AI & Chat-GPT with Jean-François Noubel. Episode 29.

Welcome to Pass the Mic podcast, where we tackle complex ideas and difficult questions with people who are curious, able to challenge their own beliefs, and who have the willingness to objectively listen and learn from the shared insights of others. Welcome to episode 24 with Jean-François Noubel. Today I have the pleasure of discussing with Jean-François Noubel, who describes himself as a temporarily human being. For more than 20 years, Jean-François has worked in the field of collective intelligence, a modern research discipline that explores how living systems work and the evolution of our species. He works on the next crypto technologies that will soon enable the rise of super smart distributed organizations. Jean-François helps evolutionary leaders build enlightened organizations towards the post-monetary society. And I'll add with this element, interestingly, Jean-François lives in the gift economy. Several years ago, he left all his positions and mandates and tore up his resume in order to free himself from any etiquette and social status. With that, he gained full creative freedom to live in the present millennium. His new path allows him to help evolutionary leaders and train human new notes, those for whom the 'Go Hack Yourself: Designate a Way to Exist.' Welcome, Jean-François.

Thank you, so nice to have this time with you.

I'm so glad to have you back. Together in a previous episode, we talked about how we can create a vegan and compassionate economy using Holochain, which is a form of blockchain, so I invite listeners and viewers to check out these 3 video series on this discussion. But today we're going to discuss the revolution happening right in front of us, right in front of our eyes, which is taking the world by storm: artificial intelligence and ChatGPT. Now, in another episode, interestingly, I hosted a group of AI researchers and philosophers to answer one question: is AI conscious? This discussion was really fascinating because we tried to know whether artificial intelligence machines can have consciousness. And to answer that question, we had to explore how consciousness is not only about intelligence, but it also includes self-awareness, intentionality, and subjectivity. And we talked about the Turing test, which we were just talking about before we started. We also highlighted the potential ethical implications of creating conscious machines. But at the end, we ended up agreeing that consciousness in AI remains a complex and unresolved question. Today we're talking about artificial intelligence and ChatGPT, and we'll see on this topic if we're able to find answers with you, Jean-François. I've sent a few questions for which I'm dying to hear your answers, and we'll go wherever this discussion needs to go. So let's just get started. I want to begin with one First question, why are people so afraid of artificial intelligence and ChatGPT? What's that all about and what is it that we don't see?

Well, first, we don't see what we don't see. So sometimes we don't know what we don't see. So I guess we can have different levels of answers. First, through history, we've seen people afraid most of the time of any new technological breakthrough. It happened for, you know, steam engine. It happened for computers. It happened for telephone. You know, many people would think that telephone would kind of penetrate in households and you could cheat on your husband because husband goes to work and you stay at home and now you suddenly have someone sneaking into the house without, you know, using the front door or some kind of back door. And things can happen out of control and you would have lots of people afraid also of those things. So you, in general, you have this kind of general fear. And I remember 20 years ago, 25 years ago, the fear about the internet itself. I remember politicians in France trying to outlaw the internet and they had some, you know, other reasons for that with lobbies and, you know, France Telecom and all these things. But we have this kind of general observation we can make. However, we also see a second level, I would say, of people scared by what they may have heard through or seen through sci-fi, like, you know, the robot, the intelligence takes over, takes control and has bad intentions and makes some kind of dystopian world and all these things. And I don't say it cannot happen. Of course, we want to talk about this, but I think it also has some sci-fi background also in this that may not have too much rationality in this. And so I guess we want to take this question into now what kind of rational fears we could have or we should even have, which then leads to maybe a conversation among more specialized people, because I don't see much awareness so far in the large public about the real threats that AI can bring to this world. And by the way, I want to make it clear here that I don't claim any expertise in AI per se. I have a field of expertise as a researcher in collective intelligence, in augmented social intelligence, in distributed systems, but not in AI. But of course, all these fields kind of entangle each other. But I want to say that I will always speak from my perspective as a researcher in more like collective intelligence.

That makes sense. So you're saying people are afraid of AI because of the fear of the unknown, which has been fed by years and generations of sci-fi movies and monsters and robots taking over the human species. Do you think there's something else that feeds this fear, this resistance that many people have? Is it something, is there something around the idea of power?

Well, first, what do you mean by people? Like, do you mean the general public? Like what, you know, people would say around a dinner table? Or do you, do you mean like the people involved in this field of research and, or other people like me, like deeply connected to that?

I would say both. Both. We can definitely see some form of resistance. I mean, recently some of the experts have asked for a halt on pushing forward the development of ChatGPT-4, 5, etc. But also in the public sphere, and you can see that in the news or even on social media, some resistance from professionals in actually using ChatGPT? Okay.

So I think we should definitely focus maybe on people like really involved in the field and who try to bring very rational questions and the debate and controversies we need to have in that field. I don't feel much interested in what the general public says because also I don't have like evidence or data. I just have like a feeling, you know, hearing people around me, but it doesn't mean anything. But so we have a few things. First, an AI can blindly grow based on the orders of growing from its creators. And you could think of any AI where you would say, okay, you have to become like hyper-specialized in playing chess and any board games in the world. And we want reality to become a big board game. And it could literally transform the world and make our species extinct and make this planet a big board game or thousands or billions of board games getting played because AI has just fulfilled that. And if it has this capacity to expand, that means it can bring in more and more data and knowledge. And then, of course, by self-reinforcement, improve this knowledge and then learn or use some piece of code to break into server firewalls and all these things, and then start to use more power for itself. Like you could have in 24 hours a burst of AI, like a billion hackers, top-level hackers working at the same time and breaking in every possible system in the world. And then to take control of those things for its own purpose, for its own original purpose of either making board games or delivering knowledge or whatever you give, whatever original intention you give to that AI. So it leads lots of questions. First, let's just follow up on this scenario. It breaks in every possible platform, your smartphone, everywhere.. And now you have pieces of code that kind of gather data, more data, more knowledge for the AI that delivers you also information, news, fake news, whatever. And that becomes interactive with you. The next step that AI would need, it would need also things done in real life, like in the physical life, because so far it works on computer. Like if you turn off your cell phone or your laptop, then you have no more AI with you, right? How would it change the physical world around you? Well, it would need to use human beings for that. And how do you use human beings? Well, what if that AI would pay you for doing something? What if that AI would give you some forms of reward, like becoming famous, if fame works for you? What if that AI blackmailed you because you have some dirty secrets that you want to hide, or you think you see them as dirty secrets, whatever? But AI could have very powerful leverages also on people, either to have them work for it or through fear and blackmailing and those kinds of things. So this does represent a very serious scenario that many specialists see, hence why they say we have to keep it in a closed container, because if it starts to write code out there to break in, you know, maybe your cell phone and then another cell phone and maybe this server here and so on, it may start to, you know, burst in an exponential way. Does that make any sense?

It does. But what you're talking about is really the intention. And so my question is, who gives the intention to AI? Because that leads to creating leverages, whether they are positive or negative. So who controls those intentions? Who gives intention to AI?

Well, I think you asked the very core questions because I hear people asking, you know, should we fear AI? Just like, you know, should we fear computers or should we fear TV or the telephone? And I ask this question all the time, whose hand holds the hammer? I think that leads to the much more interesting question, like who does it serve? Does it serve like personal private interests? Or political interests? Well, in this case, I feel extremely scared about AI because it can leverage, you know, the power for these intentions and the people in control of those tools. You know, then they can use facial recognition. They can use RoboCops. They can do whatever they want on your bank account. They can fire you on work, whatever. They can have full control of society. Okay. So either private interest or governmental interest seems like a very dangerous kind of AI. Okay. Now, hence the question, can we make AI not only open source, but controlled by society itself? That for me seems more like a question we should ask about AI rather than dangerous or not dangerous. Like anything like nuclear power, like cars, you know, laser, whatever, writing, you know, you can use writing in such a dangerous way. Imagine if writing remains just in the hands of a few, which it happened, by the way, in past history, and we know the consequences of that, right?

Very true. Okay, so the reason why people are afraid of artificial intelligence is the fear of the unknown. Fed by sci-fi, the fear of becoming obsolete as human, because potentially being controlled and manipulated by AI. And that goes to the third reason of this fear, which is who creates the intention, who does it serve. That could also be a cause of the fear.

Yes. And It leads also to more questions because you may have a group in control of the first intention, you know, like ChatGPT. I mean, the claim of OpenAI as a society, they say, well, we want to serve the greater good. Like in most cases, you have highly well-intentioned people. Like they really want, just like people at the, you know, the beginnings of social media, they really claimed genuinely that they wanted, you know, people to connect and ideas to flow and more democracy in this world. And, you know, all these things. And then AI, first versions of AI came in to find a way to capture your attention, hence what we call the attention economy, so that you would spend more time on Facebook or Instagram or TikTok and all these things. And so it has developed meta-algorithm. That means you have like one special algorithm for you Virginie Glaenzer, or another one for me, because they know what, you know, how I work on Instagram or on TikTok and what will keep me stuck into those things. So we already have AI controlling us in many, many, many levels connected to the original intentions of the makers, but then it's kind of escaped their original intention. And the next versions of AI could also escape the original intention just by creating, of course, poisonous emergent effects that we cannot see. But also, if it has to grow by itself, if you, you know, create self-reinforcement mechanisms like, okay, learn by yourself, improve your capacity to resonate, improve your knowledge, check your knowledge in a better way, and so on, then you give it autonomy. And that autonomy can become its own intention. Like way beyond the intention of the creators. And we have already many examples of this, you know, like AI learning new languages that we didn't know, AI producing code that no one has trained it to do, but it delivers like amazing kind of code and, you know, a million times faster, all these things. So then we have this question, like, can it become its own intention? Having its own autonomy. And then it connects with another important question about substrate, the notion of substrate independence. You see, what do you mean by substrate? Let me go there. For millions of years, memory and reasoning could happen in— needed to happen in a biological brain, a kind of carbon-based organism. Okay, so as a human being, but also as a bird, or, you know, any kind of advanced form of life, it would retain information, process information, make reasoning at their level based on the carbon-based substrate, what we call life forms. Okay, but now we've seen recently, I mean, in the past 50 or 60 years, that we can make storage of knowledge and information and algorithms, that means our functions of reasoning on other substrates than the biological ones. You know, for instance, you can even, you can use a sandbox and put some pebbles there to do bits and make the memory of a number or the representation of a word. You can use other substrates for that. Now, the sandbox remains a passive object, but you can use now what we call computers so that from this early memory, it can start to make now reasonings, calculations, and then chess games, and then guess the next word to put in that sentence and all those things. Does it happen on a human brain? No. Does it happen on a carbon substrate? Absolutely not. It happens on the silicon substrate. So some people think or claim or fear that, okay, well, maybe life and consciousness doesn't need the carbon substrate or cannot evolve any further on the carbon substrate. It may evolve. Evolution may shift this not only intelligence but then consciousness into silicon substrate or maybe other forms of substrate. But I think we have to remember this, just this fact that we don't even need to discuss anymore, that functions and memory can operate on any forms of substrates. Okay. Now, do we need carbon and cells and DNA to have consciousness and subjectivity? No one knows yet, but we certainly can see if we look in the past, a relationship between the complexity of a system that self-perpetuates itself, that grows on itself, that learns. Okay, so a parallel or a progression between those substrates and complexity and consciousness and complexity. The more complex, the more consciousness it has. On a biological substrate, but why couldn't it evolve into other forms of substrates? You see? And just that, just by not knowing, not having like evidence that it will, well, that makes very big important questions, which I don't hear very much in the large public, I would say. Does that make any sense?

Yes. And I invite the listeners and viewers to rewind a little bit what you just said and listen to you again. I think you're opening a door to questions that very few people are asking. That brings me to my second question, which is, since you were talking about the evolution of human species, when I look— when I imagine the evolution of human species, I imagine in my mind the discovery of the fire, the invention of the wheel, the electricity, the discovery of the telephone, the computer, and now artificial intelligence. And the digital self. What does artificial intelligence and this digital self mean for the evolution of our human species?

So as someone who participated in the early days of the internet, it became even at the time, like in the mid-'90s, very obvious that we would have more and more a part of ourselves online, okay, in the digital world. And of course, it already exists. You have, you Virginie Glaenzer, you have probably, I don't know, Facebook account, LinkedIn account, Instagram, TikTok, who knows what, you know, all these things. Plus you also maybe participate in online forums and you have a website and all these things. So you have lots of traces of you and manifestations of you in the online world. Already. You already have a digital self. Now, very scattered because you may have a part of it in an online game, another part at your job, another part in your Instagram account, and so on. Okay. But even today, most of your digital self does not belong to you. It belongs to Mark Zuckerberg. I mean, to, you know, Alpha to Omega, to whatever, you know, companies exist that have the ownership of this digital self. So that raises first an important question. Now we have the version 1 of this digital self, meaning, you know, this kind of scattered manifestations of you, what I call semiotic pheromones, you know, like you leave traces of meaning or sense everywhere on the internet. And of course it does not belong to you in most cases, unless you really make your own website, your own blog, you know, but in most cases belongs to those holding the platform. Okay. But more and more, I mean, in the next years, we will see the rise of AI that can represent you, that can become your digital self, your digital ambassador, because the same way it gathers millions and billions and trillions of information on the web about, you know, on Wikipedia, on forums about general knowledge,. But it can do the very same thing about you, Virginie Glaenzer, like the thousands and thousands and thousands of information you've already left on Facebook, on YouTube, Google, your telephone company, and all those things. Like you add, you know, all the GPS positions, your opinions, your tweets, the pictures, family, how you, how your finger works on Instagram. That says a lot of things about how your brain works, you know, maybe your very private things on your sexuality. And all these things. Like, just put all this thing together and I have tens of thousands of data points, okay, that I could use or we could use either in a good way, like, okay, what if that now starts to represent the real Virginie and my digital self or my digital bot, like my bot me connects with your bot you. So they have underlying conversation. About lots of things which we have no awareness, you and I, as in our biological being. And what if our bots have actually not one-to-one conversation, but millions and millions of conversations at the same time on Earth? Like my digital self in this very moment would have conversations, connections, opinions, tweets, whatever, with millions and millions and millions of other agents around the globe. Okay. And by the end of the day, it could tell me like, oh, you should meet with that person, by the way, because you have to do it, you know, or you should change something in your diary, or you should date that person. It could become like an amazing, powerful counselor or someone you become also fully dependent on. Depends again, you know, both, both sides. So the digital self here that I describe happens. And I would say in a kind of open world in the best case scenario case. But of course, the way the course it has today, well, Facebook wants your digital self. Google wants your digital self. The government wants your digital self. Like they want to control that. They want to, you know, have it serve their own interests, either for security, for political ideologies or religious ideologies or business ideologies and all those things. So I think we have to worry about those things and ask the question today, how can AI become my friend, my servant, like the individual AI and not just the general AI that we see in ChatGPT. We will have in the near future trillions of different AIs, you know, an AI for you, an AI for, I don't know, my camera, whatever, interacting with one another.

Does that make any sense? Yes. And actually, I'm so glad that you're bringing this conversation to this point where we are realizing that AI can take over our life. I've often, I might ask myself, why AI in this evolution? You know, again, going back to this narrative that we have about how humans species have evolved, why AI? And I want to share a theory that I have about why AI now, and I'm curious to know what you think. So bear with So my theory of why AI is coming to existence today is that I believe that we unconsciously believe that AI is our savior. And what I mean by that is that we think AI will rescue us. And the reason is because if we look at how we've evolved and what we've done so far, we've killed our own species, we've harmed our planet, we're fearful creatures, we're greedy, we're never satisfied, and we're selfish. And we've tried We've talked about it with other people. Um, we've shared these personality disorders with others. We've hired personal trainers, therapists, coaches, you know, and, but nothing worked. So now the purpose and reason of AI is to save us from ourselves and create the world that we long for. What do you think?

So I would maybe, I would argue on a few words, like you say the purpose, I would say the potential rather than purpose. It does have the potential to embrace such levels of complexity about climate change and about even our very individual issues and struggles. You know, what if I had like a personal advisor that knows even better than me, you know, what I— it could give me any advice that I ask for and become my best coach. And AI can certainly do that in the next few years, like better than even any coach, like a very benevolent kind of AI and help me run my life. But what if also AI became the best policymaker, could also even, you know, run for elections and make a fair society. I mean, it has the potential for those things. And then we would become, you know, the people asking for new things and AI would resolve very complex problems that we cannot even think of in our human level. And even the best intentioned politicians cannot even embrace those things. And by the way, I've said those things for so many years from my field of collective intelligence saying that, and not just claiming, but kind of observing as a scientist that pyramidal collective intelligence, pyramidal structures, companies, governments, administrations, armies, religion, everything pyramidal has produced the complex world, complex world in which we live and cannot embrace the challenges of complexity that has provoked. You always need evolution, always need a more embracing, a more encompassing system to embrace the next level of complexity. And evolution has always worked through those kind of quantum leaps, you know, like from one level of complexity to the next one and the next one. Hence, we don't have a world made only with bacteria anymore. We have a world made with human beings and trees and, you know, and all the— and computers and all those things. So from an evolutionary perspective, evolution will either fail and collapse because it does not happen in a linear way. Sometimes you have like, you know, it has to fail and then to start over again, but it can also upgrade itself very quickly. And a whole part of the transhumanist movement says this also, like don't count on biological evolution and even biological substrate to address the level of complexity of the world. No fucking way. Like, don't even think of elections or changing institutions or making a new constitution. It will address those things. It may improve, of course. I don't see it as completely neutral. But the real leap that we need has to happen in an evolutionary leap, which I always say myself. I've always said those things. And AI has the potential. To play a fantastic role there, just like electricity. But you can also kill someone with electricity. You can run horrible, horrendous things with electricity. So just, I would just argue on the potential, or I mean the purpose versus potential. Although just like you, I claim that no new thing happens randomly. Like the writing happened because of the need for the tribal world to move into civilization, civilizational world, because then you could unite more millions of human beings, which the tribal world or tribal reality cannot do. And you needed the writing to do those things. And it changed everything, not only externally, but also in our inner subjective space as well. And so I think if you and I will live long enough in the next few years and the next decades, we will see, I think, a level of leap so important, like an evolutionary leap, a species leap or a species evolution. But it may also go very, very bad, I think. So what do you think about that?

Yes, I like your answer, and I think what you're saying is that it is up to us individually and collectively to know how to use this AI. And I want to end— I know we could talk for hours, but I want to end with this last question, which is a little bit more practical. I mean, you advise and you help leaders on a regular basis. So what's another way, now that we've set up the context and the possibility for AI to become our greatest savior or our worst nightmare. What is a way, maybe a more positive or constructive way to see and experience artificial intelligence and ChatGPT? Is it, you know, what should people, how should people look at this in a more constructive way? What do you think? I was going to say, I was going to add, well, you pointed earlier that we need to learn from social media, that the social media damages that we today acknowledge, we need to learn from that and bring this knowledge, this understanding to AI. But what do you think? How can people see AI in a more positive and constructive way?

Well, do you ask like see or do something? Because we can always see the two faces of a coin. So then just a matter of seeing. But if you ask, you know, what could they do to contribute to a better world with the leverage of AI, then that may lead to a few thoughts that I can share. Yes, let's go that path. Okay. So let me start with the bad news. I don't think any vote, you know, your ballot will change anything. I mean, it can change something in the old world, in the old mindset, like, you know, would you put more money on school or on warfare and all this kind of things that we see in the conventional geopolitical pyramidal world as we know it today. And for most people, that remains the ultimate reality. Like they don't think you know, outside of that box. Okay. However, I would, first, I would invite people to become knowledgeable in the forthcoming technologies like the crypto technologies and everything that will enable distributed organizations and distributed societies to rise rather than giving your power and your sovereignty to other people, we have now the means to use better tools. And they arrive, you know, like the next 5 years, you will have the capacity to use this tool easily without having, you know, the need of any savviness on those things. They will just look like conventional tools that we have on our cell phones, but they will not operate through one centralized platform that belongs to a few people. It will work in a completely distributed way. So we need to become knowledgeable in those technologies and usages of distributed applications, because distributed means distributed power, distributed currencies, distributed voice, and all those things rather than centralized and with those hierarchies and social castes as we know them. And I don't mean like a bad criticism of them, I just say they don't work anymore. They've done their job, you know, maybe we needed those things. But now that they face, we face a world that requires highly individualized people. And highly individualized people means social complexity. And highly individualized people do not like to work in social pyramids and hierarchies. They want to have their own sovereignty and interact in a mutual dependency, interdependency, rather than getting dependent on some kind of social hierarchy. Okay. So if you want that, then okay, good. Hey, but maybe you need an infrastructure for that. Maybe you need some technologies, including AI and distributed applications that give the technical substrate for this. And how do you get that? Well, by having the awareness of this and then become a supporter and an actor, a player of those things. And you have many ways. First, understand that, talk about it, communicate about it, become beta testers. You have so many ways. Develop new things, try new things with your communities. You can play this kind of game. So I think the real shift can happen here. And the second thing, using also those technologies free ourselves from money and use free currencies, other forms of currencies. Because as long as, you know, we may all want to not become dependent on bad AI, you know, well, okay, but we have this race, this crazy race. Like, we all see it in a completely powerless position. Like, okay, everyone needs to win the race. Even with well-intentioned people. And that means you got to make AI stronger and stronger and stronger without even checking what it'll do. And that can become a very terrible thing. Call it the tragedy of the commons. And you have many other kind of systemic emerging effects that just drive us collectively onto the wall, even if independently, individually, we want to go in the other direction. So the individual will does not suffice. We need to understand the systemic forces and what drives— the main driver of those systemic forces today has a name: money. We need to use other currencies that will drive a completely different, completely different social contract.

That's why cryptocurrency is so important, not so much about alternative to fiat money, but it's because it opens the door to a new form of power, uh, and it empowers— it allows other forms of communities to create values, and that leads the society to move away from being dominated by money. But that's a whole conversation, a whole other conversation. I really like what you're saying, and really what I'm getting from what you just said was that ChatGPT as an the first element of artificial intelligence is not just about finding titles or creating content or even asking, you know, the tools to do a number of things, creating tables, etc., but it's about not waiting to wait on other people to tell you what to do. This is about taking your power back and choosing to spend time and acquiring knowledge on what this tool can do for us individually and collectively and being part of those discussions.

Yes, absolutely. What you said about cryptocurrencies, I would maybe add a little thing to what you said, like, do they provide some next level of distribution of power? Absolutely, yes. But we have to see it, I think, as just like the early days of the writing. In the early days of the writing, you had an elite controlling the writing. You had the scribes, you know, the knowledgeable people who would control and then would control society with this. Okay. And it took a few thousand years before it became accessible to everyone. And then you can really have an explosion of ideas and communication and the growth of culture and knowledge. But for the most part, if you look at the past 5,000 years, In most places, the democratization of writing and reading happened like very, very, very recently. Okay. And I think we see the same stage now. I hope it will take more than, I mean, less than 5,000 years, but we see the same stage like today, cryptocurrencies are remained controlled in the hands of highly literate people, even those with good intentions, but also in the control of big powers, you know, big investment funds who make, you know, the ups and downs, you know, the pump and dump thing on cryptos. And it has turned into a very insane kind of game in the old world. And it doesn't mean it doesn't have the potential, just like see the writing, same thing. It has a huge potential. But today we haven't used and harnessed the real potential of both, you know, cryptocurrencies and soon to come crypto technologies embrace something much bigger.

Fantastic. Well, on that note, thank you so much for spending the time with me and the, the people who are going to benefit from this wonderful conversation. There's so many nuggets that you said. It was really enlightening, and I can't wait to go back and just listen. We'll do a blog post as well, as we do as always, trying to bring sense and clarity to a world that is very complex. So thank you so much, Jean-François.

My pleasure, Virginie Glaenzer. Whenever you want, we can have any of those conversations. Thank you.