Mass AI-Driven Unemployment, NVIDIA Surges, CEOs' AI Avatars Unleashed

0:00Is AI going to drive mass unemployment or are urgent warnings about job loss, just marketing? NVIDIA, meanwhile, turns in stellar earnings. Mark Zuckerberg and Palmer Luckey team up to build war goggles. And CEOs are using AI avatars for their earnings calls. We'll cover it all on a Big Technology Podcast Friday edition right after this.

0:26You're used to hearing my voice on the world bringing you interviews from around the globe. And you hear me reporting environment and climate news. I'm Carolyn Buehler. And I'm Marco Werman. We're now with you hosting The World Together. More global journalism with a fresh new sound. Listen to The World on your local public radio station and wherever you find your podcasts.

0:53Welcome to Big Technology Podcast Friday edition, where we break down the news in our traditional cool-headed and nuanced format. we have a big week of news to speak with you about today starting with the prophecies of anthropic ceo dario amode that we're about to see mass unemployment due to ai we'll also touch on nvidia earnings we'll talk about this partnership between meta and andruil to build these war vr goggles which is quite interesting and might signal a shift in silicon valley and of course we're going to talk about ceos dispatching their ai avatars to their earnings calls and maybe they will come to a meeting near you.

1:29Joining us as always on Fridays to break down the news with us is Ranjan Roy of Margins. Ranjan, good to see you. Welcome. Well, at least when I lose my job to an AI avatar, I will have war goggles. So this week makes me feel better. It is interesting. Every time we talk about AI technology, there's this combination of will it work, won't it work, and this is going to be great, and this is going to be terrible. And I think we're just kind of hitting all of those emotions in one this week. And so let's start with this prophecy from Dario Amode, the CEO of Anthropic, about what is going to happen to white collar jobs.

2:07He tells Axios that there's going to be a white collar bloodbath. He says AI could wipe out half of all entry-level white-collar jobs and spike unemployment to 10 to 20 percent in the next one to five years. He said AI companies and governments need to stop sugarcoating what's coming. The possible mass elimination of jobs across technology, finance, law, consulting, and other white-collar professions, especially entry-level gigs. So let me just turn this to you right away, Ranjan. I mean, 50 % of all entry-level jobs within five years. I don't think that this is entirely out of the realm of possibility, but wouldn't you agree that this would have to just be like, it's weird to say best case, but the best case scenario for this technology, in order to come anywhere close to that, it would have to meet and exceed all of our expectations to actually be able to do something like this.

3:09Yeah, this still feels a bit like marketing hype for me. And I think to me, if we break it down to a couple of different pieces, yes, I think the jobs of today are fundamentally going to change. I'm a big believer that especially white collar work, which was not threatened in the same way as more blue collar or manufacturing type work in the last 30 years is going to be threatened. I actually, my theory on this is entry level jobs are actually going to be safer because they're going to become different. and the type of workforce that's going to be entering the market is going to be a lot more AI native than a lot of existing, especially more senior people.

3:50So if I'm junior, if I'm entering the workforce, I would much rather be entering now than 10 to 15 years ago. Okay, I hear some music in the background. Is this the, it can't be. It's the Hype or True jingle.

4:11You don't want it all. Don't just guess we'll break it through. It's your favorite game, hype or true. Listeners cannot see. My jaw is on the floor. Alex just took the worst name game of all time. And I guess we're doing this. I guess we're making a thing out of it. I ran it through Sudo. Thanks to AI. And we are in the thick of it. Hype or true. Dario Amode's prediction that 50 % of entry-level jobs will be wiped out. So let me take the other side that this isn't marketing just for the sake of conversation, right? I mean, we are seeing this technology improve dramatically. And we're also already seeing some evidence that it's harder for people coming out of grad school to get jobs today.

5:00and AI is definitely in the workplace, especially in areas like coding, for instance. And we're going to talk about it in a moment, but Amazon, for instance, is a company that has smaller teams responsible for the same amount of code, just with like less people to do it. So this idea that if the technology keeps improving the way that it is, and we go into this world where we do get agentic technology and we do get AI that can go out and do the deep research for us with fewer hallucinations, maybe it's not so crazy that Dario, who sees the trajectory, is going out there and telling us that, hey, listen, this might be an issue.

5:41No, but again, I still look at this as it's self-interest in the idea that Anthropic is a company that directly benefits if that is the truth. But to me, I mean, we just mentioned coders or software engineers or developers, that job at the scale that it is today didn't exist nearly in the same way 20 years ago. So to me, it's just going to be a shift. And maybe it's a little bit overly optimistic. I do think his call that governments or the civil society needs to recognize that things are going to change is important. And I do think there needs to be genuine conversations around this. But again, jobs change, people adapt.

6:25And I don't know, maybe I'm too hyped up from that song that you just played. But if we're talking about innovation, in what other world could we have a hyper true jingle that quickly? That's a net positive for society. That's all I'm saying. I think it's already, just because we have that jingle, it's already outweighed all the negatives of all time that we might see with AI. But Dario, in his conversation with Axios, actually goes through step-by-step how he sees this happening. And yes, innovation always creates job loss and job creation. But the idea is that this technology is potentially so powerful that it just puts this process on steroids, or as you said last week, on acid.

7:10Okay, so here's the step-by-step. Number one, OpenAI, Google, Anthropik, and other large AI companies keep vastly improving the capabilities of their large language models to meet and beat human performance with more and more tasks. This is happening and accelerating. The U.S. government, worried about losing ground to China or spooking workers with preemptive warning, says little. The administration and Congress neither regulate AI nor caution the American public. This is happening and showing no signs of changing. Most Americans, unaware of the growing power of AI and its threat to their jobs, pay little attention.

7:44This is happening too. So do you have an argument that any of those three steps are incorrect? I would actually still, again, argue with number one, but that's more of just improvements in large language models. I don't think it's going to actually see adoption in implementation. I don't think that's the issue or the problem. I think there's a lot of other points of resistance. So I would disagree with that one. I like number two, the idea that to avoid spooking the population. I mean, it's a good little conspiracy theory that seems like a good conspiracy theory. Yeah. And I mean, again, it seems completely plausible in the political climate.

8:23But again, in number three, I think unaware of the growing power of AI, this is where I would actually vehemently disagree that AI, the promises made by the industry are at such a scale that it's my again, my mom, whenever I go home, she's like, so what do I do with AI? And she's retired. And like, I mean, it's to me, it's not that there's a lack of awareness. It's just there is a lack of actually adoption and implementation. I do think it's not because of a lack of awareness. I think it's because of a lack of how focused these tools are. We've talked last week about how Anthropic and Claude are going to potentially move much more towards coding and software development, which shows that it's not working in other areas.

9:09So overall, I don't know. As regular listeners know, I can be a bit cynical about a lot of topics, but I genuinely don't think it's as much of a threat as he's making it out to be. Yeah, this is the part where I started to veer more into the hype side, where he says, then almost overnight, business leaders see the savings of replacing humans with AI, and they do this en masse. They stop opening up new jobs, stop backfilling existing ones, and then replace human workers with agents or related automated alternatives. I think this just completely underestimates how change management works in most companies and how slow most companies are at implementing even the most helpful software.

9:52And this idea that this could happen almost overnight to me is far fetched. Now, is this something that you could see happen over 20, 30 years? If the technology keeps accelerating at the pace it's been accelerating, which is an open question, maybe. But this idea that we're going to see it within one to five years, that just doesn't really seem to meet the reality test for me. Yeah, no, I completely agree because Dario came from, he had his PhD at Princeton, I think. One of his two PhDs, yes. Yeah, I mean, two PhDs, in terms of not needing to learn about change management at organizations, I think that's like a pretty clear idea.

10:32And then the other thing that - He's the CEO of a multi-billion dollar company right now, so I won't take that away from him. Okay, fair. I won't. But the part that he's missing in that whole thesis is growth. That it assumes kind of like a steady state of economic output, and then underlying that there's fewer jobs. But if we take the other side, and again, the more optimistic side that this actually unleashes new waves of growth and new types of companies, I think then new jobs are created. There's different types of jobs are created. I never thought I would be this optimistic, but on this one, I still believe, still believe.

11:13It is interesting to me that he is taking that pessimistic perspective given where he sits. He doesn't believe that if you are an entry-level developer rather than a company not hiring you, his tooling could help make you a 10X developer and then help the company grow much further than if you weren't there. So I think that is pretty telling, telling in terms of it falling on the hype perspective. And there's this kind of great slash funny CNN article by this woman, Alison Morrow. She says, if a CEO of a soda company declared that soda making technology is getting so good, it's going to ruin the global economy, you'd be forgiven for thinking that person is either lying or fully detached from reality.

11:56Yet when tech CEOs do the same thing, people tend to perk up. She says the point is that Amode is a salesman and it's in his interest to make his product appear inevitable and so and so powerful. It's scary. He stands to profit off the very technology claims will gut the labor market. But here he is telling everyone the truth and sounding the alarm. He's trying to warn us he's one of the good ones. Yeah, actually, that's a really good point. I think if you have this technology and you're genuinely worried about it and you are the CEO of a multi-billion dollar company, either you are the most altruistic person alive or you're doing it for self-interested reason.

12:36Again, soda CEOs did not tell the world that our soda will become so addictive it will unleash a wave of obesity. It just happened and people made money. Like, I think the idea that I'm going to, I don't know, is he trying to risk the entire business of Anthropic by saying this? I think it's good. We've seen this for three years now. AGI, ASI, AI, like killing all of humanity is a threat. Now white collar bloodbath. I think that it works from a marketing perspective. So we'll keep seeing this. Okay, let me one more time try to make the case for Amo Day and then we can move on. I think if you're being super charitable to his perspective, you would say he's interested in building AI that's going to have a big impact on the world.

13:26Things like he talked about in this interview with CNN, curing cancer and helping us with medical applications, creating abundance. But he also sees that there's going to be a downside. And maybe he truly believes, and I don't know, is there a 100 % chance that he's wrong in this, that he believes that his technology will have an impact? And if you do believe that, then maybe it is in your interest to go out and say, yeah, there could be problems. And in terms of being the one that's developing it, well, I think maybe there is something to be said about being the person who wants to develop the technology and steer it in a way that's impactful versus destructive.

14:01All right. I will actually, I'll end on a charitable note as well. I think on the topic of regulation, Amadei came out when there's this proposed 10-year moratorium on any kind of AI regulation. And he said, that's not a good idea. So in fairness, he is advocating for at least some kind of steering the industry in the right direction. And I think we should talk about what happens in the near term, which is not necessarily going to be job loss. But in some companies, and I think in some companies that aren't really viewing this the right way, it will be a change of job, one that will require harsher requirements on employees versus one that will sort of free them to do higher impact work.

14:52And this is the thing that I hinted at the beginning of our discussion about Amazon. This is from the New York Times. Very interesting story. At Amazon, some coders say their jobs have begun to resemble warehouse work. And the article says the more immediate downside of AI for software engineers appears to be the change in the quality of their work. Some say it's becoming more routine, less thoughtful, and crucially, much faster paced. Three Amazon engineers said that managers had increasingly pushed them to use AI in their work over the past year. The engineers said that the company had raised output goals and had become less forgiving about deadlines.

15:30It had even encouraged coders to gin up new AI productivity tools at an upcoming hackathon. One Amazon engineer said his team was roughly half the size it had been last year, but was expected to produce roughly the same amount of code using AI. So is this kind of what the world looks like where companies say you no longer have these long timelines to build the thing that you want to build? Because effectively, I know that you can use AI to build it. We've seen similar things from Shopify, for instance, you know, the AI first company episode we did a couple of weeks ago kind of hits on this idea.

16:11I have a hard time feeling sympathy. I know Amazon is a rough culture, so like an aggressive culture and many and often in a good way. So I'm not I'm guessing those were not cushy jobs. But yes, work is changing. This is happening. Like even the idea, it's like they were encouraged to come up with AI tools at an upcoming hackathon. That's not bad. That's like people like in your interview with Sergey Brin the other day, he said this is the most if you're a computer scientist, this should be the most exciting time for you. I actually would expand that, that if you're at all in interest in technology, this should be the most exciting time for you.

16:50And yes, for software developers and coders, work will look vastly different. It was a good 20 to 30 year run. It's not going to be the same for the next 20. And that's OK. Yeah, I think that's a very good perspective. OK, so but what about this? Maybe you don't have time to think as deeply about projects as you did. This is again from the story. The new approach to coding at many companies has, in effect, eliminated much of the time the developer spends reflecting on his or her work. This is when developers say, or no, sorry, this is an analyst. It used to be that you had a lot of slack because you were doing a complicated project.

17:26It would take a month, maybe two months, and no one can monitor. Now you have the whole thing monitored, and it can be done quickly. So maybe this idea, can we end up seeing less deep work or less reflection and then bigger thinking because you have to do things so quickly and be so reliant on the AI to program for you? I think that conflates two different things because on one side, almost by definition, this should allow for more deep work. This is taking care of the more monotonous, repetitive things people did instead of having to rewrite an entire code base from scratch. that like that kind of work you're able to do.

18:04The question of monitoring work, that's to me not an AI question. That's a culture question. That's a technology question that's increased dramatically, especially since the pandemic and when more work is being done on digital surfaces. So I think those are two different things. And I think some companies like in Amazon will very closely monitor And that's how it works. And other companies will work differently. But again, to me, AI, almost by definition, should free up more time for reflective deep work. And it is kind of interesting that I guess a company like Amazon will just say, OK, now you're kind of like the delivery driver with the AI systems monitoring your every move and every second optimized to a T.

18:54I mean, yeah, probably the people who built those systems to monitor the delivery drivers. are now getting monitored by the same ones. Are the ones anonymously being quoted. Totally. Sorry, guys. Can't say you're surprised. But meanwhile, we're going to have an opportunity to really figure out what's going to happen because the surge in infrastructure building shows no sign of slowing down. We saw this in NVIDIA's earnings that came out this week. This is from the Wall Street Journal. NVIDIA's business is booming despite being shut out of China. The company's revenue reached$44 billion for the first fiscal quarter, for its first fiscal quarter, a 69 % increase that was curtailed.

19:35So limited to just a 69 % increase that was curtailed by Washington's new limits on China chip sales. The company was unable to ship$2.5 billion of its H20 processors and projected an$8 billion in lost revenue for the current quarter due to that policy. So that's lots of money lost to China and or lost to not selling to China. And the company is still growing. Yet, we still see a tremendous amount of spending coming in from the tech giants, Microsoft, Amazon, Alphabet, Google and Meta. They are all going to be boosting their spending on AI. And analysts expect combined spending by those four companies to alone top$345 billion this year, up 41 % from last year.

20:25And the data center business is going to be up a surge 73 % over last year to 39.1 billion. So two things here, Ranjan. Obviously, NVIDIA is healthy. No doubt about that. Even with the China stuff, it's still growing tremendously. And then second, if scale is going to be one of the ingredients that makes these models better, we're still in the finding out stage like this infrastructure is still just booming javon's paradox welcome it it appears to still be the reality i think that the we've seen more and more about again to reference the sergey brin interview again i really liked he said it's going to be algorithms rather than compute then create the next round of advances so i think like how the actual modeling works and how these things are built could change and it appears to be changing but as of today may 30th 2025 everyone's playing the same game and it's a compute game so that's definitely not changing and deep seek but that this is a 6 to 12 to 24 month thing that we'll see rather than deep seek happened and then uh it's gonna you're gonna see a cut in spending the next day do you think that the export restrictions that I just mentioned are working when it comes to restricting the exports to China.

21:51Now, Jensen, of course, is going to be against them because there's lots of money to be made if he can sell to China. But this is from Bloomberg. He said the Chinese competitors have evolved. They've become quite formidable. So speaking specifically of Huawei, he says the latest AI chip is similar to the performance of NVIDIA's H200 chip, a component that was state of the art until its replacement in recent months. You cannot underestimate the importance of the China market, Wang said. This is home to the world's largest population of AI researchers. And he basically says, like, you know, it's important for the U.S.

22:27to at least be in there and be steering the way that they do it as opposed to restrict them. And then they're going to go ahead and do what they want anyway on almost as good technology. What is your perspective here? Yeah, I think, again, the details of our export restrictions working are difficult for me to very directly say. But I think overall, that idea that NVIDIA to expand, seeing increased competition from Chinese suppliers, it's definitely going to evolve pretty rapidly. And now we have a new DeepSeek model, speaking of which. So this came out this week. There's a new upgrade to DeepSeek's R1 model that the company says reasons better and hallucinates less.

23:18The Chinese startup DeepSeek said Thursday that its upgraded artificial intelligence model can perform mathematics, programming, and general logic better than previous versions. Its overall performance is now approaching that of leading models like OpenAI's O3 and Google's Gemini 2.5 Pro. This is according to DeepSeek in a post on the AI model platform Hugging Face. So DeepSeek's still keeping pace, it seems like. DeepSeek's still deep-seeking. I think it's a reminder, like the competitive dynamics of what's going to happen are still completely in flux in a good way. I think we're going to see more and more competition at the foundation model level.

23:59And I think this is just a reminder that like we just came off a couple of good weeks of Google I.O. and OpenAI announcements and all these kind of things. But the competition is going to be there and only increase. As long as those NVIDIA earnings are so juicy, their margins are. What's the Jeff Bezos thing again? Oh, their margins are our opportunity. Exactly. All of our opportunity. Yeah. Yeah, that's right. So just to sum up this segment, so I think we both agree that NVIDIA still has a lot of room left to run in terms of being able to sell its chips and produce them. And there's no sign of slowing down.

24:38Like if you're thinking just about the hardware side of the AI trade, and of course, NVIDIA has software that you use to train the models. But like thinking about the thing that makes them the most money, that's in good shape. It's in great shape. It's in incredible shape. That has not slowed down. And I know there's a lot of talk after DeepSeek that it would, but it hasn't yet. So there will be more DeepSeeks. And I think I feel very confident in saying that. Okay. And let me ask you this. As DeepSeek continues to rise, is there going to be any value in building these proprietary models? Returning to one of our old questions.

25:16Yeah, I still believe that overall increased competition will see commoditization at the model layer i believe that uh the one model to rule them all one giant model is going to wipe out 20 or 50 of white collar jobs like i i just don't see that happening so yeah i it's i think the model layer will get commoditized what about you yes for sure and maybe that's why we're saying it's not the model it's the product no i think better no this there's nuance here better commoditized models will lead to better products. It's not about, it's so interesting. We have this debate all the time. Actually, I think we're just going to converge here.

26:01You want better models and better products. I guess the question is what drives the improvement in the AI industry? I think, yeah, even if the models are commoditized, then better models are going to matter more. What leads to the white collar bloodbath? Yeah, exactly. Better models and better products. It's coming. Okay. By the way, just one quick note. I mean, Lord Almighty, Meta, there was a story this week how Meta's lost lots of its AI researchers. As DeepSea continues to surge, the conversation around Meta's llama models has certainly fallen off in a big way. And I mean, they've made some changes to their organization this week, minor ones.

26:38But I would watch that space very, very closely. yeah i think uh i mean yeah where metal still lives in this is it's interesting because they actually managed to be maybe i have to support it more product focused it sounds like they separate out the research conversation from the product conversation which is kind of a reminder there is they're a pure consumer company and they're able to separate the two so meta ai meta ray-bans all these things you don't really see or hear as much hype around i don't even know what llama are we on now i can tell you four okay yeah but don't we think that if their models were better then people would be using their chatbots more i mean they have bet a lot on voice and their voice ai is just not as on par with opening eyes pretty good meta ai is pretty good and i say that with a, like, it's, I think it's on par.

27:41You're the first person I've heard say that, so. Really? But maybe I'm not asking enough people. All right. Yeah, you're not asking the Metacraft. See, this is why. Maybe they're going after the everyday consumer while we're talking to the people who actually use and stress test, Anthropic, OpenAI, Gemini, et cetera. It's a great reminder for reporting on this stuff. It's important to look outside of New York and San Francisco. And to the people using meta AI. But I still think that the meta AI is not as good. Yeah. Okay. All right. So we're going to talk a little bit about this very interesting grammarly story.