in fact. Yeah. Are you familiar with Ethan Molik?
- [30:40]
Yes.
- [30:41]
Well, good. So. So he's got a new book out before his book came out, his biography. I'm using air quotes here. His biography was published on Amazon and I immediately snapped it up because it's 100 AI written. It is laughably horrible. It is entertainingly ridiculous. It is so bad that it's kitsch. And that's the difference is, yes, you can get amazing amount of output, but if you don't know what you're doing, it's. It's. Oh, what's the new word? Gosh. The Internet is being overrun with all of this AI output, and it's being referred to as.
- [31:23]
A new spam. Yep, there'll be something there.
- [31:25]
It's an AI version of spam. I just heard it yesterday on a podcast. You'll have to look it up and edit this Part out.
- [31:33]
We'Ll leave it in. But I think authenticity is, is, is a good element that we need more of. So.
- [31:39]
But it's, but my, my point is that it's obvious when it's bad. It's just like it's, it's the ever changing landscape, delving into the etc. You can tell unless the person using the tool is talented. So let's look at when cameras came out. Painters said, oh my God, that's not art. And sure enough, it wasn't. And we all have cameras in our pocket and we're taking a gazillion pictures a day and they are not art. Except for photographers who are using cameras in their pocket. Produce beautiful photography because they know what they're doing. Photographers can prompt a mid journey or a dall E to produce something really interesting because they know what a lens flare is and what an f stop is and use this kind of film and in the style of. And I don't have that. I have language because I got my degree in Shakespeare. So I have an advantage. But I can't create pictures that look anything like you would want to use in an ad.
- [32:45]
That's absolutely right. And it's the human experience and element you bring to it.
- [32:48]
Right.
- [32:49]
That makes it what it is. I am. I did see that you shared this with me in Denver. What did you start doing when you had your Shakespeare degree?
- [32:59]
What did I start doing?
- [33:01]
Where did you, where did you. When you grad, when you got your studies in Shakespeare? What was the first thing you did once you achieved that? I'm interested. Were you on stage?
- [33:11]
No, I went and got it. Well, yes, but it was in a very small stage in retail stores selling Apple IIe's. But it was performance, it was a script. What drove you to Shakespeare? Yeah, but Shakespeare. Yeah, I did, I did some community theater. Yes, I did some Shakespeare. And I did become the. Was on the board of a local Shakespearean theater company in Santa Barbara that performed all over the world. And that was fun. But me, actor. Yes, Professor, Yes. But in a business environment where my audience has, has paid good money to be there and, and wants to not just be entertained, but benefit from the education.
- [34:00]
So I've seen you on stage. Do you. Are you still tapping into the Shakespearean studies as you are communicating and presenting these days?
- [34:09]
The theatrical side? Absolutely. The language side, forsooth, none would understand.
- [34:15]
They would not understand. My, my daughter was in Macbeth and I think, I think I, you know, school play. Right. And I think I understood about six minutes of the whole thing. Right. So, yeah, it's, it's interesting. So I'm going to ask you if we can bring this down. I understand the Fortune 100, the Fortune 1000, the big structures where we have levels and we can talk about councils and best policies and spend time doing that. What about the mom and pops, the small businesses that open that virtual door of what we're talking about and look at just. Are just completely overwhelmed by what you just said, which is anything's possible if you know how to do it. And it's just overwhelming. Just overstimulated. Where, where does. Where do you advise them to.
- [35:05]
To go? Start.
- [35:06]
Begin. Boy can.
- [35:10]
Be studious about keeping track of the work that. The actual work you do. So this, this. The. Ethan Mollick says that a job is a bundle of tasks and AI is really good at some of the tasks and it cannot do the job. So your. Your responsibility is the job. Keep track of your tasks in. In minute detail. So every time I am going to fill in the blank, I have to do A, B, C, D, E, F, G. Some of those things. Well, all of them. Go to ChatGPT and ask it if it can help you. That is my entire advice and it's.
- [35:54]
So simple from a conversational point of view that most people, when I share something like that, they'll say, well then, like, well then you read what it gives you and go from there.
- [36:04]
Ask it to clarify, Ask it to go deeper. Ask it to give you an example. Ask it to. To write code to do that for you. Ask it what else can you do for me? How. How could I have asked this question better?
- [36:19]
Where do you find yourself using tools like chat these days?
- [36:24]
Ideation.
- [36:25]
Just.
- [36:26]
Just. I. I have opinions and I am a writer so I want to express those. I, I also. My output is also PowerPoint as you saw. And the. The one. The biggest go to that saves me the most time is coming up with. Instead of using the thesaurus, it's like there is this word, there's this phrase and it's like this and like that, but it doesn't mean this and it does mean that and a thesis. The source. I just. It's a rabbit hole. I get lost. ChatGPT says oh, you mean this? It's like oh yeah, that's it. Thank you and I can move on with my life. How do I. Is this argument in a. In a good format? Please look at this article. This was a. A learning moment. I wrote 12 paragraphs. Please write the introductory article. Introductory paragraph. And it was worse than bad. It was reprehensible. So I spent an hour prompt engineering to get it to do something better, and then I just spent five minutes writing it myself, so don't use it for that. And then I said, please evaluate this document for structure and flow. And it said, I should take paragraph seven and put it underneath paragraph four. But that's the weirdest thing. And I tried it, and it was a lot better. It. It provided the. The conceptual sequence of events in a clearer way. I had to change transit transitions between paragraphs, but the conceptual flow was significantly improved. What else might I have added to this article? Oh, well, you talk about this and this and this and this and this. And I went, yeah, those don't apply. Thanks.
- [38:13]
It's remarkable to me how often I am using voice, either directly to it, or I'm using a transcript from a call, or I didn't get to record the call. So right after, I'll just do a voice memo and just get everything out of my head in no particular order with as much detail as I can muster, and then ask it to organize and put it, just like you said, put it in some sort of logical order, Give me some summary whatnot. And then it's a. It's a remarkable it. You say it doesn't have logic and feelings, but at that point you start to question, because I'm like, okay, so it's reading it as it would be valuable to a human. And you almost fall into this romantic view that it's thinking with you, you know, Right?
- [38:54]
So this is. This is the. The balance. Understand that the technology is doing something very simplistic, but the output is so weird that you think it's human, and that's fine. Treat it like a human that you don't trust 100%. Right? I mean, there's. Up until you're about 12 years old, you believe absolutely everything your parents tell you. And then you go, well, I'm not so sure about the Easter Bunny. What else are you hiding? And then you learn how to get along with humans. We. Our brains are built to accept noise and inconsistency and ambiguity. And. And we look at a computer and go, oh, no, it's a calculator. It just gives us the right answer. No, go back to pretend it's a human. But don't, you know, don't fall in love. If I.
- [39:54]
This is a theoretical question, but bear with me as I ask it. If I were to come to the office or studio, wherever you're at, and you had a Whiteboard with the one or two things that are keeping up at night or the big things you're trying to solve or the thing that you're really into these days. What would be on that board?
- [40:12]
Oh, problem to solve. Time management. There's just so many things I want to do. Everything is so interesting. What am, what, what am I excited about? I'm excited about the fact that, that while there are a hundred thousand prompt engineering courses out there, I have not yet had the moment to sit down and go, all right, here's the framework for addressing a large language model in a creative way. Here are 17 kinds of questions categorized appropriately. In these situations, you should ask these things, and in those situations, you should ask those things and think of it this way and create GPTs for yourself to do that aren't about productivity and efficiency, but are about innovation and creativity. That, to me, is exciting.
- [41:08]
So as we wind down our conversation and I really appreciate you taking the time to talk to me today. You just concluded your 23rd song. Have you debriefed yet to say this was unexpected and I've got to make sure we do more of something next year?
- [41:27]
I have debriefed about the event itself. I have not had a chance to debrief about the content, about what I learned. I've got pages of notes. Yeah, you know, you go to a conference and you spend a week creating a week's worth of stuff to do while you've missed out on a week of what you were supposed to do, would you try to catch up in the following week and. Oh, yeah, you know, I do.
- [41:52]
And I, I remember being the first time, you know, the. Again, back to where you and I met in Baba with Drew McClellan's conference. The first time I met Drew, I was in Nashville at a conference with Paul Raitzer and Michael Gass. It was a content conference in Nashville. And I remember there's 150 people in the room. And I think it was a gentleman named Michael Gass. And he was talking about how to be meaningfully different in your marketing by creating content. And he was really talking about the learning center and the blog. But this was 15. 20 years. 15 years, right? And somebody raised their hand and he said, I need to ask you a question. You're telling 150 people to do the exact same thing. What's going to make us all different if we all do it right? And the answer sticks with me. Today. He look, he said, let me tell you something. There's 150 people in the room, and I'm telling you all to do the same thing. If one of you actually does this, it's a win, because here's what's going to happen. 50% of you are going to forget this the minute you walk out the door. Half of you that that didn't are going to take a phone call or have a deadline or have an email that you've got to address, and then you're going to postpone it. And then by taking back to the office, something's going to be on fire. And he went down through the whole list, you know, and he's like, so if one person actually does this, it's considered a win. And that's exactly right. You go to these things. The biggest challenge is not filling up the notebook. It's actually getting to whatever it is you put in the notebook, isn't it?
- [43:15]
Hey, Chad, GPT Here are all of my notes. How do I prioritize the to do list? Give me some criteria. What's the most important thing to you? I mean, what, what do you do as a consultant? You ask impertinent questions and, and you watch them come to their own conclusions. Well, treat ChatGPT that way. You're my consultant. Help me do this. Of course I'd be happy to help you. Tell me more.
- [43:42]
Yep. I hold up the mirror. That's what I tell people I do as a consultant is I hold up a mirror, I ask questions and watch the answers bounce off the mirror and come back to you. You probably haven't said it out loud. Is there anything that we haven't covered that you see that would be important for the audience that, you know, the podcast has to kind of hear from you.
- [44:02]
What's important is, I'll quote Ethan Malik, you got to spend the time, you've got to spend 10 hours with it and before you have a clue about what it means. And it will delight you and shock you and surprise you and scare you. And you got to go through all of that before you can say, okay, I am going to treat this like my best friend slash mentor slash professor who I'm talking to at. In. In a bar at closing time. And I am going to be. Yeah, here's an important one. Be as open and honest and vulnerable with it in ways you. You would feel really awkward talking to a human about this. Now, we. We've been taught not to do that because Google remembers everything. But you can hit that switch on ChatGPT and say, don't you know this is a temporary question. Don't remember this. Don't attribute this to me, but I'm really worried about blah blah blah, my relationship at home or my boss or this client or my place in the universe. Like get get real with it about everything that you're doing and it will surprise you by coming back and saying tell me more.
- [45:27]
That's a fantastic piece of advice and a great way to close. I know there's some great value here and I appreciate you spending the time. Thanks for thanks for joining me and sharing some nuggets. I appreciate it.
- [45:37]
John. My pleasure, Carrie. Anytime.
- [45:39]
We'll catch up soon, I'm sure.
- [45:42]
Kerry Weston
All right folks, that's a wrap for today's episode of the ChatGPT experiment. Thanks for hanging with us and diving into the world of ChatGPT. We hope you found it helpful and fun and informative. And speaking of helpful, we'd sure appreciate you taking a moment to rate and review us on your favorite podcast platform. By the way, five star ratings are our favorite. Till next time. Keep experimenting, keep learning. And remember, ChatGPT may be smart, but it can't make a decent cup of coffee yet. See you soon.
- [46:24]
Now.
- [46:24]
T-Mobile Advertiser
AT T Mobile get four 5G phones on us and four lines for $25 a line per month when you switch with eligible trade ins all on America's largest 5G network.
- [46:35]
Minimum of 4 lines for $25 per line per month with autopay discount you using debit or bank account. $5 more per line without autopay plus taxes and fees and $10 device connection charge phones via 24 monthly bill credits for well qualified customers. Contact us before canceling entire account to continue bill credits or credit stop and balance on a required finance agreement due bill credits end if you pay off devices early. CT mobile dot com.