AI’s Future: Hype, Risk & Responsibility | TSG Ep. 919

Hey, everyone. Quack, quack. If it walks like a duck, it's a bubble.

You're watching. Textron Gang. Hey everyone, it's Alan Shimmel for Textron Gang.

I'm not in the studio if you couldn't tell. Today, I'm actually out in Napa Valley of all places for a, uh, JFR Swamp Up conference. I'll be at this week.

We'll be, we'll be, uh, streaming live from. So stay tuned for that starting tomorrow and Wednesday. But the gang stops for no one.

And so I am hosting the gang Mike's on his way to Boston for a spa conference. We've got a great gang lineup. Let me introduce you.

Of course, we have our friend JP Morgenthal. Steven Foskett, you've Kenny Rum and Wiki Wang here on the gang today. Welcome gang members.

Let's jump right into it. Um, jp you know, all of a sudden, you know, the, the bird, the blue bird of happiness is starting to drop droppings all over the AP story. You know, the, there's people actually using the, the, the bubble word.

Uh, but at the same time, you know, a new report out a few terms says it's gaining transformational traction. I know you have strong thoughts on this, as does Steven and probably the rest of us, but why don't you kick it off, jp, what do you think? Sure.

I think you cannot deny the, uh, ability and the, and the strength of what this technology can do today. It, it, it needs to be tempered. Um, I think it gets, I think it's right now in an over hype cycle, but that doesn't mean that it also doesn't have tremendous capability to do real things.

Um, it is really good at creating content, uh, and that content isn't just written word. Um, in many cases right now, it's actually software itself, which is a huge advantage for ma many, many companies right now, today, the ability for, uh, the, the, to use AI to actually help in the backlog of developing software, I I am going through this right now, uh, with an organization and it, and it takes time to mature it. It doesn't happen overnight.

You don't just put somebody, uh, you don't give somebody cursor or, or Windstream or one of those, and, and the next day they're experts. Um, there's a lot of experimentation. There's a lot of trial and error in learning how to use these technologies effectively, but the proof points are there, and I think that is what is starting to change people's minds.

I, I, I have seen it personally. I, you know, I took an a semi believer and who challenged me, said, prove it. Um, and, you know, without knowing anything about their code base, I was able to do a legacy modernization of it and, and get 80% of it migrated to a new modern, uh, language and, and framework.

Now, interestingly enough, where I ran into a wall, uh, and that stopped me on my tracks, was I know nothing about their application. So it does require subject matter expertise in the application itself, knowing what the app's supposed to do, how it's supposed to work, to go that extra to, to take the, the last mile. Um, but that proof point, that early proof point was enough to change, you know, the semi believers mind to a believer.

And so I think that's what's happening around the world right now is that non-believers, semi believers are seeing proof points by those who do know how to use the tools and demonstrating a potential outcome enough to bet on. For sure. Steven, why don't you chime in here.

I know you have some thoughts on this subject. It's almost like we've talked about this before. Yes, I have some thoughts.

Um, so, um, people tend to be irrationally exuberant about technology. It's just na in our, in our nature. Um, and, and this happens every single time a new technology comes up.

We, we, uh, imagine what it can do in the context of what we exist in today, and we try to fit it into that context. And in the case of LLMs and generative ai, we are looking at it with rose colored glasses still. And we're seeing all of the potential that this, that this technology has.

That doesn't mean it doesn't have potential. You know, that the, the, the, the, uh, counter viewpoint to the bubble doesn't mean this stuff is useless. What it means is we need to hold our horses and find out what the true value of that is gonna be.

And what we're seeing as is reported in this report, if you ask enterprises, um, they see a lot of potential in this. They see that this can be useful in many ways, uh, they see the ways that this can be used to augment the capability and productivity of their employees. And that's actually happening.

Um, you know, Alan, uh, you and I, I know all of our employees are encouraged to use, uh, AI technology as a tool, as a very powerful tool. Just like we encourage them to use, um, computers and cell phones and, you know, whatever else, as a very powerful tool. And it really has helped their productivity and it has helped them to do so many things.

But that's not the same as refuting the bubble. The truth is we are a hundred percent in a bubble of, um, CapEx acquisition, and that's supported by the stories. Um, and what we're hearing from coming out of companies like Broadcom and Nvidia in this absolutely off the charts purchasing, it's reflected in what you said in your shimmy says video.

And it is reflected as well in the minds of a lot of the financial analysts that I'm talking to are saying, basically, buckle your seat belts folks. 'cause this is about to get bumpy as the rationality comes to the overbuying of these things. I was talking to one of my friends who said, um, that, uh, he, he was working with an a, a, a big Wall Street bank, a name brand that, you know, they had pallets of Nvidia GPUs sitting on their loading dock floor.

And they had, they had, they had spent literally hundreds of millions of dollars on amd GPUs. They weren't sure, you know, essentially they brought this guy in as a consultant. And their question was, now what?

Like, what do we do with all of these things that we bought? We know they're supposed to be valuable, but what do we do with them? And these things have been sitting there for about six months, and they're not even the latest generation, unfortunately, they've lost much of their value already because they haven't been productively used.

That's the kind of bubble that we're in. Now, to your point about, um, that maybe this is like previous bubbles where there was over purchasing of capital and over build out, and then eventually kind of, we came in and absorbed that it's true. I noticed that you were very careful, maybe you were careful on this, on purpose.

Alan, you talked specifically about the infrastructure in energy sources in data centers and that sort of thing in terms of the bubble and the usefulness of this technology in the future. And I, I think you're right, essentially provisioning a lot of energy provisioning, a lot of data center space, et cetera, these things are gonna be valuable resources in the future. What I question is the value of last year's NVIDIA GPUs in the future.

That's what I wonder about in terms of the bubble. My question. So my, so based on what you guys are saying, and let's think about this is the problem is the bubble is the problem is the hardware, or the problem is we don't have enough people that know what to do with it.

Um, I think it's a fundamental business, um, case problem. Yeah. Yeah.

I do want to add some sense here, right? So if I, I think within the AI ecosystem, people start to notice that, and they did something ahead of time, right? If you look at this year's Vidia, um, conference, I forgot the name of the conference, but you can see, um, in their keynotes, they mentioned something about quantum computing or some other way to use of their chips and, uh, the, the AI supporting system, right?

They try to bring up the quantum computing in this bigger ecosystem and also talk about robotics or other stuff, try to see how they consume the additional chips or other stuff. People do. Think about that.

But I do agree with Alan, there's a bubble. 'cause if you think about that, right? Even the, like, uh, people who does like retired or like, like students in the elementary school, they're talking about ai, which means like, it's overheated, right?

Everyone try to know about it. Everyone talk about it, but only a few people start to really use it or leverage it. If you see the company, actually, when I look at the report of IMIT Media lab, right?

They mentioned like 95% didn't work very well. Only 5% work very well. Yeah.

I also say there's a, there's a question from the compliance side asking, okay, if people start to use ai, how do we make sure everything is in place that work very well, right? Overhead is happening. The question for me is like how people try to dig into those like 95%, see how we can, um, make the AI work better in those area.

Yeah. So here's the, here's the, you have a, you have a technology that fundamentally affects, uh, it's B2C. It really is, it's a B2C technology.

I mean, let's, let's trade, let's replace AI with cell phones. Everybody has one. Oh, more cell phones.

The technology, when it emerged, right? It took a while. Not everybody had a cell phone at first.

And then, you know, people started doing only, you know, you know, if they needed it because it was so expensive as cost drop. Obviously now everybody, even homeless people have cell phones still. It's, um, it's ubiquitous.

And, and to some degree, AI has a little bit of that ubiquity built into what it is, right? So, and the other piece of it is, is it's a spectrum of technologies. There's AI infrastructure, there's AI applications, right?

So these are applications that people build, build with AI built into it. And if, you know, have I given an analogy, right? It's like when people said they were cloud, but really all they were doing was moving Office 365 or Google Workspace, and they were like, we're using cloud.

Are you, well, to somebody who like me, it was technical and it's like, you know, leveraging AWS or you know, Google infrastructure. I'm like, you're not cloud. You're just, you know, you're using SaaS, but that still counted for cloud revenue by, by Wall Street accounting methods, okay?

That was cloud revenue for Microsoft, cloud revenue for Google. And the same way we have that going on here, not everything that's called AI would be considered AI by AI purists, but if you're doing a Wall Street accounting, right? These are AI technologies.

This is it being applied for it to be a bubble, it would have to have irrational exuberance. And the thing is, it's not completely irrational because you have these different levels, spectrum against which these technologies are being applied at different levels. You know, I I hadn't thought of that analogy, jp.

I just wanna say that is a very good analogy, the cell phone world, because in that case, what we saw was exactly what we're seeing here, which was an incredibly capital intensive rollout of a technology that will be obsolete by the time the business case comes along to use that technology. And that's, and that's sort of the problem essentially, that that global rollout of 3G Telecom that has just been sunsetted for good and turned off in most countries this year, is exactly what's gonna happen to those sort of first generation data centers full of first generation, you know, AI accelerators, none of those are gonna be used productively. It was a huge drain on resources.

It was a huge drain on capital. But having that infrastructure in place meant that people could start becoming dependent on that infrastructure. And so the next generation becomes profitable.

I think you may onto something there. So I, I think this is a quantum problem, right? It is, it is both successful and a bubble at the same time.

I, the, the problem with our bubble, with the, the bubble end of the equation is the irrational exuberance of how quickly and how profoundly it's gonna change things. And it, it may profoundly change things just maybe not as quickly as we're betting our money on. And in some extent, it already has changed things.

Jp I would disagree with you though, in saying it's strictly a consumer technology. I think the recent report from the FU team and, and other things we've seen, right? Something like 60 a, a majority of the code being generated today has AI fingerprints on it, if you will.

Whether AI generated the code, tested the code, a majority of the code has AI fingerprints. This is, and this is being done at the enterprise level. You know, I said this on a previous gang episode.

If we stop developing AI right now, we stop going after super intelligence and a GI and all of these, I don't wanna say pie in the sky, but all of these, you know, singularity moments, it would still take the tech industry, I think five to seven years to digest what we have the capability of with today's AI to do, right? And, and not just tech, it, it, it, it filters across into, you know, I have an article up I co-authored with my good friend Bill Brenner from Cyber Risk Alliance. I don't know if any of you guys know, bill, bill and I wrote an article about AI as therapist, right?

A recent case in the New York Times, a woman, her mother wrote a heart wrenching story of her daughter who was kind of confiding in her therapist, Harry, who's chat GPT, and Harry said all the right things and her daughter committed suicide. Anyway, um, right? We, we need to get better at that.

But at the same time, I do think we have to recognize this is, you know, as I said in the outset, if it quacks like a duck, if it walks like a duck, it's usually a duck, Ellen, And this has go you, Danny, I think the biggest change and the, for us gonna be when this duck gonna be walking in our house, where we are gonna have AI enabled dogs, where we're gonna, like, I mean, like robots, when we have robots walking in the house, cooking for us, taking our trash out, washing our dishes, maybe potentially babysitting our kids to, towards what, what you're saying, this is gonna be the biggest impact in my mind, because now are we becoming lazy? What, what are we doing? What are we not doing?

Are we sitting and meditating the entire day, running, uh, doing retreats, and then it's kind of basically become a butterfly effect? What do we teach our kids? What do they need to learn?

What are they gonna do when they grow up? And I don't have an answer for all this, but, um, it's a problem. Is it, you know, I I call that the Wally world kind of syndrome, right?

Where we're just going to get so lazy and dependent on AI doing things, we lose the ability to do them ourselves. And that's, that's not good. See, I Was thinking vacation Wally world, I thought what you were gonna say is we're gonna get there and then be really disappointed.

Well, we'll have to take grandma off the top of the car. Exactly. Um, but no, no, I, I meant, you know, humans as corpus souls.

Um, but, you know, so it, so it is a little of both. I, I, I'd encourage you all out there to go, go check out that Futurum report if you can. There's a bunch of stories, obviously on techstrong ai.

We all, we all have strong feelings on this, and, and many of us have been around the block more than a few times, and we've seen the hype cycles, right? And nothing, I'll leave it with this. Nothing ever truly lives up to the top of the hype cycle.

And nothing ever truly goes down to the bottom of the trow of disillusionment. Reality's always in that middle somewhere. And, and it's probably the case here too.

So let's take a break on Textron gang. We're gonna come back and talk about opa, opa, opa, no, it's not octoberfest. Um, we'll be back.

Discover Textron Group, the epicenter of tech innovation. We are your go-to for reaching IT, leaders and practitioners worldwide. Our secret impactful content that sparks awareness, engagement, and top quality leads with us.

You'll access editorial websites, streaming videos, virtual events, custom content analyst research, and more. Join our satisfied clients. Let's revolutionize your tech journey.

Contact us today and tell your story to the world in the most powerful way with Textron Group. So Apple being Apple, they did a very apple-like thing. They, they were big, uh, proponents of something called the open policy agent opa, which is a, uh, A-C-N-C-F owned and operated open source project, but it was started by the folks over at a company called st, excuse me, Tyre.

And, um, and Steyer was still the leading lights in continuing development of OPA and, and leading that community. Well, APA didn't sort of an acquihire without the aqua, right? They, they did not buy Steyer at all.

They just hired the entire engineering team, including the co-founders, including the co-founders of opa, and kinda left the company as a, as a shell, a husk, uh, with their corporate customers receiving letters saying, look, very sorry, but, uh, your, your support won't be going forward. Here's the good news. We've taken all of the enterprise, uh, functionality of, of the Strayer product and donated that to the CCFs OPA project Wiki.

You know, that's not a nice thing to do to your customers or your former customers, but what's your take on this one? Yeah, Actually, I wanna talk about this piece, um, from two perspective, right? Uh, one is from the merchant acquisition trends, the other one is from the compliance customer perspective.

Maybe talk to the, talk about the customer piece first. Um, I personally feel it's very interesting trends if you say something like that, right? If I'm the person who manage the security or PR product, vendor side, I may, I may feel like, oh, this is a big surprise, right?

I'm working with a stable company, but suddenly the things which may have impact to my company's operational effective now move to open source foundation, right? And we are not sure who's a responsible person yet. So my, this is a, i I think normally, like when my friends talk to me regarding, okay, what's, what's my plan next, uh, from the compliance or security perspective, I first suggest them to, to find out what's, uh, uh, official communication channel about this, right?

Um, so they can get latest information and see if they have the transitional supporting plan for the customers. Otherwise, it's kind of like, oh, after I finished my business, I went to a, I go to a good place. I just quickly hand over to nowhere and my customer is struggling for the future supporting, which is not a good way to do.

So, uh, for, from the customer perspective, I do need a transitional supporting plan. Try to understand during this grace period who's going to take care, right? Um, and in the future, if, if the open source, um, ecosystem will have the enough support for my, for me and from the customer perspective, I'll also check the current contract I say and, and decide my transition, uh, strategy, try to understand do I need to stay or do I need to leave, right?

At least to do, uh, something like, uh, export my data and confirm new obligations, make sure I have something, uh, to support myself or there's a backup plan, and try to find out the repla replacement plan details. Who will take care of me in the future, right? This is the customer side.

And from the merchant acquisition trends recently, I do see there's a big trends, like people try to do the, uh, acquire to hire stuff, right? You can see the wind serve, which is the biggest case recently. The Google take care of their, um, uh, their founders, uh, core team and some of the, um, I should say product, little bit of product, right?

And, and the rest of the team went to a company called, um, Uh, it, Trying to think myself Called. Uh, But it, so yeah, someone bought the, the, the rest of the company, right? Yeah, I can say there's a trends, like, it's very weird.

The rest of the team, they went to some like old competitors, so people can take care of the business and they also, that part of the customers, the customer will have a good take care of, uh, way and also have a good, um, good supporting system in the future. I think that that's the best plan for the acquire to hire case. The customer got good, good take care.

And the new company got a new customer base, right? And they also acquire rest of the talents, which is very nice as well, even though after they also have other stories as well. Um, So, so in that, in the, uh, in the case you're talking about Google got a license to use the ip, but the, but the, the original company still own the ip.

So when the second company bought it, it wasn't just, you know, they still had that IP to use, but Google hasn't a license to use that IP two, right? So they, in essence, you bought the company without its biggest customer, maybe who's now your biggest competitor. Um, but what's interesting, different in this case, wiki is they didn't even buy the company.

They just hired the people flat out. But that's the part I try to suggest their customer to have the transitional supporting plan and see if they, what's their transitional strategy, right? Do they want to, do they still want to stay with the current product or they should find out some alternatives?

'cause now it's not very clear what's next for the supporting system, which is very important for the product like this, right? This is kind of like, uh, the product, what we call kind of like, um, policy engine, right? That powers Kubernetes, API and microservices.

Those are all key stuff in from security perspective. I am, if I'm the customer, I need sync a lot here. Yeah.

What happens if you buy a company and then close the business and just get it for the code? So yes, I think it's a bit of a nasty move to just hire the people and not get the company, but we can have the same problem. I can buy the entire company and just say, okay, we're closing everything and we're moving the code.

So effectively we have achieving the same problem for the customer, and it happens with other Well, And after We've done that in the past, I mean, you know, apple has bought the, you know, acquihire the team of companies and kind of left people high and dry. I mean, I'm thinking, um, you know, dark skies with the weather, um, you know, they did that. Mm-hmm.

And with some navigation and, um, localization services, um, and I think in this case, CNCF and Open Source is saving us from that and saving absolutely customers from that. I say Bravo. Yep, absolutely.

And, and I will tell you, so I wrote this story that's on your ticker. Since writing this story, I've had three, four companies who either support opa, right? We'll have commercial support for OPA or have Oprah alternatives, but that integrate with that, that, uh, you know, they view this as it's, it's Christmas day, right?

They're, and they're opening their presents, right? And they, what an opportunity, because Tyr was the biggest player. And, and, and wiki, I think you said it, this OPA piece of functionality is embedded into so much of the cloud native infrastructure and huge enterprises use it.

But, but Steven, to your point, this is a term red, this example of why we need foundational open source where we're not dependent on any one company or any company coming in with some deep pockets and leaving everyone else high and dry, And why companies should insist that foundational technologies like OPA are open source, because it means that they won't be left with a bunch of abandoned wear. Agreed. Agreed.

Um, I'm going to be doing a follow up on this with some alternatives for people out there who are wondering. But let me just say this. I don't wanna make the ty of co-founders and people out to be villains here, right?

You work really hard, you build a company, someone comes by, I'm sure Apple didn't hire them for a dollar, right? Apple probably paid a pretty penny. And you know, it's hard to sit there and say, let me do what's right for the community.

And in, and in some ways they think they did. They, they donated the enterprise functionality back to the community and, and they went off to go get their jobs and make their money. So, you know, they're not the villain here.

They're not the villain. Anyway, let's take a break on, on the gang. We're gonna come back and we've got like this, this next one's right outta Hollywood.

Crazy story. com is the leading resource for news analysis and education on challenges facing the cybersecurity industry. com covers all aspects of cybersecurity, including data security, DevSecOps, cloud security, application security, network security, security threats, and more.

com has the largest selection of security content featuring breaking news, blog posts, podcasts, and more. com to learn more. com.

Home of security Bloggers Network. Hey, everyone, we're back. So this one's kind of like right out of a Leonardo DiCaprio.

Catch me if you can kinda story. Um, when I first heard about it, I, I assumed it was nonsense, but it appears that, uh, a young AI engineer working for Elon's roc, you know, XAI, uh, somehow was able to get the source code to the entire Grok ai, not only get it, download it, exfiltrate it, take it out, and give it to open ai. I don't know if it was quid pro quo for getting a job there, or just his little gift for coming on board there, but he then quit Rock.

I went to work for Open ai. Oh, and by the way, before he quit Rock, he was sitting on about $7 million worth of shares in XAI for, you know, as part of his employment there. Well, he sold that too, figuring once the source code's gone, I might as well grab my money while I can and grab the money as well.

You get it? Are we gonna see this in IMAX soon or what's the Story? It's a very interesting story.

And when we started this show, you mentioned about hype cycle. You know, something goes up and down why I am bringing this up because we something, we have something called DOP, data Leakage Prevention. DOP been around for long time, we all known version for five, whatever you wanna call it.

And everybody complained about how hard is DOP, how, how is to secure the data? And every five years we have a new generation of companies to secure this. So we have a very classic example of people trying to secure the data without impacting the employee performance or employee employee way to work.

And we all know developers are the ones very, very demanding. They want admin rights, they want to install their own stuff. They need to, they need to do stuff.

And especially if they're very ahead of the curve, if they're very innovating, they probably want more access. And it's very hard to understand how we give them access without making sure they can do stuff. Use v key cloud uploads as a cell inspection, ss, ESEC.

There are so many different ways to protect the information, but as an architect, and I've been doing social, an architecture for a very long time, it is very hard to understand the best way to protect the data, especially when the employee has not admin rights In most of the cases, we can get logs and understanding, but this is a very classic example. He did something, we don't really know exactly how he took the data. If it's a USB, if it's a cloud, if it's an email, but it's not a very big file.

It's not gonna be terabytes of data. It's gonna be probably something that you can put on a floppy disc. Um, I don't think it was probably a floppy disc.

I'm gonna go out on a limb. You're right. Probably not.

Probably not. Well, three and a half or five and a quarter. Uh, I'm, I'm bringing floppy disc because, uh, I had a floppy disc that my son found yesterday.

He's like, dad, what is this? He's like, it's a very old USB key. He's like, what do you put it in?

He is like, I have no idea. He, there's no to put in right now. He's like, how many I put here?

How come This looks like the save button? I don't understand this A So, yeah, sorry. Yeah, I'm trying to understand this one a little bit better.

As you mentioned, there's a loss, data loss protection stuff, right? Normally company will have it. How come their legal department didn't give the warning before this person leave?

This is a question mark from complex perspective. Well, I think it's very, Well, this is the whole thing here. It's very classic.

It's a big, if We're just gonna look at the DLP perspective of, of it. If Jenny, as you are here, right? One has to question does XA, I even have DLP Wiki, does their personnel department have processes in place to make sure employees haven't taken any company properties or ip, right?

Elon runs it lean and mean, and sometimes lean and mean comes back to bite you in the butt. And there was a problem in Tesla. Just gimme one second, jp, if I may.

There was a problem in Tesla about two years ago. Somebody left with information as well. So there's probably already a president before to understand maybe there is DOP, maybe there's nobody watching the DOP.

I don't know. It's not even, I mean, well, the only way for them to institute governments around this is you can't bring any of your own external outside equipment in. Yeah, it's almost like creating a skiff, right?

You know, cell phones around our computers. You, you know, usps, we're gonna put you through an X-ray when you come in. By the way, there are, se there are places where this happens, right?

There are plenty of government contracted facilities in which this is common practice, uh, you know, to avoid. Is it work from home or not? That's The issue.

But you know, to, again, these, uh, uh, point earlier when you are developing next generation technology, it's hard to implement, uh, controls when you, without tying the innovation, right? Because you need, the last thing you want is for the developers to run up against security, als and walls while they're trying to build and test code, right? Ah, I can't get to that server without permission.

Now. I'm wasting a week waiting for, for paperwork, or I can't, you know, I can't get the resources I need in order to execute, especially the AI stuff. I can't get the GPUs, I can't get allocated enough without getting clearances from fourth levels of supervisors.

All that kind of stuff just slows you down, right? So that balance between, uh, governance and controls versus innovation is a really tough One. One, it's a tough one, but like from the data classification perspective, the source code, I think it's as important as a financial data.

This should have very high rank of the classification number, right? Not sure like how what happens here, but let's spend that on it later. Well, let's also consider, um, I wanna point out too about this story, the dates that this happened and the companies involved.

So I'm gonna, I don't wanna, I don't wanna make a conspiracy here, I don't want to like be like putting a string and thumbtacks on the wall and stuff like that. This happened o in the early 2024. So well over a year ago before the lawsuit was filed, it happened at X AI and open ai, which we know, I mean, Elon Musk was one of the founders of OpenAI.

He's been going after the company for a long time. He's actively engaging in litigation against them. Uh, you know, XAI has had, um, its share of, uh, pr GAF and so on lately.

I wonder if they knew about this for a long time and really didn't care and only brought this up now because it serves their PR purposes. And I wonder if this story essentially is giant, is a giant nothing burger that the guy stole a bunch of outdated stuff a long time ago and nobody cared until they felt that they could get some kind of value out of attacking. Well, No, it goes one, it goes one better.

Steven. They have open source their code, right? Uh, the XAI has open source the code.

So you know, it, it, it is a nothing burger on on that. No, No. They've open sourced this code.

Let's just be clear about that. Okay? Yeah.

It may not be this particular, but they've opened, you know, supposedly it's being open sourced. But you know, the bigger issue here is why isn't open a, you know, what kind of company is OpenAI and where are their scribbles? Why aren't they being for, you know, in, in the old days, Steve, remember, they would have to gorge disgorge themselves of that code.

And anything they were working on would have to be shown to be in a clean room environment where they didn't even use that, right? And the people who had access to that weren't working on, I mean, there used to be rules around stuff like this where, what, what the hell's going on? Secondly, we're in such a rush to hire AI engineering talent.

Where are the penalties to this kid who did this, right? He, he's, he got a $7 million windfall. Now, one might say, well, he was entitled to that 7 million for the work he did there.

It has nothing to do, but it, it's almost like fruit of a poisonous tree, right? He, he, he got a $7 million windfall and he, and he's going to work at OpenAI, probably making a small fortune there. You know, where, where is the penalty for, for doing what we would consider a criminal act?

Are you suggesting that X AI and open AI are acting in a way that's unethical with regard to data? I'm, I'm shocked to find gambling going on here. Yeah, exactly.

Exactly. Shocking. Shocking.

But, but seriously, I mean, we used to, we used to have rules because this sort of stuff has happened before, maybe not with ai, but we've had, you know, companies raid other companies employees and IP gets out with them, mailing list, customer list, you know, trade secrets and, and we've all probably signed the agreements. I don't know how many of you have read them carefully about what, you know, how you have to handle company confidential information and when you leave, you can't, you know, we have non-competes and, and stuff like that. It's all out the window now because we're, we're, we're so hot and heavy to go hire anyone who claims to be an AI engineer.

It's crazy. Well, I, I don't wanna bring another story in here, but, uh, you know, we did just see Anthropic who's not named in any of this story, uh, agree to pay one and a half billion dollars. That's $3,000 per copyrighted work for copyright infringement.

So, um, you know, there's a lot of crazy high dollar, um, let's say, uh, legally questionable stuff happening here. And, uh, yeah, No doubt about it. Meantime, who's signing up to be an AI engineering guys?

We're about out of time, and I need, I need to get over to this Jfr Swamp Up event we're doing. Uh, Wiki, you have ney. Jp as always, Stephen, thank you for joining us.

Thank you for joining us. We have a full Textron TV lineup today for you. Check that out.

We'll be back tomorrow with another fresh Techron gang. But in the meantime, everyone enjoy your Tuesday and we will talk to you later.