Are we too reliant on technology? Questions arise after Microsoft's widespread outages

Meanwhile, this week's Microsoft outage highlighted the risks of relying on a single technology provider for essential services. For those outside of the U.S., it also reignites debate over dependence on American businesses. The disruption lasted nearly 24 hours, impacting millions of users across more than 70 countries, from government agencies to airlines to hospitals and small businesses. Well, for some analysis, we can speak to Marcel Saussure, a visiting professor at the University of San Diego and the founder of Nestor AI, an AI-powered learning tool. And he joins us via Microsoft Teams, which thankfully is. Good morning and thank you so much for taking the time to speak to us on a weekend. Thank you so much for your invitation and I'm happy to speak on Teams. We can prove that it's working now. Yeah, they finally fixed the problem. It took them a while though. Can you start perhaps by just giving us your reaction to these disruptions at Microsoft? Were you perhaps personally impacted in any way? We perhaps spoke too soon about teams working again because it seems we are having some sort of connections issues. We'll hopefully get back to Marcel Saussure later on if we can. Well, after failing to catch its Starship rocket booster on Friday, SpaceX recaptured the rocket's spacecraft for the first time on Saturday. That's an even bigger milestone in the company's decades-long pursuit of fully reusable rocket systems, which could drastically reduce the cost of space missions. Daniel Quinlan has this. It was a world first. The largest spaceship that's ever flown returned safely to Earth. After a day in orbit, the chopstick arms of Launch Pad 2, also known as Mechzilla, safely closed around the spacecraft Saturday in Texas. After decades of dreaming of reusable rocket ships, it was a major milestone for the space industry. That was the holy grail of spaceflight, plummeting through fire to be snatched out of the air by steel. We caught a starship and now everything changes. Despite failing to do so yesterday due to an unknown error, SpaceX has captured the rocket's boosters three times, but they fall from a much lower altitude at much slower speeds, making the catch of Starship significantly more difficult and complex. The spaceship re-enters the Earth flat on a route carefully mapped out over sparsely populated desert, then maneuvers with aerodynamic flaps a little bit like a skydiver, losing orbital speed before the Starship uses its rockets to flip itself upright. And at least some in the crowd gathered to watch share the belief of SpaceX that this is a major turning point. Amazing, amazing, it's incredible. I really do believe this is the dawn of the space age and everything changes after today. That catch, it was incredible. I am ecstatic to have been here today to see this happen. Since its IPO in June, SpaceX has invested heavily in launch capacity and rocket production, claiming on Friday it can now produce a spaceship every two weeks and a booster every month. I'm going to go back now to speaking about this week's Microsoft outage and reconnect to Marcel Saussure, a visiting professor at the University of San Diego and the founder of Nestor AI, which is an AI-powered learning tool. It seems we were a little too quick in saying that Microsoft Teams was back to working because we had a little bit of a bug there, but we're speaking to you now again. Can you start by giving us your reaction to the disruptions at Microsoft that we saw this week? The first reaction is that we can see the interdependence that we have from the different actors of the tech. It's starting now with Microsoft Teams, but it can be one day with Google, so Google Maps, Waze, Meta, Facebook, Instagram, WhatsApp, et cetera, et cetera. Can you talk more about that interconnectedness and why it's such a problem? I mean, I saw a lot of people joking on Friday when Microsoft went down that it was sort of like a snow day for workers who couldn't access their email, but it could actually cause serious disruption. Yes, of course it can, because the problem is, the problem goes back a long time, because basically the main actors of AI and softwares and computer industry have started a race where they need to collect a lot of data. So they had to use techniques for making us dependent from their tools. So the technique can be the autoplay when you're... Guys watching Netflix, for example, or Amazon Prime. But there is also the open source technology when you are using a free tool. Once again, like Google Maps or WeChat or WhatsApp, et cetera, et cetera. And more and more, we are going to be interconnected, specifically when we're going to use our phones for paying. Like in China, you certainly know that in China. If you take your American Express, it's really difficult to pay. You need to have an account with WeChat and using WePay or Alipay. Because you're saying that services are interconnected, which means that if one kind of goes down, it could stop all of the others from working as well. In the future, it will be more problematic. There will be more and more problems with these specific interconnections. That's right. And for the companies, the governments, but also the overall job. And is that really what happened? Because this was an issue that, from what I understand, really just impacted Microsoft systems that are so widely used that it impacted a lot of people. Was interconnectedness the problem here? There was a small problem, so I did not hear the question. I just heard, is this what happened? In the question, is this what happened? Anyway, Microsoft would answer better than me. This is their problem. No problem. I was just wondering what the issue was. I'm pretty sure that you will not have the real answers for the next 48 to 76 hours. Trust me on that. From what we do know, could this really be considered an error on Microsoft's part, or is this just sort of the kind of thing that happens? Amelie, it's a mistake in the AI industry is always human, okay? It's nothing related with machine. You know that you have two kinds of AI, the weak AI and the strong AI. The weak AI is basically the chess game that you are using to play with. So it's extremely efficient, but super stupid. The strong AI does not exist yet. So this is the AI with singularity, Terminator, Matrix, et cetera, et cetera. So with the weak AI, the only problems that we can have are only human. If there is a glitch, it means that the coder who did write the code was human. If it's an AI who coded the code, it means that the mistake came from the original coder who coded it. If it's not a glitch from a human coder, it can be a cyber attack, for example, or somebody from Microsoft who's not happy with something. And from what I understand, this was an issue with Microsoft's servers. Does that mean that one of the solutions perhaps for avoiding this in the future would be for companies to host Microsoft services on their own private servers? Yes, of course. This is the basic. And this is the key of the future of AI, you know, private servers, giant data centers, etc., etc. You know, certainly that in the United States, the real estate for offices is slowly collapsing and the real estate for data centers with private servers is booming. And is having a private server sort of really an option for a lot of smaller companies? I mean, well, what does that entail for them? Yes, it's an option, but it depends on your capacity of cash flow, because, you know, it costs a lot of money for having a private sector and a closed environment for protecting your data. But with the rising of cybersecurity problems, which is the next big problem of the securities of companies, it's going to be soon a real issue. Even if you know 80% of all the cybersecurity problems of the world. What advice do you have for government agencies, organizations and small businesses who want to protect their technology infrastructure and sort of aren't sure which companies can be relied on and what a good strategy is? Which government? Because I have advised my government for the last 10 years to do the right choices and the good decisions. And if you invite me to another interview, I will explain all of the things that I have seen for the last 10 years for explaining them that we... We were supposed to take, you know, the race of AI, the race of robots, the race of cyber security, et cetera, et cetera. So which government? In this case, I suppose the European government, you know, Europe, European countries, France, for example. I'm so sad that you're talking about the European Union government and France. Thanks, James. I would tell them, listen to me now, because I have talked to you for the last 10 years. You know, I have been audited by the French National Assembly and I went to many, to many groups of reflections, reports, et cetera, et cetera, all the time. I warned them, you know, for the last 10 years, more than 10 years about the boom of AI. And I think that we are in Europe totally out of this race. Let's forget the battle of AI. It's totally behind us. You know, it's a race, AI. The first who starts has a big advance. But the problem is that he has also all the energy. The energy is the data. So how can you compete with a Google who has accumulated for the last 20, 25 years, billions of billions of data? As a reminder, per day, we are collecting 2.3 quintillion of data. Quintillion. It's 18 zeros. After the coma. So anyway, the battle of AI is over. If really now we want to do something, we need to win the next battle, which is the battle of robots. When we are going to link the AI with the robots, we need to... We have the industry here in our countries, the manufacturers, the factories for creating, for making robots, which are going to collect the AI. Europe has a long way to go indeed when it comes to tech sovereignty of all kinds. Thank you for giving us your insight this morning. That's Marcel Saussure, a visiting professor at the University of San Diego and the founder of Nestor AI, which is an AI-powered learning tool.