Have you ever wondered what the future of AI holds? How will AI be used to transform the way we work and live? Mark van Rijmenam, a world-renowned AI expert, explores these questions and more on the Data Scientist podcast.
About the guest
Dr Mark van Rijmenam is a (virtual) speaker and best-selling author on the future of work. The Digital Speaker has deep expertise on how AI, big data, blockchain, the Internet of Things and other digital technologies disrupt your business, government and society.
Transcript: ChatGPT and the future of AI
Dr. Stylianos Kampakis: Hi, everyone. Welcome to The Data Scientist podcast. We’re very happy to have Mark here with us. Mark van Rijmenam. Did I pronounce this correctly?
Mark van Rijmenam: Yes, you did.
Dr. Kampakis: Okay, perfect. Mark, would you like to introduce yourself and say a few things about your work?
Mark: Sure. Well, thank you for having me, Stylianos. It’s great to be on your show. My name is Mark and I’m also known as The Digital Speaker. I am a strategic futurist, which means that I think about emergent technologies and how they change organizations and society. I have written 5 books on this topic, with my latest book being Step Into The Metaverse and Future Visions, which was actually written in collaboration with ChatGPT, in just five days and that’s basically what I do. I really like to practice what I preach to really understand all these new cutting-edge technologies and how they are impacting our lives. What I then do is use these technologies and I apply exponential thinking to them to understand how they’re going to impact our future and how they’re going to affect our society.
I’m also a global keynote speaker. I deliver keynotes for Fortune 500 companies around the world and large-scale events. I’ve just recently started a new initiative called the Digital Futures Institute, which focuses on ensuring a thriving digital future for everyone because I’m not necessarily convinced that we’ll end up in a fair digital future. I think there’s a lot of work to be done to ensure that we all understand what’s happening, how technology is evolving, and what we can do about it.
As a futurist, it’s my job to stay up to date about cutting-edge technologies, and even for me at times, it’s nearly impossible to stay up to date because everything is moving so fast. So yeah, I can’t blame anyone who has difficulties trying to keep up with all the changes that we see today.
I think it’s very important that we are staying up to date because otherwise, we will be sleepwalking into this digital society where we don’t really know how it works. So that’s the next venture that I focus on.
Finally, I also run a media platform called Datafloq, which focuses on emergent technologies.
Dr. Kampakis: Nice, I’m curious, how did you find ChatGPT How big is the book? How many pages?
Mark: The book is about 150 pages. Let’s put it this way. It’s not my best book. I wrote it very much as an experiment and literally, I changed like, 5 or 10 words in the entire book, but my whole objective was to have it written by AI, edited by AI, designed by AI, and recommended by AI.
So basically, I asked ChatGPT literally 2 days after it came out because I thought this is really cool technology and I want to investigate what you can do with off-the-shelf technology. So I asked it, “I’m going to write a book,” or basically “You’re going to write a book and tell me which technologies you think are going to impact our lives.” That’s the question I posed. That is my starting prompt. I listed several questions. I added a few of my own and removed a few of ChatGPT’s.
And then for each technology, I asked, “Now that we have this technology, which questions do I need to ask you in order to answer and understand how this technology is going to evolve?”
And then I used those questions to get answers from ChatGPT. And I literally copy-pasted the answer into the word document I was creating, only moving different sentences up and down to make it more readable and correcting incorrect answers, which there were quite a few of them.
Then I edited via Grammarly. Once that was done, I asked, “Okay. I have written the book, give me a title, give me a subtitle, describe the cover,” I used a description of the cover and I plugged it into Stable Diffusion to get the cover design, and then in the end, I asked, “Write me a recommendation for this book,” and it also did that.
It was a great experiment to get an understanding of where we are with these technologies. As I said at the start, it’s not my best book, because what I noticed from writing it is the book is quite shallow. If you write a book yourself, it has more peaks and troughs. There’s a lot more happening in a book.
It’s quite shallow in terms of depth and there’s a bit of repetition in the book because once you start using it at 450 pages, you start to see how ChatGPT works and it often gives the same structure for an answer, the same paragraph structure, not the same answers, but it’s a completely same structure. And that, of course, doesn’t read very nice.
So, I think as authors, we don’t have to fear for our jobs just yet. But it’s a great tool to kickstart your creativity and I use it daily now. I almost can’t see my life without it anymore. It’s funny how fast it works. But it was an interesting experiment.
Dr. Kampakis: Yeah, that’s interesting to hear. What do you think is going to be the impact of ChatGPT on society? It’s already passed a medical exam, an MBA exam, I think, and a bar exam. So it’s quite impressive. It seems to be amazing at generating logical, factual responses with a little bit of creativity. Obviously, its intelligence is not unlimited but it’s doing things that people wouldn’t expect. So, what’s your take on this? What do you think is going to be the short-term impact?
Mark: Well, it’s doing things that we don’t expect, but it also does very stupid things and makes up some things that don’t exist. So for my book, I asked, “Now tell me how did AI and robotics converge?” And it stated that “in the 1980s Robotics and AI converged to create self-driving cars that completely changed transportation in the world.” Well, if only it was the case, we would have been in a completely different place. Unfortunately, it wasn’t.
Later on, I asked it, “Who is Mark van Rijmenam and which academic papers has he written?” and it said that I had written 2 books, which is incorrect because I’ve written 5 books. It said that I had written 3 academic papers, which is correct, but then it came up with like 3 totally ridiculous, non-existent academic papers. All 3 had the same structure. So I think, from that perspective, it’s not the Holy Grail. It’s just a tool.
Now, of course, when we see that it can pass MBA exams or bar exams, or other exams, it basically means we need to ask different questions. Obviously, we had training ourselves. We are teaching our children, our students to think in a certain way. Well, that way of thinking is out of the door and we really are experiencing a paradigm shift and we need to change our perspective of how we teach stuff. How do we learn things? How do we deal with emergent tech? I think it’s best compared to when Wikipedia came about. Everyone said, “Well, we can’t use Wikipedia, because it’s terrible,” and nowadays it’s pretty okay. You can use it. You still can’t cite it in academic papers, but it’s reasonable. I think it’s the same thing we should apply to these kinds of technologies. We cannot forbid them because people will use them and the ones who’ll use them are always smarter in using them than those who want to want to forbid them.
So I think we should teach ourselves and train ourselves how to use these technologies responsibly and how to make the most of them so that it has a beneficial social impact on our society. I think that that’s where the main role is. So if we see that AI can pass our exams, we need to change our questions.
Dr. Kampakis: Yeah. Yeah, that’s a great point. I think the whole model of education needs to change. It’s funny that the initial reactions of teachers and everyone is like, “Oh, we need to have some kind of AI detector to see whether this was written by ChatGPT,” but these tools are here to stay. So let’s say that you go through the whole higher education and you learn to write your own papers, then you’re out of the door, and you no longer have a teacher above your head. What’s the first thing you’re going to do? You’re going to use ChatGPT. No one at work is going to tell you, “Don’t use ChatGPT because you should write your own,” Do you know what I mean?
Mark: Yeah, 100%. When I was doing [inaudible] in high school, you had to learn facts, you know. And then of course we don’t have to learn those facts anymore because we have a tool in our pocket that has all the information on the internet. So now we need to learn that now that we have all the information of the world, we can also prepare it and offer it in a way that’s easily digestible for what we humans like. So how do we deal with that then? How do we go about that? And I think that is something very important that we need to think about. This whole generative AI movement is a lot more than text. We shouldn’t forget that. It’s also images. It’s also 3D assets. It’s complete virtual worlds that are being generated. It’s music. It’s audio. It’s voices. It’s literally anything digital can now be created with generative AI. That’s a complete paradigm shift, where we create this synthetic future of ours that we need to think about.
I think we have to take this different approach from now on. Now that we have a society where this is possible and this will only get better, or worst you could argue, in the coming 5 to 10 years. How are we going to deal with it? Trying to forbid it is not an option. So what are we going to do? How are we going to be teaching our children? How are we going to educate the organizations on how they can benefit from it instead of that it will bankrupt them? I think that’s something that is very much necessary. Y
Dr. Kampakis: Yeah, absolutely. What’s your long-term view on AI? I mean, you’re a futurist. So what does the future look like? “Future” I guess I’m talking about 5, 10, 15, or 20 years down the line.
Mark: Look, as a futurist, I try to stay away from making predictions. Predictions are very difficult especially when it’s about the future as a famous saying goes. But jokes aside, I think what we see happening is there is a convergence of technologies. And that’s what really is going to change our lives and not necessarily for the better, to be honest. Because I think the way technologies are converging and with the level of understanding of the vast majority, let’s say 99% of the people, of these converging technologies, which is quite limited, to be honest. We are in for a wild ride and it’s not necessarily positive, I think. We should really pay attention to how AI and quantum computing together with the internet of things create data. How are they converging? What’s going to happen? How is it going to affect our society?
We’re going to have an internet, which is first in the next 5 to 10 years, a very immersive internet. In the next decade, it might move into becoming some kind of personalized or personal and is one internet using AI where everyone has their entire internet in the pocket, which is completely personalized with an AI coach that helps you and that’s there for you. In the next generation, in a decade, we might move into the thoughts internet, where we integrate with brain-computer interfaces and we can control everything that we do in our lives just with our minds. By that time, laptops and smartphones will have been long gone and they will completely change our world. Now will it make our lives better? To a vast extent, absolutely. Will that make our lives worse? To a vast extent, absolutely.
I think as we have seen how much the world has changed in the past 25 years, history has not always been repeated but it often rhymes as they say. I think we can see the same thing happening in the next 25 years when all these different technologies are converging. They’re going to completely change how our society is run. And we are going to move from a world where we have to make a conscious decision to go on the internet to a world where we are on the internet. Where we are part of the internet. Where we communicate seamlessly, literally with each other at any given moment. And what that world looks like is going to be quite interesting, I think.
Dr. Kampakis: Yeah, that’s true. I think what’s very interesting is that everyone’s been talking about, “AI and the future,” this and that, and it feels like ChatGPT took everyone by surprise. Because many of the things that we believed AI would do at some point in the future suddenly are possible. And I think that’s very interesting. What’s even more interesting is that if you look into how ChatGPT was built, there’s nothing fancy about it. It’s just deep learning with a bit of reinforcement learning and [crosstalk][inaudible].
Mark: It’s not intelligence.
Dr. Kampakis: Yeah, exactly. It’s not like some kind of quantum machine that is fueled by cold fusion and is on Mars or something out there, you know what I mean? This makes you wonder, it looks like we have the right tools there to create, something which looks like artificial intelligence. Probably many other discoveries. I don’t know what’s that, you know. Obviously, as you said, it’s difficult to predict the future. Don’t make predictions about the future but it feels like we might actually reaching… I don’t know what your opinion is on this… some kind of singularity effect maybe over the next 5 years. Maybe it’s not completely out of the question, what’s your take on that?
Mark: Well, that’s a very good question. I think it’s almost philosophical if you go down that route. If we talk about singularity, of course, AGI, Artificial General Intelligence, often people say that AGI is also conscious. But then, if we want to know if AI becomes conscious, we don’t even know what consciousness is in humans and human intelligence, let alone how can we figure out what consciousness will be in artificial intelligence. I think the singularity is still quite far away. Then again, I also thought that what ChatGPT is capable of today would be further down the route as well. So, it might come a lot sooner. I think it’s still far away because having an AGI which is as intelligent as human beings is quite far away. I am convinced that we will end up there. There are people who say we won’t, but I’m 100% convinced that we will end up in a world where AGI and super AI will become part of it. I think it’s just the next step in our evolution. Every intelligent life that evolved on Earth or anywhere else in the universe will eventually move to an artificially intelligent life.
Now, the way I look at it, AI by itself l will beat humans. AI, in a lot of domains, is very superior. In very narrow domains at the moment. But I think a collaboration between humans and AI will beat AI because I noticed already, when you start using these kinds of tools as they are tools, not intelligence, then you really augment human intelligence. I think that’s how we should look at it. But in order to use and create a human-machine collaboration, we need to be able to understand, and we need to be able to talk the language of the machines. And that’s something I think our society is lacking. Where I think we as a society have some really quick catching up to do. If we don’t do that then I think we have a bit of a problem when AI reaches singularity, we might be too late.
Dr. Kampakis: Yeah. So I guess one thing that we can all agree on is that the future is going to be very interesting and something which fascinates me is what’s going to happen if we merge ChatGPT with neural networks which can come up with discoveries. Like this neural network that Deepmind developed called AlphaFold, which would discover proteins and all this similar work being done. And they discover pharmaceutical drugs. So if we merge everything together, what is this going to look like? You might have a network that can discover all sorts of things. I don’t know, but it’s definitely a thing within the realm of possibility.
Mark: I completely agree. A couple of years ago, we saw that Facebook developed an AI that developed a completely new language that the developers of that AI couldn’t understand anymore and they pulled the plug. That was 5 years ago. Now if we use this technology and indeed if we start combining the different AIs that we have with quantum computing, then the sky’s the limit. We really have to be very careful how we’re going to do that and have the right boundaries present.
There’s a book called Superintelligence by Nick Bostrom which is a bit of an academic read, but absolutely one of the books on superintelligence, which I can definitely recommend reading to anyone listening. It sort of explains know what can go wrong if we don’t think about this kind of stuff. It’s definitely useful to read that stuff at this time of day.
Dr. Kampakis: Yeah. I’m familiar with him. I think he’s a philosopher at Oxford, right?
Mark: Yeah.
Dr. Kampakis: He has some interesting views on the philosophy of AI. He’s actually one of the few people who is a non-technical person, but his philosophical views on AI are very interesting and highly non-trivial because most people just repeat the same things they read on media. Very intelligent.
Great, amazing. So I think our time’s up. Where can our listeners learn more about you? What’s your website?
Mark: I’m pretty visible online. So, it’s thedigitalspeaker.com. You can find me on LinkedIn, and on Twitter @vanrijmenam, which might be more difficult than The Digital Speaker. If you just Google “the digital speaker” then you will probably find me.
Dr. Kampakis: Amazing. Thank you, Mark. It was a pleasure.
Mark: Thank you very much for having me, Stylianos. It’s a really great and quick conversation.
Dr. Kampakis: Likewise. Thanks, everyone. Thank you for being here with us and make sure to go to thedatascientist.com for more content around AI, data science, and blockchain. Thank you.
[END]