← Deep Questions with Cal Newport
Deep Questions with Cal Newport
Do Managers Actually Understand AI? (I’m Not So Sure.) | AI Reality Check
00:19:56
Loading summary
Transcript1 lines
- [00:00]
A
Do you ever get that sense when you're hearing business executives talk about AI that they seem to be living in a completely different world than the rest of us? I mean, I get that feeling a lot. I want to give you a recent example. Last week I came across an article from the New York Times business sector that interviewed multiple executives from the German software company SAP, as well as several European business and economics analyst. Now, what struck me was how confident everyone was that AI had already started massively disrupting the economy and would continue to do so at an accelerating rate in the near future. Here, let me load the article on the screen here. Look, the title was, can you embrace AI without layoffs? This company says it's trying. And look at the subhead here. The German software giant SAP says it's betting that employees can reinvent jobs instead of eliminating them. Experts are divided on whether it will work. All right, so let's think about that wording for a moment. That article, this interview with the SAP executives takes for granted, right? The people they talk to take for granted that AI, if left unchecked, is poised to eliminate large number of jobs. The only question is whether or not there is something we can come up with to stop that by having maybe new jobs to replace all those that we are for sure about to to lose. Now, these are incredibly bold claims, and yet they're made so confidently and so casually when you hear business executives talk. And to me, this raises some key questions, such as, why is the business community so confident that AI will definitely eliminate lots of existing jobs? What is it that executives and SAP are seeing about their own employees and their clients that makes them think that these massive changes are imminent? And most importantly, are they right? Well, it's Thursday, which means it's time for an AI Reality check episode of this podcast, which is the perfect opportunity to go seeking some measured answers. All right, so here's my plan. I'm going to dive deeper into this article and try to better understand what the business community over there in Europe actually believes about AI and then I'll step back and give you my take. I want to give a trigger warning here. If you're an executive at one of these companies or a business analyst, you might not like the conclusions I end up reaching. All right, anyways, let's dive in. As always, I'm Cal Newport, and this is Deep Questions, the show for people seeking depth in a distracted world. All right, so I want to return to the article here and read another quote from it that I think sets the stage for how executives at SAP are thinking about AI and its impact. So I'll put this on the screen here and read it. Its leaders concede that AI is eliminating the need for humans to do many of the tasks its software engineers did until very recently. In a round of restructuring two years ago, SAP cut nearly 10,000 jobs, some as a result of AI. The company said, all right, I want to stop right there. That's an incredible claim. So the executive cited in this article said that the job cuts they did two years ago were in part in response to AI. So in the spring of 2024, they claimed they were laying people off because of the power of AI. Well, look, I've been reporting and writing about AI since the beginning of ChatGPT. Let me bring you back to the spring of 2024. The state of the art generative AI in the spring of 20 was ChatGPT, powered by GPT4O, which its main claim to fame is that there was more input types it could accept. In GPT4 you could have like for example, audio as input. So it was like a multimodal thing. There were no real programming harnesses back then. The very first serious programming harness, Devin, had just been released that spring in a highly restricted experimental form. So there's no one using AI in any systematic way to program. We were still mainly talking about the potential for this technology in the future. We were still dealing with relatively limited chatbots. So this sort of this claim. Yeah, those layoffs we did were also part of AI. I mean, it feels like that's just essentially made up then. We have other sources we can turn to here. As I covered in an AI Reality Check episode of this show back in May. In the last few months, many technology leaders have come out to say really clearly that that narrative that people were laying people off because of AI just hasn't been accurate. Let me remind you of some of these quotes. Nvidia CEO Jensen Huang said claims that AI would eliminate large number of jobs were, quote, ridiculous. He also said the following, and I'm quoting him here. It was just a way for executives to sound smart and I really hate that. I think they're scaring people and that's irresponsible. Venture capitalist Marc Andreessen also has this spring ridiculed the idea that AI could be responsible for software development job losses already, given the fact that mature AI coding tools were only just became available this fall. He notes that cuts in development jobs, like the cuts that SAP made two years ago or more, a combination of factors such as Correcting for over hiring of tech companies during COVID and increase in interest rates, which requires them to actually reduce expenditures to keep profit numbers where they need to be. The most recent data backs this up. There's a new report out that reveals that tech job openings rebounded sharply in 2026. The number of software development job openings at the moment is at its highest level in three years. Even Sam Altman in a conference last May said his fears of mass job losses had turned out to be wrong. Let me read his quote here. I'm delighted to be wrong about this. I thought there have been more impact on entry level white collar jobs being eliminated by now than has actually happened. All right, so let's step back. I think. I think this is really critical. Basically every major figure close to the current AI boon has said this idea that layoffs were happening in the past couple of years due to AI is just wrong. They also are largely eliminating the idea that we are going to face massive job losses due to AI in the near future. And yet the executives from SAP quoted in this article casually just threw out oh yeah, in the restructuring we did a couple years ago, some of that was AI. We had to replace people because of AI. So they were presenting that claim as if that was obviously true, even though a closer look at the reality shows that nothing like that is true. All right, let's read further in the article to get some more insight into the way these business executives think about. I'm going to bring another quote up here on the screen I'm reading now from the article. I'm not sure if here someone in two or three years will still code software. Christian Klein, the chief executive, told me in a recent interview. He said he did not expect to operate with a smaller workforce, but with a very, very different workforce. Such a strategy could be a source of hope for Europe. Economists say it might prevent large scale job losses in some sectors while plugging unfilled vacancies in others. Europe and aging continent is facing a long fate of slow economic growth, has a chronic lack of skilled workers. I want to break down that passage from the article because there's some key leaps of logic in here that I really want to emphasize. It starts by the chief executive of SAP noting that AI tools are now much more involved in the actual production of computer code. This is true. Programmers now, as of 2026 this last quarter, now nearly ubiquitously work closely with AI agents. They, in interactive conversations with the agents, specify exactly what piece of code they want. The agent produces and tests the code the humans look at it and then go back and forth if it's not quite what they want. So the actual typing of computer code has been greatly reduced. There's more and more of the actual typing of code that's coming out of a model that's working interactively with computer programmers. Okay. But as we just noted with all those quotes we read before, this has not led to software jobs disappearing. There's now a record high number of job listings in software development. We're still trying to understand exactly the best way to work with these agents, but it is certainly not replacing programmers. Maybe it will, but it's not right now and it's just as likely that it won't. But that reality doesn't show up in this quote because if we go further, the chief executive of SAP goes on to imply that if AI's coding capabilities are not addressed, those software development jobs will disappear. Right. Because remember, he said, and I'll bring this back up here, he did not expect to operate with a smaller workforce, but with a very, very different workforce. Right. The implication there is that they would be operating with a much smaller workforce because these jobs are going to go away. But he will avoid that fate for his programmers by finding them different types of work to do what he's calling a very, very different workforce. They get into it later in the that it's going to be more about being a forward deployed engineer and orchestrating AI or something like this. So his implication is if we don't invent new jobs, we're going to lose all these software development jobs. Then comes the even bigger leap of logic. The quote goes on to say that actually there will be, I'll put this on the screen, large scale job losses in other sectors of Europe due to AI. That's the default assumption. That's what's going to happen if we don't do anything else. But if the rest of the economy that there's some silver linings here is that actually if we plug in the SAP strategy of inventing new AI aware work for these workers, all those jobs won't disappear. And then it goes on to say, and also the fact that a lot of these jobs of course can be automated is maybe not that bad of a thing because we have a labor crunch coming. As we age out workers, there's not going to be enough skilled workers to fill the current job. So maybe it's not a bad thing that a lot of those are going to be automated. So at the core of this sort of logic here from the chief Executive is this idea that software jobs, if nothing else changes, will go away, that at SAP they will find other work for those programmers, new types of jobs that involve coordinating AI. And in other sectors unrelated to computer programming, they're also going to lose a lot of jobs. But some of those they'll be able to redirect in other places. And in other cases, automating jobs will be okay because we don't have enough workers anyway. So again, there is just this casual, confident assumption that AI is in the process of eliminating lots of jobs or soon will eliminate lots of jobs, not just in computer programming, but in multiple sectors across the European economy. All right, well, you might be wondering maybe like what's really going on here is that internally at SAP they have found really innovative ways to use AI and that's why they're so sure that all these other sectors are about to be eliminated. Well, fortunately we can get some insight because in this article the SAP executives go on to summarize how they are currently using AI within their company. We're about to see here some, like truly, I'm assuming, very impressive, super powerful automation. So let's read what they said here. SAP workers use AI tools to try to improve applications for patents and to handle some customer support request. It's AI agents code prototypes for new software or add new features to existing programs, speeding up the process for developing products. Okay, so within SAP, the way they're using AI is to generate or clean up some writing on applications, to do some in quotation marks, customer support tickets, the vibe code prototypes, and the developers are using agents to help them when they're adding new features. I mean, I guess us humans are obsolete now. I mean, I'm being sarcastic, of course, because these are the same narrow use cases we've heard for a long time. The only interesting one is programmers are now ubiquitously using these tools. We don't know what that means, but they're ubiquitously using those tools. But this idea of it can help you write drafts of things or clean up text, or maybe it can be like a chatbot for some customer service. We've been saying that for years. It hasn't done anything to the economy. And so if that is what is happening within SAP with AI, how do they get from that to multiple sectors of the economy are going to lose huge numbers of jobs because of this technology. But that's okay because we have a labor shortage anyways and we don't have enough people to fill those jobs anyways. How do you jump from these? There's nothing new here. This is the same narrow use cases we've been hearing about for a long time. And yet they've jumped from that to this confidence of like, yeah, software development jobs are going away, but we can find new jobs for them and other sectors will lose a lot of jobs too. But you know what? It might not be so bad because we don't have enough people to fill those jobs. And younger people, we can invent new jobs that are like orchestrating AI, so maybe it'll be okay. And the headline of the whole piece is like, yeah, but the experts they interviewed are like, maybe, but maybe that won't work at all. Like, it's a 50, 50 chance about whether we're basically about to go through a massive economic depression or maybe we can kind of muddle our way through it. There is such confidence here that economic devastation is coming. And even though we just looked at the last three months of the technology press as all these leaders saying, that's overblown, that's executives trying to sound smart, we don't really expect there to be massive job losses. There is a huge disconnect here between the casual confidence of the business community, as exemplified by the quotes in this article, and reality. Well, what explains this? What explains what's going on here? We don't have HAL 9000 right now. We're not in an Asimov iRobot short story. We have a cool new programming utility that we're still trying to figure out and some truly mediocre, narrow business applications. That's the reality. How do we get from there to this casual confidence that the economy is going to be destroyed if not for some very creative reinvention and pivoting by business leaders? Well, here's my emerging theory. Many business executives and analysts don't really understand AI much at all. They are operating on vague vibes or reacting to things that they've sort of heard endlessly repeated on LinkedIn that seem directionally true. I also think that a lot of people in these situations are more afraid of missing a disruption than they are about being wrong about a disruption that doesn't happen. There's just more credibility and feeling like you're very AI forward to say, oh yeah, we laid up people. Wasn't interest rates, it was AI. It makes you sound like you're AI forward to be like, I bravely know, as some C suite executive at a massive software company, I bravely recognize that the economy is about the crater, but we're going to pivot and be smart about it. Makes you seem AI forward. It makes you seem high tech. But I keep hearing this again and again from individuals who are talking to their managers like, I don't think they understand what AI is at all. They're repeating like slop from LinkedIn. That makes no sense. You better be using AI or someone who uses AI will replace you and like to do what though? What part of my job is AI going to make better? What do you want me to. And they're just like, AI, good job, go away. Me see person on LinkedIn say that all software dev go away. Me not want that. I guess the managers talk like cavemen. Is that how that works? But I think this is putting us in a difficult situation, right? Think about it as just the American public or the world public here. On the one hand, we have the people who arguably know most about the technology, which are the AI executives that have this weird addiction to doom trolling. So we have a hard time listening to them and they're kind of going back and forth. They seem to be a little bit better recently, maybe because of pending IPOs. They're a little better recently about like, yeah, we're okay, it's not going to take jobs. But then every once in a while, like, but the AI is going to take over, kill us all, and then they go back. So they have this weird addiction to doom trolling. So it's hard to sort of pay too much attention to the AI executives. But now we have the people who are most impacted by the technology, the people who run companies who are dealing almost exclusively in vibes and sort of like things that they heard on LinkedIn that sound directionally true. They seem completely disconnected from the reality of what's really happening with AI. So how are we supposed to understand what's really going on? We can't trust the people who know the most about it. We can't trust the people most impacted by it. Of course we're all confused about it. I think one of the solutions here is journalism. I think journalists, here's my idea. I think journalists, they need to change their frame when they think about AI and they cannot approach it like you would approach normal technology reporting or business reporting. I think we have to bring a frame of skepticism and criticality to AI reporting that's similar to what we do for political reporting. It's very important what happens in this sector. A lot of impacts are going to happen. But we don't just trust or take for granted or accept whatever the people we interview happen to say, because they all have their own agendas or biases or blind spots. So and so says this. Someone else says this. This claim was made by this CEO. These other people completely dispute that claim. We talked to east coast computer scientists who don't wear robes and go to X risk cult sacrifice meetings about what this technology actually means. And here's what they said. We have to bring the same skepticism you bring to like a presidential or the congressional midter terms that are happening now. It's incredibly important. The impacts are big and we don't trust a word out of almost anyone's mouth. And that'll change the way I think the public understands AI as well. We'll be less susceptible to this doom thing we saw over here, this confident prediction here, the skeptic that dismisses everything over there. If we put on our political reporting blinders or glasses or whatever we want to call it, I think we'll be able to navigate this important moment. Now obviously I want to give a caveat here. Many, many business leaders and managers know a lot about AI and they're very savvy about it and they're very clear headed about it. I've talked to a lot of these managers, but there is a class of people right now in the business community who are very vocal who really have no idea what they're talking about and they're causing a lot of confusion for people who read their quotes and they're causing a lot of pain for the people who work under them who are like, I don't know what you're saying. This is like a Dilbert column cartoon from hell. All right, so that's my plea. Be wary. Just as wary about what business leaders are confidently and casually saying about AI as you are about the doom that's coming out of the mouth of AI tech CEOs. All right, that's it for this week. I probably will. I may or may not have another AI Reality Check episode next week. I'm traveling again summer. I travel a lot so we'll see if I can get one in, but otherwise I'll be back to week after. And on Monday though, there will be an advice episode of the show, so definitely tune in for that as well. And until next time, remember, care about AI but not necessarily everything you read about it. Hey, if you've made it this far, you must be ready to join my fight for depth in a distracted world. Now the best way to do this is to join over 125,000 people who receive my email newsletter each Monday. You can sign up@calnewport.com ideas and when you do, I will send you a free guide to my seven best ideas about cultivating a deep life. Sign up today. Calnewper.com ideas.