Qualcomm’s AI Chip Deal Headlines a Big Week for Enterprise Tech
Qualcomm’s AI chip deal is the biggest story in enterprise tech this week, but it’s not the only one. Meta launched a new autonomous AI agent, and Oracle is fighting battles on two fronts. Here’s what the panel broke down on the show, and why each story matters for enterprise leaders right now.
Meta Launches Muse, an Autonomous AI Agent
Meta just launched Muse, an autonomous AI agent built to handle real-world tasks on its own. The move signals that agentic AI is moving from pilot projects into production use. Muse can complete multi-step tasks without constant human input. That raises a real question for enterprise teams. Governance and oversight practices are not keeping pace with how fast these agents are shipping. Read the full story from Techstrong.ai on Meta’s Muse launch.
Qualcomm’s AI Chip Deal With Amazon
Qualcomm locked in a massive AI chip supply deal with Amazon worth up to $60 billion. The agreement puts Qualcomm squarely in the custom-silicon race that hyperscalers are running to reduce their dependence on Nvidia. It also raises a bigger question about where chip supply concentration goes next, and whether more cloud providers follow Amazon’s lead. Full details are in this report from Techstrong Semi on the Qualcomm-Amazon chip deal.
Oracle Faces Spending Cuts and a Congressional Subpoena
Oracle had a rough week on two separate fronts. Enterprise spending cuts showed up in the latest ETR survey, pointing to softening demand. Separately, a House panel subpoenaed Oracle executives over a $17 billion VA health record system overrun. Together, the two stories raise pointed questions about vendor accountability on large government IT contracts. Read more from Techstrong.it on Oracle’s spending cuts and DigitalCxO on the Congressional subpoena.
Mike Vizard hosted this episode of Techstrong Gang, joined by Jon Swartz, Teri Robinson, and Elizabeth Safran. Together they unpacked what Qualcomm’s AI chip deal means for the custom-silicon race. They also covered why Meta’s Muse agent matters for enterprise governance, and what Oracle’s tough week signals about vendor risk. Watch the full conversation for the panel’s take on where each story goes next.
Transcript
Hello, everybody. Welcome to Techstrong gang. It's Wednesday, AKA hump day, and well, it's amazing.
We have a lot to talk about even though it's a short week. John Schwartz, as always my friend, good to see you. How are you?
I'm good. I'm a little hot right now. It's supposed to be 95 today here- Oh.
in the Bay Area. It's gorgeous New York September weather here, so we cannot complain. Oh, wow.
I'll swap you. In fact, Liz, how are you? Are you going to get out and enjoy the weather, or what's going on?
Well, we seem to have an issue from Liz's mic. So we will, yeah, I guess tell Liz to hop on and off, and we'll come back. Terry Robinson, how are you?
I am good and enjoying the weather, sort of. I'm actually having mostly to work from my roof deck because they're doing repairs in my house, and there's drywall everywhere. So if you want to breathe, you have to go outside.
It sounds like a rough assignment. I'll await the invitation for the roof deck party. Well, you are absolutely welcome to come.
Come work with me sometime and have a drink and whatever out on the roof deck. It'll be just like old times when we worked in an office room. That's right.
Exactly. All right. Well, I want to get started with Meta Muse, which is this AI agent that Meta has been telegraphing a little while now, but they've kind of rolled this thing out.
And they're saying to folks, "We're going to give AI agents," the very thing that we've been talking about on this show and other folks elsewhere that are somewhat problematic because they just won't do as they're told, and they kind of want to do whatever they want to do. Terry, what could go wrong here with an AI agent that we're handing to what might be, I don't know, millions of, shall we say, less than highly educated consumers when it comes to AI? Yeah.
So they're not so fluent in AI, I imagine, than most of the population. The good news here is it's only for people 18 and over. Glad we're not giving it to teenagers right now, though I imagine they'll find a way to hack in.
Yeah, there's just so much- I have my fake ID from high school. I don't understand. I have one of those.
I was from Arizona. A friend of mine had one that we all used then when we tried to get in the same club at the same time. That was pretty interesting, to stagger ourselves in the line and pass it back.
So anyway, that's the way that went. Yeah, so there's just, I think, a lot that can obviously go wrong here. After all that's happened more recently, we don't really have a grip on how to control these things, right?
I mean , does anybody feel comfortable with that? No. And then you have it in the hands of people who really aren't fluent in AI, and I think we're increasing an attack surface at the very least.
Mm-hmm. I think we could create some really bad situations where these agents go out and just do whatever. I don't want to say like War Games thermal nuclear war, but I'm starting to...
The bad things are going off in my head about this. As we were talking about this before it was released, I was wondering how much latitude it was going to have, and I think quite a bit. And there was something, too, in the story that I thought it could go beyond your basics to, what was the line?
I know I can't find it now, but not only executing real-world tasks, but it sort of marshals together your assets and your being and makes decisions on your behalf. I don't think that that's a good idea, and I really think this is going to be problematic, and I'd be interested, have they said more about what they're going to do to make sure this thing doesn't just go off the rails? I have no idea.
But Liz, my father-in-law is on Facebook, and he's also a retired data analyst, and so he still dabbles. And I can't wait for him to get this thing in, because eventually he's going to take down the entire IT environment for the Department of Labor as he goes after Maybe that's some sort of karmic justice on one hand. Is it going to just be a bunch of lone wolves out there now?
Is that what's going to happen? No, I don't feel good on both of us. There's a couple of precepts that I don't get about this.
Well, first of all, to your question about what are their safety controls, if I understand, they're saying that it's going to be executing in a secure- Oh. Not maybe getting into your WhatsApp. Oh, we've heard that before, haven't we?
We've heard that from Meta before. Yeah. Yeah, exactly like that.
Yeah. We've heard that from a lot of people. No.
The other thing is that it's going to be execute-- The platform's going to be WhatsApp, and apparently, eventually Meta Glasses, which we all know is the big privacy icon of their product suite. And so I wonder, will it compromise somehow, either intentionally or unintentionally, WhatsApp's supposed privacy if you're allowing all these hooks to open it up? I just see so many vectors for disaster, and that's even not taking Meta's record on privacy and safety into account.
Yeah. You think about the timeframe. This timeframe is really kind of unfortunate for Meta in a weird way, because remember last week they settled the social media addiction trial for $18 billion.
And then they make this announcement at the same time we have an Anthropic researcher resigning saying that AI's going to kill us all, right? So the thing about this, and again, I think this always comes back to Zuckerberg and what he wants, and he has this blindly ambitious vision of super intelligence, and this is part of that. Plus, he's also inhibited, and he's also pressured by his competitors.
So he's competing with Microsoft and Google on the consumer side to bring AI agents to Meta. And in his haste to do this, I think this is going to lead to all sorts of problems, honestly, in terms of privacy, in terms of data protection, security. I don't think they're ready to handle this.
Can you imagine the scale? Safety and everything else. Safety.
You've hit the nail on the head. It's like this rush, right? And he wants to stay competitive.
But this is always the way it is. Let's rush to it and then figure out the security. We'll fix it after we run it.
Yeah. Well, I think that's the whole doomer thing comes from, right? Is that it's like this runaway train and everybody's hopping on it.
Another thing is that since they're rolling out to consumers as sort of a equivalent to OpenClaw, which was basically a developer toy, right? And so what's going to happen when these users bring it into the enterprise, right? And so then you have these issues of corporate governance that are adjacent to their target audience, and there's so much shadow AI already.
Who thinks this is going to be any different? Well, and I just- This mirrors OpenClaw a lot, and Instinct as well. And you think about it, and Mike, I think, mentioned this earlier, and maybe Terry, when you have so many people who are not familiar with AI thinking, "Oh, I can do these personal tasks," it can really escalate out of control really quickly once the agent takes over.
Well, I have a question about that, John, and I'm going to punch this over to you, but did I read this correctly? They want like 200 bucks a month for this? What in God's name do I do with Facebook or any of these other things that they provide that I would remotely even think about using an AI agent for in the first place, and then paying them 200 bucks on top of that for the privilege when I have a hundred other AI agents out there?
I don't want to pass judgment on people who use Facebook because it is still extremely popular, especially among people over 40 or 50 I think people are so dependent on it, I think they're willing to do that. I think they might think it might be convenient, it might save them time and money. I'm not sure a lot of people are going to do it initially.
I'm sure Meta's going to drop the price. I can't conceive of anything remotely to that extent. Well, especially with all our prices going up on everything else, I don't think that people can afford to shell out 200 bucks.
But also, that might be within that population on Facebook, or other social media, that whole idea of fear of missing out also, right? It doesn't just happen with corporations and whatever. It's people there thinking they...
Sorry, there are people there thinking they... You wouldn't know I was on the fourth floor, would you? With the window closed.
But, yeah, I think people will be maybe convinced to use it eventually or maybe pretty quickly. I don't know. I don't know, because here's my issue.
So my kids are on Instagram. None of them go near Facebook, which they call it Boomer Book. Yeah.
Yeah. They know it better. Yeah.
And the only people that are on there that I know of are the folks that I grew up with and hang out with who are all my age. And while most of them are tech literate, a substantial portion of them could care less if there was an AI agent and will never use an AI agent. So I don't know.
Who is this audience that we're playing for? Or John, is this all for show, for Wall Street, just to show that I made- Yeah, that's exactly right. This is their AI play.
This is their agentic play. This is to mollify Wall Street and let them know that they are a player alongside everyone else, that they have some sort of plan. Maybe they-- I don't even know how committed they are to something like this, honestly.
Maybe if the revenue comes in, I think for them, they're going to try to figure out a way to monetize it through advertising. But, can I just mention one thing? When we think about all these companies, and it always comes back to me to this now, the more I think about it.
Apple, when we think about Apple, it's think different or whatever it was. Computer for the rest of us. Google was do no evil, whatever that meant.
Yeah. At the time, what it meant. Yeah.
Microsoft, excuse me, Mike, I keep comparing Meta to Microsoft, which- Microsoft. Metasoft, right? Yeah.
There's something that Zuckerberg proudly said that will always haunt them. " And I think this is just the latest example of that. And I think it's going to land them, and I'm going to make a prediction, I think it's going to land them in some sort of hearings when something goes awry, because this thing is like a train coming off the tracks.
In fact, if they're just doing this to soothe Wall Street or whatever, it sounds like a train now, I'm sorry. If they're doing this to soothe Wall Street or whatever, now we're looking at courting disaster for all the wrong reasons, right? Mm-hmm.
They're trying to make a play and be the Wall Street darlings or whatever they are, bolster their position, and now they're willing to open the floodgates and let all this really bad stuff happen. Yeah, they're- I know it's not so negative, but I just can't see- No, you're right, though, because as we said before about Meta, they've poured so much money into CapEx spending on AI, and they did this with the Metaverse, right? And they're always trying to rationalize or justify the amount of money they throw at these things before they abandon them, or try to figure out a way to reposition them or pivot them.
And I think this might be a case where they're playing to investors, but they're also trying to play to investors the idea that they're going to create some sort of revenue from this. Well, if they are, then is it going to be about our agent is the one that's transacting on behalf of the customer, so we get a piece of that? They'll be fighting over infrastructure stuff, where all these companies will be having turf wars over who owns what downstream or down stack of the users who are using it, who are clueless, but they are the ones that might have something to lose, or they are the ones that have something to lose when something goes wrong, as it inevitably will.
Especially given Facebook's Boomer audience, right? That aren't exactly-- Who's reading the EULA? And I'm sure there's something in that terms of service that's going to absolve Meta of any- Absolve is the right word, yes.
So wait, Liz, pick that up a little bit more, though. " Well, I think that issue is not necessarily specific to Facebook because there are other, like Perplexity and Amazon are in a lawsuit about who owns a transaction that is agentic transactions. And so I think they're just trying to get the low-hanging revenue fruit.
I think they're going to do what they always do, which is move fast and break things and just do stuff until they get stopped. And I think, again, because their audience isn't necessarily, or their user base that they're selling this to isn't the most tech-savvy, take the money and run. I don't know.
Maybe it's training data for their LLMs. I don't know. But I think suffice to say, they have a high tolerance for bad PR and legal blowback, so I assume that they're going to be aggressive.
So- And it's worked for them, too, this kind of reckless approach. Exactly. Until they paid the settlements, they kind of, in a sense, got away with it because Zuckerberg would just go before congressional committees and tell them, "We'll get back to you.
" And he's been doing this for a decade, more than a decade. So how can that be stopped? " Or- Well, Zuckerberg's a master chameleon.
He's so good at cozying up to the people in power and getting on their good side. Look at his new best friend, Trump. In a sense, he got on that train as early as anyone, and I think in a sense, whispering in his ear, or maybe shouting in his ear, "We don't need regulation of this.
" Do you mean the people who are spending millions on re-election campaigns that go through social media are friendly with the guy who owns the platform? Right. Pretty shocking, huh?
Shocking. No. Maybe if they move fast and break things, they'll finally break customer trust.
Yeah. But the customers always need something they have, or so the customers think. That's right, so that keeps that whole bit going.
But, do the fines need to get bigger? Because we've talked about these pretty big fines, but they're not so big for them. They're big for us, but they're not so- Yeah.
What are there, 3 billion monthly active users? They're still growing. They're not growing as fast as they used to, but they're still a tremendous platform.
And to his credit, Zuckerberg knows how to throw a lot of things at his users, and they eventually adopt them, even though there's initial resistance. He's done that throughout the history of Meta or Facebook. Okay.
Here's a new feature. And eventually they adopt it. I'm going to leave this here, but I would just remind everybody that if there's an AI agent on your machine that you don't own, then think about who owns it and why they put it there and what they're after.
And why they want you to use it. Yeah. There you go.
All right. Shifting the gear, but talking a little bit more about AI. But there was a significant alliance between Qualcomm and Amazon Web Services, and they're talking about $60 billion worth of potential chips that Amazon's going to use instead of theoretically NVIDIA chips.
And of course, there's a financing aspect to this. What a surprise, because it turns out that there's some equity being exchanged in return for this particular agreement. Liz, as you look at this, is this just one more in the great circular financing, or- Yeah.
The latest installment of the great AI investment circle jerk. I didn't coin that phrase, I just borrowed it. Wow.
But points for using it. They're very good. Yeah.
So, was it just a few weeks ago that Anthropic and AMD just announced something that was different, but the same, right? Mm-hmm. So this is just another example.
Instead of- Like I think the Anthropic and AMD one was that Anthropic was kind of subsidizing, or the other way, AMD was subsidizing Anthropic buying their chips. Right? In this one, Qualcomm is giving Amazon equity for the sale, in addition to scaling up their use of AWS.
And so I think that is good for Qualcomm and that it gives them a $60 billion foothold into a game that they've been a negligible player in. But I think it's just another layer in this house of cards, and maybe it's not quite as Jenga-ish as it appears, because there's real revenue that's happening here. But I just think it's so fragile, one misguidance.
Something happens with open source models and then Frontier Labs have a really bad few quarters and all this hard spending could just fall apart. So I think it's sketchy. I think it's good for Qualcomm, seems to be good for AWS, and everybody's hedging their bets with everybody else.
But everybody is so intertwined. Amazon is in bed with a whole bunch of other people, so there's a conflict here somewhere for sure, in some way. So I just feel like, unlike the Meta thing that's targeting consumers, this is just another group of Frontier Labs behaving badly or just being driven by greed.
" Can you blame them? No. The concept is so ridiculous.
You know what I would love to see? Maybe Alan could do it. I'm going to get this in his ear as one of his projects that he does or in special investigative projects.
So we get this $60 billion number, and we try to figure out actually what amount of money is exchanged between the companies, like what the real figure is, and do that for a couple of these, and also try to put together almost like an infographic that shows how all this is intertwined. Because in a sense, I think this is all overkill by at least a factor of 10 in terms of money amounts. Although I will say, I will give Qualcomm credit because in a sense, what they did with AWS, no matter what the amount of money actually exchanged is, it does establish them as an incredible contender in the AI inference- Right ...
and custom silicon space. So, kudos to them for that. But, in terms of the deals themselves, I look at the deals in terms of the partnerships, and I almost now gloss over the numbers because I don't believe the numbers.
Right. They're certainly high. Yeah.
They're all ridiculous. Yeah. Does this sort of move- It goes back to Stargate, going back to that.
These are nice round numbers that, isn't it amazing or coincidental how all these numbers are very neat and tidy? It's like 100 billion, 60 billion, 50 billion, 500 billion. That would cost you an IRS audit right there, all these- Yeah ...
nice round numbers. I'm just rounding up and down is what they'll say. So did- Go ahead, sorry.
Well, I was just going to say, does this move us away from NVIDIA just sort of owning the whole... In theory, right? So here's how it could play in a way that wouldn't upset the NVIDIA apple cart.
Right? So you've got Qualcomm, which has chips almost everywhere on the mobile side and has a lot of stuff where AI can run at the edge, and then theoretically, you're going to have a bunch of Qualcomm processors up in the cloud running larger AI models than they'll be up on AWS. And these things will be much more, shall we say, easier to integrate running across that.
And we'll have some federated and highly distributed models and some capabilities that go with that. And I wouldn't be surprised to see a lot of the folks that make endpoints based on Qualcomm suddenly cozying up to Amazon saying, "Hey, we'd love to be your partner and we're going to build some app and some service together," and away we go. The issue, though, is that Amazon, as far as I can tell, has no footprint on the client.
So, it's still going to be Amazon trying to talk to somebody running on top of a Qualcomm processor, so there'll be some integration work, which may not be any different than what NVIDIA tries to do when, assuming right now there's no NVIDIA client processor to speak of, but if they own the server side, they've got to go partner with somebody on the client side to have this highly distributed AI experience that we're all hoping for. That might ultimately reduce the amount of data center capacity we need in obscure places like Texas and Idaho and wherever else we're building these things. I don't know.
Liz, you think that could play out? I don't know. Sure.
Yes. It could play out. But I think that before anything starts hitting the data center, I think there's going to be a big economic explosion.
I think the floor is going to fall, and we'll see what happens. I think I brought this up previously, like some Depression-era event, and I got poo-pooed. But I just think there's just too much money floating around, too much overlap and overlay between all the vendors and their deals.
In January, Bloomberg did do an infographic of the circular investment trends. Mm. But that was in January, and so much has happened- It's changed ...
since then. Yeah. This deal is not the latest in an ongoing series of deals where I think the complexity is going to make it really hard to figure out what you were saying, what's really being spent over how many years?
What are the terms? Yeah. Liz, I just become numb to this stuff after a while.
I'm just like, "Yeah, whatever. " I don't believe it. But going beyond this, I don't necessarily disagree with your assertion of some sort...
I wouldn't use the word depression, but I would say a severe recession of some sort. We've been hinting at this for- Economic calamity, yes ... several weeks.
Economic calamity, yes. So the counterargument to that is they'll sit around and they'll say, "Well, look at the backlog for demand for these processors," and it's huge. And so therefore, we have these billings, and those billings are indication of future manufacturing capability, and away we go.
And that's what justifies the whole Wall Street hoopla. " Right. The secondary markets.
Right? Like, "We'll just buy a bunch- Right ... " So you've got to wonder what's going to happen here and what is actually happening, and shall we say there is some additional scrutiny there, John?
What do you say? Absolutely. I think part of the story, and I think James McGuire wrote this story, is that Broadcom has this goal data center push towards a $15 billion revenue target by fiscal year 2029.
So when I look at these agreements, like the announcement for Meta that we talked about earlier, this is all with an eye towards the Wall Street investor community because I think as each quarter moves along, there's going to be more and more scrutiny over the revenue that's being brought in versus the amount of money that's being spent. And I really do think there is going to be this obsession with prove to us that all these investments, all these numbers you're throwing at us are verifiable, that they're actually warranted. And I think right now it's just almost like this accounting game that they're playing.
Right. Ball game. Right.
This is Enron of- Yeah. Yeah, totally. Well, there are weaknesses in the bond market where a lot of the money comes from these days to fund these data centers, so that may be something that everybody should keep an eye on.
Yeah. Didn't Treasury just make an announcement today? Yeah, they're going to purchase up to $6 billion of the- Right ...
yeah, long dated debt. So think about that. Is that a subtle way of the United States taxpayers subsidizing the AI market, at least the funding for the AI market, which will come through that bond market.
And at some point, maybe we are at a point where AI has become too big to fail and it's bigger than the banks and everybody's- Yeah ... 401 is wrapped around- Yep ... these AI investments.
So maybe, it will be the United States taxpayer that's on the hook for all of this, like it or not. What do you think? I think that's right.
Well, do you think also just on the positive side, there might be a positive for some of the security stuff? Like that Dragonfly C1000, isn't that built to sort of secure AI workloads and- Yeah ... just to throw out a positive not to ruin the- No.
Somebody should ... How dare you be positive, Tharja. Stop it.
Last time I checked, the AI agents weren't paying attention to anything, so great. " Having religious experiences, yeah. We will happily wave at your safeguards as we drive right by them.
Well, I believe that's roughly what's going to happen, but- All right ... at least there's a safeguard out there on the horizon. Yeah.
Well, we were talking about this on the show yesterday, and I got to say, the one thing I love about these AI agents, they're kind of like, "Let me get this straight. You built a technology that's going to come along and rip apart my entire company and invest its AI into all my systems and take all my data. " Not quite clear that's how that's supposed to work.
Well, possibly that model will change. Don't know how, but you know. Well, what the hell?
Let's go for the 10 yards here. So maybe, using the circular financing metaphor, maybe the AI people should be paying the companies out there to let them use their AI because, well, after all, we're paying customers now to use stuff, so everybody should take that AI money they have and just invest it in their customers. We'll just create this whole- Imagine that.
Imagine a world where that would be true. I mean, god, Nvidia takes 20% stake in Ford because now Ford's going to use AI agents everywhere, right? Right And spend money to secure its systems, I think.
It's convoluted, but- Very convoluted, but it just might work. Just might happen. All right, speaking of convoluted, we're going to shift topics here.
All right. So there's a new report out about what's going on at Oracle, and there's some indications that there's some weakness in demand. Hiring customers that have been stalwarts from Oracle for a long time are a little put out, and Oracle, meanwhile, is getting called in front of Congress about what's going on with some of their projects, such as the DA.
And of course, Oracle is spending a fortune on AI data centers, and no one's quite clear if that investment's going to pay off because while everybody else is investing in AI data centers, so it's not clear what makes Oracle's any bigger, better, or badder or different. John, you're out there in the valley. What's going on with these guys?
So we saved the worst for last today. So- ... Oracle, oh my God.
So I live in the shadow of Oracle actually, unfortunately. Two things going on that you mentioned, and there's actually a third that I'll get to. So our friends at Enterprise Technology Research did a study into Oracle, and they found that there's a, as you mentioned, Mike, a broad-based pullback in spending because of large reductions among its largest corporate clients.
They do something called a net score. This is what ETR does. It's a key metric that measures customer spending intentions, and they are in the negative range, and they have been declining year over year.
And this has been going on and accelerating the last couple of quarters. So one of the overall numbers that ETR found was 20% of enterprise customers intend to increase spending with Oracle, but those gains are outweighed by 17% who plan to decrease spending significantly, and 9% who intend to replace Oracle products entirely. At the same time, and I'm not sure if this is-- In a weird way, I think it might be related.
The House Veterans Affairs Committee has voted to, not unanimously, to subpoena Larry Ellison and the CEO of Oracle over massive cost overruns and delays- ... over our VA's electronics healthcare records overhaul, and there was an initial estimate of $10 million to a projected lifecycle cost of $48 billion. In any event, Oracle's being subpoenaed because they didn't agree to a scheduled oversight hearing.
They said they couldn't make it, or they had some sort of disagreements with the panel set up. It's just all kinds of they're losing some of their larger customers, or their larger customers are decreasing. And at the same time, they're being looked at for inflating costs of a contract.
I don't know, maybe that compensates for some of the lost business. That's just total conjecture on my part. Here's another thing.
Here's a third thing that I want to mention. Larry Ellison's personal fortune, I think, has been shaved nearly in half- Oh ... the last year.
It was an ungodly amount, like 400 billion to something like 220 billion. Those numbers are not exact- Oh my God ... but it's in that neighborhood.
Mm-hmm. Their strategy's not working with AI. The data center strategy, I don't know where that's going.
We hear about pullbacks and about pipebacks. They're laying people off. I think it's very interesting and bears looking at with more scrutiny because this is not the end of their travails.
Right. Larry might be sliding cash out the back door of his personal fortune to help buy movie studios, right? So- That's true too.
That's part of the equation for sure. Not as possible, because I was going to say it's very unusual for somebody in Trump's orbit to lose their personal fortune, even a penny of it. So let me tell you a story.
I've been covering and dealing with Oracle for 30 years, and I can remember when they were a brand-new relational database company, and they were competing with Ingres. Okay. They grabbed a bunch of meanie guys and threw them on a supersonic jet, and we went for a trip to nowhere.
We actually left JFK, flew around through the Bahamas, and came back and landed at JFK. And it was all just to prove the fact that their relational database from Oracle was really, really fast. " And it's been that vibe ever since.
As long as I can remember, there's always people out there who are on the IT side or the customer side that are always ticked off at Oracle, but nothing ever really fundamentally changes. They keep buying it. They're either locked into the data where there's no good alternative.
So John, 30 years later, is something different afoot here? Do people have options that they're actually going to go and execute? I think they do.
I think before they were locked in, as you said, because remember, they were innovators. They were the kings of the relational database. Larry was on top of the mountain.
I'm trying to be careful what I say. They were tolerated, but they were not liked. They still are not liked here.
They were arrogant, but they had a home field advantage that they took advantage of and ran as long as they could. And now I think they find themselves in a situation where the chickens have come home to roost. There are other alternatives, and people will gladly switch to those other alternatives rather than deal with Oracle.
I think Oracle's living in the past. They have been for several years. Larry Ellison's still the chief technology officer.
I'm not quite sure why. Maybe because he still wants to be in power. It's like the Vince McMahon- Pride ...
thing. Pride. Same thing, right?
Exactly right. That's his whole identity is based on that. And I'll tell you, Mike, that's interesting you mention that story about the flight to the Bahamas and back.
I remember a few years ago, Ellison has an estate at Woodside. It's like a Japanese compound. And I remember they invited some of us to this event, and it was all for ostentatious show.
It was all to show us how powerful and important he was. It was all a complete power play. They used to do this all the time at different events.
They had an event at his home in San Francisco for a book event for his ex-wife, and I always thought, I'm going to go back to a really old hamburger commercial, where was the beef? Where is it? Right?
I mean- It's all flash ... they're being exposed. Yeah.
I think there's two things at play here One is the open source community around Postgres is a lot stronger, and you hear and see a lot more deployment of that database as an option. And so especially for new workloads, you're also seeing other classes of databases, whether they're some variant of NoSQL or a document database, and a lot of workloads are being dropped over there instead of a relational database. " And as such, then they make their investments elsewhere.
On the other side of this, Oracle has this portfolio of ERP apps that they've been pushing for a long time, but SaaS applications eating away a lot of that base and, or at least limited that opportunity. And I think in the age of AI, a lot of folks are trying to figure out, well, what is my ERP app? And is it just going to be some sort of system of record that sits in the database and therefore am I paying a fortune for an app that I don't really need?
I don't know. Can I mention something that was really jarring to me? Was several months ago, NVIDIA had an event in San Jose at the SAP Center.
That's where the NHL team plays. It was packed. They turned people away at the door.
There were, I'm going to estimate, at least 15,000 to 17,000 people there. I went later that day to a Java conference on the Oracle campus. There may have been 100 people there, tops.
I went to a reception for reporters. Most of the people there were Oracle employees. I'm just thinking, to me, it was such a stark vision of reality in this day and age of AI.
And I know it's just probably an unfair comparison, but to me, it spoke volumes. Just for the record, SAP has a longstanding relationship with the Sharks because, well, Hasso Plattner was one of the early owners. I don't know if he still owns it or not, but they've been trotting out the Sharks as a customer of SAP for as long as I can remember.
Gotcha. So I'm always like, "Don't you guys own these guys? " Yeah.
There you go. Here we go. It comes back to circular financing.
I used to live walking distance from that arena. Yep. So, but what Oracle's going to take away from what John said is we just got to go buy a hockey team and everything will be fine.
Well, so- More streams ... I remember interviewing Larry Ellison years ago, and he told me that what his plan was with the NFL is he was going to buy a team when LA didn't have a team, and it was going to be called the LA Stars. And he told me this was a plan in motion.
It never happened, of course. But he also tried to buy the Warriors. He overbid for the Warriors, and Joe Lacob, who didn't bid as much, got the team.
And the reason why is that the league didn't want to deal with all the histrionics and the mercurial nature of Ellison. But going back to this article real quick, isn't another, they were talking, didn't Oracle agree to cover any cost overruns that was part of the contract, and then they were reneging, and then there was all this being a Trump bro or whatever, that anybody else probably would've been held to the terms of that deal, because even Republicans are questioning, well, it's 17 billion. How come the taxpayers, I thought you said you were going to cover it?
But somehow they're not covering it. It is going to be passed on to the taxpayers. So what's that about?
So just for grins, what percentage of the dollar in my pocket is going to fund AI as this evolves? I mean, is it going to be 10%, 20% of the United States value of currency gets tied to productivity and GDP and all this stuff at the end of the day? So, have we tightened this so tightly per se that no matter what we say here, we're in, right?
Yeah. Especially, I don't know whether another AI overlord like with Doge has anything to do with that health data, right? But it's veteran, it's federal data, right?
Right. Wow. So what happens when something goes wrong, right?
Why... I don't know. It's crazy.
Yeah. The government is adopting this stuff faster than anybody, as it turns out. That's what I- Because remember it was laying off thousands of people ...
captured last week from- Yeah. Yeah. I've seen that from Broadcom.
" There's the key in Doge right there. There you go. But at the end of the day, it's pretty clear that the United States government is very now tightly dependent upon a small number of what would be called private public companies, and these relationships seem to be getting deeper and deeper.
And unless there's some new massive technology curve that changes that equation, doesn't look to me like we're going to be able to back out of this no matter who's in the White House next time out because we're just too far down the path. Yeah. Well, and is there a hope of increasing oversight?
Because I don't know that that actually works. Not anymore, no. No?
Not in this administration. The weird thing, the irony is that AI is taking over our lives and frightening us all, and yet we're also paying to help subsidize what it does to us, whether we like it or not. Well, if Oracle becomes the Bear Stearns of its time, I won't be sad.
I don't know how I feel about that. Wow. Nice.
Yeah. Wait, and then who's Lehman Brothers in this play? Yeah.
But at the end of the day, too, it can go the other way. Think about it this way. So the United States government is so deeply invested in all these companies, and if you have somebody who may be is running things who's a little more aggressive about regulations, well, now the United States government will be in a much better position to enforce those requirements because, hey, it's a shareholder.
Yeah. I think actually that we're laying, we're moving in that direction, actually. So I'm going to give this show a lot of credit for being ahead of the curve in terms of what we- ...
seriously, I mean, I've said this before, but we predicted and said what was going to happen with the data center pushback, and that would become part of the midterm discussion. And I think actually what you just mentioned, Mike, is also going to become a next logical step, right? Mm-hmm.
If we're tied into this, shouldn't we have some sort of oversight of what's going on? I mean, that's inevitable. There you go.
So the question early, I guess we'll end it here, but yeah, it's a lot like a lot of things in life. Be careful what you ask for. Mm-hmm.
So, if the Republicans are out there saying we're going to get rid of the filibuster, there's some Democrats out there that are just fine with that idea. Right. Right.
And the same- So, write that on ... and the same thing is going on with all these investments in private companies because there are some Democratic socialists out there that are probably sitting in a back room going, "We're just fine with that idea," because think about it. That's what we think.
All right. Hey, I want to thank everybody for sharing their knowledge and insights as always. Guys, you guys are awesome.
I want to thank everybody for spending time watching today's show. Please stay tuned for the rerun of the Techstrong TV lineup, and of course, hopefully everybody, we'll see you tomorrow.



