Jason: Okay, we are now going live. Hello, Jim. Thank you so much for coming as a guest to do this skills training for all the Engage AI customer and users. Super excited to be hosting you for this event.

Jim: Thanks for the invitation. I’m happy to be here.

 The pleasure is ours. I read that you have published a couple of books.

Jason: One of them being the AI for marketer. And when I look at the date, I realize that you published a book probably quite a few years ago before

Jim: Seven years.

Jason: Before ChatGPT was introduced, so you were ahead of the time.

Jim: Yeah, the book is about machine learning. Supervised, unsupervised, reinforcement learning, classic or traditional AI, and now we have ChatGPT with generative AI.

Okay, time to come up to speed on that.

Jason: Wonderful. I will be starting this training very soon. Thank you. So I’m going to make a few intro about what we do and also share with people. And then once I finish that, I will get you started. How does that sound?

Jim: Sounds great.

Jason: Wonderful. It is coming true. Now, while we are waiting for the file to upload, maybe you want to give us a sneak peek about what would you be sharing with us today?

Jim: I want to talk a little bit of history. So how did we get here? What is the difference between classical machine learning and this generative AI thing?

I want to talk about how it works. Back in the early 1990s, I had to explain to people what packet switching was. So that they understood the difference between telephone broadcast and internet. Internet was a brand new thing. And if you didn’t understand the technology, it didn’t make sense. So I want to explain some of the technology of generative ai so that What comes next?

What you can do with it makes sense. I love it.

Jason: I’m super excited to hear about it, especially I come from the world of the AI where I used to do algorithm pricing. Let me get my thing started and then we can get you to share. What you have got so I’m probably gonna have to remove you from the screen just a little bit and I’ll get you to come back when I finish this one here.

Thank you, thank you so much for everyone who dialed in for this webinar and the training; it is about helping you to do better with the sales and the prospecting in the B2B business, especially if you are prospecting on LinkedIn and the whole series of the training is brought to you by Engage AI.

We understand that sales and prospecting is more than just using LinkedIn. It is about what you do as you warm up this prospect and how you book the meeting and when you book the meeting what you should say to them. This is what we are doing with this training and the webinar. Now what you can expect from Jim today is how do you grow your business visibility?

What are some of the actionable strategies that Jim will be sharing with you and most importantly what are the true and proven methods that Jim has done over and over the years from the market we’ve sustained for the last 20 25 years in bringing some of the businesses to worldwide global stage.

If you but anytime if you want to Get our AI prompt library. These have got some of our best prompt on the, that we have created that we have. crafted after speaking with user all of these take a screenshot or photo of this event Say something that you want and then share it on LinkedIn, Instagram, TikTok or any social media that you really using and then tag us there.

We’ll get notified and we’ll send you our secret AI prompt library. This is a little gift from us and also stand for a chance to be one of the 10 lucky winners to win a one year worth of the Engage AI Pro subscription. So what is Engage AI you may ask? So Engage AI It’s about helping you to start conversation that will get responses from the prospect on LinkedIn.

A lot of our customers tell me that they are, a lot of people tell me that people no longer read messages. People no longer read email. People no longer pick up the phone call. But one thing that they have found to be so useful is that they can still get responses if they engage with their prospect on social media.

And Engage AI is about helping you to achieve that and helping you to craft an engagement that will get you responses. But more than that though, another thing that we do at Engage AI is what we call the monitoring. One of the biggest pain that a lot of the LinkedIn users has got is when they go on to LinkedIn, They have to scroll endlessly just to find the people that are most important to them just to find the people that they are interested in engaging.

And a lot of times the news feed just simply would not show it to them and they get distracted. So what monitoring do for you is that if you tell us exactly who the people that you are interested in, We will monitor their posting every day, every single day. And whenever they have a new post, we’ll bring it back into a single webpage.

for you to engage. And what on the screen is an example of that. So no more distraction, no more endless scrolling. You can see all the target opportunity and the prospect within a single screen. Now, super excited to you is that we are going to listen to Jim Sten. He is the president of the Target Marketing from Santa Barbara.

He has been in the industry for two for 20, 25 years now. And I say in the early, in fact, he published a number of book and one of them is Artificial Intelligence for Marketing. He was really. Expected and know how all of this is coming. So he’ll be sharing with you About how to use gen ai and how to use ai for marketing your business and drive business growth so without further ado I will be going to have Jim on stage. Jim, I’m gonna let you take over and super excited to hear what you have got to share

Jim: Thanks very much. And you’re going to need to push my slides onto the stair. We’d go. Thank you so much.

Jason: Thank you for reminder. I wish that I could have done that for me. I’m gonna get off the stage.

All right. Yeah, we’ll see you again toward the end when we do Q and a.

But for right now, I want to welcome everybody. I did write a book called Artificial Intelligence and Marketing in 2017. That was about machine learning, not about generative AI. So what is generative AI? Large language model, GPT. What I want to cover is what is it? Technically, what is it? What does it mean?

What should you do now? And what’s going to come next? I had the same reaction that everybody else did when the first time we saw a chat GPT, you tried a little bit, you ask it a question, and it gives you a funky answer. And you realize, Oh, wait a minute. This isn’t Google. It’s something else. And when you start using it in a different way.

Like me, I’m sure it blew your mind because it’s a different animal altogether before we start three things number one There’s an awful lot of hype about this stuff Some of it is legit like there is a reason that everybody’s talking about it There’s a reason that it’s the big thing the best things in sliced bread However, it is not going to take your job away from you Somebody using AI will take your job if you don’t.

So a job is a bundle of tasks. AI can do some of those tasks. It can’t do the whole job. So you’re safe if you pay attention. Now paying attention is going to be tough because I’ve got a lot of ground to cover. It is absolutely drinking from the fire hose. So I want you to sit back, relax, and just let this wash over you.

It is an introduction. Some of this, already some of it you don’t. But I need to bring everybody along. So for those of you who know nothing, you’re in the right place. Let’s start with history. In fact, I’m going to go a little old school because we’ve seen so many a I created images. That I’m going to go completely old school.

I want to talk about the history of computing. It starts with code. It starts with programming. And this is absolutely deterministic. You tell the machine what you want. It does exactly what you want. If you didn’t write it correctly, you get an error. And if it really messes up, you get a blue screen of death.

So we don’t want that, but we do love to build applications that do exactly what we say. It’s a spreadsheet. You put in variables and the relationship between them and it does the calculation. You change one of the values. It changes the result. It’s magic. That’s deterministic. Then we go to probabilistic.

Just not just code, but predictive analytics. That’s where we take all the data that we have and we predict where it’s going to go. What is going to happen? And if our predictions are a little bit better than flipping a coin, hey, it’s worthwhile. It pays off. Then we get to machine learning, which is the high end of probabilistic.

This is where we give the machine the data. We ask the machine to look at it and it creates structure about the data from the data. It finds where things are out of balance. It finds relationships, it finds clusters. And if the data changes, it changes its mind. That’s why it’s machine learning. So we’ve gone from deterministic in code to probabilistic.

And now we’ve moved into this weird new area. That’s linguistic. A large language model is completely unlike. the rest of computing that we’re used to. This is guessing the next word. So if I said the cat in the it could be in the window or in the tree or in the chair, but most English speaking people know that it’s actually cat in the hat, which is cheating because it’s the title of a book that we all grew up with.

So if I said, you could knock me over with a the common parlance is feather. All of them are correct, but the most likely next word is feather. That’s what this does. That’s all it does. It’s not measuring, it’s not adding up, it’s not remembering, it’s just generating. So this is linguistic. We can also talk about artistic.

This is meta. ai that works so fast it changes while you type. So imagine a horse on a unicycle playing an accordion, smoking a pipe. It happens that fast, right? Okay. Images, not my thing. My thing is language. So I want to talk at the example I’m going to use is chat GPT. That’s the one I’m most used to.

And we start with, okay, what is a GPT? What does that mean? It stands for generative pre trained transformer. It generates things. That’s its job. Not remember or calculate or analyze, just generate. That’s what it does. Okay, what’s it, what’s the pre trained? It is pre trained on everything that can be found out on the internet.

That all goes in there. And oh, by the way, hundreds of thousands of books. That all goes in there. And it changes those words to numbers. And the numbers can be calculated and they become vectors. Alright, so what’s a vector? Here is a specific word. It’s a cat. A cat is an animal. It eats and breathes and sleeps and has claws and fur, but it’s also a house pet.

So it has a relationship to dogs and fish that are also animals, but it’s a different way of talking about cats. It’s also part of our language. You can be a cool cat or a crazy cat or a hip cat. It is also we know that these cats are trying to kill you. That’s a different vector. We’re talking about these animals in a different way.

And of course it’s an internet meme as well, because the internet is absolutely filled with cats. There’s no getting around it. These are different vectors for the same word, right? It is, and this is two dimensions. When you take, when you give these numbers to a computer, they can calculate the vectors in not just two dimensions or three dimensions, but a thousand dimensions, which we can’t even picture.

Okay, so generative, pre trained, transformer. What does transformer mean? The sentence that begins, we need a tie could mean a couple of things. We need a necktie to have tea at the Ritz. We need to have a tie in our tennis match to stay in the game, or we need a tie to keep our hair out of the machine.

When the computer sees this. It goes back and transforms what it thinks the word tie means. It’s transforming meaning. I went to school this morning, and to get my degree, and my parents were very proud. To pick up my daughter, and all her friends were there to teach a course, and one of the students gave me an apple.

Or to meet with the teachers about my daughter. What I went to school in the morning means changes based on what comes next. So that’s what transforms. So GPT, generative, pre trained, and transformer. Now that’s where the base, that’s where we start. But next we get to fine tuning. This is where the people who make these models give it some instructions and rules.

This is the chat GPT model spec that they published a couple of months ago. And it says, Don’t break the law. Don’t tell people how to build bombs. Don’t do anything that’s not safe for work. Be cognizant of people’s privacy. These are rules on top of the model. So all of that goes in as well. And you can fine tune for whatever purpose you want.

For human resources, or IT, or marketing. That’s up to you. Then we get to embeddings. Embeddings are where we take an additional set of vectors. That are specific to the task at hand. They might be medical terms or legal terms or the jargon of your industry. So that gets put on top and then we get to retrieval, augmented generation that says, I want you to take these rules that I’ve written in the fine tuning and the embeddings, the jargon, and I want you to apply them to these particular documents.

I’m going to ask you questions about these particular documents. And now finally we can do our prompt engineering. That’s where we type in what we want it to do. The thing that’s important to remember is that this is not a database. It’s not a search engine. It’s not a logic system. It’s not a calculator.

It’s not an encyclopedia yet. People are working very hard to make it more deterministic. instead of linguistic, but it doesn’t act like those things. So don’t treat it that way. So that’s what it is. What it was. So what’s the big deal with the big deal is that even though all it’s doing is guessing the next word, the output is actually mind blowing.

Now it doesn’t, it’s not human. It doesn’t have logic or judgment. Or understanding or memory yet they’re working very hard on this, but right now it’s just repeating or inventing the next word. Now, we humans, we have thoughts and feelings and we’ve struggled to find the right words. These large language models.

They have all the words. They don’t know what they mean. It’s just a mathematical relationship to the next word. Large language models are tools that we have to learn how to use. And you just change your mind. You free your mind. It’s not computing the way you’re used to unlock. How you think about computers in the best way is to just go play with it.

Try things, see what works and doesn’t work. Experiment with different methods and find out what it’s bad at. For instance. It’s bad at facts. I was at a conference and I took three separate photographs and uploaded them to ChatGPT and said, these are three different views. Of people in the room. How many people are in the room?

And it said there appear to be 10 rows of 15 people each. So that’s about 100 people. That’s not good math. It said the second same thing about the second picture and the same thing about the third picture. And therefore, there are 100 people in the room. Don’t ask it for facts, right? It’s bad at that.

It’s also bad at bias in the data. It’s it if it’s You ask for a story about a pilot and a flight attendant, the pilot will be male, the flight attendant will be female. Doctor and nurse, the doctor will be male, the nurse will be female. The bias is in the data, so we have to be aware of that. And it wasn’t designed to do math, that’s not what it’s for.

It’s not a calculator, it’s not, it doesn’t do what we expect computers to do. What it does do really well, is come up with an opinion. It’s a great tool. Teacher. It’s a great coach. It’s a great brainstorming partner. It’s awesome at coming up with new ideas or starting on a blank page, or I’m writing this article about this subject or I’m creating a landing page where I’m trying to write an email.

Give me some ideas. Where do I start? So what does that mean to you? It means that you need to do your prompt engineering. This is using the English language as a programming tool, but you don’t treat it like it’s software. You treat it like it’s an intern. Really smart, but not educated. This is somebody who you.

Got a lot of book learning in school, has no practical knowledge. They don’t know, but they’re really smart. It’s like having a conversation with a fascinating stranger in a quiet bar around closing time. So we’re not sure if what we’re hearing is a hundred percent true. It’s the trust, but verify.

Now you’ve heard all kinds of things about what prompt engineering, how it works. So I’m going to cover those really quick. Just to refresh your memory, give it context. It doesn’t know you have to tell it specific, tell it that you want to ask it a question about this person regarding that topic, and you’re looking for output in this format.

You want it to think in this sequence, you want to give it some examples. You want to tell it to go on at length or be brief. That’s where you need to give it all the context you can Context is in a couple of different categories.

Who is it? You are talking about a customer, but you need to tell it clearly What you mean all of these words might be used in the same company about the customer But it’s going to be very confusing the machine. So make sure That you’re sticking with specific definitions and you use those terms very intentionally.

All right, now we get to the point where you have to give it a persona. You, ChachiPT, are a fill in the blank, and I want you to do this thing in this way. A meeting. I want you to be the meeting note taker to tell us what the takeaways are and what the action items are. I want you to be a career counselor and help me figure out what my next job is.

I want you to be a fill in the blank. I want you to summarize this meeting or format this spreadsheet or create an outline for one to one meetings or review my budget. That’s the task. And I want you to be professional or serious or silly. or helpful or friendly. That’s entirely up to you. So here’s an example.

I want you to be a digital analyst and the task is delivering a presentation about the differences between last month’s numbers and this month’s numbers because the client wants to make the website more customer focused and I want you to do it by using the on site search keywords report. So I’m giving it as much and I’m telling it that it is a master digital analyst instead of just saying, look at this report and tell me what’s interesting.

Which you can do. You can I had five tabs open with five different large language models and I simply said, write a paragraph. And they were all happy to do it. And they were all equally as useless. But without context, it’ll still take action, right? It’ll, it will do things. So give it context. Now, here’s some interesting hints, some advanced prompt engineering.

Think step by step. Show me how you got there. What is your chain of thought? So why are you saying it that way? And you can do chain of thought. If you really want to get into the research, there’s chain of thought, there’s tree of thought, there’s graph of thought. And yeah, it gets wild and crazy.

But here’s like super advanced. Flip the script. Instead of you prompting the machine, you ask the machine to prompt you. After every prompt, give a normal response, then criticize yourself harshly, tell me why the response is bad, and give it a score from 1 to 10. If the score is below 9, improve it, critique it again, until you get a response of 9 or 10.

And it’ll go off in the corner and crunch away and give you its best answer. It will give you a better answer. Ask things from a different perspective. What innovative principles from The private sector might be applied to this government problem. What are potential future scenarios and how can I prepare for them?

What are some unintended consequences? How can I deconstruct this problem? How are these concepts connected? What are some alternative perspectives? Ask it to think backwards. Ask it to think upside down. Help me think clearer. What else should I have asked? What does a more creative solution look like? So here’s an example.

You’re doing B2B marketing and you’ve got not just one person, but three people, and you’ve got them to be willing to take a meeting and you need to persuade them to buy or trial or whatever it is. Three people from the same company. Hey ChatGPT, I need to persuade these three people. Here’s their, I’ve scraped and copy and pasted their LinkedIn profiles.

Here’s a link to their blogs, some videos, some speeches. This is who they are. Create a persona for each one. Show them my PowerPoint deck. And here’s a video of me presenting the deck. Simulate their questions and comments, so I know what they’re going to ask, or I know what to add to the presentations they don’t have to ask.

Have them discuss it after I leave the room, and show me what that conversation is, and then tell me what language, what vocabulary. Should I use to be more persuasive to these specific individuals? How can I relate to them? That’s the kind of thinking that’s the kind of there’s no perfect answer.

It’s not a calculation, but it will give you its opinion and it will role play and it will help make you a much more persuasive salesperson. And then we get to the point where you can create your own. You can use chat GPD to create a version for a specific purpose. You give it a name, a description, you give it instructions.

You, you tell it that he wanted to do that. You upload files. I want you to consider these documents. You want it to be able to browse the web and create images and use the code interpreter and you create. Your own. So here’s one that I created. I want you chat GPT to respond with advice. One sentence of advice from each of these people.

I’m going to put in my persuasive email or my PowerPoint slides and I want response. What would Einstein say? What would Machiavelli say? Oh, and you want to get further? Hey, let’s create a whole bunch of people. Let’s create social experiments and using artificial people to see what might happen.

This is, cutting, this is this month’s research. Okay, that’s prompt engineering from, yes, you’ve heard it before too. I hadn’t thought about that. So hopefully that’s helpful, but rest assured you will always be needed. There are three things we need humans for. Number one, what problem are you solving or question do you want answered?

Number two, what data do you want to consider? Here’s a spreadsheet. Here is a book. Here is the annual report from the company I’m selling to. What problem, what data, and does the answer pass the smell test? This is where it, the absolute human part of it is critical. It’s where your knowledge and your experience comes in.

The smell test is, you know it when you see it. The computer says, oh, it’s a tiger. The human says, oh, wait. There’s some context that’s missing. Oh, it’s a dog. Okay. Got it. It’s a dog There’s some things that a human can look at and go. Oh, yeah, that just doesn’t work That’s no somebody should lose their job.

That’s just wrong. I was in a very fancy hospital brand new building Really impressive clean as a whistle and I went into the men’s room and I found this

I couldn’t believe my eyes like How many times is that going to happen every time? Yes, it does every time I did it about 10 times Oh, then I have to take a video of it and I came out and my wife said you’re in there a long time Are you okay? Look at this video That doesn’t pass the smell test. So what problem, what data, and does the result make sense?

So you have a job. Those are your job. Now there’s three things to worry about. We already talked about bias in the data. It’s there. It lives on. There’s no way around it. You have to know that it’s coming and be aware and pretend, sorry, and protect against it. Number two, assuming that whatever comes out is correct, because this is what hallucinations are, it’s going to make things up.

to satisfy the algorithm trying to figure out the next best word. It doesn’t have reason. It doesn’t have logic. It doesn’t have common sense. It just knows that light as a feather is the next word. But it could it be light as a hot air balloon. So you are responsible for the output. And then the third danger is serious.

It’s bad people doing bad things with power tools. That’s just, anything can kill you. You can use a baseball bat in bad ways. People will use this in bad ways. And that remains to be seen, especially in the United States with an election year and the ability to do. Deepfakes. It’s pretty scary.

It is a tool. I don’t think it’s cheating. I think it is a tool that we learn how to use. So I look at it this way. There’s time and effort that go in and a tool lowers the amount of time and effort. There is a value coming out the other end and that value depends on how much how long you’ve been in the business, whether you have talent, right?

The results, the value of the results.

That’s great. So if I use a spoon or a whisk or a blender or a food processor, The value is I have a cake that tastes good and I did it faster than I could have by hand. Same is true of whether I use a campfire or a microwave oven. Can you tell that I cooked it in a microwave? You can’t tell. It heats the food.

Do I use an abacus or a spreadsheet? Do I use a pencil or a typewriter or a word processor or ai? If I use AI and the result results are blah, or the results are wrong. That’s my fault. If I use AI and the results are amazing. I was the one that identified the value and said this is amazing and therefore valuable.

Whether it took me half the time or not, that’s internal. You don’t need to know that.

I think that the most powerful new capability is AI. Is not automating the mundane and unlocking our creative potential now. Yes, we can automate the mundane Yes, you can take rote processes and automate them and that’s great. I’m all for productivity. I’m all for expediency That’s terrific, but I don’t think that’s the most powerful thing.

I think the most powerful thing Is not looking at this like we look at computers But to change our approach to computing to open our minds All right, so So far, GPT stands for generative pre trained transformer. We’re moving from code that’s deterministic. To probabilistic to linguistic. So it’s different efficiency, productivity, optimization.

Awesome. That’s great. And you can do it with terrific prompt engineering techniques and you can make your own GPTs. But what’s really interesting and really cool is what comes next. What comes next, what comes after, and what I think is we’re on the verge of is everybody owning their own agents. So it’ll be my agent.

It’ll know everything about me, and I’m going to give it certain access to certain people or companies for a certain purpose. So I want my doctor to have access to all my medical files, and I want my CPA to have access to all my financial files, but not vice versa. I want Amazon to know that my credit card number changed.

I want Facebook to know that I want to be logged in on every device I look at it on. I want that new store that I’m buying from have some information about me, but not everything. So I’m going to have different levels of access. And this will be my Jarvis, the thing that knows me and is my personal digital assistant the way Siri and Alexa never figured out how to do right.

It’s time for them to up their game and they’re going to. What happens is that I’m going to give it the ability to be proactive on my behalf. And here’s the result. My robot is going to be listening to that and watching what’s happening in the house. And it’s going to say, Jim, your dishwasher is going to fail in about 30 days.

I can hear the gears grinding. There’s an error. I’ve checked the troubleshooting guides. They say that this bearing is going out and it needs to be replaced. And if you don’t, you’re gonna end up with a kitchen full of water and the repair price is too high for the value that you’re going to get out of it.

You need to buy a new dishwasher. I found 12 models that will fit in your kitchen, of which five of will fit in your style. Three will fit in your budget and two will fit in your schedule and can be installed when you are available. Cause I know your calendar and I’ve negotiated the best price and they’re within pennies.

So here’s the last decision for getting it installed on Wednesday. Do you want the best reviews in Wirecutter and Consumer Reports or the most stars on social media? Pick one or the other, okay to have it installed on Wednesday, yes, no. That’s where we’re going. So that means we have to think different about computing to free your mind.

We’ve been doing tabulating and calculating and quantifying and archiving and searching. That’s great, but now we’re going to strategize and brainstorm. And collaborate this morning. I spent a half an hour doing about two weeks worth of work trying to do research doing industrial research on competitors.

I asked the machine as I would have asked an intern, actually not even an intern. I would have asked a well educated experienced consultant. I want you to look at these companies. I want you to go find their competitors. I want to find out what their pricing models are. I want to know whether they’re growing or shrinking.

I want to, I want you to do a SWOT analysis. Strengths, weaknesses, opportunities, and T, the other one. And I want you to bring back, give me a table that lists the top 10 competitors with all of these things. And then an analysis about which one has the, is the weakest and that I can have the best opportunity for going after their marketplace.

Boom, it’s magic. It is a collaboration tool. The output is mind blowing and yet this is the dumbest AI you will ever, this is the dumbest AI you will ever use. The next, GPT 5 is going to come out and it’s going to astonish people. It’s going to be amazing. So what does it take to win in this environment?

You have to ask better questions. We think that it’s all about asking about buying the best technology, the latest software. We think that you have to have the best algorithm. We know you need the best data. That’s fair enough. You also need really good strategy. You also need really clear vision that you can share with everybody in the organization.

But, The best questions wins. The one with the best questions wins. Is the winner. Now I warned you I was going to hit you with a whole bunch of information. So I’ve done that. I hope that’s interesting. I hope it’s valuable. I hope you’ve learned some stuff. And now I’m going to open it up to see if anybody, and even if it’s not the best question, you can just go ahead and we’ll open it up for questions.

Jason, are you going to pop back onto stage and help facilitate this for us?

Jason: Absolutely. Absolutely. That was really great. I think I have to say as the practitioner in the data engineering and the AI. That was still a lot of good information for me to learn. And I think the most important thing is about how do we think about the future? Not only now, but how do we think about the future?

So certainly have some questions coming in already. And so the first question that I have got for you is that in what ways can the prompt engineering be optimized? to yield the most relevant output from the Gen AI? I think we probably, even for us, we actually get this question a lot. So we’d really love to hear your perspective.

Jim: I think three things. Number one, context. It imagine that you have an absolute genius walk into your office, But they don’t know your industry really and they don’t know your company and they don’t know your prospects So you’re going to ask that genius to do something or come up with a process or write a document You have to give it give that person lots of information so they can give you what you want.

So context to start with number two is Playing with it like trying stuff and then go back and ask it a different way in a different way in a different way And three is make your own GPT. So I have about a dozen little models little gpts that do specific things for me This one goes out and researches companies this one compares and contrasts marketplaces and stock prices This one is to help me write in my own voice I need three paragraphs about this thing that writes the way that I do and so I uploaded a bunch of My books and it says okay.

I’ll write them the way you do now You What it comes out with, I can’t copy and paste, but oh boy, it saved me. Hours. I’ve done, I’ve accomplished my task in half the time. So context matters, try things, play things. And then when you create your own GPT, it’s not like programming, like it is a set piece of software.

You have a conversation with it like, oh, that was good, but now I want you to always do it this way. And don’t forget about this. And please include that. And you can improve it and improve it. And over time, it becomes the ideal assistant to do that task.

Jason: 100%. And the other thing that you spoke during the presentation is use it as to unlock the creative potential.

I really like that. I think that is also the way that I use it. And that’s the way that we think about it as an Engage AI. It’s not only just automating the mandatas, but it’s about unlocking that creative potential. Can you share with us in terms of like how the businesses can leverage AI to unlock this creative potential for their business growth?

Jim: It’s really wrapped up in this idea of, it’s not a computer, it is a collaborator. If you collaborate with it, it becomes a member of the team. It is. I’ve written a white paper on how to measure the business value of generative AI. And it’s, the usual things make more, spend less, make customers happy.

Can I can measure those things. But now I want to measure creativity. Measure not just how many new ideas. But how good are they? So what is a good idea? A good idea, a creative idea is one that is as different from the others as possible, and it is as viable as possible given budget or given politics or given the time of year that we’re in.

And it’s opening up your mind to what if I had an expert in every field that I could just ask questions of? What if I turned the computer into a mentor? Now, mentors are great. They’re full of ideas. They don’t really know your situation. They don’t know that, oh, by the way, you have three kids at home, and you’re taking care of an elderly parent, and your car just died and.

But they’ll give you some good advice. This computer will do that forever and it’ll never get tired. So open up, you be more creative by opening up your mind to asking more types of questions.

Jason: On that note, that really bring up the, one of the thing that a lot of startups and also a lot of companies are starting to try now is about building their own AI agents.

And it, it, it’s the reason why I ask that is because when you talk about building an expert in every field. area of the expertise, right? Do you, can you foresee that one day we, a company or businesses could have a special AI agent who is the expert for the marketing AI expert in the HR?

How do you think this how this whole thing is gonna turn out?

Jim: Let’s take it, let’s do the progression. It starts with, I’m going to build a GPT to help me to be an expert at doing this task. So every time I need to format a new sales report or create a new PowerPoint presentation, this GPT helps me in this particular way.

And this one helps me in that particular way. So I create all of these things. All of these GPTs that are their assistants to me, and then I can start stringing them together so I can create an executive GPT and tell it I want you to use the outline GPT to create outlines for my presentation to that company, and then I want you to ask the script writer.

GPT to write a script based on that outline. And then we’ve got the image generator to look at the script and create images and then, and the, and I’ve linked them all together. This is absolutely doable today with creating your own GPT in chat GPT. It’s you just call one from the other it’s programming in English, if you will.

So that’s moving up the chain. The science fiction writer in me would say that it is possible that somewhere in a college dorm room, there are a couple of people who are creating these specific GPTs saying, look, here are all these business books, ingest all of that stuff. So you know all about good to great and the five dysfunctions of a team, all the best business books.

Yeah. And then I want you to help create a new business with me for me. And it’s going to string all of these together with an executive layer and they’re going to create an automated Company that’s that is where we’re headed that is going to happen.

Jason: I love it. And the other question that the viewer is sending in is that so we have seen quite a lot already in terms of like how to use chat GPT, how to use jet AI for the marketing in terms of writing blogs, writing article, creating social media posts in a lot of those is about creating and generating content, whereas what we do at Engage AI is about seals and perspective.

Where else do you think Gemini AI could potentially play its role in terms of the marketing? What do you see? How is this heading?

Jim: I’m gonna, I’m gonna run over to my which tab it’s open in to find SmarterX. I want you to write down SmarterX AI slash JobsGPT. SmarterX.ai slash JobsGPT.

This is a GPT created by this company that you put in the title of the you’re a content marketing specialist. You are a search marketing specialist. You are a head of sales. I tried this one this morning with the head of sales, and it will list out. For that job here, the tasks generative AI can help you with those tasks.

It’s limited by your imagination. You tell it, thank you. There it is. Go there and don’t click on the image, but go down to, I guess it’s an orange button that says launch. And it’ll tell you that for what marketing manager role, their strategic planning and content creation and market research, and here’s what they can do for you.

Now, think about this from two different ways. Number one, this is helping you understand how large language models can help you. On the other hand, you can create a GPT that does this yourself. What do you want to create? This is the problem that we face, that it can do anything. So this is like being, on the internet in 1995.

What can I put on the internet? Anything you want to. That’s not helpful. What kind of app can I make for my smartphone anything you want? That’s not helpful What can I use this for? Now this smarter ai Has this jobs gpt that tells you how it can help you.

Jason: I love it! On that note, speaking of jobs, one of the questions that Abhilash sent in is that what are the suggestions do you give the professional in every field for upskilling in AI? That’s really good question. And ensuring that they are skilled. Don’t become redundant in the near future.

Jim: I think when I started selling computers in 1979, you had to learn programming. There were only a few package software things you could buy, and they usually did accounting and there was word processing. So yes, you have to learn how to do programming in order to make use of this computer. But once you do, you can program it to do anything, right?

Now we can buy applications, but now we can make our own applications. The same thing applies to prompt engineering. I think prompt engineering is a momentary skill. It’s important to know how to use it today. And everything I talked about with context and give it a persona and all of that stuff is valuable today, but it’s not going to take long before people realize that instead of telling it exactly what you want, You’ll say you’ll type into chat gpt I am a sales professional in the pharmaceutical industry And this is my territory and these are the kinds of people I talk to What sort of I want you to ask me questions?

about what I know and what my skill level is so you can judge how capable I am so you can create a training program to help me be more successful. So I’m not prompting the machine, I’m asking it to prompt me. I’m asking it to be my tutor, to be my trainer. What can I do? What’s the next step I should take that’s going to result in more leads?

What’s the next step I can take that will move these leads further along the sales funnel so I can close more contracts? Get the machine to advise you. That’s a skill that will never go out of style.

Jason: I so hear you. And one of the things that we observed and we came to hear your thought is that what we observe here is that we, at this stage, we don’t quite, I think there are two parts of it.

So the first part is we don’t quite see of the redundant of the jobs. But it’s the shifting of the workforce where a lot of the workforce can Will within a lot of the works was are shifting and will be shifting even more to oversee because What is happening is that the offshore resources? Can now use the ai to help them to stay to become Even more relevant to the domestic audience and to speak in a way and converse in the style that is even more relevant to the domestic audience or the customer.

And I think actually the job will be even shifting even more. That is really the first part that I would be here. Interested to hear that. The second part of that is more of our own opinion is that it’s similar like the industrial revolution. It does Will make some jobs redundant, but at the same time it will create a range of different jobs as well It’s the same like how the internet did to the jobs it Remove a bunch of jobs Like the cloud computing but also create new bunch of job as well So I think that part is definitely is happening.

But the first part I think is the shifting of the workforce and accelerating the workforce to offshore is where I feel not the best thing though, depending how, which, what position you are in. It’s just the nature of the dynamic of the work environment.

Jim: One of the advantages of age. I’ve seen this movie before, so I’m old enough that I, when I started in sales, I had a secretary, I would hand write letters and the secretary would type them up and then I would correct them and then the secretary would retype them with a carbon sheet to make a copy for the file and put it in the mail.

And that’s how we communicated. Then we got computers. And we didn’t need secretaries anymore, but we didn’t fire anybody. Those secretaries became administrative assistants. They became, they went into human resources. They became salespeople themselves because they understood what we were doing.

They knew how to sell. They’d seen all of our letters. They heard us talk on the phone. They were trained already. So the company not only did the salespeople do better because we had computers, but the people who used to be secretaries, Became much more valuable

Jason: Exactly spot on. Thank you so so much for this and that bring us almost the end of this, training and webinar Once again, thank you so much.

And so for anyone Who have missed it this will be recorded and put out online And so you can watch it anytime on demand and just as engage AR is really about helping You For the seals and prospecting and we have a special discount for you If you want to join engage ai get it 50 discount and use the code that we have shared with you So again, if you take a screenshot of this any time of the presentation share it with your social media Tag us tag jim and we’ll send you our super ai prompt library And once again, thank you so so much Jim, it has been an honor to have you sharing with us and also telling how we should learn from the history and then look forward to the future.

Jim: Thank you for inviting me. As you can tell I really enjoy talking about this stuff. I look forward to chatting you again.

Jason: Thank you!

Jim: Thank you.