[00:00:00.240] – David Maples
From zero to 100 million in under two months, ChatGPT’s artificial intelligence represents the fastest adoption of technology in human history. That is six times faster than the previous record holder of TikTok. That’s right, today on The Buck Stops Here, we’re talking about artificial intelligence and machine learning, and how it will reshape how you do things in your business in the next 12 to 18 months. Artificial intelligence will reshape entire industries in the coming decade, and the future of business owners and businesses will be working hand in robotic hand with these intelligent machines. The only question is, will it look like Terminator or WALL-E? Today on The Buck Stops Here.
[00:00:53.800] – David Maples
So, today we’re going to talk about artificial intelligence, what it is, what it means for you, how things have shifted. And this is one of those indelible bright line moments where everything is shifted in human society. Now, this has been going on for years behind the scenes. So, anybody listening to this podcast used to use, let’s say, Google Translate, Google Translate has used a type of artificial intelligence, and translation matrices since 2016.
[00:01:23.780] – David Maples
So everybody has the same baseline of definitions here, artificial intelligence is the development of any computer system that would be able to perform tasks that normally would require human intelligence. And human intelligence, because it’s what we’ve always had, doesn’t have a cool moniker like AI. There are no movies named HI. It’s just, you know, hi. But what it really comes down to is anything that would normally require human intelligence, such as visual perception, speech recognition, any decision making processes, translation between languages, anytime you make extrapolated leaps from large data sets, generally speaking, that would require human intellect or human intelligence, for lack of a better word. This is akin to the Wright Brothers, and by the way, many others who were working on powered man flight. Imagine if that stuff had happened all behind the scenes, and make it even larger then. Make it hot air balloons being used in the late 1700s. Because if you look at the inception of artificial intelligence from just a few years ago, these machine learning systems, this has happened in the blink of an eye. Overnight, we have gone from a technology that you didn’t even know you were using or utilizing to now there’s an airport in every small town in middle America.
[00:02:52.090] – David Maples
You can go to any airport now and catch an airline flight. And if you’ve not been asleep, you’ve heard about ChatGPT or GPT-3, you’ve heard about these technologies that you can go try out for free right now. And so, to take a step back on this and say, how does this affect things in your business? It’s going to change everything. It’s going to change the ways we do business, the ways we communicate with other people, and just like any new technology, there is a litany of ways you can use it for good and there’s a whole minefield laid out in front of you, you can walk through that you’re going to have to traverse to figure out what actually works on the other side, what works in your business and what doesn’t. And so the purpose of this first episode is to give you a taste of it, to give you a couple of examples on it. And this is a multi part series. The third part is going to be the pitfalls, and we believe the second part really should be meat and potatoes, like how you can use this in your business.
[00:03:51.080] – David Maples
Now, people may already be using AI against you or for you, and you may not even realize that. Did you realize there are people right now who are using these artificial intelligence machines to generate their cover letters that they’re submitting with resumes to come work for you as an employer? Now, there’s a big question there, right? Because the cover letter was always supposed to be one of those things that you looked at. It’s nice to have. You wanted to see that people had it, but it was such an easy thing to provide. I’ve actually hired somebody based on the strength of their cover letter. Now, I’m going to have to look at those things with a more skeptical eye than I ever had to look at before because I don’t know if a human actually did it. And you’re going to be hard-pressed to tell those things apart if you’re not really paying attention. Now, what it is not, currently. It will change everything in a lot of industries and stuff like that. In its current level of sophistication, cannot and does not replace high-quality or even mid-grade quality work by a lot of individuals.
[00:05:04.360] – David Maples
But there’s a lot of people who may find out there are not jobs for them in the next 12 to 18 months, if they don’t produce anything better than the baseline material. And this is beyond the scope of this podcast to talk about the existential threat to society as a whole, if eventually it can replace those high-level things that humans currently replace, which is what they’re aiming for in the long run. But again, this podcast is not therapy. It’s literally just supposed to help you find ways to work in your business and not give any crap about it. So, just to introduce the different players in technology. There’s a whole bunch of different ones, and they’re the standard big tech companies you would expect to see. You’ve got Google, you’ve got Facebook, or Meta now, you have Microsoft, you have Amazon. There’s a whole bunch of different people, and the challenge with it is that there’s going to be a big gold rush on AI. There’s a lot of fool’s gold out there in them hills, too. By and large, there’s going to be a lot of things that are offered to you that are “AI-powered” that just ain’t AI-powered.
[00:06:11.030] – David Maples
Already seeing that. I’ve messed with 40 or 50 different applications in the past three weeks alone in my job and experimenting with different parts because I’m really trying to push the envelope on some of these. And some of these are not powered by the things that they say they are. I’m not here to blame people right now, etc., because there’s a lot of money in Silicon Valley going around. If you just slap on AI to the end of what you’re doing right now, you can attract more dollars and money. And so, there’s a lot of people operating under that. They just literally slapped the moniker on the things that they’re artificially intelligently powered, and they’re just not. I ran into a scheduling piece of software the other day that’s like, completely changes your day time, backed by all these people at these big incubators and stuff like that. And I looked at the software and I was like, “Oh, man, this sounds really good. Look at all this!” I got in there and I was like, “This isn’t AI-powered. It’s literally a set of if-then statements.” If you have a three-hour block of time, insert this thing in there.
[00:07:09.920] – David Maples
That is not really artificially intelligent. It is definitely machine powered, but it’s the difference in the machine making decisions for you based on a set of inputs and not. So, to let you know about artificial intelligence, at the baseline here, these large data learning systems, generally, they’re fed a ton of inputs, a ton of data, but then they have to be trained by human beings. And if you’ve been watching, there was a New York Times article the other day, or Time magazine article, that talked about, apparently, GPT-3 or ChatGPT was trained using a whole bunch of really inexpensive or poor labor in another country because they needed to train it on inputs and systems. So, what happens is, for example, if you use ChatGPT, and we used it today to pull some things, you can ask it to produce content for you. You can ask it to produce articles, blogs, a whole bunch of other things. And the company I work for, Catapult, we get paid to produce those things. I’m going to go ahead and say, for the record here, that none of this stuff, none of it, not a single bit of it we’ve generated now, can stand on its own.
[00:08:20.840] – David Maples
It’s grammatically, usually, correct. Factually, it’s somewhat suspect, but a lot of it is very boring, at the current place and time. But that being said, it represents an okay first draft. In some cases, you might say it’s good, but if you work with this stuff a lot, you’re going to see that you have to have people who really know what they’re looking at, etc, to make it better. But one of the challenges with this is that you’re going to have a race to the bottom. You’re going to have a lot of people trying to generate all of their content using these systems, and the systems produce a lot of the same sounding content. And nobody knows what Google’s response is going to be on this because ChatGPT and GPT-3 are powered by a piece of software called OpenAI, and the major investor in that is Microsoft. And they are running this on Microsoft servers because Microsoft gets paid for people running stuff on their servers, and they’ve invested several billion dollars into this company already, and they allegedly are investing 10 to 11 billion dollars more. These systems are very expensive to train and to run, but they’re pretty neat.
[00:09:30.490] – David Maples
I definitely recommend everybody out there to go out there and try them. But what you’re going to find is that it represents just… I’m going to give you some examples in this one about some things you can do with it. It’s not bad. If you give it some information, it can generate a first draft of a proposal if you give it enough inputs. It can write job description first drafts. It can write some general web content. It can write low-level pages for websites and that thing. Now, the content is not great and it’s not strategized. It hasn’t optimize for search. It doesn’t do a lot of the other things you want it to do if you’re really trying to build a website. But if you’re trying to just get it to generate something, it’s great at doing that. OpenAI has a piece of software called DALL-E that actually, like Dolly, the artist, and it generates some very interesting photographs, etc., from inputs. You can ask it to generate a picture in an oil painting style of dinosaurs playing poker, and it will generate an interesting picture for you, and it does it very quickly in a matter of seconds, in many cases.
[00:10:47.290] – David Maples
You can have it do those things. It doesn’t do any of the much more advanced things like that very well at this point in time. But people have made some really interesting pieces of art with it. But depending on who you ask, they had to spend a ton of time programming it and getting it to generate these inputs. But it is, nonetheless, very impressive. It can summarize documents for you. It can generate first drafts of software code in some instances. However, that software code, you would never use it in a system without having a software programmer look through it. But again, it’s only as good as the inputs being put in the systems, and these systems are going to get better and better and better over the next 4, 6, 12, 18 months. And you better believe it’s a really big deal, and you need to think about how you’re going to use it in this business. So, I would go ahead and recommend that everyone use these things. And in fact, some people out there, on the internet, are saying this will be the biggest threat to Google in its history, as a search engine, because it could replace everything that Google does with search.
[00:11:57.870] – David Maples
Well, that’s possible. I think it’s very unlikely, and I think you would be backing the wrong horse, so to speak, if you bet against the company that had a $250 billion profit last year, and we haven’t even seen their AI yet. And that being said, Google has just fired the first shot across the bow of Microsoft by announcing that their challenger to ChatGPT will be released in just the next few weeks, and it’s going to be called Bard. So, I think it’s going to be a very interesting world to live in over the next 12 months and seeing how these different pieces of software work.
[00:12:40.470] – David Maples
So, this brings us to the No BS segment of the podcast. The future of all businesses on planet Earth will be those that work with AI and those companies that don’t exist anymore. What it’s going to do is fundamentally change stuff about entire industries virtually overnight, at least in the next 12 to 24 months. It’s going to change how every industry works. It’s going to change the way how you generate proposals. It’s going to change the way how you generate content. It’s going to change the way you write job descriptions.
[00:13:11.720] – David Maples
Eventually, there’s a way that this thing could actually help you make decisions about your business. Imagine it as a business therapist per se that you can bounce ideas off of. It doesn’t mean it has good ideas, but half the time, you as a business owner, when you’re talking to other people in the room, you’re just trying to see if they’ll give you some feedback on what ideas you’ve been giving out to them. And that’s how it’s going to change the way the world functions. This is a bright line in the sand, and you can decide if you’re going to exist in the world and live and work with these artificially intelligent machines, or if you’re going to live in that previous world. And you’re going to find that that previous world, those people go the way of the dinosaur and no longer exist anymore. And that is your No BS segment today.
[00:13:54.810] – David Maples
Now that we know that AI is here, and it’s here to stay, we want to talk a little bit about how you need to look at things. So, in your particular company, what you need to look at is think about… Okay, are the things I’m doing here, do I have the proper…
[00:14:12.510] – David Maples
You need to play around with the technology first and see what things that could possibly save you time on because it could be a huge time saver for you and your company, right? You, as a business owner, if there’s a lot of things right now that you haven’t given to a third party individual in your company to help you out make these things happen, in a timely fashion, are there things you could actually ask the AI to do for you? For example, you’re putting up a new job post, right? Maybe before you generate that job post and you go give it to your team and say, “Hey, make this job post better for me.” This is just one case in point. Sometimes your team needs something to go off of, right? You could actually go to ChatGPT and say, “Generate a job post for me for a sous-chef at my restaurant who’s got two years of experience and trained at a culinary institute in my local area. Go.” And it will spit out, in about 30 seconds, it’ll spit out a complete job posting for you. Then you could take that as a skeleton or a framework, and you could give that to your team and say, make that better.
[00:15:17.200] – David Maples
Now, you as a business owner, I don’t know how often you do these things or how often you write your standard operating platforms, but these are things right now out there that already exist. Yes, you could hire a third person to make them better, and you probably should. But, at least when you go to that person, you can’t just say generate this from scratch. You can, but what usually is better is you give them a framework to work with them. And I think, for a lot of business owners, at least at the very, very beginning, you figure out, “Hey, how do we do this for a framework? How do I get my people a starting place to work on something, and how can they make it better?” You need to think about things that could represent in your industry as a real threat. It’s got a lot of information it generates, and some of that information just isn’t correct. So, to give you an example, we asked ChatGPT to answer a question for us. Could Teddy Roosevelt, President of the United States in the early 1900s, who was born in 1858, could he have worked on the Gettysburg Address with Abraham Lincoln?
[00:16:23.870] – David Maples
And that address, though not written in 1863, was delivered in 1863 by Abraham Lincoln in the American Civil War. Now, it told us that no, because they were people who lived in different eras. Now, that is technically correct, but woefully inadequate. If I asked another human being if they could have worked on it, you would have said no, because Teddy Roosevelt was just a kid. He was a child. He was five years old when the Gettysburg Address was delivered, and I don’t even know he knew how to write back then. So that’s the answers a human being would give. Now, ChatGPT gave us the answer saying, oh, they were people in different eras. Well, that’s technically true, but they did overlap. They did live during the same time, for at least a period of eight years. So, the question is, how does that work? They just talked about one of the challenges with ChatGPT. Apparently, CNET was using it to write some of their financial articles, and it was talking about the interest earned on something at 3% interest rate on a $1,000 investment or something, or a $10,000 investment is what it was.
[00:17:34.560] – David Maples
And it said, oh, after a year, your earnings will be $10,300. Well, that’s not correct. The earnings actually would only be $300 on $10,000 at 3% interest. So, that is wrong, and it’s not just a little wrong, it’s completely wrong. The amount of money is correct, but what you said is factually inaccurate. And the other thing is, one of the other challenges with it is, ChatGPT currently is trained on stuff back in 2021. It’s only as good as the inputs, it’s only as good as the prompts you give it, and you’ve got to know you’ve got to work with these systems. The reason I’m encouraging people listening to this podcast to start working with these systems now is you’re going to have to do some learning. You got to do some training. You’re going to have to learn how to ask these machines questions to give you feedback that at least is usable. And here’s the other thing, you got to know there’s opportunities for you in your business. It can be a real time saver. I read something the other day suggesting that ChatGPT is like having an endless supply of interns, who you don’t have to feed or pay.
[00:18:38.460] – David Maples
You’re going to have to pay them soon, I promise you that. But, that being said, I don’t know about you guys out there, but interns require a lot of oversight. Even the plucky first draft on these things, a lot of those things need a lot of work. So, you’re going to have to be aware that a lot of this stuff is required. You’re going to have to do stuff on it. You’re going to have to make it sound non-human. And we’re also going to talk about challenges in the third part of this. We’re going to talk about challenges, trademark, copyright. There’s a whole set of legal challenges and problems for you if you just try to… I think it’s like handing a five year old a knife. You need to understand, there are not ethical guardrails that have been set up on these systems. They’re having to train these systems how not to be racist and not tell people terrible things and not turn into very… These are computers. They have no ethics, they have no morality, and there is no legal regime these things currently work on. So, there’s a whole host of questions as we unlock the power of artificial intelligence, what it means for us.
[00:19:50.430] – David Maples
And we’re going to talk about that because this definitely deserves a couple more episodes to talk about. But, looking at the positive sides on it, the things you can make it do for you, there’s undoubtedly a lot of opportunities out there for business owners. But, remember, as I said earlier in the episode, there’s a lot of fool’s gold out there. There’s a lot of these pieces of technology they’re going to say, powered by AI. You’re going to have to try it. Your mileage might vary. You’re going to have to find out if you get the answers that work. The company I’ve been working for, Catapult, who helped sponsor this podcast, full disclosure, has been using artificial intelligence in the things we’ve worked with for the past several years. And one of the things we do, as we have writers and things like that, is we pull up drafts of this because sometimes it adds in information that we had not thought about. Generally speaking, we find that we do a better job on our own generating this content, but we’re trying to generate content that doesn’t sound like it was written by a machine because we know, just like when there were content mill farms on the internet back in the early teens, like 2011, 2012, for those of you guys who don’t know, there were companies who generated a lot of what Google would consider very thin, not authoritative, useless, regurgitated content for websites to try to get ranking on the internet.
[00:21:13.100] – David Maples
And it was damaging the power of Google search because you were getting a whole bunch of responses that weren’t what people were actually looking for. And Google, with two updates, in particular, they were code-named Penguin and Panda, basically killed those things overnight. I’m guaranteeing, as I said earlier, Google made $250 billion last year, I guarantee you that though they may not penalize AI content, if AI content sounds a lot like other content and doesn’t provide real value to the user, if you’re just regurgitating everything else that’s been said by another machine somewhere else… Because remember, these AI learning systems are not generating stuff from scratch. They are literally cobbling this stuff together from things that have been written before. And if you think that content is what you need to power your websites and your business, you need to think again, my friend, because at some point in time, Google will figure out how to detect it or work with it, and they will actually penalize those particular things, and you will find people losing ranking for those things. So, it’s something to be very aware of moving forward. It’s a cautionary tale and we’ll talk more about that in part three.
[00:22:20.120] – David Maples
I’m not saying you can’t generate good first drafts, but go add to them, make them interesting, add something new to the mix because that is what people are craving. They’re craving new ideas and just because AI will rewrite someone’s ideas for you doesn’t mean you shouldn’t be adding new things into there. So, that sums it up for The Buck Stops Here podcast today. Your three takeaways today are number one, AI is here and it is here to stay. There will probably be legal regimes that govern it in the next few years, but be aware that you need to know how to live, work, and love it on some level. Number two, you need to think about the opportunities in your company. If you’re only generating low-level content right now, if you haven’t thought about how you can use it to help give you some ideas for email responses and things like that, go out there and look at those things. If you’re providing low-level work for people right now that can be replaced by artificial intelligence, you need to get on that with a quickness because that is going to become an area of your business that is, literally, an exposed wound.
[00:23:29.700] – David Maples
The third thing is to know is that there will be a whole bunch of pitfalls and consequences of using AI, and we’ll talk about that in another episode coming up very shortly. I’ve been your host, David Maples. Thank you for listening. If you like what you heard today, please give us a thumbs up or give us a five star review on Spotify. Like and subscribe and tell your friends. Again, thanks for listening to The Buck Stops Here. Go out there and make it an awesome week.