To Ai or To Not Ai, That is the Question

I wanted to jump on here today, and I want to talk about artificial intelligence. And the reason I want to do that is because I had a conversation today with a dear friend and we were discussing, you know, some of the things about artificial intelligence and what it can and cannot do. And as I have done quite a bit of studying on AI and what's going on and using it in our business at Swing Point Media, one of the things that, that I was kind of blown away with was its intellect. And for the first time in human history.

We are not the smartest beings on the planet. That's right, the chat GPT4. When it was tested for its IQ, it was at 155. Now to keep that in perspective, Einstein was 165 and the average person out there is like 120, 130.

So up until now, I should say now, maybe up and say up until 2023.

Intelligence was really kind of that, that cornerstone, that kingpin, the thing that said, hey, this is, this is what determines the top of the, the food chain. Well, now AI has the ability to think and act and, and grow just like a human being. And I shouldn't say just like a human being, but similar to what I'm talking about. In other words, it is a, it's called a LLM, Large Language Models, and it's learning all the time.

And so, you know, we use it a lot for our clients in digital marketing. And the reason I want to step on today is, is really just to, I want to talk about the positive aspects of it and how it can be used. In this conversation I was having with a dear friend over coffee, I had to drive into the, into Redlands today, and we were chatting about this and, and she was mentioning that, you know, she uses it some, but sometimes she doesn't even know what to ask. And I said, well, here's the cool thing about it. I said, sometimes you can ask Chat GPT, what question should you ask to get a certain outcome? Because it's often, you know, sometimes people, you know, when they talk about digital marketing, they're saying, I don't even know what to ask. And I say, well, let me tell you a little bit more about it, and maybe that will help you solve or come up with some questions that you might have after you learn more about it. Well, that's kind of the way artificial intelligence works. It allows you to ask questions, almost to interview it and, and have a, an interactive conversation with it.

And what this means is going back to its level of intelligence, it's like, and this is just feedback that I'm getting from there's a product called stable diffusion out there. And the inventor of that said, you know, for Chat GTP or AI is like hiring a thousand college graduates that do research. And that's why we can get information to come back so rapidly. Because when you put a question in, in Chat GPT or Bard, any of those, the answers come back very rapidly. And how does that happen? Well, it's like having a thousand, you know, college students that are graduated on staff giving us an answer. So the answers come back. Now, I want to point out that some of those answers are what they call hallucinations. And sometimes the artificial intelligence will hallucinate an answer. It's no different than if you're having a conversation with somebody. They might make the situation a little more grandiose than it was. And so with that, it's important to look at this information and say, okay, is it accurate? Is it truth? And that's where I think, you know, for right now, you step in as a human and whatever response you get back, you want to review that and make sure it's accurate. And I know there's been a couple of instances out there from, you know, the attorney who turned in his arguments to the court and they cited a couple of court cases in his, in his submission that didn't even exist. Well, shame on him for not looking over his, his document. For the most part we've noticed it, it does an incredible job and it just gets better and better and better. Now we use, we use a different system that actually combines three different AI models. And so it's, it's connected to the Internet where chat GPT 3.5 is not. It says, hey, we stopped putting information in learning UP and through 2021. Well, the three systems that we use, it's current, so it will go out and grab sources and it'll cite that information. And so when we put information, it allows us to go back and check, hey, was this the right link? Is that, does the article actually exist?

Here's the thing that can happen.

And in this, I'm going to go back to this conversation we had that I said, I, I went through this process and I asked Chat GTP GPT, I said, could you create a test and grade it for me and you can tell me what my learning style is? They said, sure enough. So it spit out about 16 questions and after that I took the test and it graded it and it told me what kind of learning style I have and I said, now that you know my learning style, and if I asked if I wanted to learn something, could you actually create the curriculum? And it said yes. Now what does that mean that could happen? Well, imagine a teacher, and this doesn't matter. It doesn't matter what age level, but let's say you have a teacher who is teaching. A student is teaching a classroom of 30 students. And those 30 students all have variety of learning styles. And it's up to the teacher to try to create or convey the, the teaching in such a way that.

Covers all those learning styles. And unfortunately, some of those students are left behind because they missed the point. They did. Maybe they're visual learners or auditory learners or kinesthetic learners, and that teacher, in their best of, best of their ability, maybe didn't get that across. Well, imagine what this, what this process could do for those kids in school. It could actually give them individualized teaching. It could be as simple as, hey, there's a monitor in front of them. And as the teacher is giving them their lesson, what's showing up on the screen is in their dominant learning style. Well, now what would happen to those kids? Well, let's face it, they would learn faster. They would retain more information. And so there's an op, an opportunity to have individualized teaching in a, in a group classroom. So I think that could help accelerate the learning that we see and our education system, because, you know, most people, and I'm not saying it assists, but I'm just echoing it, that, you know, know, our, our education system is broken and this could be something that could happen. So there's another nice benefit of chat, GPT or AI when it comes to individualized learning. The same could be true for things like medicine, individualized medicine. You know, if you think about it, whenever there's a study done on any drug, there is a, there's a success and a failure level. And the failure level, there's a, the percentage of, hey, it didn't work for some people. And so what happens is there comes out, I'm making this really, really simple, but it turns out that there is a standard for this particular dosage or drug, whether it's one a day, two a day, three, whatever it is. And that's really for an average. And so some people get better, some people don't get better, some people get worse. But when they do these clinical trials to say, what's the tolerance for not getting better or getting worse, and if it passes that it's okay. But now imagine with AI, they could take that and say, what is the best dosage for that individual? And again, this is now getting past my, my knowledge level, but I guess I could ask ChatGPT and it may be something to do with their blood tests or other tests that are run and that's fed back into artificial intelligence. And it says based on that situation, this person should have this dosage. And while the drug is very effective at that dosage for that person, it may not be for this other person because it could be individualized for them and maximize the benefit of that dosage or that medication and the outcome would be much better. So there's just a couple of things that, that are going on there now from a standpoint of, of creating a, are using it from a marketing standpoint. And this is the fun part because I get to do this every day is let's say for instance, you're a company, you don't have a lead magnet. There's nothing out there that's attracting new leads where you can actually prompt it. In other words, put in, put in information into ChatGPT and say create a lead magnet based on these, on this criteria and it'll spit out lead magnet. And then you can take that lead magnet and say, hey, let's do some follow up emails for that. And you can say, I want to, I want four emails, I want six emails and it'll spit that out. I gave that example to the young lady I was with today and she said, well, like how long did that take? And I said, well, here's the thing. If we, if we sat down and did that, you know, with a human, I said based on how it turned out and what we would do, it would probably take between 40 and 60 hours to complete that task to do the lead magnet, do all the research, get all the data and then create the four to six emails. She said, yeah, okay, how long did it take with AI? And I said it took about five to 10 minutes. And I said the scary part is I said it was really, really good. But part of the reason it was really, really good is because we gave it really, really good information to build off of. It's just like, you know, garbage in, garbage out. If you ask real simple questions, you're going to get real simple results and you're going to say, ah, that's no big deal. But if you give more in depth information then it will give you a really, really good answer. Now I followed up by saying, you know, my concern was this, how will future clients look at that and say, well yeah, it took you five minutes instead of, instead of 40 to 70 hours, I'm going to pay you the, the, the five minutes. And I said, so I have to look at this and say, I'm like that plumber says, you're not paying me for the fact I have a hammer. You're paying me for my 25 years of service. And I knew where to use the hammer on that pipe to fix the problem you had. And that's the investment our staff at Swing Point Media are making. I'm making in my staff to say, hey, we're going to make sure they're on the cutting edge and knowing how to use artificial intelligence. As an example, our writers now are what we call aio. They're artificial intelligent optimizers. So after the articles are written, and again, this is where that attorney, if he had an AIO in his back pocket, he wouldn't have, he wouldn't have gotten in trouble in court and maybe lost his license or whatever, whatever happened to him. But I said, you know, our writers have now turned into a iOS. And so when the, our, our combination of three artificial intelligence writes an article, then they go back and they make sure it reads and writes or reads as if a human had written it. They make sure all the grammar, they make sure the, the sentence structure, the keywords, all of that is in place before it ever goes on our client's, client's website. That's just another great piece. Now, the cool part about this is we can now generate this at scale where maybe one article a week, because the amount of time it took, three or four days. Now it's a fraction of that time and the quality is very high. We haven't gotten into using things like Dall E or Mid Journey, if you're not familiar with that. Those are AI programs that create images. We haven't quite gotten there yet.

And part of it probably is.

Refining the language that we use to get the image results we want. Now I'm excited about, there's an image, AI image with the Adobe suite called Firefly. And I like that it seems like it's a little bit more intuitive and it's not as difficult for me to use and easier to get into than something like Mid Journey. Dolly is part of the AI or excuse me, Chat GPT platform. So it's right there to be used. And I really haven't dug into that much. Just because we're using, we still have graphic artists and we still use other platforms to build our images, but it doesn't mean we're not going to do that. I can kind of see.

When we're done with an article that will then ask chat GPT or whatever platform we're using, hey, what, what images would you.

Recommend to support this article and what prompts should we write to get those results? And I suspect it's just a matter of doing just what I said because that kind of popped in my head right now. I was like, yeah, why aren't we doing that? That would make sense. And then go to one of those platforms and spit that out.

That's it just keeps evolving. The reason I bring this up today is hey, should we use AI or should we not use AI? Well, if you have access to a faster car versus a horse, you should use it. I think there's definitely benefit in that. Some people are going to say, hey, you know what, I'm too old, I'm about ready to retire. I'm not going to mess with that. I think you should still understand it, but you have that option to step away. But if you're still conducting business, I think you should understand, hey, what capabilities does it have which really are only limited by our imagination? And can I use it to do what I do better? Because here I think is what's going to happen. I'm going to go back to the gentleman who.

Created the program I mentioned earlier, stable diffusion. And he said, in his words, he said in the next three to five years there will be no jobs for programmers. Now think about that.

41% of everything we see out there in the marketplace on the Internet is generated by AI. It's already building. Most of what we see, we just don't even know it. And the, you know, if you go to chat GPT and you ask it to write a code for something, it will spit it out and, and you know, it may take some extra tweaking here and there, but it gets better and better and better all the time. So what does that mean? Well, it means now if those coders who could lose their job in three to five years learned how to use AI, or let's say they started leveraging, I'm not going to say they don't know how to use it, but they started using it more and more. They are more equipped. They have the horse and, and the fast car and they can do both. So I'd recommend that you, you do both. That's what, that's what we decided to do. Maybe, I guess it's probably been, you know, 12 to 15 months ago, we started playing around with it and seeing its capabilities. When we started creating articles, at first it was terrible. It really was. It was really bad. But in a very short period of time, it evolved into quite a machine. And I don't even say a machine is into quite a partner in what we, what we create. So I just wanted to hop on today and share that with you. I thought it was very valuable. I thought it was. I thought it was one of those things where, you know, it kind of makes you think, should I use it or should I not? And if you're not using it, go check it out. You can go to Chat open open source.AI and you can set up a free account. It's called chat GPT and it's 3.5. Sorry, hit the mic there. 3.5 and it's. And you can check it out. I'd recommend you get out there and try that. And again, when you don't know what questions to ask, ask it. Say, hey, I'm thinking about doing X, Y and Z. What questions should I ask to help me achieve that? And you'll be surprised what happens. All right, thanks again for joining. My name is Jeff Harrison, Swingpoint Meat. If you have any more questions or you would like to learn more about AI and how to help you use it in your local business marketing, check us out. We are@swingpointmedia.com and you can get a hold of me at jeff@swingpointmedia.com.