OpenAI Cuts Prices, Huang Calls Out CEOs and Cybersecurity Hits a New Scale

Hey everyone, happy Tuesday. I was here fixing my sleeves, and all of a sudden, we're on Techstrong gang. Man, this week's flying by.

I don't know if AI is moving ahead our calendars on us or what it is, but time is just moving. We've got an interesting show today. Of course, a daily dose of AI information as well as some other stuff, and we've got a great gang to talk about it.

Let me introduce you to who we've got here. The man from up north, Chris Blask. The man from out west, Jon Swartz.

The lady, well, she's from up north too. She's a Yankee, Kate Scarcella. And well, he's a Yankee too, but he's still basking in the glow of the Knicks' Mike Bazark.

Gang, welcome to Tuesday's Techstrong gang. We've got a lot to go into. Mike, I'm happy to see that the laws of physics don't change, though.

Even in the age of AI. Maybe in quantum they will. Do they?

Because the whole thing leaves me scratching my head. But let's just jump into this. So there are reports that OpenAI intends to start something that, I guess you could call it a price war, and they want to gain share, and they're in a battle.

But there's also reports out that they spent $34 billion last year generating $13 billion worth of revenue. God knows what exactly how much of that turned out to be anything that looks like a profit, but it looks like there's a big gap between 13 and 34, so I'm thinking that they're probably on the loss side of that thing, and they continue to spend money on R&D, and so there's some question of how sustainable it is that they might actually threaten a price war. That said, you have an article, Alan, talking about how well the cost of AI might continue to come down, and there's a lot of open-source alternatives starting to emerge, and this whole focus on the cost of tokens might be misplaced because, well, they're going to be dirt cheap commodities.

Make sense of these two things. Well, look, one way of competing with Anthropic is for OpenAI to lower their prices because Anthropic has kept their prices high and continues to raise them. Another way may be a rumor I heard that OpenAI paid the guy at Amazon who reported the jailbreak to the government- Wow ...

which caused them to shut down. Nah, just a rumor. I don't know.

I was making that up. Just making that up. I was going to write a story about it, damn it.

Yeah, but that being said, I made a reference to the laws of physics. When you've been in technology as long as some of us here on this panel have, you get to recognize patterns. It's like we're AI ourselves, right?

We're pattern recognition experts. And this is what happens. Things, when they come out, there's a high demand and prices rise.

Over time, though, for these things to get to scale and normalize, generally, they move towards the great commoditization trough. Right? And they become commodities.

AI intelligence seems to be no different. And even more than that, in most technologies, it's following more of what I call the public utility model. Right?

Where AI becomes the great enabler, much like electricity was the great enabler of countless... Think about General Electric, 500 companies that really had nothing to do with each other. You know what they had to do with each other?

Most of them revolved around using electricity. But think about all of the industries electricity has spawned, right? And the thing about the electric utilities is they're solid businesses, they're regulated by the government, and they're allowed to make a reasonable profit.

Mm-hmm. Reasonable profit doesn't give you multi-trillion dollar valuations when your business is costing $50 billion and throwing off five, but reasonable profits. But the Sam Altmans and the Dario Amodis, and the AI tech bros are not satisfied or going to be satisfied with reasonable profits.

Mike- Go ahead ... I was going to say, I'm sorry, Alan. The reason why I was going to mention about OpenAI, the reason why they did it, I think, was more motivation because Anthropic's going to do it.

They're going to drop their prices, and they're killing it in the enterprise. They're all going to do it, John. It's the way of the world.

That's what I'm saying. Yeah. They have- Because they're still amassing customers, and it's also the way these things go.

It gets cheaper as it moves towards commodity status. But the real question is, okay, let's say it gets down to commodity. Let's say even that the government or governments step in and declare it some sort of utility with some reasonable profit margin built into it.

That's not going to be enough to keep Sam Altman happy. It's not going to be enough to justify Anthropic's trillion-dollar plus or Elon's $2 trillion market cap. They've got to do things that are going to be higher margins.

So lowering prices and making your product a commodity is generally not the way to make higher profits. Yeah. But I think also they're acknowledging this, what we're calling it, what, a severe corporate budget burn?

So these companies are-- They just refuse to spend, and it's become a huge issue. Altman said it himself. So they have to acknowledge that in the short term to keep their customers and not lose them, and there are examples of the founding of it.

It's not a refusal to spend. They spent their whole year's budget in three months. Right Uber did in four months, yeah.

Four months. I have a question. So here in New York, I have a utility, it's called Con Edison, and I got to pay them, and there isn't an alternative, really, so I'm locked into that there utility.

Is that really going to hold here in the land of AI, or is it going to be that there's going to be multiple utilities, and I will dynamically select from each one as I bounce around from one AI model to the next? Or am I going to sign something that feels like an enterprise license agreement and I got to consume a billion tokens by the end of the year? Think of cable TV.

But Chris, you got your hand up. So you guys, the utility model, and we've talked about this in the last couple of weeks. This is utility-grade spending.

It's $100 quadrillion or whatever, paving the earth with what, tokens and compute? No, it's infrastructure. Right?

And to answer your question, Mike, even in the grid, the evolution of the smart grid over the last couple of decades is a fascinating analog because you actually can buy power from someplace that isn't running a wire to your house. And, look, we've talked about this week to week on this very episode on every Tuesday, right? And we're coming down this path where the IPOs are going on, and we're looking at the valuations of.

SpaceX is ironically, being in space. That's one where I see physical infrastructure. It's lift, it's orbit stuff, there's revenue there for lots of reasons.

This lays on top of that, but a centralized model, Mike, the models don't matter that much. What matters is what you do with them, right? And that's still unwrapping, right?

And this year, four months using up all your annual budget because last year when they were making budgets, they didn't know what we all should know now, right? Mm-hmm. The value is not in models and tokens.

The value is how much work can you get done, actually done per task. I have another question. Go ahead.

Will the AI providers of these models take a look at my prompt and determine that maybe they could fulfill it using somebody else's infrastructure less expensively, and therefore they'll just reroute that prompt to wherever the most efficient place to do it is? Which for all I know, could be in the middle of South- So theoretically- For all I know ... so we use Perplexity Computer here- Mm-hmm ...

a lot in the office. Theoretically, that's what Perplexity's doing. It's picking the best tool for the job because on the back end, it does plug into Claude and OpenAI and Canva and a bunch of others.

So it's using the best tool. So that's already happening. But you know what?

Who was it the other day who was talking about Si Sims? Was it you, Mike? Yep.

Educated consumer is our best customer. I think it's up to the customer to figure out. I do it, right?

I say to myself, "Is this something I really need an agent for, or is this more of a generative task? Is this something I really want Opus for, or can I get away with Sonnet? " I try to pick the appropriate model for the job.

And we have people talking about harnesses, right? That's kind of the notion behind the harness is, look, I've got a team of these things. I'm going to pick the right one for a given job that's most efficient.

But make no mistake, in order for AI to reach the pinnacle of our dreams, or some of our dreams of what it can do and how it's going to transform things, it has to be ubiquitous, and it has to be affordable. Yes. And so the price war to me is not the big story.

That was inevitable. Absolutely. And if it's not OpenAI clashing with Anthropic, it's those two clashing with open source models or models from other geopolitical areas.

Right. But in any event, intelligence, that base intelligence of a large language model becomes ubiquitous and cheap. It's what you do with it that counts.

You know what I mean? Yeah. And that, I think, is where Sam Altman and Dario Amodei and the tech bros want to be.

They don't want to be Con Edison. Con Edison isn't sexy. Right?

They want to be General Electric. Kate wants to jump in here. Yeah, what, Kate?

No, I actually loved your article, Alan, because I do think that it's going to be a commodity. This whole AI stuff is just going to be commoditized just like our utilities and everywhere where that's going. The business that wins is the one that will take this AI and actually generate some really exciting business models.

I think that this is a time to be really excited about, and I think I'm so tired. I almost feel like they're whining. The Sam Altmans and stuff.

Like they're just such whiny little punks. Just at the end of the day. Kate, that's his reputation.

He likes to whine a lot. Really? Because I'm just like, what the heck?

Sorry. I was going to use another word, but I won't. This is infrastructure.

Let's use it as infrastructure, and let's really start to build remarkable technology and new ideas from what... Think about what electricity did for us. How many exciting businesses were created because of electricity.

That's what I think. I think this is such a new ... new time period, and we should be excited about it.

And I'm so tired of these tech bros, man. Wait a minute. I'm tired because I'm tired of talking about models, right?

You use the electricity-- We're arguing whether AC or DC is the big thing. I want to talk about refrigerators. Yes.

Things that go in people's home and do things. Right. The most exciting thing of my week right now is the Standards Council of Canada working group on systems of AI systems and AI innovation.

And you get this great group of folks, getting together, looking at how you put this together. And Mike, your answer is no, none of this works if what you said could happen. If your inference can be routed somewhere you don't know about, none of it works.

That's like saying we're going to use DC for long-range transmission and power grids. It doesn't function. It's not about a good or bad.

And I get talking about whose model has the most training and whether Mythos is more dangerous, I don't care. Right? What's interesting to me is how all these things go together when we assume we're not going to act out dystopian things that literally won't function, right?

Right. But that's where most of our conversation is wrapped around models and- But let's take for given what Kate says, right? We are going to use this new electricity to invent the next television, radio, refrigerator, you name it.

All of the above, and then more. Yeah. Right?

That's great. Who works on those? Is it the same company that's supplying the juice?

No, probably not. No, I don't think- And I think that may be what keeps Sam Altman whining and Dario Amodei- He sees where it's going. I mean, we're in this transition, as Chris put very well, from model capability towards maybe affordability, utility.

It's not just OpenAI and Anthropic. Google just made its budget AI subscriptions plan a lot more affordable, the AI plus. Yeah.

No, there's no doubt. This one is going to go on- But that's the way these things go, right? Yeah.

I got a subscription- As you scale, they do get cheaper ... " Assuming someone's going to pay for the $8 trillion in data center build-out. I see an explosion, the age of innovation sort of thing.

That's what I see. Yeah. Anyone can do this, frankly.

It is a utility. Anyone can use AI right now to achieve all sorts of things that were inconceivable before. So, I think, look, I try to come on shows like this and not downplay the big players.

They have trillions of dollars. However, are they the rulers of the future? I don't think so, unless they're really smart.

I don't want to have to think quite as hard as Alan does about whether to use an agent or whether to use this versus that. I kind of want- Mike, let's face it, could you think quite as hard as I do? Yeah, I know.

And- I'm kidding ... and of course, since we're on that subject, yeah, one of us engages more in overthink than the other, for sure. Yeah.

Yeah, for sure. Well, look, I got nothing better to do. But I digress.

I got nowhere to go. I just want it to sit, whether it's from Apple, Microsoft, or somebody, I just want a little widget that sits in front of the thing and says as I'm typing in, this is what I want the AI to do, that it will recommend to me which one to go to and what to do and what's- So, but I think you're on the wrong track there, because I think the more likely scenario, Mike, is you're going to have Mike's model. Hmm.

Right? That part. That's tuned into Mike.

Right. And it may not live out there. It may live on your phone.

It may live in your house. It'll live on your watch. I don't know where it lives.

But it's- It's going to be your best friend ... Mike's model, and 95% of what you need, Mike's model's going to do. And then for the other 5 or 10%, maybe it does go out into the public grid and grab stuff.

But that's where I think we're going with this, right? I want to be clear that I didn't pay Alan to say that, because that's exactly where I really is. Well, no, Chris, but let me hear you say Alan's right.

Alan's right. That that's exactly where we're going. All right.

That's all. Okay. Alan can say it.

Alan's right. You will have a little box in your house with all your stuff on it. And saying it too.

And it will go out and figure out what to deal with that. You don't ask Apple to tell you how to do these things. No, you're not.

You can have Mike's model. You have your little box, your little appliance at home. Yes.

It does your stuff. It's called a node. All right.

But I don't come on here to push stuff. Assuming that that's how that works out, will we therefore combine and mix and match our personal models with our friends and families to kind of trade a little? Oh, absolutely.

It's called mesh canon in our words, right? It's sharing slices of your own canonical stuff, like putting it inside your node, to Alan's point. You'll have a system at home that's your stuff, and you work with Textron, so Textron has its AI stuff.

There'll be a slice of that you bring home as part of your own stuff because that's part of your world. That's how we share things, not go out to some central repository and ask Apple to tell us what Alan thinks. That's baseline intelligence.

Anyway, hey, speaking of baseline, I don't know how intelligent I am or we are, but I do know we're over time on this segment. Yeah. And we got to move on.

Mike, what's our second-- We'll leave this for now. We'll come back to it, but what's our second segment today? Well, our second segment is Jensen Huang said out loud what many people have been thinking and speculating on for a while now, and he basically said, "Hey, all these people who are laying folks off are, well, it's just lazy Business management to begin with, and a lot of these folks were over-hired and people built empires before COVID.

And then there's an interesting little survey from the Linux Foundation to back this up, but it basically says that just about everybody in the world is looking to hire IT talent, unless you happen to live in the Bay Area, where many of these companies happen to be. So, Jon, I know you talk to folks all the time who are looking for work, but maybe they should just move to another- They should move. Yeah.

Or they- ... or maybe they could work for a company on the East Coast remotely. So I'm going to editorialize a little bit.

Kate, I'm going to be speaking to these tech bros. It is very maddening, but I got to say, based on what Jensen said and some of the stuff that Satya Nadella has said, there are some voices of quasi-reason starting to surface. And I think what Jensen is getting at, and it's so obvious out here particularly, Nvidia doesn't have a lot of layoffs.

They're hiring. They're pretty smart about how they do this. Meanwhile, we look at other companies, oh, let's say Meta or Oracle or Salesforce who are out here, who in a sense did a big buildup of hiring, then they slashed like crazy.

Amazon does the same thing. What they're doing is they're blaming AI. They're using it as a scapegoat for their over-hiring.

So they're framing the layoffs as this technological evolution that allows them to present downsizing as this forward-thinking strategic pivot, rather than a failure of planning. And look no further than Meta. There was just a story, I think, in "Wired" where they got ahold of a memo where the CTO there admitted their AI reorg was atrocious.

And they did a terrible job of that, which they are very apt at doing, by the way, when they go through their waves. And I'm glad that Jensen said this because he carries a significant amount of weight, and I'm glad that he's challenging this growing corporate habit of using automation as a scapegoat for broader financial missteps. And- But it gets better But he could have well understood-- I'm just going to point out one quote he said.

" Well, there's that. But it gets better because these same companies then go visit Wall Street and investors and tell them how great they're managing their companies because they're using AI to eliminate all these people as part of a, shall we say, bold-faced attempt to boost the stock price. And Wall Street falls for it every effing time.

Just like they fall for, I don't know, a deal is imminent in the Strait of Hormuz or- ... oil prices are going to go down. You just tell them what they want to hear, and they'll buy.

And then the next couple of days, they'll be thinking soberly and thinking, "Oh my God, we got to sell. Like, what the hell? " And it happens every- Fools and their money ...

freaking time. Fools and their money. Yeah.

I'm sorry. It p****s me off. Anyway, I'm sorry, Alan.

Go ahead. But it's fools and their money. But look, there's a lot of things we could talk about.

The so-called Iranian deal, where Iran will get their billions back and who knows what they're doing with nuclear, because as far as- Maybe they'll get phones ... our government's, it's done. It's a done deal already.

They don't care what happens. " We could do all of those things. But let's bring it back to the subject at hand here.

Right? And Mike, what is the subject at hand? Leaders.

The subject at hand at the end of the day is, it seems to me, even in the age of IT, that, or the age of AI, that IT folks, software engineers, and all the people who make it work are more valuable than ever. And still in demand. But here's the other thing, though, all kidding aside.

As far as I could tell, these so-called AI layoffs fall into two buckets. One is no one wants to admit they hired like drunken sailors during COVID. Yep.

It was during COVID that they hired- Mm ... we hired like mad. All remote workers.

We didn't even know these people had two and three different jobs. I know several people who were collecting two or three full-time paychecks. Some of them worked in North Korea.

Just saying. Yep. That's number one.

So there is that correction in play. And then I think another big correction is, I call it the Oracle phenomenon, which is they want to be an AI data center player. They want to be an AI infrastructure player, and the only way they can afford it is by cutting humans.

And so they cut 30,000 people, not because they're saying those jobs are being replaced by AI. They're being replaced to pay for AI. And the question is, is that going to pay off?

And even are the clerks down on Wall Street smart enough to see, oh, what the hell's Oracle going to do with... Were they that fat that they could afford to cut off 30,000 people to build a couple more data centers? " Mm.

And- You know ... there is that. So that's the other bucket.

And speaking as a CEO, and we have all of 25 people, so it's not quite at Oracle's level. But I see I'm not replacing people with AI, I'm enhancing people with AI. And I'm reminded of my conversation with the CEO of Automation Anywhere, who said, "Anyone who tells you they're laying good people off because you want to replace them with AI is crazy.

" Not 10X, mind you, but 3X those people with AI. Chris, you got your hand up. I'm glad you ended with that because I was going to interject and say, this isn't just about markets and market cap and economics, right?

This is about leadership, right? Mm-hmm. Now, people are out there saying I am not a very good manager in the first place, so I can replace the people that I'm not managing well with AI that I don't manage well.

Right? And that's really the dynamic we'll see play out. Right?

And in these big companies that we talk about on our shows like this, yeah, they do it to do market manipulation on their prices or whatever, cover up mistakes. But this affects everybody. Right?

And there's more people watching the show, hopefully, than just the Fortune 500. Right? Management actually matters.

Leadership matters. And this is not so much exposing or causing leadership problems. Exposing leadership problems already exists.

We've all seen office space, right? If you do not manage your people well, you're at risk. And if your company has existed for 30 years while you've managed people well, this may be what kills it.

If you can't get human leadership and efficient to the point... Again, we're a little company too. We're not laying off anybody.

" But then again, I believe in actually human leadership and so forth, and frankly, a lot of corporations just don't. I am sympathetic to these companies who are trying to kind of fight their way through AI or adopt it. It is a hugely changing situation we're in.

But there are certain companies, and I'm going to pick on Meta again. They did the Metaverse. Remember when they lurched in that way?

They hired like crazy, and then they decided, "Oh, wait. No, we're going to become an AI company. " Just this rationalization that these companies go through.

And you know who always suffers are the guys at the bottom, the guys, the men and the women who lose their jobs. And I just talked to a friend of mine yesterday who worked at Amazon for eight years, and he said, "You know what? Once we made our pivot to AI, I was toast.

I'm gone. " And that's just the reality and it sucks. And you're right- John, that's Amazon, not AWS Yeah.

I'm asking. What, sorry? Is it Amazon or AWS?

It was Amazon. He worked within Amazon. So guys, I'm glad I asked that, and I'm glad you answered it, John, because that's a pivot here.

" No, he worked on the, without revealing too much about him, he helped develop their book sales, their book department online. " I think this is where you see something we're not talking about, which is the effect of AI on non-tech jobs, on non-code developers. You got a business like Amazon, they were already using robots in the factories to grab items and box them to ship and everything.

You're going to see these companies increasingly use a knowledge skill kind of roles with AI as well as physical as that becomes more developed over the next two years. And I don't think the focus is going to be on how AI killed the IT jobs. I think the focus is going to be where it belongs, which is how is AI disrupting and rippling through the broader economy?

Yeah. I think because we're in IT, when you're a hammer, that's the bubble we live in, and so that's what we look at. " And that may not have anything to do with the AI other than the fact that- Sounds like Tuesday to me, Mike.

That's been going on. Don't blame AI for that. Yeah.

Well, AI is a scapegoat for that, too, right? Right. It accelerates the- That's the thing.

It's become the scapegoat ... it's more acceptable to do that now. Yeah.

Mm-hmm. Terrible. And again, that's an indication of leadership, right?

If you don't understand you're running a company with people, and over years they change, and you want them big and valuable over time, then you replace them with younger folks. And when that works profitably, I see markets that have opportunities for others to change the model. The Logan's Run theory there, huh?

If you're of an age to remember Logan's Run. Farrah Fawcett was in that movie with Ryan O'Neal. Hey.

Hey, hey. I was actually- Michael York What, Mike? That was not Farrah Fawcett.

She was in the second one. Jenny Agutter who- Oh, Jenny Agutter. Another name from the past.

Erich were both in London. Yeah. I see that glint in John's eye.

I'm not quite sure what it means. I'm sorry, Kate. You go.

I'm sorry. Kate, go ahead. And I didn't even need AI to pull that up, so there you go.

I'm sure you didn't. We'll talk about it later. Kate, what do you have to say?

Well, just include me in that conversation for sure. But anyway, usually I feel like I'm the Pollyanna in this show, but I have to tell you, these articles that you brought up, Mike, I feel like again, it's just whine, whine, whine. Everybody's whining about this.

And I'm glad that they brought up this. It is lazy leadership. And I'm so tired of hearing about what people are having issues with within their company.

It is a lack of this leadership. And if they can't lead right now, just go home. We don't need their weight on us.

This is an exciting time to be alive with all this phenomenal technology that is just about to-- We are about to give birth to so many new things. This is an exciting time. Yeah.

And if you don't want to be a part of this, go home. Just go home. You know, so Kate, it's funny.

So I was in Las Vegas last week, and my wife and I went to dinner, and unbeknownst to us, we were sitting next to the Rubrik table. So the CEO of Rubrik was there- Vipul, who I know pretty well, and I thought about all the companies like his. We spend so much time talking about five or 10 companies who demand or command all this attention.

But for all those companies, there are scores of others who are doing the right thing, who are led by good leaders like Vipul- Yeah ... who are hiring prudently. And it frustrates me because all these other guys, the pros, are getting so much attention, but there are so many other people in tech who do the right thing or trying to do the right thing, and they're the vast majority, I think.

Yeah. Agreed. I think the point of all of this is that- Yeah.

They're like 35% of them. You're right. " Right.

I think the point of all this is that there's a lot of leaders out there whose pants are on fire. I'll leave it at that. Well, and I'm going to take the last word prerogative, Mike.

Jensen Huang, this guy just gets up there, says what he feels, feels what he says, says it as it is, let the chips fall where they may, no pun intended, and he did it again here. So kudos to Jensen. Let's move on, though, Mike.

Segment three. Usually this is where we take a left turn into cyber or something, but what have we got today? 6 compound annual growth rate.

Now that's down from previous projections, which is interesting in itself, but some folks are saying, well, the market's just getting so big that it's hard to sustain 15 to 20% growth rates. But Kate, I got to ask you this, are we spending enough on cybersecurity? Because it seems like we're starting to get into some sort of rhythm that says, the amount we're going to spend is this as a percentage of the whole thing, and we're going to hope for the best and see what happens.

So let me just start by, I can remember the days when I was standing in front of a C-suite, just literally, I can probably say begging them to take cybersecurity seriously. Let's put some money into this. So the numbers at the end of the day shock me.

Yes, for sure. But however, I don't think that we are doing this right, and I say that if we were doing this right, we would not see the breaches that we're seeing today. So we continue on the whack-a-mole strategy that ...

I wish it didn't bum me out so much to have this whole, we are still chasing threats, and it's crushing almost. And again, I just remember standing in front of a SOC center, because that's where I came from, and then having a new CIO be all excited about the new SOC center that they are building right next door to the old SOC center, and for sure that this was all going to be different. So I should be happy about this spend, but I'm not.

Like I said, I'm usually the Pollyanna here, but yeah, no, this has been horrible. I am not today. Go home everybody.

Just go home. You're crushing us. With more alerts and more tools and more dashboards, it is not helping us at all.

People, wake up. We need to change the way that we think about cybersecurity. Yes.

Yep. Good. Being grandpa age right now, it's kind of handy in this.

When I get this feeling, it's like when you see what's going on, and this, to be clear, it's going to take a year or two to work out, but there's a bunch of young folks running around, and they got chainsaws, and they're starting fires, and you know there's going to be accidents and somebody's going to the hospital, but they won't die. And there's nothing you can do about it, so you stand back, and you do the things, you move the sharp thing out of the way when nobody's looking. Because yeah, this spend, it's fine.

Is it going to be spent on the right things? No. Right?

But people are going to be forced to learn the lessons. We, again, in this show, over this last year, it's been a amazing little clock. You can't solve Typhoon.

The adversary knows all your gaps, and they can exploit them at speed and individually. We know that. Mythos, all this other madness, filling in articles in the last couple of weeks.

Everybody can find all the vulnerabilities. Yeah. This is where it is.

But there's CISOs who've gone to SANS, and they've learned a lot of good things, and they've worked a date. They'll do them for another year or two, and they'll put budgets in the wrong place because that's not going to help. But evolution is a b***h, and it will push everyone that direction, and we will see.

As long as we do keep doing this show for another 12 months, I guarantee a year today, we'll have this conversation. " Because that didn't work, but it was a lot of money, and vendors were happy, so why not? So I'm just going to say it.

Same s**t, different day. And let's stop blaming AI for this. This has- Mm-hmm.

Yeah. I've been in security since 1997, '98, right? Newcomer.

Yeah, I know. Because you're old, dude. But the thing is this, security's always had a shiny trinket syndrome problem, right?

Yeah. We're always looking for the next magic bullet, the next thing that's going to cure the evils and kill the vampires. And now with AI, what's cool about AI, it's like Eddie Murphy.

It's playing a dual role in this movie. It's the villain and the savior, right? You like that, Chris?

Yes, I like that. That was great. Because AI is going to create all these new security conundrums for us.

But wait, AI can help us solve some of these things we've been trying to solve. Let's spend more money on it. Right?

The cloud. Oh, my God, the cloud. What kind of cloud security issues are we going to have there?

Next-gen firewalls. There's always a new security kind of thing- Yeah ... to spend your money on.

I don't think security budgets have gone down for more than the one year in a row for all 20, 30 years that I'm doing this. Right? They always go up, and they continue to go up pretty high.

And yes, we will use the AI boogeyman to even push our budgets up. But at the end of the day, the hard lesson to be learned here is it takes more than money- Yes ... to do security right.

Alan- It's a classic ... you and I are on the same page today. " You're right.

I'm taking all these things. You are right. I bring them on, and I play them over and over again.

So let me ask this question, though. As we continue to spend more money, and the attack surface is expanding. But Kate, is this whole notion of security by design kind of failing because we have to add all this overlay of cost around it to make up for the fact that the thing itself might be fundamentally flawed?

Just saying. Yes. I think at the end of the day, and I'm going to say this over and over again, we're not doing this right, and we need to rethink the way that we're thinking about cybersecurity.

We need to change this game, and everybody- Can I ask a quick-- But what will ever change it, though? I really think- So this has the potential to- Okay ... because not in the short term.

The short term, it's just a question of moving from finding vulnerabilities to fixing vulnerabilities. But as Jen Easterly wrote, and I think, Kate, you agreed with her- Yeah ... if I remember back when we talked.

This hopefully will force us, finally, to develop more secure code. Because those of us in security have known for 40 years, now 35 years, that what we really have here is a bad code problem. If you make the code better, we won't have these problems.

That's a big part of it, right? Software bloating and reachability. It's really time that we start to throw the bad code out.

It's time. I don't understand. That is the simplest thing that we can do.

" I don't know. It's craziness. Get rid of the bad code, people.

No, I'll tell you why, because you've got a $500 billion industry that's grown up around doing the same old, same old on the status quo of doing what we've done here. 7 billion? I think that there's more- So where's the incentive?

I think it's almost there's a better way to do this, and we will make more money at the end of the day. That's what I believe. So- Yeah ...

" But that's where we are. The humans aren't writing the code. I was just put my finger up at my ear when I heard humans.

Yeah, this is something we just- And that is the promise, right? As part of AI writing code. Maybe we do a better job eventually.

Yeah, but the AI is writing code that it learned from humans, so. Yeah, but it iterates and makes it better. Chris, you have your hand up?

The entire IT world, not just the internet, is a high school project-level quality thing, right? In the OT world- Yeah ... infrastructure is actually industrial stuff.

And we see where you take the IT stuff and put it in OT, the weak point is IT, because it's not nearly as-- We've gotten away with it this long. And I think what's happening here is, again, we touched on it multiple times today, all this is turning into utility-grade stuff. And that means a lot of things are going to change.

So what we've been spending on security, it will change. Right? We don't have time to reiterate again what that'll be.

We don't really know. But this is not something that works perfectly fine on a bench, and you can get grants any longer. It's got to go out in the field over rugged terrain and not fall apart.

Yeah. And that's all of it. I get it.

I'm agreeing with everybody today. This is awful. What's wrong with me?

I like it, Kate. I like this new Kate. Well, I just want to thank you, Alan, for sharing your implants.

At the beginning of the show, you were telling us about them. Oh, wow. No, those are dental implants- We all going back to the bug now ...

but they talk to me inside my skull. So he said. " Okay.

Hey, we're about out of time, though. Is there a Wi-Fi connection in those implants? Just checking.

Yeah, but he keeps playing, what do you call it? Who's the Italian tenor? Pavarotti.

No, not Pavarotti. No. I forget.

Botticelli? The bald- Botticelli? Botticelli ...

guy. Andrea Botticelli. Yes.

There you go. Ask the Irish guy. He knows.

Yeah. Of course he does. Wow.

He knows about these things. He's a big music fan. Guys, we've got to pull the plug on this stuff.

We're at the end of our time. We will be back tomorrow with the gang. But look, if you're paying attention day in and day out here, we are documenting historical times where AI is evolving and changing and pushing and pulling the world of IT and the greater world beyond that.

And we'll continue doing so, I hope. You can check us out on Techstrong TV, on our Techstrong OTT app, on no matter what, just about any screen you watch this stuff on, or on any of our great sites, including techstrong-semi, our newest site. But until then, on behalf of Kate and Chris and John and Mike and myself, thanks for watching, everyone.

Have a great day.