Episode transcript The transcript is generated automatically by Podscribe, Sonix, Otter and other electronic transcription services.
Hi everyone. Welcome to the 5 Minutes Podcast. Sorry for my voice. I'm a little bit tired. I just arrived in Lisbon, coming from Austin from the South by Southwest, and my flight was canceled and delayed. So, it took me almost 50 hours to fly from Austin to Lisbon. So I apologize for that. And, of course, these 50 hours and the past seven days were very transformative for me. It's like the knowledge and the exposure I had in the past days was like trying to drink water from a fire hose. So it's very hard because it's just an avalanche. So, I recorded 26 hours of talks. So I'm trying to, you know, summarize using AI to summarize and help me figure it out. But I want to share with you some of the key takeaways I shared recently on social media, five quotes that I will not repeat here. That was, I would say, something mind-blowing for me, but now I want just to summarize what I saw in terms of basically four trends. The first one is related to AI, the concept of natural language and general intelligence. It means that many people think that the path to a more sophisticated artificial intelligence is through large language models. I would say that ChatGPT 4.0, etc., and this is not true. Most of the talks there were saying that the path to artificial general intelligence, this human-like intelligence, is not true large language models, but it's much more true reasoning models. And this is why there is such a massive investment in reasoning models. Because reasoning models, they they go through that network many, many times to reason about their answer. This is why, for example, it's slow many times compared with the other ones, and it's more dedicated to more complex thinking. So these will be the path. And there are people saying that, you know, maybe AGI will come in five years. Other people say that AGI will never come, or AGI stands for artificial general intelligence. So, you know, I don't have the competence to say. But what I saw, it's the evolution that we are seeing in terms of AI, is just mind-blowing. Okay. Ai connecting to physical objects. AI connecting to robots. You know, so it's a very, very powerful thing. And two points also on AI. It's one is the dilemma between power and efficiency. If you look at the brain of a child, this brain consumes around 24W of work. And, you know, you show ten mugs to a child. He or she can easily understand what is a mug just by learning from ten mugs. The computer is on the other side. Instead of using 24W, it will use 24kW or even more than that just to to understand and reasoning on that, maybe it will need to train on 1000 or 10,000 mugs to identify a mug. So there is still this dilemma. So we talk about that. Okay, a ChatGPT, for example, is writing like a human. Yes. But the consumption of energy is not efficient. So there are a lot of discussions about small language models to make it more efficient, okay? We do not need to have this huge data service, data centers, and all of that. The second is about continuous learning. Instead of learning, you create a large language model with a specific date. The challenge now is how I can talk to ChatGPT or Gemini about something that happened on the news half an hour ago. So, this will be the concept of continuous learning. So this was very, very impactful for me in terms of, for example, environmental scanning. For example, if we want to analyze risks for our project, we can do that with this technology in a dynamic way. The second point is biotech, okay. And this new paradigm of life genetics extinction. So I saw the talk from Colossal, a genetic firm that is trying to bring to life animals that were extinct. And it's just crazy. Honestly speaking, of course, I'm not an expert on the topic, but the way they are using technology was really mind-blowing. I would say to rebuild the DNA of, for example, a mammoth or others. For example, they already discovered pieces, for example, like the fur of the mammoth, and they implanted this in the DNA sequence of a mouse. And then the mouse grew this, this kind of fur. So it's really, really disruptive. And also this biological computing. So, for example, the use of brain cells to power things. So, this integration between biology and technology will be very disruptive. So this is another trend. And these will generate many, many projects for us. The third topic I want to discuss is quantum computing. We had many talks on quantum computing, and this was probably one of the topics I heard more about. I attended just myself. Maybe 4 or 5 talks on quantum computing and quantum computing are moving. Of course, it is still a massive amount of money, a massive amount of challenges because of temperature, because of size. But we are moving towards that. When this becomes accessible, we will be able to make forecasts and decisions, and we will make very complex decisions using them. And it's something that is already happening. IBM is building a massive quantum data center in Illinois. So it's a it's a very, very. And just to give you an idea, one large quantum computer is not the size of a traditional computer or a traditional server. It's maybe the size of a shopping mall with a lot of stores. I'm saying it's big. It's big. It's it's blocks and blocks, uh, in the area. So it's not something, you know, that we think it will make fit in a traditional building. You need to build a whole facility just for that. And it's it's gigantic. I'm talking about hundreds of thousands of thousands of square feet. Okay. Maybe even more than that. So this was also really, really something that I thought would be much further away than it is now. So we may see something on that very soon. And last but not least, and this is the topic I'm talking about a lot, is the concept of ethics and the future. So, how will we organize privacy? For example, we had the CEO of Signal that the communication platform talking about the challenges of privacy and how you will control that. For example, we have the problem of hyperconnectivity. We have the problem of exclusion. We have challenges with AI agents in automated decision-making. For example, if I have an AI agent in my project that will be able to automatically decide on the supplier and hire the supplier, how do I know that the agent will choose properly what to do? So this was a very big dilemma, and we had several talks on that. Last but not least, the social impact of these technologies and how we can use these technologies to promote and reinforce something that is good for society. Okay, because these changes are already happening. I would like to finish this podcast by suggesting that you go to the YouTube channel South by Southwest. S x, S W, and go there. Many talks are there. Not all, but many. I would say there are a lot, including this from Colossal, Scott Galloway, and Amy Webb of MIT. So, all of the IBM CEOs are there. Do not miss the chance because this costs nothing, and it's full of insights. Again, I'm recording this. I will put this in my newsletter at the end of the month, and I'm working very hard on this to make sure that people who were unable to attend can understand what is happening and prepare. And this is at the core of what Antonio and I are doing with the concept of AI. So how we can help people to understand and visualize that this was one of the key reasons for me to go there and try to learn and try to share as much as I can so everybody can prepare and get ready for that. Because the transformation that is coming, it's massive, and it's, you know, it's very complex. It's everything happening at the same time. In terms of geopolitics, I would say that the challenges in terms of global governance and technology are advancing at lightning speed. So let's think about that. Let's be positive and get ready for that. I hope you enjoyed this podcast. I apologize again for my voice. Now it's almost 2 a.m. here. And I'm recording this because I want to make sure you all have access before, you know, this topic becomes a little bit cold on my brain. So enjoy that and see you next week with another 5 Minutes Podcast.
