ChatGPT as a performance booster for your success

This image has an empty alt attribute. The file name is itunes_badge_md-300x110.png

This episode is about what the ChatGPT is and what potential this AI tool has in practice. Have fun listening.

Further links

Train in Time

ChatGPT

OpenAI

Jasper.ai

Merlin

GodInABox

Introduction

Great to have you with us again for another episode of Digital4Productivity, and if you’re listening to the episode on Monday the 16th or Tuesday the 17th of January 23, when it came out, then check your calendar. Because you know, I’m also involved in innovative online formats for further development and one of the formats where I’m a co-founder is the Train In Time business expert pool. And the idea of the Business Experts Pool is that we have a pool of absolute experts who are available to answer your questions live online for an hour once a month. That’s the idea behind the Business Experts Pool. Some of you have probably already taken part in my monthly technology consultation. This means that you can ask me about video conferencing, the iPad, Microsoft 365 and new tools such as ChatGPT for an hour, and that’s definitely what this episode will be about. But in terms of topicality, this week we’ll be focusing more on the Business Experts Pool.

Thomas Skipwith on 17.01.2023

On Tuesday, the 17th from 10 to 11, Thomas Skipwith will answer all your questions about how to present professionally. You may know that I have written a book with Thomas Skipwith on the subject of online presenting. So Thomas Skipwith, Mr. Presentation and Rhetoric. I am Mr. Digital. In other words, we did it together.

Emmanuel Koch on 18.01.2023

Yes, then another one of my buddies, as you know, Emmanuel Koch. He has a new consultation on the subject of events. In other words, hybrid events. In other words, if you have any questions about live streaming, video conferencing, sound, image and video transmission, the first consultation hour with Emanuel Koch and myself will be on Wednesday the 18th from 10 to 11 a.m.. In other words, we will be happy to answer your questions, make of and also tell you a bit about the mistakes we have already made that you don’t need to repeat.

Hubertus Kuhnt on 19.01.2023

And on Thursday the 19th at a new time, 15:00 in the afternoon, dear Hubertus Kuhnt on the subject of meetings. You may have noticed that Shopify has banned meetings internally.

And I found this discussion very helpful, because the problem is not necessarily the number of meetings, but the quality. It’s often not the medium that’s the problem, but how it’s handled. It’s the same as with PowerPoint charts, where many people always say, Death by PowerPoint. PowerPoint is not the problem. The problem is the people who can’t handle it.

If you look at Steve Jobs’ keynote speech on the iPhone, for example, which is legendary. Yes, what does it have? A whole lot of not PowerPoint charts, but keynote charts, but ultimately also a bunch of slides. And I’m sure you’ll agree with me that he built them wisely.

The topic of digitization

And with understanding, we’re right in the middle of the topic of the episode. As you know, I’m always dealing with the topic of digitalization. Look at me, what challenges are there on the one hand and what tools are there for this? That’s, so, first turn on your brain, then technology. But the second thing I always find interesting is to look at new technologies and say what can you do with them? So in one of the next episodes, we’ll be looking at the topic of virtual reality, because I now have a weekly jour fix with a group of absolute professionals on the topic of virtual reality. And we discuss new platforms every week. And one thing you’ve probably already heard about isChatGPT.

ChatGPT

And ChatGPT is a tool, an artificial intelligence tool, a chatbot that is very intelligent and with which you can simply communicate, and more or less clever statements come out afterwards. And the interesting thing is to take a closer look at the topic. And I’ll start with the topic of ChatGPT. What is it? The first question is that everyone uses the term ChatGPT and doesn’t even know what GPT means. GPT stands for Generative Pretrained Transformer. And the idea behind it is that it’s an advanced artificial intelligence model that was developed to understand and generate natural language.

And the company OpenAI has trained this model and uses a technique called Unsupervised Learning to learn on the basis of a huge amount of text from the internet. And it has the ability to generate text in different languages and also different writing styles. Quite simply, if you write something in German, a German answer comes out. If you write something in English, an English answer comes out. That’s the simplest way. But there are even more possibilities.

What can it do? It can merge texts, it can translate, complete and it is a further development of the GPT3 model and has the ability to generate more complex and natural responses to queries. And now some people are saying, yes, ChatGPT is replacing the writing industry, it’s totally brilliant. Others say, boah, what’s coming out of that is ridiculous. It’s just copying existing texts. Perhaps a few thoughts on this.

The first thought is that ChatGPT, like other AI tools, is a tool. And let me translate that into the analog world. Michelangelo also had a tool in his hand, namely a chisel, with which he carved David out of marble. And I think you will understand that you can have the best chisel in the world in your hand. But hardly anyone will succeed in chiseling a David out of such a block of marble. So the question is, what is the bottleneck? The bottleneck is not necessarily the tool. On the other hand, if someone says, like Michelangelo, that he has a great chisel with which he can really realize his talent perfectly, then the whole thing is an excellent tool to become even better. And if I now bring the whole thing back not to Michelangelo, but to a more grounded basis, you can also symbolize this with a simple image. If you have a hammer in your hand and you are supposed to hammer nails into it, there are people who hit the hammers at an angle, they don’t hit them at all, they hit their fingers. And there are people who manage to hit a hammer in one go with the right technique.

And that is exactly the issue. And a professional can achieve good results even with a slightly inferior hammer. And if a professional has a much better tool, then of course he can do it even faster and easier. So it is a tool. And with tools, it’s always good to understand how the thing is constructed. How was it created? And with ChatGPT, vast quantities of texts were analyzed there and it was done with a so-called non-supervised model, i.e. the non-supervised means that there were no specifications given on how to analyze it, but simply many texts were analyzed, similarities were worked out, word repetitions were worked out, so in the sense of a language model with very, very much. The whole thing was completed in September 2021, which also explains a limitation, namely the limitation that you cannot answer all current things with ChatGPT there. And also answers the question, does ChatGPT replace Google? If you want to analyze current topics on the Ukraine crisis, for example, or on political issues there, if you want to write articles about them, you won’t get any decent results. However, if you want something on the history of the Ukraine conflict, for example, you’ll get a lot out of it. So it’s a tool. And just like in a toolbox, you rarely have just a hammer, but usually a handful of tools that you can combine intelligently.

Incidentally, there are also different systems in the field of AI text systems. ChatGPT is currently very popular because it is free of charge and is also very popular right now – the model has even made it into the Tagesschau. There are paid tools, but in my opinion the most powerful comparable tool is Jasper.ai. That’s Jasper.ai spelled with a J. Here you pay money. In return, you have, which I think is very nice, a series of templates, some of which you can also generate via ChatGPT if you write the commands intelligently to ChatGPT, especially since Jasper is also based on the same machine on the same artificial intelligence. However, it is an add-on that sometimes comes on top, where you have templates, for example, for chat articles for Amazon product descriptions, i.e. for certain types of text. This means that you can work with it in much more detail. You can also say, should the result be short, medium or long? Should it be in a “you” or “you” style? These are things that you can do there with an interface that I think is somewhat optimized and easier to understand. In terms of performance, however, both are based on GPT3, i.e. on the same machine, on the same logic. However, as always, it also makes a difference how I communicate with this machine.

Yes, from practical experience. What can such a model do? What do I use it for, for example? It can do more than you might think. So what you can do, for example. You have to write an article on a certain topic, a magazine article or a blog article. Then you write, for example, give me an outline for an article about this and that. And then you can also add that the article should be positive but critical. Or you can say, write me an article that is very negative and very critical of the metaverse and write it in the “you” form. Or you can write, write me an enthusiastic article about the metaverse to remove all doubt that this is the latest great trend, once and for all, then a different result comes out. So you, the system can’t read minds, nor does it have any knowledge of its own. It’s just an amalgamation of knowledge. These are also the biggest points of criticism. But ChatGPT also says that it has no knowledge of its own, it can’t develop any profound new ideas, but if you read the newspaper, if you watch TV, there are rarely any profound new ideas. Our scientific work is always based on previous work and then new knowledge is usually generated through empirical research. Of course, ChatGPT has empirical data based on the texts it has analyzed, but of course no new empirical data. That’s why we thought it would be a good idea to combine it with different things. In other words, the idea here is that it has been trained with huge texts from the Internet, which then have a structure that corresponds to the meaning of the words read, and it has no knowledge of the world, but it is based on the texts that you have included there. The thing that is particularly good is to say quite human-like creative texts, even poems, stories, jokes are also possible, but boah, with jokes, you realize that everyone has a very different kind of humor, it’s very digital. So you can give it a try. But what’s very, very good is simply when I say, for example, give me an outline of a podcast episode on the subject of ChatGPT, like this podcast episode, I can explain it wonderfully using this example, I simply said, give me an outline of a podcast episode on the subject of ChatGPT and then it came out, first something English, then I said please in German, so sometimes I throw it out in English anyway, that is to say, please in German. Zack. It worked. Then out came the outline, firstly the introduction with the points, then I’ll say roughly again, secondly what is GPT, thirdly how does GPT work, fourthly what information was used and to train GPT. Fifth, what is GPT good at? Sixth, what can GPT not do. Seventh, how do I interact with ChatGPT. Eighth, the future of ChatGPT and then a conclusion with a summary. I don’t think it’s so stupid. And when you say that only humans can do that. So a lot of the articles we write are not necessarily rocket science. And then I can say, man, write more about A) explaining what GPT is. Write more on this point and then more will be added. And what I’m telling you now is a mixture of what I got out of these results and my own knowledge and experience. So this image that I have of Michelangelo and the hammer, for example, is an idea that always comes to me in the current discussion when I look at it on LinkedIn, and that’s also a thing for me, I think that if you just ask a stupid question, you get a stupid answer from people and also from AI. And I often don’t just ask one person when I’m dealing with the topic, but several. And that’s why it’s a good idea to use it as one tool, but not the only one. So what can you do well? I can say, man, give me an outline on this and this topic. Then give me more on that topic. Which there is. Wonderful. I’m currently writing a new book and this book, which will be published by Gabal Verlag in the fall, is really about saying, what are the biggest mistakes in dealing with digital work tools and how can I avoid them? And I can simply say, please give me ten provocative book titles on the subject of the biggest mistakes when using digital tools and how to avoid them. Out of ten, four were really, really good. Then I can say, give me another four, I can say give me another four based on the second one. In other words, he always reads what is written above it. And if you then have ten titles given to you, there won’t be ten great titles, but one or two are really great. Or to say, for example, that last week I sent two new seminars to a seminar provider at the Chamber of Industry and Commerce in Reutlingen, one on the topic of digital self and team management and one on the productive use of Microsoft 365. So I said, give me ten seminar titles on this topic, give me ten seminar topics on that. And I came up with much better seminar titles than I had initially thought. In other words, the box is really, really good at reformulating things. And then I could say, give me some possible seminar content on a topic, and there were a few points where I said, that’s a bit of a stretch, but then there were one or two points where I said, gosh, I wouldn’t have thought of that. And that’s always the case, you know, I always say switch on the brain first, then the technology. And that also applies when using tools like ChatGPT. They can be very powerful tools if you use them wisely. Then they can complement good human intelligence wonderfully. I also always say that before we start with artificial intelligence, let’s start with common sense and not just work with it blindly. That starts with very simple things. On Thursday or Friday, we had a sales booster event with Karin Burger, who manages my speakers, where we met as a team. And we had already met in the evening, and then Karin Burger had reserved a table for us in a restaurant. And what did she do? She put the address in the WhatsApp chat and I just looked up where it was and took the ICE to Munich. From Munich I said, okay, Gräfelfing, I’ll take the S-Bahn. Then I got off there and did the following. I switched on the technology first and not my brain. So, I tapped on it and the WhatsApp chat said and that’s very close to the S-Bahn station. I tapped on it and it said 500 meters and I just started walking and kept moving. Then at some point I saw the car of my dear colleague Martin Senger and then I kept walking further and further away from it, so I thought to myself, wait a minute, it can’t be at S-Bank. What did I do? I didn’t enter the address that Karin Burger had typed into the map application, but simply googled the name of the restaurant and bup, I saw that it was right at the first traffic circle I had passed earlier, just before I saw Martin Senger’s car. A typical example when it comes to navigation. I know that the younger ones among you no longer know how to navigate with a map. I’m still familiar with folding maps in the car doors. So I always had folding maps of Munich, Stuttgart, Berlin, Düsseldorf, Cologne, Frankfurt – in other words, of the major cities – in my car, in the field, always in my pockets on the left and right. And then I used them to navigate. And I sometimes find myself thinking that if I know half the routes and I’m only driving by sat nav, I think, okay, I’ve got there now, but it wasn’t necessarily the smartest way to get there.

And why am I telling you this? I’m telling you this because I sometimes fall into this trap myself, where I say, man, I’ll switch on the technology first and much, much more important is to switch on the brain first and ChatGPT, if you don’t know it, you should definitely use it. There are also a few very clever additions, which is great. There is Merlin OpenAI, and if you have that with Merlin OpenAI, you have access everywhere via a key combination, correspondingly again to ChatGPT and there are two more very intelligent extensions on this topic. One is an extension that allows you to integrate current web results, called WebGPT, into ChatGPT. You’ll notice that it’s currently not as smart as ChatGPT. And if you have installed it as a Chrome extension, you can do the whole thing by activating or deactivating it. So that’s really, really good. There is also a Chrome extension from Jasper.ai. So that’s really quite good that you have that with you. WebchatGPT is the other one. And then there is also a very intelligent extension called WebGPT-Boss, which allows you to communicate with ChatGPT via WhatsApp. Here you can send up to 40 requests to GPT via WhatsApp per month free of charge, and any more will cost you money. And it’s much easier to use on a cell phone than in a browser. So it’s a really good idea and you can see that if you use it, you’ll be happy to use it again with practical tips and practical extensions. And you can see that it’s in line with my motto “Switch on your brain first, then the technology” and then “Simply use the technology”.

Conclusion

Yes, if the topic is interesting for you and you have questions, then please feel free to contact me. If you would like to have the show notes or have any questions, you can simply tap on the link to the episode, then an email will be sent to me and then you can write in, “Gosh, I would like the show notes, I would like to be added to the info mailing list or I have a completely different question.

And if you perhaps also want to give presentations for your customers, your employees or your sales partners there. I am also frequently booked to give customer presentations at the Volksbanken and Raiffeisenbanken. And if you say there that you would like a presentation on artificial intelligence or metaverse from a practitioner who can bridge the gap between technology and the implementation of meaningful digitalization and who will whet your appetite for the topic, then I would be delighted to hear from you.

Yes, with this in mind, I wish you a lot of common sense and use IT as an intelligent tool. Then you will become the Michelangelo of digitalization and be able to fulfill your full potential. Because for me, digitalization is above all a tool with which I can become much faster.

Yes, with that in mind, I look forward to seeing you again next week.

Yours, Thorsten Jekel.

Also available in: Deutsch

Antworten

Your email address will not be published. Required fields are marked

{"email":"Email address invalid","url":"Website address invalid","required":"Required field missing"}