I just don't think AI's been around long enough for it to have gotten people comfortable enough with it yet for the jobs to go. But that's got to come. It's as clear as the sky to me. I think that's what's going to be the real political hard decisions. When you've got rising unemployment, we're going to be in a recession cutting rates, not hiking rates.
The McKinsey Global Institute is saying 770,000 workers per year will need to switch occupations. And that's a lot, Michael. The historical average is 215,000 as an example. So, we're talking about 11 million total workers or 7% of the entire US labor force, okay, over the next 10 or 12 years. It's a massive amount. Should we be worried? Do you think that's exaggerated? You think that they're not thinking about this correctly?
This is going to be one of the big issues that governments around the world have to face. People ask me all the time, "What should your kids study?" And I keep telling them,
Welcome back to All Things Markets with me, Anthony Scaramucci, and I'm Mike Novogratz. Michael, we are recording this at 10:00 am Eastern Standard Time. It is Thursday, October 1st. We are about to start.
Rev your engines. We're about to start the fourth quarter. GDP didn't look so great in the second quarter. Michael, they revised it from 2.2 to 1.5%. It's a pretty big swing for the first half of 2026.
Which we definitely have had a steady period of expansion. And so we got to give the — despite the war, the energy shock, so on and so forth. Where do you think things are from a macro perspective? And plus you got a good inflation print this week. So where do you think we are macro and where do you think we are Fed related?
You know, it's so interesting. It's the same story that just keeps repeating itself and you go a little bit left and a little bit right, but the economy seems to be reheating a little bit. It is still being driven primarily by this giant infrastructure build and all things that go with it. You know, that's not going away and this K-shaped economy is not going away and it's going to become a big deal in five, six weeks when we have the elections — actually five weeks now.
And you see, you know, you see that in the numbers day in, day out. And so what I believe is this inflation is going to be transitory from the capex side, but the money — you know, the debasement of money side, which you talked about in your book even, is not going to be transitory. And so, you know, it's getting more and more expensive. It continues to be. It's the biggest political issue for most Americans. But I don't think in the future there are going to be as many hikes as there are priced into the two-year or the five-year. And so —
So you're thinking what, two hikes in '27? I mean, people are thinking three, four hikes in 2027.
Yeah. I think too — like we haven't seen job losses due to AI yet and maybe we never will. We've seen them in some areas only, right? The young coders are not getting hired. The AI-forward companies — I was talking to a brilliant guy yesterday, he's gone from 55 to 13 engineers and his company's bigger. So you either are doing it or you're not. And yeah, and I would say very few are doing it to that level yet, but that's just engineers and coders.
Right. Now you've got this consumer app that's come out and — so it's pretty darn good.
There's no question. I mean, we're using it. Lawyers are using it. The people we use for travel are using it. I agree.
I just don't think AI's been around long enough for it to have gotten people comfortable enough with it yet for the jobs to go. When you've got rising unemployment, we're going to be in a recession cutting rates, not hiking rates. But we've got this huge demand for capital, right? Hyperscaler demand, data center demand, infrastructure demand, and government bond demand. There was a lot of steepeners of the two-year versus 30-year. It went all the way to 50 basis points. And I looked at that and I thought, I should buy that thing with both hands. And then I was like, yeah, sometimes things go further than you think. And I think the more I think about this, probably the best macro trade in the next six to 24 months is going to be 2s30s. And Anthony, I remember a part of my career when that thing was at 320 basis points. 2s30s could literally in this cycle, next five years, get back to 350, 400 basis points, right? We have 41 trillion of debt. We did it back then.
Well, I mean, yeah, that's the problem, right, Mike? I mean, look, I mean, they're front-ending the deficit, you know, in terms of the interest rate expenditure. If you had a 5%, you paid 5% on the 40, that's $2 trillion. That's more than the Social Security expenditures for the government, right? So we're in trouble if we don't figure out a way to get the rates back down. But I want to go to the October question. Last week we both said in our bull-bear discussion that Powell was going to raise rates in October. Do you feel that way still or does this new data mean he may pause?
I think he still raises, full raise. Okay, I'm with you on that. I think you got to do a little bit like — this is moving a battleship or an aircraft carrier. It's not moving a speedboat. And so you can't say, "Well, yeah, we thought we had to do 25 or just fine-tuning it." They've started a small cycle. It will be a big cycle, but a small cycle. And you at least got to go twice before you pause.
Yeah. And let me provide some more evidence to our viewers and listeners. You know, the AI infrastructure buildout is now explicitly cited as a driver of GDP growth, which is simultaneously inflationary, right? So there's another data point that the Fed has to work on potentially containing. But let's go to AI for a second. The president is — I guess he's changing the name. What is he calling it? Super intelligence or strategic intelligence or something like that. But —
Super intelligence, changing the name.
There's a fair answer to like debate: should we actually even call it intelligence? Like, you know, is it intelligence or is it a spectacular processing machine, right? Humans have intelligence, nature has, biology has intelligence. Do computers really have intelligence? Are we making a mistake by humanizing computers as opposed to having a different vocabulary about them? This is my wife's big issue. She screams at me all the time, "Stop calling it intelligence. It's not intelligence." But she has a point.
But it talks to us, Michael, like a human, you know, and that could be another thing. Maybe that's another thing.
By calling it — like by talking about it in human terms, I think it's a slippery slope to just assuming it's going to be human. And that's where job displacement and seeding what it means to be a human to these — this new form of thinking, right, of these machines that think. It could be a slippery slope. And I was on with a guy yesterday. He was an early Warner about the dangers of social media. Right, he was ringing the alarm bell five, six years before everyone else was. And did he know a lot about this AI idea? And he wasn't an AI doomer like, stop, stop, stop by any stretch. You know, the Bernie Sanders, just shut it all down. That ain't gonna happen anyway. Like, what's the rational man's middle? The smart man's decision that should be bipartisan of how to think about pacing AI. You know, all these swarms really are in some ways unplanned, unwanted excursions from these bots. You know, how do you ensure that things don't go wrong? Well, there are ways to ensure it, right? And I think that's really where the debate needs to happen. It's not yet, but I think that's going to be a big, big part of the next two months.
When you see a group of senior executives — and these are some of the most talented people in our country, whether it's the heads of OpenAI, Anthropic, Google, Meta, xAI, they're all there. Elon Musk, etc., Jensen from Nvidia — and they're signing a quote-unquote morally binding compact on AI safety. As a student of the world and a student of human nature, is that enough regulation for AI?
No, not close. Not close. Listen, I think we have to be very certain that unlike in the internet where you can't sue people for, you know, because of that section — whatever, Section 230, Section 230, y — right? We got to make sure there's not a Section 230 around these AI companies. But even the ability to sue them — like if something catastrophic happens, something catastrophic happened. And so I think there needs to be a better watchdog. I don't know what that is yet, but think about that table yesterday. We didn't even know the names among a lot of those people four years ago. And so like the trust, you know, the American public's trust in Sam Altman — we've only known Sam Altman three years. You know, I mean, his colleagues have known him a long time. I mean, those people in that 300 — you know, that elite group of AI researchers have known them. But I think it's really a lot to ask, just trust me on this, when the same guy said this could be catastrophic. Like, it's not Anthony Scaramucci yelling this could be catastrophic, it's Dario and Sam. And so I think there's — you know, again, I haven't figured out exactly what the way is, but there has to be a way where there's a little more teeth. And some of this is monitoring, you know, like this idea that Dario wrote about where we monitor these things with what is functionally an fMRI. And it's not that effective, but it gives us what we get. Like investing more in that technology, making sure that their own monitoring keeps pace with the growth of what they're building. You know, again, I'm not a science guy, a computer science guy, but there's got to be something in that line.
You own Anthropic privately?
Well, I'm already exposed via xAI. And listen, we have a big data center build. If Anthropic really works and changes the world, my data centers be worth a whole lot of money. Let me ask you this though, Michael. Would you buy it at two trillion? No. No. Listen, you know, it might be a great buy at two trillion. I think investing is getting harder and harder. Let me tell you a story. I have a small investment in a company — actually, Galaxy does — in a company called Etched, which is making a faster chip by a future Nvidia competitor or some of Nvidia will swallow up. And I'm going out to visit them and I'm sure I'm going to see the most impressive 500 young guys that are, you know, eating through steel and building this amazing company. The acceleration in learning that these models are doing makes in some ways us being able to predict what the political situation is, what the technology situation is, harder and harder. Is AI itself going to solve some of the problems that are making the cost of these things so expensive? You have to think yes. And so I just think when you get a huge valuation, a lot's priced in, you got to take some chips off the table, even if you still love the story.
Yeah. Yeah. Listen, I mean, you know, they're putting in their perspectives that they expect potential risks to include catastrophic or existential risks to humanity. I mean, it's pretty outrageous prospectus. I got from our last podcast when I was actually just quoting the letter, you know, across Bloomberg, "No regrets, says it's the end of the world." And I'm like, I didn't say that. And you know what I said is we need to make sure through regulation that that 10% chance that Dario talks about is less than 1%. Right? Like that would be, if I was the president, the mandate I would give, or if I was Congress, the mandate I would give those executives: Hey guys, you're not going to go to zero and then nothing's zero. But your own words were 10. Let's drive that thing down to one. You know, look, I mean, thank God we've done a pretty good job with — I don't want to jinx this, but we're 80-plus years out from the introduction of atomic warfare. We've done a pretty good job. We have to figure this one out as well.
Well, you know, we put limitations on, right? We knew we were going to have an atomic world, but we put limitations and verification that slowed that pace of the arms race and kept things in check.
Yeah. But also the good news is you have had state actors, even in authoritarian regimes, you've had state actors that want to preserve humanity, right? So that's the goal. You know, the goal is to keep the preservation of humanity tantamount to everything that we're doing. But I want to go to what McKinsey is saying, Michael, and get you to respond to this. Okay. McKinsey's Global Institute is saying 770,000 workers per year will need to switch occupations. I mean, that's a lot, Michael. The historical average is 215,000 as an example. So, we're talking about 11 million total workers or 7% of the entire US labor force, okay, over the next 10 or 12 years. It's a massive amount. Should we be worried? You think that's exaggerated? You think that they're not thinking about this correctly?
Let's both admit that we don't know if they're going to get that number right, plus or minus, but they're the right direction. It's really clear to me that both parties will have a political platform in this next presidential cycle, which starts in five weeks, about transition, right? With the best AI optimism, there's going to be a lot of need for transition and that's going to be a lot of different pieces to that. It's going to be financial, you know, transfers of money, right? A buffer. It's going to be job training. It's going to be — but I think governments have a sense of responsibility to their people. And this is going to be one of the big issues that governments around the world have to face, assuming McKinsey's right. Right. This is — you're just not going to, as a government, say, "Hey, we're going to just let those people sort it out themselves." You know, it's not the instinct how governments function. And so certainly if you believe the Democrats are going to come into power because of all these polls, that's going to be part of the future going to have to be. You know, people ask me all the time, what should your kids study? I keep telling them the humanities. Learn how to read, to think, and to write. And as the world unfolds, you're going to pivot into — but vocational things are going to be harder and harder. It's bizarre that this is like the greatest bull market for liberal arts educations. You know, people don't intuitively think that, but if your whole job was — you know, I mean, think about if you're a bariatric surgeon, right, doing the stomach stapling, and then GLP-1s come out, you're out of business. And I talked to one recently, he said he used to be booked.
You know, one of my buddies is one of the largest candy distributors in the Northeast and he's saying that the GLP-1s are killing them. You know what I mean? I mean, everybody not on a GLP-1 still loves candy. But Michael, this again, it's not a political question, but it's more of a humanities question. Okay, so let me see if I can address it to you this way. If those people that are in the White House combined net worth of one and a half to two trillion dollars, the people that were sitting with the president, how do we reduce the potentiality for greater wealth concentration as a result of AI? Because if you're having this level of job disruption, okay, you could also create a wage crisis in the country contemporaneously to the productivity increase.
Yeah, there is going to be an AI dividend that is paid to the American people somehow. And there's a philosophical rationale, right? AI is getting its juice from the combined knowledge of this planet, and therefore it's intuitive that there should be a dividend. It's going to come in some form of equity in those companies for the government. You know, Trump's done that a few times. A token tax. Again, I'm spending a lot of time trying to think what if I was in charge, what would be the best way? You need that emotionally to bring the rest of the world, the rest of the country along with these guys that are doing so well. But there's going to be some redistribution. It's going to happen. How it happens, I would pray it happens in a really efficient and fair way. But it's going to happen like it always does. And so, you know, we'll see if this California wealth tax passes. I hope it doesn't because I don't think statewide wealth taxes is where to go. But that's where the direction of people are going. And for our country to really heal itself and move forward in a great way, this idea of government assistance, right? You have people that are on SNAP for food or Section 8 housing for housing or Medicare or Medicaid for medicine. Often the recipients don't feel great because they feel like we're welfare recipients. The people that aren't receiving them feel like, "Oh, I'm carrying these people."
Right. There's paradoxes to it. Yep.
Right. What I really think needs to happen is a reimagining of what the government's role is. Like if you're born in Norway, you don't feel guilty that there's this giant sovereign wealth fund that's there to support you. You don't. You feel it's like, I'm lucky to be born in this country or I've immigrated to this country and that is what it means to be Norwegian. We need in America to say, I'm glad to be born in this country or I've legally gotten citizenship in this country and one of the privileges of being American is I have this safety around housing, medicine, education, and food.
Yeah, it would reduce anxiety, base layer protection, also let people actualize what they really —
A lot of it is the narrative around it. Like if it's seen as a bad thing to receive or to give, it's just not — the politics don't work.
What you're basically saying, Mike Novogratz, is you want America to treat all our citizens the way you and I treat our kids, Michael. That's basically what you just said.
All right. That's the Scaramucci party and I'm going to be his campaign manager.
Well, trust me, my kids would be voting for that all day.
