AI valuations depend on new revenue or major efficiency gains

Transcript

From the transcript The Diary Of A CEO publishes with the episode.

1. 4:52

   I've got one here from Elon that says AI and robots will replace all jobs, working will be optional, like growing your own vegetables instead of buying them from a store. The challenge will be fulfillment. How do you derive fulfillment and meaning in life? And Sam Altman said, by the end of 2028, more of the world's intellectual capacity could reside inside data centers than outside them. And when I went through Sam's blog several times, I mean, he's talking about a fairly dystopian future. And many of the CEOs, I mean, even Amidea and Thropic talks a lot about, you know, his bare case for AI and the job disruption that will follow. And so I think I'm trying to get clarity for myself really. I'm like, what is the truth? What's marketing and what's the truth?

2. 5:17

   And I want to add another point, which is I hire a lot of people. And I'm already finding that off frameworks, specifically for entry-level roles, has quite radically changed. And so I didn't believe this stuff. Like a lot of people told me that, oh, Sam Altman, these guys are doing marketing to make their company seem powerful. But then when I noticed a change in my own behavior, when I'm sifting through these CVs, I thought, oh, maybe things are going to change.

3. 5:40

   I think it's mostly bullshit and catastrophizing in a means of fundraising. Every technology in history goes through a similar art. There's some catastrophizing. There's some job loss. That increase in productivity results in additional margin, new business opportunities, and employment growth. I don't see any reason why this would be any different. And I think that catastrophizing and talking about this massive destruction in jobs is a way of saying, we're justifying the massive investments. These companies want enterprises to make in their companies. Because if you look at the amount of capital they have committed in the valuations of these companies, one of two things needs to happen in the next five years, at least in the US. There needs to be a trillion dollars in incremental revenue from new products

4. 6:22

   from companies that have licensed AI. So Lori L. does a big-site license with say OpenAI. Does it come up with new products? Like, how does it justify that investment? We're not saying a lot of AI moisturizer or cars that are built by AI. I mean, there's AI has hit industrialized robots in Amazon, but there's not a lot what you call new AI-driven or AI products. AI plays a role in the background, but it's very hard for companies right now to point to incremental growth or revenues from AI. So in order to justify these valuations, then they have to say, all right, there's going to be massive cost savings or efficiencies. And there's some examples. I think meta meta has just announced new layoffs because of efficiencies they're getting with AI targeting.

5. 7:04

   So one of two things needs to happen in the next three years. Either these companies' valuations need to be cut by 50 or 70 percent or you need a massive destruction in the labor market that creates tons of efficiencies that their customers can then flow to the bottom line. And I think what the CEOs are saying is that there's going to be massive efficiencies. I also think that you sound more interesting and it makes your technology sound more seminal when you say it's changing the world and we don't know how to control it. And quite frankly, I find it a little bit obnoxious that some of the founders in key figures in AI catastrophize just about the time they take a second area and piece out to the Cote de Zour. That's not helpful. I'm Dr. Frank Dessignan. I've created this monster, but I don't know how to deal with it. So I'm going to go piece out to San Trope.

6. 7:48

   That's just not very helpful. They always talk about the peril. So what do you mean by that? So the first big fear of catastrophizing is this unprecedented job destruction. They've been talking about that for three years, but let's look at the data. The unemployment rate in the US is 4.5 percent. Among youth, it's 8.8 percent. That is slightly below the historical average. The number of new business permits, our new businesses started per capita in the United States has doubled in the last 10 years.

7. 8:19

   And we hear about, well, what about meta? They announced they were laying off, I think, 8 or 10,000 people yesterday, 2019 to 2025. They went from 16,000 people to 80. So even if they go back to 60,000, it takes them back to, it takes them back about 24 months. So I don't doubt that there'll be a real dip or a severe V down in certain industries, customer service, probably the legal field. But I believe over the medium in the long term, it's actually going to create more jobs in the destroys. And I don't, I think some of the catastrophizing is nothing more than thinly veiled attempt to say, my technology is so devastating that it's going to shift society and you should invest at this crazy valuation.

8. 9:05

   Could you be wrong? Oh, I'm not out of the time, Stephen. But I mean, what would have to be the case for you to find out that you were wrong in terms of how you've reasoned up from the first principles of your thinking there? Job destruction, layoffs. So what would you have misunderstood for that to be the case? Like what, what would you miss? Would you have, uh, misestimated for that to be the case? Ground zero for the job that was going to go away was radiologists. You could scan billions of images and radiologists. Kid mama don't let your, your daughter grow up to be a radiologist because that job's going away. The new job listings for radiologists in 2026 is up because what it ends up is that while scanning the image is important, it's a small part of the job. The value out of a radiologist is diagnosed in the illness and then coming up with a treatment

9. 9:52

   plan. And that's as important as ever. The number of coders job listings for coders year on year is up 11%. So people who understand how to use these technologies and come up with different prompts or different means of vibe coding, that demand has gone up because now AI can be applied to almost any startup where I will be wrong is if there is sustained job destruction and the new jobs created in new businesses and the employment those jobs create doesn't keep up with new jobs and there is a scenario. You don't need 100% unemployment like Musk is predicting at 20% unemployment, the French had a revolution in Weimar, Germany turned very ugly at 20% unemployment, especially in my youth, especially young men tend to get very angry and take to the streets. So you could see a V that's so severe, even if there's a job recovery, if it hits 20% unemployment, that could cause

10. 10:43

    real civil unrest. But I just look at the data, I look at the employment numbers. If you didn't know there was this seminal technology, the very smart people were predicting a job apocalypse around and you just looked at the data, you wouldn't know there's anything going on. The data doesn't reflect that there's some big exogenous meter media coming for the employment market. Is there a case that this technology because it's built on the internet so it has