NVIDIA OpenAI Financing Guarantee: Inside the $250B Ohio Data Center Deal

Hey, everyone. Wow, it's Wednesday. This week, hump day came.

It's here. It felt like I just came in from the weekend yesterday, but we're already going to the second half of the week. It's been a busy week at the gang.

We've got so much news. The cycles is compressed and deep and thick, and I've got 12 different writing projects I'm working on, so it's crazy. We're going to- It's supposed to be why these people are supposed to be on vacation, but they- I get that, but someone didn't get that memo.

There's- These are ... probably it's because... No, really because we're living in a boundary time.

Mm-hmm. The beginning of one epic and the end of an era. And so that displacement, you got a lot of dinosaurs kicking.

I'm writing an article about that now, especially in the analyst business. And there's so much money at stake. The amount of money in this data center, I just got a new thing we're doing at Techstrong.

Be on the lookout for it the next day or so. The first edition of a Techstrong investigative report. Ooh.

That sounds cool. On data center lobbying money, where it's going and who's paying the bill. So stay tuned for that.

But- How much Trump and family are making on all of this? You know, Jack, I try to stay out of politics. Not.

Why? Yeah. But, no, you know what?

Just follow the money. Yeah. That's all I could say.

All right. But let's introduce our gang today because we got great stuff to go over today, man. We've got our favorite lady in front of the brick wall, Terri Robinson, with a pretty picture in front of the brick wall now.

I wish you could see it. Yeah. Well, we could.

We also have joining us my friend, Jack Gold. Another friend, we've got a lot of New Yorkers on my panel today. Yeah.

Yeah. I love it. Liz Safran.

Representing. Yeah. Representing.

Not from New York, though, out there in the valley, though he does come to New York to visit his daughter. I just did. Hello, Jon Swartz.

Yeah? Yeah. Brooklyn, nonetheless.

Go Brooklyn. Hey, Jon Swartz. And kind of the quintessential...

Well, he's not really from New York because where I grew up, we call that the Bronx, not New York, right? But the man from the Bronx. The Bronx.

We call that the beautiful borough on the mainland. Exactly. The only part of New York that's actually part of the mainland United States, Mike Bizard.

Hey, guys. Welcome. It's a crazy day.

Mike, what a tangled web we weave. It's kind of insane out there, but the latest insanity is that NVIDIA's coughing up, allegedly anyway, $200 to $250 billion to help fund a data center project in Ohio for OpenAI. And to me, it feels like all these relationships are getting to be so intertwined that it's getting hard to distinguish who works for who out there.

And in a lot of ways, I spend some time in the channel, and this is starting to feel like distribution agreements where if NVIDIA- Too much ... if NVIDIA ponies up billions of dollars, maybe they wind up funneling more business to OpenAI. It's not quite clear to me how those relationships will work.

But Jon, you're following this a little more closely than I am, so start us off here. What the heck is going on, and who's working for who? Yes.

Follow, I know we were mentioning earlier, follow the money, it's almost impossible. This is a, I guess, for want of a better term, it's something we were going to call, and Alan, I think among them, is calling it circular financing. So, as you mentioned, Mike, NVIDIA's in talks to guarantee up to $250 billion to finance this OpenAI mega, massive Ohio data center campus, which will cost half a trillion dollars.

And NVIDIA's also potentially financing $350 billion of GPU purchases. So effectively, what I think, NVIDIA is helping fund the customer that buys its own chips. I think Alan, in his column that I read earlier today, argues that OpenAI is becoming too interconnected to fail.

And that's kind of, to me, when I look at this, a step back, I think, is this smart vendor financing, or is it systemic risk level? Mm. What happens to the chip sales, the cloud commitments, and infrastructure debt if OpenAI stumbles, which I think it will?

This makes sense from a strategic point of view because NVIDIA's leveraging its incredible balance sheet as a pragmatic long-term hardware demand, architecture. But the risk is, by back stocking $250 billion in leases while separately financing $350 billion in GPU purchases, it creates this vendor financing feedback loop. And so, Alan, what could possibly go wrong?

Right. What could go wrong? So let me be clear.

NVIDIA's not paying $250 billion. This is more akin to when my kids call me up and say, "Dad, I want to buy a car. I want to get an apartment.

" Right. " Right. And that's what this is.

But there's more to this web than just an... And then, as you mentioned, Jon, in exchange for that, maybe there's a quid pro quo, as they say in politics, that OpenAI is going to buy about $350 billion, right? So NVIDIA's guaranteeing $250 billion, and OpenAI agrees to buy about $350 billion in GPUs from NVIDIA.

Right? And so everybody's balance sheet goes up this way here, and who gets left holding the bag? But there's more to this web, because if it was just those two, I would say it's not even a ménage à trois.

It's just the two of them. But there's more. But wait, there's more.

You've got SoftBank Energy, right? So SoftBank's coughing up, I forget how much, tens of billions of dollars through this subsidiary called SoftBank Energy. Where are they getting the money from?

Well, the Trump administration, there they are, Jack, they made a deal with Japan that they would give Japan favorable tariff rates if Japan invested in America. " They sent Masayoshi Son over to America, and they invest in Masayoshi Son, and that's their investment in America, SoftBank Energy, owned by SoftBank. So they're putting up another couple of hundred billion dollars.

There's another piece to this, though, Alan, that I-- I agree with everything you just said, but look at this from NVIDIA's point of view. If I backstop OpenAI and SoftBank and whoever else building out these massive, massive, massive data centers, buying 300, 400, 500,000 GPUs from me at 40, 50K a piece, and my margins are 60% plus on those things, then that $250 million I'm giving them or- Billion, Jack ... sorry, billion.

You're right. I'm backstopping 60% of that, well, perhaps not all of it, maybe half of that is chips. 60% of that half goes back to my bottom line.

It increases the need for other companies to get in line and buy those chips because I'm going to tell them, "Hey, they're going to be available. These guys are buying them all. " So- But Jack, you're missing the indispensability trap.

This script has played out before, and let me tell you when. The dot-com era, we saw this with the fiber players. They played the same game.

WorldCom- Yeah ... Enron, all these guys that went to jail for malfeasance. Right?

This is the same Ponzi scheme they played there. Sure. They did this back and forth with their books.

I give you some, you give me some, the money never leaves, and I build it up. It happened before that. There's a guy named Samuel Insull you may not have ever heard of.

Samuel Insull was a 5'3" guy from Britain who came here as Thomas Edison's secretary. Edison invented the modern electric industry. Samuel Insull made it a business.

He went out when JP Morgan took over the Edison companies and formed what became General Electric. They didn't make Insull the president. He went out to Chicago and bought a public electric utility out there and then consolidated the whole Midwest, became the electric tsar.

Baron. Yeah. Made untold dollars.

It happened even before that. He did exactly what you're talking about. All kinds of subsidiaries, cross-licensing.

This one guaranteed that, that guaranteed this, this guaranteed this. Boom. And then the stock market crash hit.

Right. Samuel Insull died penniless on the Paris subway of a heart attack. Yeah.

Well, and even- He went down ... before that, Alan, if you look back at the railroads- Same story in railroads. They all went bankrupt.

Right. They all gone- By the way, all covered in my book coming out. So check it out.

Hold it up. How about that. Hold it up.

I don't have it to hold up yet. I'm waiting for the guy to finish the cover. But it's all in there.

But that's what it is. But this is going on through the whole AI industry. This isn't just OpenAI.

Anthropic and AWS. I'll tell you the company that's most leveraged in all of this, Oracle. Oracle, 300 billion.

Oracle's debt rating has gone down a step or two above junk. Yeah. So let me ask you this question, then.

And I'll throw it to John, actually. Is Wall Street starting to figure this out? Because- Yes ...

there's certainly a lot of volatility in the stock market around this mega company. I hope so. Yeah.

As squeamish as these analysts are on- John, your mic's losing on us Oh, how about this? Is this better? Yes.

Okay. So as meek and as squeamish as Wall Street analysts are when they have a post-earnings call, they're starting to ask the question. And I think what they're going to start looking at, especially because OpenAI and Anthropic are going public, so they're going to be under more scrutiny.

They're going to look at the expenses going out. You see Meta, which has faced the same questions, has whacked their employees, their staff. They're going through these multiple changes.

Amazon's doing the same thing. Oracle's doing the same thing. So is Microsoft.

I'm going to go back to this deal. And in a sense, what did Judge Judy say, remember, about co-signing loans? Never ever do it.

It was always every court case came down to that. Please tell my kids that for me, please. Yeah.

Well, everyone should live their life according to Judge Judy. I do a little bit. Yeah.

Yeah. No, but let me tell you where the real tell is. Alan, can I- Go ahead.

Here's just my takeaway. So the structure of this deal transforms what I think should be an independent commercial transactions to a single point of failure. Right.

So you're making OpenAI, NVIDIA, and their infrastructure partners too interconnected to fail. It's not just- Well, OpenAI certainly is at the center of a lot of these webs. But guys, here's the telltale sign that you know, and this is what happened again in the dot-com.

Up to now, most of these companies' capital expenditures, meaning the Mag Sevens, the AWS, the Google, the Microsofts, the Oracles, the Broadcoms, all of these companies that are financing this, have financed these CapEx out of free cash flow. Meta, well, not Microsoft, because they still have a ton of cash flow. Meta, Google, for the first time ever, Oracle, all of these companies are now negative free cash flow, which means they're spending more money on CapEx than they're throwing off.

And that's the red signal- So it used to be- ... for Wall Street ... so it used to be the big question over the last couple of years was, "Oh, when's the AI revenue coming in?

" That's the whole narrative just flipped. Yeah. But now it has to be financed.

And now that means the bankers come in, and the bean counters come in, and they're ruthless. Yeah. They look at our credit cards.

So how long before this just becomes yet another bill handed to the US taxpayer, and we have to bail all these people out? Exactly. Couple of years.

Well, no, beyond a couple of years, because our politicians are all for it. Remember, Slippery Sam has already offered to give 5% to the government. So that's to the present administration on this side of the spectrum.

" We want 50% of all the stock of any company going over $200 million in value on AI, right? So the government thinks they're going to line up at the trough as well to feed. But- You know- It's not going to be worth a hill of beans, though, I think- Yeah ...

eventually. You know what's really interesting, though? I was reading of regime change.

It has all these really interesting dynamics between Trump and the tech executives. The one guy who escapes all this, who is really the Trump whisperer, who's influencing him every step of the way, is Jensen Huang of NVIDIA. Well, and it's funny, he's the master web spinner, because when you've got $5 trillion, even if he lost $250 billion, all right, wouldn't be the end of the world, right?

Even more than Elon. But so the question becomes, is this a short-term phenomena or a long-term phenomena? One of the issues that happened, we talked about the internet, right?

The early web days. AOL spent, I was going to say something, I won't use that word, a bunch of money. A something load of money?

Yeah. Yeah. Something like that, right.

And for them, it all went away. But we can't live without the internet today. Is AI going to hit some sort of bump?

Which they could, or pothole, but it's not going away, guys. No, I don't think it's going away. So this is the point of my book, though, Jack.

It could only reach its potential if it becomes so ubiquitous and commoditized that everyone could use it. Sure. And if it does, then you got to ask yourself, is $8 trillion being invested in this the same as the trillion dollars that was invested in fiber that didn't get lit up until the cost of fiber dropped from $1,200 a megabit a second to 50.

And there's still dark fiber out there. Yep. And so are we over building this?

Yes. Are we overinvesting in this- Yes ... AI capacity?

Because then we see the open weights, and they could run on personal machines, and we don't need these factories. So that's the indispensability trap. When it becomes that indispensable, the value goes upstack, it doesn't stay there.

Alan, this is the mainframe versus PC world all over again. Everyone had- Mm-hmm ... went out and bought IBM mainframes, then the PCs came along and guess what?

Well, first mini computers, but then PCs. Over about a 10-year period, we went from mainframe domination in computing to minis and then to the PCs. How many mainframes is IBM selling these days?

They're still selling a bunch, but certainly not what they were. Well. Not where they was, yeah.

Not what it was, but it still floats IBM. Anyway, Mike, we're over time here. Right.

What are we going to do? But to Jack's point, the irony of all of this is Zuckerberg's in the journal talking about how centralized AI is a bad thing and it needs to be distributed out to the edge, so. Not just the journal, he's on a whole media blitz.

There's Reuters, there's a Times article he did an interview for. He is on a campaign. But I've got an article on tech strong AI coming out.

If it's not out this morning, it'll be out later, around what's behind this with Zuck? Oh, man. What's the game?

We won't go there. We're not going to go there. All right, well we're going to shift the gear to our next topic then, but it's related.

But there's this whole concern around the chips and the supplies, and there's a new deal between Broadcom and Samsung, where Samsung was going to be the foundry and kind of an alternative to TSMC. And Jack, I was just curious to know, how viable is this? Do we have another source?

And if there's another source for this, what will be the impact on pricing and pressure and all the things we just discussed? Well, look, we're already seeing the main foundries, TSMC and Samsung, the two biggies, right? Already full.

If you want to bring a new chip in, unless you're Apple or somebody, a big company from their supplies, you ain't getting in the queue anytime soon. And so what Broadcom is basically doing is saying, "Okay, Intel isn't quite ready yet. " So that really leaves TSMC and Samsung as the two main players in building out these advanced chips.

And Broadcom has a lot of customers that want to use Broadcom as basically their supply house, their development house, their creation house. So Broadcom is looking for as much capacity as they can. They're also looking for price leverage.

I mean, TSMC this week, I think it was, said they were going to increase pricing about 10% or so, depending on which chips that you want to buy from them, and that's because they can. If you don't want to buy from them, there's the guy behind them that's going to buy it anyway. So we're going to see a lot more shopping around, if you will, if they can.

The problem with smaller producers, or sorry, smaller suppliers of chips that have to go to these big houses, is that they don't have a whole lot of leverage. Broadcom is a huge company. They've got a lot of leverage that they can play with Samsung.

" I think there's a lot of that going on, and Samsung's going to use also the Broadcom support. Samsung is not just a foundry, they also make chips, and so they're going to be learning a lot of stuff from Broadcom coming in. So this is a benefit to both Broadcom and Samsung.

It's a benefit to the marketplace. Mm-hmm. Now, the problem, of course, is most of the supply constraints are driven by AI.

If AI tanks, as some people think it will, then there's going to be an awful lot of spare capacity out there, and these guys are going to be hurting. Hey, Jack, can I ask you a quick... Oh, sorry, Mike.

I'll ask one quick question. Does this deal meaningfully close Samsung's foundry gap with TSMC, or is it merely just a memory deal dressed up as a foundry deal? I think it's both.

Samsung certainly makes a lot of memory, but they also make a lot of chips for other people. They made their own chips for a long time for their smartphones. They never really went anywhere.

Qualcomm killed them. But, if they can get more other stuff from Broadcom, particularly around Broadcom's networking and connectivity options, then that gives them a leg up in a more broad-based environment than just building out memory. How long will it take for this to have a meaningful impact, and are there other foundries at play here, or have all the players been accounted for?

Well, good question. And so Intel is the up and comer. If they can pull it off, which I think they will, in about a year or two they're going to be fully competitive with TSMC.

There are the Chinese players that have been so far behind, but there's a lot of news coming out of China where they're developing. There's a new memory company that just went public in Shanghai, I think it was. SK Hynix or whatever?

No, CXMT, I think. Oh, yes. The CX.

Yeah. And it went up 500% because of the memory shortage, even though they're not up to par with the high-end- Yeah. But these are sub-two nanometer chips.

I don't think China's there yet. No, no, they can't. But they just announced this week that there's a new company in China making ultraviolet semiconductor machines.

Not the- But not the- That right ... not the EUD, the other one. The D.

D. Right. The D.

So what? When you're talking memory, Alan, you don't need- No, I know, but it's not for these sub-two nanometers. Sure.

These are four-seven nanometer chips- Of course ... that the lithography was. But when they get into networking, when they get into some of the memory issues, when they get into some of the security chips, you don't need the sub-two.

No. And Jack, from what I read, unless I'm wrong, 85, 90% of this deal is good old-fashioned memory chips. Probably.

Right? And that plays into Samsung. I think the gap between TSMC and Samsung is on the wafers, right?

Yeah. Not on memory. And it's only maybe 10% of this deal that's Samsung wafer.

" Yeah. And there's also the issue of if it is memory now and they're doing well on memory, Broadcom is going to increase the other stuff, because I've got a relationship with Samsung now. So this- So I have a question for you guys.

Go ahead. I have a question because I don't know a whole lot about the chip space. But given that this is an AI-related deal, I kind of like the fact that it's like a new player's coming in, but it feels like it's still so kind of contingent on the AI economy booming, and this whole thing could blow up if there's a hiccup with AI.

And so, okay, so now you have another couple of oligarchs at the table, and that's probably good. Or new companies that could hedge some of the magnificent seven other players, right? Which is probably good, but I just feel like it's more of the same kind of circle.

It is. I don't know if you know what I mean. Well, I have another article coming up tomorrow called "As the AI World Turns," because it's a soap opera.

It is. There's a lot of characters and everything else. But to your earlier point, Alan, if a lot of the AI processing shifts to the endpoint, does that not favor a Samsung and a Qualcomm over- An Apple ...

traditional players? An Apple. But here's the thing, let's not kid ourselves.

AI is not going away. A failure of AI may mean that, look, out of the $8 trillion we're investing in data centers, my goodness, only $6 trillion will be used- Oh, right ... or $5 trillion will be used.

What are we going to do with just $5 trillion? Right? The monies here involved are such that failure is relative.

It's not going kapooey. And there's an acknowledgement among some of the biggest players, even they acknowledge it, the AI pace has to decelerate. So they know they're going to hit their targets.

It's just a question of when. But what will be the disruption? Oracle's really laying off tens of thousands of people, so it doesn't matter.

So people- Well, they've been doing that for a while ... have their doubts, right? Yeah, but no, Oracle's laying off people not because of AI, but to pay for AI.

But I understand that. Oracle is laying off those people so they have the money to go buy AI factors. I understand, but there's still an economy underneath.

These guys don't operate in a vacuum, right? It's just like so- No, no. Be very clear.

Larry Ellison bet the farm. He could be this generation Sam Insull. Here is the danger.

It could happen to a nicer guy, though. Yeah, exactly. Here's hoping.

Here is the big danger to the average person. Too much of your retirement accounts are wrapped up in these AI companies, and if they take a tank- That is my point. That's what I'm getting at.

Yeah ... all you guys who think you're retiring because your AI stocks are doing so well will be going back to work. Yeah, but guys, think about it.

I think you're right. AI is here to stay. But what is the depression-level event that's going to happen before everything balances out?

Because something's got to give, right? Yeah, but it's not necessarily a depression-level event. Look at any new marketplace.

PCs, railroads, automotive. There were dozens and dozens of companies that started down that path, and at the end of the day, there were a handful that made it. Mm-hmm.

The same thing is going to happen in AI. It might be open AI, it might not. Meta probably not.

I hope not. There were dozens. It's the companies that move up the stack, and that's what you're desperately seeing- And didn't you write about that?

ChatGPT. Not Chat, Anthropic. There was an article, I think it was in the journal yesterday, that they partnered up with Figma.

Yeah. Right? The design company.

Yeah. " What'd they come out with? Claude Design.

Because they're looking to move up stack. They know basic intelligence is going to be commodity. It's what you build on top of it.

AWS, Google, Azure, are all built on- Microsoft ... the skeletons of the corpses of the dotcom era. And not to start a rumor, but does this ultimately mean an Nvidia's got to buy an Oracle at the end of the day?

Oh, God. For pennies on the dollar. For pennies on-- Yeah.

No, they got so much- It won't work. It won't work. Think about what you're talking about.

A chip company buying an enterprise software company. I can't find a successful- You know, Jack, I remember when they said that Broadcom would never buy a software company. They bought a few ...

look what happened to them, Alan. You don't even see the software guys anymore. Symantec?

Where was Symantec? Wow, Symantec. Computer Associates?

Let me just say, for all my VMware friends out there, I'm still here. We do business with them, Jack. Fine, but they're a shadow of their former selves.

No, no, but what Hock did there, right, is he brought the software to even out the hardware cycles. To even out revenue around the hardware cycles. To get the money, the margins on software are much higher, and he took what he could and jettisoned the rest.

Yeah. But here's the other thing. Quite frankly, Oracle has a high debt ratio or a high exposure to AI.

Broadcom has also made deals with the same Anthropic and OpenAIs that we talked about in the web. They put their balance sheet on the line for some of these deals they did with OpenAI and Anthropic. Sure.

They're exposed. Where's another company like that? Right?

They're all exposed, every one of them. Intel- I feel like this is just a tempest in a teacup. Since Liz's point, though, isn't there going to be some sort of event that- Yeah, there's got to be some kind of record.

This seems like it's a house of cards or whatever that... Where's that game where you're just putting things up? Jenga.

Jenga. This is like a high-tech version of Jenga, except it's involving trillions of dollars. I mean, I- Right.

I didn't have- It's so huge ... I think it's more like three-card monte, where it's all stacked. But what do I know?

Russian roulette is probably better. Wow. I learned that three-card monte game firsthand in New York City.

Yes. But anyway, we've got to move on to section three, Mike, segment three here today. What do we have for the fine people at home?

All right. Well, truth is stranger than fiction, but apparently in New York State, somebody decided that they were going to test out an AI robot teacher. And I'll be honest, if you asked my 12-year-old self to design an AI robot teacher, it would look something like this.

But apparently- They did a movie like this once, didn't they? Yeah. They made a couple.

Yeah. They did a movie. It turns out that the company that makes these things also makes sex dolls, and apparently they use that as part of the design, and, well, the local people decided that this was not a good thing for their children.

So- No. Why? Let's start with Liz.

His reputation. Okay. So I love this story because I have a high schooler.

I mean, obviously this is upstate New York, but I'm on the school leadership team in AI. The responsible use of AI in high schools is a big thing in New York City in particular. They just passed a guidance.

Teachers don't know what to do. But they feel like they need, like everybody else, "We need to do something, so let's get AI in the classroom," and they don't know how. But aside from the sordidness of the backstory here, which I love, which I think makes it cooler, it's the only thing I like about this, is that it should've never happened.

I mean, the fact that this school was rolled out in some reservation school in upstate New York just smells like corruption to me. Like somebody's palm got greased. $57,000 robot.

Yeah. And it's like, come on. It's less than a teacher.

Yeah. Oh, God, don't say that. Or what the teachers make.

Yeah. But, I mean, there was no oversight. And of course, the kids like it.

And I mean, I think that was supposed to be for STEM or for coding. So why do they have an AI robot teaching kids how to code when one of the things that AI is best at is coding? It makes no sense to me.

And given the sensitivity of kids in school and AI and how is this going to be used, it seems like such a poor use case. Especially, I understand this isn't New York City, and the New York City DOE passed the guidance. But it's like there's so many educators that are struggling with what is the appropriate use of AI in schools for kids whose brains are still developing, and there's so many train wrecks.

I got involved in AI safety because my kid's work got flagged by an AI checker, which was legitimate, and then it got flagged again by a teacher when it wasn't legitimate, and he had to do the assignment again. So there's real consequences for how kids are evolving, and I feel like somebody paid somebody. " Right?

That's a distraction. We are forever and a day talking about how AI is trained and who's training it, and then we have this robot with a past. So- Which questions- Must've been some interesting training, huh?

Yeah. And what's her name? TransTech.

There's another angle to this thing, though, that I want to get Terry's input on. One of the things that occurred to my 12-year-old mind was that I could hack the teacher now, and I could just cheat on the exams because I'm going to find a way to figure out what's on that exam by hacking into the robot, and I'm pretty sure there's no security in this thing. Well, sure.

I mean, look, they didn't vet this thing. So the chances that the security is locked down in there are just next to nothing. And you've got these enterprising students, especially if they were in front of STEM classes.

Those kids are bright, and they know what they're doing, and it doesn't take too much of a leap for one of them to hack this. I think that is a real problem. And then get the answers to the test, maybe start having the robot do some things for you.

I mean... What kind of things did they do for the sex teaching thing? Oh, come on, stop.

Stop. Yeah. No, I mean, that's Mike's 12-year-old mind in real mode right there.

But- Not to mention, I could just change the exam. I could just alter the whole thing and make the exam a joke. Well, right.

But you could also start making some money off of it, your enterprising 12-year-old self. Sell the tests to your friends, the answers to whatever. There's some opportunity there.

I think this is interesting, and I was interested to read about this because in New York, when this robot, Sally, was on TV quite a bit more recently when they were talking about having this in front of kids or classes and the response to it. And I think there's another robot teacher in maybe Long Island. Liz, do you know anything about that one?

I think I had her. I think I had a couple in- Yeah. Hey, Alan, in the 1800s, we didn't have robots.

Yeah. No. Who knows?

Come on, you don't know that. But yeah. So everybody was kind of excited about this, but no due diligence.

There was a great movie not long ago, a few years back, called "Robot & Frank," if you guys haven't seen it. And the basic premise of it was, there was a retired guy whose son bought him a caretaker robot, and the robot was put in the house to take care of this guy and do whatever he needed. And over time, he convinced this robot, because he was a retired thief, that the best thing the robot could do for him was help him steal things.

And so it goes right to the hacking issue. If you train a robot to do what's best for me, and I convince it that what's best for me is to do something illegal, it's going to do something illegal. Well, do you need a full-on robot, though?

Can't you just have an AI agent or something? Well- No, I think the robot's a nice touch. Here's my take on this ...

it's a nice touch, but it's unnecessary. " Then I read that this is at a Seneca Reservation school. And I'm saying, "Was this from the Bureau of Land Management?

They paying for it? Like Department of Interior? Is this coming from the federal gov?

" That's not a bad idea. Get gambling money, and the next thing we buy is a robot from the sex farm. Wow.

It all kind of goes together. Right. Side hustle.

But then here's the thing. And then you build a data center on the reservation and- Maybe that's what it is. It's tied into that.

They're already talking about that. But there's more. At the same time, the federal government, I don't know if you saw this, just signed, I don't know if it was an executive order or how they are enforcing it, but I think it's through the Department of Commerce, again, our favorite Secretary, Howard, who likes to send letters in the middle of the night, who says, "No Chinese robots.

No Chinese robots. " The robots from the sex place, no good. Chinese robots, no good.

Elon's robots, because after all he's got a great moral compass. Right. Elon's robots are great.

Let's go make babies. Right? You know, this whole thing sounds like a bad John Hughes movie.

This is a freaking soap opera, guys. It's crazy. It sounds like a bad John Hughes movie.

This is right out of a movie from 2018, "Weird Science," Kelly LeBrock. Yeah. Mike, I got bad news for you, my friend.

That was not 2018. That was 20 years ago. That might've been 1998 or something.

Okay? " Yeah. But you know, the thing going on- Okay, boomer.

I know. We're all too old, guys. I know what movie- ...

everybody's going to be watching on Netflix this week. That's what- There you go. Go check it out.

Kelly LeBrock. That was a great movie. I do think that was early '90s, though, Mike.

Yeah, it was a lot longer ago. Was at least 30 years ago. Yeah.

A long time since Kelly LeBrock. I just wanted to hop back to the kids hacking the robot again, what you were saying, because you're right, if they're STEM kids, they're probably smarter. But that's all the more reason to have more oversight about...

What? What were you going to say? So I have a serious thing, Liz, I want to ask.

So this kind of shows this fundamental gap in how local school districts bought emergent AI and robotics vendors. For instance, they're evaluating textbook publishers and software tools, but there's a quantum leap to physical AI or robots. It just happened overnight.

But that's why it smells like corruption to me. What happened? Because the stories that I've looked at, they don't explain how it got in there.

They said what the price was. Well, so let me throw something in here, because it's happened before, and it's going to happen again. " Pick a number.

I don't know what the number is. " And reference. Yes.

And frankly, most school districts in this country are so cash poor- Yes ... that they're happy to do it. And so- Not in Florida, though.

They'd be afraid the robots might give out vaccines. Yeah. Or Fauci.

He's on it right now. Or, yeah. Just wanted to throw that in there.

Or get a worm in their brains, yes. Yeah. Have their lifestyle.

Well, I think it's interesting. You're banning things like "The Bluest Eye," right? There are all these books that you can't read in these schools and whatever, and then somebody throws a robot in- Right ...

and really realizes- A sexy robot is cool. Yeah, a sexy robot. Right.

So how much of this, though, do you think is going to play into the politics of the day? And districts are going to start having robots that they program to be a certain way because they want the kids to learn XYZ as dictated by the local politicians rather than by a human teacher. Mike, you're assuming that the school districts are programming them.

In all likelihood, they're not. They're getting them over the counter, whatever that happens to be, from- Yeah, but the school districts are already censoring the textbook authors. They could censor the robot makers as well.

Some school districts are. Yes. Yeah, but- I can't help it.

I live here in- Sure. Florida, yeah. I don't want to use the word.

But you're assuming that they have the expertise to do that, and technology generally, in most school districts, they're not exactly at the level that we're talking about today. They're usually a decade behind when it comes to technology. They actually usually rely on parents coming in and volunteering.

Okay. So therein lies the bigger problem. The school districts themselves don't have the expertise to do it.

Right. They lag behind or whatever, but there's somebody out there that does, that might very well come in and- And you just hit it on the head, Terry. Let me put the bow on this.

Here's the facts. On the whole, our schools suck. Yep.

Right? You've got a federal government that is wiping out the education department. Yeah.

Right. Cutting budgets. Local governments don't have the money to spend on students.

Liz, you don't realize how lucky you are in New York City- No, I do ... when you look at what they spend per student compared to the desert I live in. And in a time- Hey, Alan ...

when the rest of-- I'm sorry. Yes, John? I was going to say, Linda McMahon, who's the head of education, she refers to the technology as A1.

A1, like the steak sauce. Well, because she's a wrestler. Like the steak sauce.

Yeah. Or a wrestler's wife or whatever. That's where it starts.

She said that from the stage. Yeah. She said that from the stage at an AI conference.

Yeah, no. Well, if you could put Robert Kennedy in charge of the health department, why wouldn't you put A1 steak sauce in charge of education? It's true.

Right. But let me go back to this thought. Sorry.

The rest of the world is lapping us, laughing at us and lapping us, and at a time when we need this next generation to be educated, AI may be our last best hope, and we're going to fritter that away because we're worried about someone's sex dolls or where it was made or what have you. And who winds up losing here? Our kids wind up losing.

Right. And that's the thing to take home. My 12-year-old brain has one more question.

Oh, no. Kelly LeBrock. Let me hear it.

2018. Yeah. I don't know what calendar you're on, man.

So what does the sex education class of the future look like, enabled with robotics? Oh, geez. That is what's on here.

We're going to have sex. It's not a problem. You know what?

Mike's still living his fantasies. Yeah. Still living the fantasy.

We have the male gym teacher teaching it. Oh my God. I think we- Hopefully not too many Asians ...

all right. We've descended into the- This is just a- We've descended into the inferno. Let's stop.

What was the robot's name, Sally? Sally. Sally.

Oh, by the way- Now we got Sully. By the way- Sully and Sally ... here's- Sully and Sally.

Okay. "Weird Science" 1985, John Hughes. 1985.

1985. Mike thinks it's 2018. 30 years ago.

Oh, boy. In my mind, everything was 20 years ago, so there you go. Yeah.

Yeah. Oh, God. All right.

Guys, I got to pull the plug. I got to pull the plug. But Liz, Terry, Jack, John, and Mike, thank you.

Thank you for watching. We hope you've enjoyed this rant and round robin- ... through AI.

But I do think all these three things are connected. For sure. Right?

And it's showing that the edges of our society are fraying in terms of how are we going to deal with this? There's so much at stake. The stakes are real.

The games are real. The people here are real. The money is ultra real.

There's so much riding on this. We're talking about words like depression, right? And somehow or another, we'll have to muddle through it.

Regardless, we'll be here on the gang, documenting and talking about it, and we invite you back here every day, Monday to Friday at noon, to do it. tv, the TechstrongTV YouTube channel, or our OTT channel for Android, iOS, Amazon Fire, Roku, and Apple TV as well. Any screen.

Check it out. We'd love to hear what you're thinking. But until then, on behalf of a great gang today, have a great day, everyone.