Techstrong TV August 6, 2025

Hey everybody. Welcome to Textron Gang. And yes, we've been waiting for it forever now, or at least it feels like forever, but a different way of thinking about AI and processors and how we're gonna run code.

Stay tuned. We'll be back in a minute. Hello, everybody.

Welcome back. We have some of the usual faces for the show today. Mitch Hasley.

How you doing, Mitch? Good, Good, good. Mike, thanks.

Guy carer from the Futurum Group. Dy And Chris Blak. How you doing buddy?

Loving life. And Kate Scarsella, who's joining us from Kate. I forgot where you are again.

Southern Maine. All right. Awesome.

I think I haven't been in Maine in a very long time, but I was close. I was in upstate New York last week, so that's relatively in my mind, somewhere nearby. All right, guys, I'm gonna jump right into this first topic here 'cause I'm kind of excited about it, but I might be the only one.

So we'll see. Um, there is a, uh, new approach to, uh, creating silicon wafers, I think in terms of how we stack the chip together that uses AI accelerators, but depends less on GPUs, at least is how I understand it. And it's a different architecture and it winds up being among other things, enabling the ability to run 2000 tokens in, in a timely fashion so that when you're writing code, it feels more streamlined or smooth as part of the whole process versus today it's a little jinky as a technical word for you.

But Mitch, you looked at this a little bit and you know, what's your take on this? Is this the beginning of something? Are we gonna see a lot more of these things?

Or is this some sort of weird one-off thing? Well, I think what what one of the things that's interesting about this is they recognize as the patterns of how we use tokens for different types of work. Um, software developers do are very iterative type process.

It's not just writing code. There's a lot of check this, try that, what is that? Let's, let's change this here, see what the result is, test it, whatever it might be.

It, it's a highly iterative process. And if you're constrained, it's, it's kinda like if you're driving a car and you had to stop every five, five miles for gas, you know, 'cause you're got such a small tank where you only get a little bit and you have a really big tank. And that's the whole idea of this generating a lot more tokens per second that you can consume.

Of course, you can argue, are you using those tokens efficiently? And a lot of people have kind of focused on how do I write more efficient, uh, prompts to be able to use less tokens. But then again, if they're kind of coming to you, yeah, you high and fast, you may not be thinking about, uh, saving money on 'em necessarily unless the cost is right there in front of you.

So it's kind of a double-edged sword, but I think the main point here is it, it addresses the flow issue of, it's like reading a book and you have to set it down for five minutes after every page right before you can pick up the next page and start reading again. No, that is how I read. And I'm, I wasn't putting on you specifically, Mike.

Just a generic reader. Sorry, I mouth the words too as I go. Oh, no, I'm just kidding Out out loud Guy.

Courier, what's your take on this whole thing? 'cause I know you follow a lot of the hardware stuff out there, but is this something that has legs? Uh, I guess what you're saying is, are we gonna see this as a, a new model for hosting training ai?

Um, I don't think so. Um, so you, you said this is a new way, uh, this is 10 years old, roughly. Um, and it's not necessarily a new way.

Um, so, uh, the, the company, um, uh, celebr cbra, what are they, what are they called it? I I always wanna say Cereus, so I do Too. Sure.

So, but let's go with Cereus for now and we'll apologize. Yeah. Uh, they, they, they've been around about 10 years.

Mm-hmm. Um, and um, what they really did was add real estate to a monolithic silicon, um, uh, uh, production method. Um, I don't wanna get into like all these, you know, different types of manufacturing, uh, chips, but kind of have three main ones right now.

Um, a sort a monolithic one, um, a chip lit one, and then mainframe, which is a whole different, you know, kettle fish. Um, chips are what a MD uses video uses them. Um, they are making smaller chips, assembling them together.

There are some speed trade offs there, but there's great flexibility. And you can do things like load tons of memory onto a chip, um, or have the same basic architecture, but less memory and a lot more processing. Um, what, uh, CEUs has done, uh, CEUs, uh, what they've, what they did starting 10 years ago is they said, um, well, um, why don't we just make a giant honking single microprocessor or chip?

Um, so these things are huge, and this is the third generation. And, um, how many, uh, how many AI processors does it have on it? Uh, That was a lot.

The cores, I mean, yeah. 90,000, 900,000 on a bumper. Um, here's the point.

The point is that, um, it's a whole specialized fabrication process. It has not taken the world by storm in 10 years. It's, um, you know, generates its terrifics feed.

It allows you to put a ton of memory, but they've had three chips in 10 years. Third generation, it's a specialty chip, but it's not a programmable specialty chip, which is, um, what a lot of specialty chips are right now. Like if you've heard of neuro processing units or NPUs, they tend to be programmable field, programmable gator arrays, um, that are programmed and then put as a chip into another chip.

So a lot of wonkiness aside, um, what you would need for this to take the world by storm is more fabs that can make giant honking chips, um, and more companies making specialized chips this way. So why is this one company doing this? Well, uh, I'm not sure, since it was founded 10 years ago, that this is a response to the AI craze.

I think their announcement and their ability to produce this high volume tokens, it's very interesting. Um, but it's not what they were built for. What were they built for?

I'm not entirely sure, but seeing as they are, uh, something like 70% owned by an Emirati company called G 42, which is owned in turn by the Emirati Minister of Defense and their one supercomputer that they're putting together, which, uh, put together, I'm sorry, it's called Condor Galaxy. It's based outta Dallas, um, is has one customer, which is G 42. I'm not exactly sure that generation of code tokens was what they had in mind when they founded the company and started building it.

So I think, Mike, you put together today's show with three segments, right? And one's the AI segment, and two of them are the security segments. But I have a feeling that's actually three security segments.

And this is the first one. Chris, you're nodding your head a lot. So jump in here.

Yeah. Hold my coffee. Right.

You know, so Mitch is right, right. You know, this is the, you know, we are going down this path. Speed, speed, speed, right?

And it is not just writing the code, it's iterating and testing, feeding back in. So whatever numbers you're at now, Moore's log and, and there you go. And, and Guy, you're exactly right as well.

You know, that was the point I was gonna bring to this. I mean, we've done speed, right? You know, what matters now is traceability.

You know, we can write the world faster than we can secure it. So what's the provenance of your logic stacked? Who attested it?

Right? And velocity, you know, without integrity is just another intac attack service. So I could take either threads, you're going just the sheer speed or guy or guides, chips, infrastructure, what's underneath that?

Where are the decisions? Who's keeping track? You know, largely the answer is nobody.

So if you, if we think that is sustainable over time, I've got some land to sell you. So We'll see. And I've got a bridge, Right?

Just from a guy that lives on a boat. So, you know, is this, is this Arizona? Is this in Arizona where I can build the data center?

No. Hold on. Yeah.

And that's one of the things that we, that from a cybersecurity perspective, right? That we keep coming across as, you know, cybersecurity comes as an afterthought and it keeps biting us, and it's gonna keep biting us at the end end. You know, we just, um, as we will see, I'm sure that this, um, theme will continue to come up in the next few segments is the idea that that it, there is such a speed with ai, but with that, there's also, uh, that the surface just expands enormously as far as the attack goes.

So yeah, Exponentially, It's, it's amazing how many more security vulnerabilities you can create with 50 times the AI cores on your chip and all this SRAM available as well. Yeah. It's more than 50 times as many, right?

Because it gets exponential rent 500. I don't remember. It's just, Yeah.

But I mean, even if it's 50, you know, we're getting into the, it, it, when we start adding vulnerabilities at that speed, it, it hairballs, right? Because larger and larger exponentially, you're adding zeros everywhere. And, and Kate, right?

You know, as a security person, I love when necessity drives security. And I do not believe security is aside. You can operate, we can operate these systems much longer without attestation systems keep keeping track of things.

io, which is still out there, there are ways to do this. There's lots of driving factors been coming along for particularly the last five, 10 years. And this, no, a year, 1, 2, 3 years from now, we're not gonna be doing it this way.

Not in stacks that aren't falling over all the time. All right, let's come back to the beginning here for a second. Security is great guys, but there's a goal, we're trying to get to what Mitch was talking about, which is, can we find a way to cost, effectively get to 200 tokens per second?

This, just because it's 10 years old, I would argue doesn't disqualify it. It may be too costly, I'll buy that. But I would remind everybody that GPUs were 10 years old before we figured out that they work for AI as well.

So I'm not seeing GPUs are really optimized for AI either. So I'm pretty much sure that we need processors that are gonna be optimized for ai. The question is, Mitch, um, can we get that outta GPUs in a cost-effective way?

Or do we need some other type of AI accelerator? And maybe we got too much betting on, you know, Nvidia is the god of all things for processors, when in reality it's a moment in time. Well, there's, there's one path we, we may head down, and that is all prompting goes through optimizers before it actually is executed.

So, you know, today we do meta, you can do meta prompting. There's something called contract prompting, which is structurally laying out, uh, you know, ways that you do a prompt to, to keep the, keep the AI from going kind of na you know, native, native and going everywhere that shouldn't go. Uh, but I think that's, that's one of the ways is let's just use AI to economize how we're, uh, accessing or using tokens, you know, kind of use the same tool, what they're, they're trying to use smartly for other tasks.

I think that's a pretty obvious path that we'll head down. Not that I'll invoke an optimizer. I think that'll become part of the front end of any of these tools, maybe built into the id, BE maybe built into the LLMs, that type of thing.

I'm gonna add another thing onto this. Uh, am I the only one who thinks tokens are a bit of a scam? The whole pricing model is artificial.

It's kind of based on an input and an output. And I get charged for the input and the output, and each one of those is called a token. And it looks like a reasonable price till I start adding up all the tokens.

And the next thing you know, I'm spending a fortune to go build something. Isn't there a better way to think about pricing these AI services than just using tokens? 'cause right now it seems like it's kind of a, a money pit.

Well, I'll go where Mitch was taking it, right? Mm-hmm. You know, I, I, I, we've talked about this in the show in the past.

I do not believe in the near and particularly midterm future, we're gonna be using AI the same way and asking every, every single, to your point, uh, Mike, you know, I can you add two plus two, you know, that actually uses tokens out the wazoo because we're asking an LLM to interpret what English two plus two even means and do a bunch of other things. So I think, yeah. What, what optimizers, like I saying, Mitch, I think the inputs we we're putting into an AI or into a computer system are gonna parse themselves out and use AI for AI things.

'cause it's, it's ridiculous, uh, economically, energy wise, resource wise, the way we're using the bottles now doesn't make any sense. Mm-hmm. And to that, I'm Not sure the tokens are any better or worse than any other method, Mike.

I, I, I, I'd rather think they are better, but I think it's the same. It's the cloud cost problem all over again, only in a different sphere, which is yeah, avoiding blood, avoiding, uh, uh, over provisioning, uh, or whatever. Like all that sort of stuff that we started to deal with.

And I'm sure a new pocket industry and monitoring your tokens is gonna appear just like, you know, cloud service monitoring, you know, has appeared and become its own industry. Yeah. And I agree with guy on that.

Totally. It like the cloth model, it's, you know, um, I I think it's just becoming, um, it's, it's just a way to monitor it. But, you know, in the long run, I, I don't think it's the way to go.

I Can, I, can I bring this kind of back home here though? I think, Mike, your initial postulate was, wow, this is impressive. Maybe it's something that'll spread and we will be able to gain these kinds of amazing speeds, um, uh, you know, through competition or through growth of this particular model or that sort of thing.

Um, and I think that's an important question for the, the, the viewers of this, or listeners here, because they're trying to figure out how to make decisions. So, um, I, I'll just say that you knows super computers aside, you know, like, like this supercomputer in Dallas conduct Galaxy. Um, I wanted to know like, hey, what, what would it cost me if I'm a developer or development team?

What do I need to do to, to get this? Maybe I can just buy one power with this thing in it and stick it under a desk and all my developers can use it, and you can do that costs about $9 million or $8 million, nevermind maintenance and all that other sort of stuff. So then I asked myself, okay, can this scale?

I don't think it can scale because there's a whole supply chain question and issue. There's a completely, uh, different model than all the other, like the, the advantage here is conceptual. Instead of it being a GPU or a field program gateway, or any of these things that follow a certain model, these cores are custom made and have been for 10 years for ai, and that gives these incredible performance abilities.

But I just feel that the market is limited and, um, uh, TSMC amongst others are getting better and better and better at getting more performance in all these various ways out of slightly more general purpose chips. So, I, I just don't see it as having likes Mike, not in that way. And I'm really trying to burst your bubble because when you get optimistic about something, I get extremely scared because that is such an anomaly.

I tell, I tell you why I am jumping on this though. I don't really care too much about the processor one way or the other. I do care about that 2000 tokens per second.

'cause that's what developers need to have a, a, a, a streamline workflow process. And that's the goal. And I'm just dubious that we're gonna get there using the existing GPU architectures in a cost effective way.

So I'm kind of looking around the mobile world here saying, Hey, there's gotta be a better, smarter way to do this thing. You know, Mike, this is also in addition to the hardware side of it, the, they use the, uh, Quinn three model, which is created by Alibaba specialized doing software development. So it's an optimized model for software development.

How much of that op, how much of that, uh, consumption is, uh, is more efficient because of that model. That would be an interesting exercise, right? So I can generate X thousands of tokens per second.

Do I need to consume less? How much less do I need to consume when I've got a model optimized for software development for that specific task versus medical diagnosis or other types of models? So we, we always talk about small language models.

This is still an LLM, but it's optimized for a specific type of class of problem information it relies upon, et cetera. So that, that's gotta also be factored into the equation of how efficiently are we Using these. Yeah, I think there's efficiencies to be found everywhere in this.

I think you're exactly right. All right, I'm gonna leave it there on the, on, on this political slogan. 2000 tokens per second are bust.

Here we go. All right. All right, folks.

We'll be back in a minute to talk about, well, some actual cybersecurity Discover Techron 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. All right, folks, we're back, and the folks over at Cypress Data have a new report talking about, well, just how often are vulnerabilities actually being put into code and that we know about, nevermind the ones we don't know about, but the ones that we just deliberately ignore and add to our technical debt. And we've been talking about DevSecOps now for many years, and you would think that this was getting better, but the report suggests that maybe not, and I guess we'll start with, uh, Chris here on this one.

But, um, what's going on here? I mean, are we just deliberately ignoring all this stuff be and the name of speed, and how can it be speedy if I gotta go back and fix it later? Anyway, it seems like it's a vicious cycle of being counterproductive no matter how much I think I'm being productive.

So the audience doesn't know this because Taylor's gonna cut it out, but we're talking about software billing materials in the break, right? And, you know, we're working with side beats. You know, my, I'm, you know, clarity cl you know, I'm vice president of the strategy with site meets having been for a number of years.

It's an s om platform, uh, tool. And I've been pleased both with the company and the platform, but you know, to our purpose here of the market in general, s oms and that sort of visibility in supply chain, you know, has actually help dev, uh, DevSecOps work with the rest of the company, you know, help break through some of those, you know, cliche cultural barriers. Um, I think what we're talking about here is the opposite, right?

We're, we're once again, running up against misalignment of teams and motivations, and we're getting bad results. And I would argue is perhaps usual that we can force the issue and force people to do the right things, or we can recognize the problem, right? And we've got brittle governance, you know, evolving roles, and we need, we have a agent AI coming into this, and we need a flexible, flexible, participatory framework that can all work together.

Or DevSecOps, you know, God bless our community, we'll stick their heels in the ground and and generate articles like this one. What do you think would happen, Chris, If, um, uh, and Kate, if, um, instead of, um, addressing, uh, security breaches or vulnerabilities that come up because of this code, um, you just fired a person who reported it to you and pretend that it never happened. We had the same, we had the same conclusion on another story yesterday, but yeah, so I can understand where that's going.

If You, if you don't test for it, either, you won't, you have your vulnerabilities Go down. So, I mean, when someone, when someone damage sues you, you can just say, that didn't happen. That's why I fired the person.

Just didn't take place. I I my entire career, right? You know, I've, I, I've worked in security for 30 something years now, and you know, I'm a pragmatic person.

We can do what we can do, but at every opportunity, be really clear test to what you're doing, do that, right? You know, because you want your adversaries to see it, your friends to see it, whatever, right? And it's not just, you know, hippie philosophy.

It's a structural part of security. You know, either you're going to believe that you can hide everything from everything and keep that hidden forever. So nothing bad happens to you, or you'll make it as clear as possible.

And this is the, you know, this is literally a, a, a, the, this is this acting on in real time. We have, you know, for brevity, I'll say a standard corporate philosophy that we can just, you know, fire the people and force the, the issue and in, in security and dev sec and so forth, that you have people that really know and really care. And no, they're not gonna do the, the what they see as the wrong thing.

You need to actually address across the, in the, uh, as they say, with, so with supply chain, getting SBOs in there was helping DevSecOps explain to, you know, operations and dev and so forth, why they're being so insistent about things. You know, this is sort of doing the opposite, you know, but we can fix this, you know, put the visibility into the system, get all the players to, to agree. And, you know, without getting all AI about this conversation, like every other one, I think we have the tools who can speak to us and help us speak to each other so we can say what we need to do and do it.

Write the stuff down. Yeah, go ahead, Mike. Sorry, I was gonna come to you anyway, Kate, with this one.

But, um, are we just overemphasizing the wrong things here in DevOps all these years? Because it seems like it was, you know, right code and get code and deploy, code dev will be dimmed that, but the quality of the code, and can we strike a balance or we just too far gone on the, we gotta be fast and everything else is the secondary consideration. So I'm always hopeful.

So I definitely think, uh, that we can, um, we, we can all play well together. And I want that. It's one of the reasons I think it's so important to talk about why it's important.

You know, it, it, cybersecurity really is it, it, so often we are looked at, we are looked at as the bad guys, right? Chris? I mean, we are the ones who are stopping the speed.

We are the ones who are stopping, you know, true development and true innovation and everything else. But the truth is, is that I have found that even if you provide them with tools, even if you provide them with the, with the knowledge that they are writing vulnerabilities into code, there is a resistance of, you know, I like, I don't care, you know, I just wanna get this out. And, and this whole idea about speed is so, it, it is so important to business that everything else we're gonna, you know, throw the baby and the bath water out.

You know, we don't care. Um, well, you Sorry, go ahead. Go ahead.

You know, I, I, I, I feel very passionate about this. To be clear. I have sold billions of dollars worth of security product all over the world for decades by understanding the story.

And the story isn't that you need to secure your digital certificates. So the story is that you need to get code out there so your customers can achieve something and like you, so they'll keep buying your stuff so you can keep paying your employees and keep doing this, you know, all of it. So in a scenario like this, I just feel a disconnect that the people telling DevSecOps what to do, don't understand their own story.

Why is the company even here? You know, we're not buying a firewall because we need to keep be bad people outta the company. We're buying a firewall because we're a shirt company and we need to keep selling shirts and spending our time thinking about shirts.

So the button manufacturing people should be aligned with our customers out there, not our internal process on thread weaving, right? We get lost in the details and we forget why we're doing anything. Yeah, you're absolutely right.

Absolutely. But I think this is a deeper problem. And, and Mitch and Mike, I, I wonder what, I wonder if you agree with this.

I mean, uh, uh, a software vendor CEO many years ago, um, a real smart one told me, or, you know, revealed to me something that was obvious once he said it, which is that there are two types of developers, two types of coders, there are perfectionists, and then there are ones that release. He was kind of trying to say that, uh, he doesn't want any perfectionists around 'cause he wants to release. But, um, it's very easy as a perfectionist never to release because there's, it's a bottom.

Uh, coding is a bottomless pit of, of, of opportunities to improve, uh, an application, whether it's the ins of it or the user interface, or anywhere in between. And I think that I, I, I, I'm trying to wonder if I've been wrong all along to say the developers are like everybody else, and they don't care about security. I kind of think that developers that release don't care about anything, that security is just another one.

Another part of that, um, and and willful ignorance of security goes hand in hand with willful ignorance of, uh, uh, data hygiene, willful ignorance of the user experience, willful ignorance of whatever, even of regression testing so that they, they release and they, and those are the folks that get promoted. Those are the folks who make their name because they got, they got more and more features out they can. So anyway, Mitch might keep you guys, I'm gonna stand up for the developers here.

All you communist, whatever you people are I what your, I what your Problem is. I, I'm a Democratic socialist. Yeah.

Communist Supports communism. So Not true, not true. Well, first of all, you say one thing.

What I've found, and I've run many product development teams, software teams, the best developers write the least amount of code. Interestingly enough, it's not about creating lots of code, it's actually creating less code solving problems efficiently, meaning efficient with your time and your work. 'cause every, you know, every line code is tomorrow's technical debt.

So why do I want to create stuff that I've gotta worry about, maintain, and my technical debt could be an hour from now when I gotta go back and fix this damn thing. So it, you know, developers, developers are, are stitching together more than just code, right? They're designing an approach to solving a problem.

They're in some part of it, they're embedding it in a design of code that they're creating, but more so they're leveraging other code that they didn't write. Matter of fact, most of the code that's running in the stack is not their stack, even in the application stack is not their code. So when we say developers are introducing vulnerabilities, they, it could be as simple as the library I used has a vulnerability in it.

I didn't know that. Is that my fault that I didn't go examine every library to see if there's a vulnerability in it? Yeah, you could say, well, yeah, it's your code.

So you're, it's your job to do that. Well, if I did that for everything, then I would never do anything else. So my point is, is it has, it's a systemic problem.

It's not a person problem. If auto complete fixed vulnerabilities when we wrote 'em, we wouldn't have vulnerabilities If auto complete and includes of libraries or other code scanned for vulnerabilities, the second that you included it and it could fix it, then you wouldn't have vulnerabilities. So I'm saying ask asking people, it's like saying, let's pull over to the side of the road and check the, the air pressure on our tires every five minutes.

No, let's put an air pressure sensing system in our car to tell us when we've got a bad situation to go fix it. Right? That's what we need.

So developers unite, stand against the communist hate developers. I, I always, uh, I always said that, um, a good coder knows where to get their code. Okay?

With that being said, it brings maybe a totally different topic that we'll have to, you know, talk a different time about. But software de bloating, like what if we actually, from a cybersecurity point of view, came at this from a different side and we actually went after the code that is just, that we know is bad, you know, and just finally get rid of it. I mean, we have so much code out there that just really needs to go bye-bye.

And, you know, and then get coders, like, can we attack this from a different angle and actually provide coders as they go to look for code with safe code Instead, instead of going to the developer, Mike, what do you think? Oh, guys, I gotta come bring this to an end. But I wanna also bring in this one last point.

So when I talked to people about this story before the show, one of them looked at me and said, so we're shipping known vulnerabilities in code? And they were like, yes. And there's gambling in Las Vegas to use Mitch's phrase.

However, one of the things that is gonna happen out there is, if I look at it historically, maybe 5% of vulnerabilities are actually exploited by some cyber criminal. But in the age of ai, they got all these fancy new tools, and it's gonna take 'em a third of the time, maybe a 10th of the time, they're reverse engineer and exploit for vulnerabilities. So it seems to me a much greater percentage of those vulnerabilities out there are about to get exploited and all that old bloated code that Kate's been talking about, Mitch probability assessment, Oh, don't ask me for that.

That's, you're gonna put this, you're gonna plaster it all over every article for the next three weeks. If I do that, let me, let me answer a question, a question you didn't ask. We don't, we aren't providing developers with the right tools.

If we provide developers with the right tools, they wouldn't have to fix the vulnerabilities in the first place. That's my premise. I like it.

Mitch. There you go. 55% probability of something happening.

Wait, i i 25% that it won't, whatever that something is. Wait, I mean, New York, Yankee and I made an error and it must be my glove. Is that what you're saying?

Yes, that's right. This is totally their fault. Totally.

Well, it is, Mitch is describe, I feel, Mitch, you're describing an arms race, right? So, you know, uh, agentic, AI and ag agentic takes that 5%. Mike's talking about the 15 to 55 or whatever it is, but the tools evolve or actually are developed to counter, um, my take was culturally, this is impossible to bring it into, and Mitch said, I'm insulting developers.

So, um, I object to that. And then Kate's saying, fire the code, don't fire the developer, fire the code. So I like all of these takes except for mine, which is apparently communistic.

So, so I gotta close on this point, Mike, if, if we're gonna generate, pick a number, 10 times, 50 times as much code, it is impossible to have developers fix vulnerabilities in the code that's generated, period. It is just, you're, you're, you're, you have an escalating, uh, curve of which there is not enough people on the planet to fix vulnerabilities. So you've got to generate code that doesn't have vulnerabilities in it, that that's what we have to do.

Otherwise, the whole thing crumbles, and we're gonna reach a tipping point where people will see, I have to use models that generate secure code because I'm not gonna spend my developers at the, the rate I'm paying them. I need more developers than are on the planet to support this code. So we've gotta find a better way.

That's why I'm saying we don't give developers the right tools yet. All right, guys, they're taking Betson Vegas on Legacy Code, Armageddon because all these AI tools, and we'll see what happens from there. We'll be back in a minute.

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, folks, we're back and kind of continuing this conversation, but we have an article on Security Boulevard talking about all the ways that Gentech AI might be insecure.

And I gotta tell you, it's a really long article, so you might wanna just jump in and take a look at that. But, um, Kate, I know you looked at this and, um, and, and you've been thinking about this subject for a while, but it seems a little bit like we're way out ahead of this one and maybe not thinking about the security implications once again. Uh, absolutely.

Uh, one of the things that I have talked about as we talk about ai, um, agent AI systems, and this just goes right into, um, just this massive scale, you know, changing the attack, um, vector and using, um, this new threat landscape from an AI perspective and now identity based vulnerabilities at scale. I, you know, I, I hate to say I love it because there's just parts of this that just, I've had so many issues with. Like, I've had issues with zero trust forever, and this, for me, um, really highlights the zero trust, um, issues that I've had.

And one of it being with identity, you know, identity aware attacks. So, I'll, I'll hand this off to Chris very soon, but I will say that the top three, um, areas that I have always been asked about when talking to C-level people have been, um, privilege escalation, uh, command and control and lateral movement, and with identity-based vulnerabilities with, um, agent AI systems. I mean, this plays beautifully into that.

So, wait, I just wanna check one thing here. You love this in the sense like of a tornado has awesome power, kinda love, or what are we on With that? You know, it just, um, so after coming back from this, from my sabbatical and looking at, it's the same attack methodology.

It's just bigger. It's just faster. It is just so, I love how, how it, it threads through how it, how it, how it's connecting.

And you know, we have to be proactive, relentlessly a proactive security strategy. Forget about resilience. I, I forget about, let's just throw that out too.

I believe in having, if we're gonna win, we have to have a proactive strategy. So I'll leave it with that. And thank you for predicting that I couldn't leave this one alone, right?

So I I was saving this from the last segment. So, uh, uh, you know, uh, in Socrates and thymus in the Republic, you have this great argument about whether might make, right, might, makes right, or whether role equals responsibility, right? And I've always loved that one.

You know, if you are, you know, a, a a teacher, your focus is not on making a living and everything else, when you're being a teacher, your focus is on the subject. And so I will stack the responsibility all the way back, starting with investors, you know, looking at you really clearly boards absolutely no excuse whatsoever, uh, c-suites, you took the job, right? You agreed to it all the way down before we can start picking on, you know, Mitch, you know, you know, think about the dev folks.

You're exactly right. You know, they're motivated to not get fired, you know, keep their job, feed their families, you know, get through life and the business structure they've been put into leads up to all these problems. Absolutely not their fault, right?

It goes all the way down. And this one, yes. You know, we salt typhoon, right?

We all know salt typhoon. We've heard the phrase, did we, we were doing a lot of analysis over this over the weekend, and this is to your point, Kate, right? So, but all this, all typhoon is the difference between what, you know, the elders in, in our security community said we need to do all along.

And what we actually did, you know, what percentage of vulnerabilities are exploited? When I look at salt typhoon, I just assume a hundred, you know, how many configuration files in A PLC in a water system have been already compromised and not actually triggered all of them. All of them.

So the only way to deal with this is, I think it is where you're going, Kate. You know, I, I love it because I think the solution to salt typhoon in all of this is more ai, you know, let you look at these systems, find that gap, enunciate it, say that we know about it, and deal with it instead of pretending it's not there, because people will always use it against us, which is obvious. Let me take that, Maria.

Hold on. Mitch. Mitch, I'll let you go with, but Chris, go ahead.

Did you just, Chris, did you just blame capitalism for this problem? I think you did. Go ahead.

Stop. I'm a capitalist pig, right? This is the, this is point that I've been trying to make in business.

All along all this, you know, being nice, you know, ethical business stuff. I've been arguing all my life. To be clear, billions of dollars, 90 plus percent, you know, has, has worked multiple times because it is better, faster, cheaper.

I like capitalism. I think the way we're running it now is stupid and inefficient and expensive. And we're losing business opportunities because we think just being cold and calculating works, it does not, I'm just saying, I just need one more.

And I have a majority communist podcast. I don't know, man. I, I'm watching, you know, Chris, the Reagan publican arguing with the social democrat over there.

So I'm not quite clear where this is going to jump in here. Yeah, go ahead. So, so let me throw out a a an odd way to look at this, but it's actually a positive.

And you said more ai, Chris. So in the theory of constraints, when you remove a constraint, right, that creates flow through a system. And in, in effect, if our constraint is how much code we can generate, how fast because of humans and how fast we can do it, and AI can do that in, in, you know, exponentially or multiples of that.

Okay, well, that removal of that constraint now moves the problem elsewhere. As I was saying before, the next con, one of the next constraints down the line somewhere is people can't fix vulnerability problems. Um, so, so in, in a way, in maybe a perverse way, but in a way, us generating so much volume of code will force us to, to fix this problem without involving humans in it.

It might even forces Kate to implement more of the zero trust framework. 'cause we just don't have a choice. There's not a human way around to solve it.

And the current methods of incident response and everything else, you know, the three, the three things, three deadly sins, we continue to, to commit that you talked about, you know, can't continue at that kind of a scale. So it may force us to look at very different solutions. Yeah, I live in, I live in the New York City area, and I would tell you that, you know, breaking things doesn't lead to forcing a solution.

I said it was perverse way. I'm like, I'm trying to look at a, this could be a positive. Yeah, I actually believe that that's possible.

Well, Mike did this, An overwhelming thing, right? You know, so this is, this is why I'm always ran about inevitability, curves. Everything you said, Mitch, you know, is happening regardless, we can just wait a little longer or we can, you know, you know, and if I was in depth right now, I might just do it on purpose.

Let's just amp up the amount of code because very soon, I would say in the recent past, we can't do it the old way anymore at all. And in the future, we're absolutely not going to be able to stop trying and I'll finish my ran and let you get in, guy. But you know, Kate, you touched on zero trust, you know, love it, hate it, you know, it's a stupid idea, but it's what we had to work with.

That's not how any of this works. We need to have appropriate trust, and that has to have appropriate attestations or records so that automated tools can be relied on and not just optimized by, again, investors and board making bad choices. Okay?

And, And, you know, to, to circle back and to con or to continue with zero trust, the idea that, um, and something that I've seen it forever. I mean, stolen cred credentials, manipulated tokens have not been enough. And we continue to push that story where it is.

But the problem is, is that the, these AI agents actually become double agents. And, you know, and totally, you know, we can exploit the entire system with that. And the reason why I love it is because it, it finally breaks down that myth, the myth that this is good architecture.

It, it just, it's not sustainable. That's a problem. We have non-sustainable architectures, and that is something that we need to change and we need to actually change it.

You know, to your point, Chris, with, um, a and Mitch, I mean, we talked about ai. We, we do need to change this on how we think about it. Well, and it, it'd be cliche to say it, but, but again, you know, I i, I make the point quite often that, that someone with my inclination and skillset will get your information, will always right, highly motivated enough.

I'll just stare at the organization, find the weakness, go in and take it it, you're right, Kate, with these tools today, and you know that, 'cause every time I said that's not literally true. Probably, you know, you should expect it absolutely true today. You know, highly motivated, skilled adversary with good AI skills and good tools will absolutely exploit any vulnerability you have.

As before you can get lunch today. Oh yeah. So you have Colonel Sandra's secret recipe.

What's where you doing with that? Yeah, I'm not a good cook. I wouldn't, I wouldn't know what to do with it.

What There's gambling going on here. Yeah. Hold on.

I know. Just a second. I know.

That was My First flood from Casa Blanca. Yeah. So, so it's all well and good to stand up and say we need a new architecture, Mitch, who's in charge of building that new architecture?

Who's gonna wake up in the morning and go do that? And how do we get that done across, you know, thousands of enterprises? So, so in a systemic view of it is, if you, if you believe that we're gonna be generating so much code that it's generating so many vulnerabilities that it's, it reaches an intolerable point.

The people won't buy your product or won't use your product. 'cause they can't afford to introduce that many security issues into their environment. Guess what?

The vendors have to respond to it and they just haven't had to yet. Right? It's, it's kind of to Chris's point of we don't need security yet, so we're not gonna fund it, right?

We haven't needed it yet. And so Alan talks about this all the time. You know, the, the, when, when security is important is when customers start asking for it, right?

That's one of, one of his investor friends said, it's kind of the same thing when the market says we have to have this, um, we wanna use your stuff, but it's just in, I just can't implement that at that scale, at that speed. I'd have to throttle it down and then why am I paying for, for a, a tool that'll go a thousand miles an hour and I'm running it at 50. So I, I think, I think the dollar is what solves the problem.

May maybe not solves it, but moves the needle to addressing it. Because the market won't buy products if they can't maintain what's being put out there, even if they're using AI to maintain it. All right, guy then to Mitch's point, how long do we have to wait?

Again, gambling in Las Vegas for this cataclysmic AI agent that gets hijacked and takes over some businesses processes and results in, you know, a I guess these days it would have to be what a $50 million error that to, for anybody to take notice and not treat it just as if it was, you know, the cost of doing business. What's that number? I'm gonna answer your question, but I want to ask you a question first, which is, did this story, um, make you feel sad or neutral or good?

Like remember the first segment that made you feel like hopeful? What, what, what was your reaction to this story As, as, as a journalist, I'm with Kate man, I'm like, wow, this is gonna be one hell of an awesome traffic accident and it's gonna be, you know, this is going for months. Yeah.

So, but as a, but as a, you know, as, as a person invested in our technological process, you're like, oh no, this is gonna be a huge traffic accident. I think that this is a neat opposite day because I disagree with you yet again. And I'm gonna signal an unusual, uh, uh, ring of hope here.

And that answers your question. This traffic accident is going to happen very fast compared to, I mean, MCP itself is like not even a year old yet here we are. And I take the speed with which there is recognition of the dangers to be a very hopeful sign.

And I don't think it's due to the dollar, Mitch, I hate to say it, but I think it's due to like culture distrust. Nobody wants robots to be in charge. Everybody's getting suspicious of ai.

Now I'm using these terms, I don't like to use nobody, everybody, but don't you all feel this general sentiment? And I think there's survey data to back it up that people don't trust where AI is going a mere two years after Gen ai, like less actually splashed upon all of us and started taking over the world. So I think that this has the potential, it makes me hopeful to be one of the few areas of technology, maybe the only major area of technology I have seen in my career where people are gonna start to get proactive because their friends and neighbors and everybody else are do not trust where any of this is going.

And I So you're saying The future looks like Blade Runner? Is that what you're saying? You know, we're gonna hunt down the Replicants, I think, I think it's our best chance to avoid the future looking like Blade Runner really?

Well. Have you seen who's in charge lately? 'cause I might consider the robots.

I mean, go look at umpires, you know, the strike zones are all over the place and after every game everybody's like going, well, maybe we should give robots a shot. I'm just saying, But I'll riff on that the same way I was going to because that's, you know, to, to Diane, uh, and Kate's points, the forcing functions we're getting forced into this point to make these decisions, right? And you know, it's not about the security perimeter, you know, to zero trust is about trust apology, right?

You know, everything you just said, guy, literally your neighbors. This is not just tech people. Nobody trusts anything.

These, and, and to your question, Mitch, at the rate of things I do not expect, uh, one way or the other, you know, not to have that, that the, the, the question you asked not resolve itself before this goes to air tomorrow, right? That's the world we're living in. 12 hours, 24 hours.

Yeah. That's as likely to happen in the next 24 hours in the next week or two. And it's not gonna be a year from now that's gonna happen this summer.

Maybe today. Yep. It's gonna happen.

Announce it Sure could. Yeah. Yeah.

Yep. All right. Well I'm gonna leave it there.

And you know, what was that like from the sixties? My kid was just in the production of hair at summer camp and so they were doing the whole thing, but trust no one over 30. And now we're just like, trust no one.

Right? That's where we're Trust. Trust appropriately, right?

That's, that's the thing, you know, don't trust because of this, a digital signature that doesn't mean anything. All right. Hey guys, I want to thank you all for being on the show, as as awesome, great conversation.

I wanna thank you all for watching this latest episode of Textron Gang. Please stay tuned for the rest of the lineup for the Textron TV is right behind us, and I promise you it's gonna be just as awesome. We'll see you guys next time.

Hi everyone. Welcome back to you to Techstrong tv. Glad you can make it.

Um, our next guest is no stranger to text Strong tv. I always love having her on though. She's Christine Yen.

Christine of course, is the co-founder and CEO of Honeycomb. And Christine, welcome back to Text Drunk tv. It's great to see you.

Thanks, Alan. It's always great to be here. Thanks for having me.

A pleasure. So, you know what, Christine, not everyone watches every, I don't understand why, but not everyone watches every episode of Text Drunk tv. Maybe they haven't seen yarn here before.

Maybe they're not familiar with Honeycomb. Why don't we start with you and transition in a honeycomb and get them up to speed? Happy to, uh, my name's Christine.

I used to be a developer today I am the CEO and co-founder of Honeycomb. Honeycomb is an observability platform that helps you figure out why your code is not behaving the way that you expect. That's the simplest way I know how to describe it.

I love it. Can I play a therapist for a second? Sure.

Have a seat on the couch. What do you mean? You used to be a developer, do you not consider yourself a developer anymore?

Oh, Alan, uh, there was a point in time when I thought about how much more time I spend in docs and slide decks and emails and, uh, You and me both my friend, you know, I know I Finally, I'm like, I've carved out some time this weekend actually to really get to go play with Claude Code, and I know part of the promise of of these tools is like, anyone can be a developer. So, so maybe it's just an easy way for me to get back in. But, um, you know, I think of used to be, or still like, I, I still see that as so much of my identity for why we're building honeycomb, why this is important to me.

Um, I have had those experiences breaking production and having the ops team be very upset, and then me as a developer not knowing how to interpret their graphs or how to, you know, work with their workflows. And so I think of, I certainly think of past Christine as, um, a real motivator for making sure that observability tools are accessible to everyone on the engineering team and are a real channel to bring people together. That was just the response I'm looking for.

Oh, Oh, good. Right? Good.

Because it, it's fine. I, I was talking to someone the other day, so I'm at the, at the end of the day, I'm a gadget boy. I just like diving into technology and playing with things up to my ears, right?

I roll up my sleeves, I want to get dirty with it, you know, and I spend so much of my time looking at financial projections and pipelines, not the kind of pipelines we'd like to look at, you know what I mean? And, and all of this stuff, instead of talking to people like you or writing about things that I find really interesting. And yeah, it kind of, that's the downside of being a founder, being a CEO running a company is you have that pre CEO person who's still like the little child in you, but you now you're an adult with all these other things you gotta worry about and take care of.

Um, it takes getting used to, but it, look, you're doing a great job of it. Thank You. It takes getting used to, but it's also then, you know, at least then you get that emotional satisfaction when you're like, oh my gosh, this would've helped past Christine so much.

Or like, yeah, you know, past Alan would be so excited that you get to nerd out on this today. Um, so I, yeah, I mean, we all change. We all grow.

Yeah. Um, yeah, I've always, I've always had in my head like the important things that you're having fun and are happy with the outcome of how you're spending your time, but agreed, now our therapy hour is getting, getting a little bit long. All right.

That's it. io, is it? That's right, yes, that's what I remember.

Um, so you could find out all about Honeycomb if you've ever watched not only Christine, but Charity or anyone else from Honeycomb here on Textron You or read of articles, we where we cover them, it's all on there, but go to Honeycomb. I you get smart about it. Christine brought you in today because you guys recently announced an MCP server for AI agents, which MCP servers usually are for, uh, available on the AWS marketplace.

Talk to us about it. Yeah, so I know on the surface that's, uh, a while to think about it, but it feels like a pretty table stakes thing right now. Everyone's got an MCP server, you know, you take your API, you make it available and readable by LLMs.

Great starting point. There's a couple different different angles that make me really excited about what it means for honeycomb observability first, um, sort of from a user perspective, the team that built out the M-R-M-C-P server was really thoughtful about not just exposing our API but adding a bunch of different tools that LLMs will need to be able to orient themselves and navigate schemas and make connections that we are so used to. The AI is doing magically, and so, you know, there's a lot of the, this early exciting thinking about new ways to work with schema, new ways to work with telemetry data that secretly, I'm hoping and expecting will be feedback into our UI experience, because there's just a new way to think.

Um, one other angle to think about it is one of the things that's become very clear as we've played with these tools and really worked with LLMs to interrogate existing services, is that there are two qualities that make an MCP server really pleasant to work with one speed, right? I think we're all sort of getting back into this habit of firing off some tricky question to an LLM and going and getting coffee and then coming back and being like, oh, are you done yet? We know the faster we can get your LLM answers to whatever sort of incremental questions it's trying to ask your service, the more effective and pleasant that interaction will be.

Well, since the beginning, honeycomb has really, really optimized for speed. Uh, one of our sort of internal engineering mantras is FAST is a feature because we know that no root cause analysis, known incident response, no curious engineer wants to be sitting there for five minutes waiting for a query to finish. We left that back in sort of the battle days, and it's translating really beautifully to having a better experience through LMS and cps.

The second thing thing that really contributes to a rich MCP interaction, not just Demoware, but once something that like really feels right is the richness and the depth of the data that you're working with, right? We know by now in an AI world, data is the oil that drives everything along. Uh, and with Honeycomb, because so many of our customers have been able to send data to Honeycomb that is rich, that maps to their business, that use, that talks about their customers or talks about like aspects of the business, in their words, it means that they're better able, the, the, the l even the off the shelf l LMS can do a better job of mapping an English level question to the telemetry data that they have in Honeycomb way better, uh, and much more naturally than they might with sort of a traditional more constrained observability tool.

And so with that speed and that flexibility, um, it's, it's turning, playing around with something like honeycomb into a really exciting and fast experience. Absolutely. It's just the beginning.

I, I agree with you. I mean, look, it's changing. com, something along, I forget the title now.

Is DevOps becoming an AI first DevOps, or is AI first DevOps now the kind of norm is AI first honeycomb, I think, Is honeycomb an AI first? You know, I think we are internally, we are definitely leveling up rapidly across the, across the organization in our AI use. Um, I think this question of whether DevOps in general is becoming AI first is, is a good one.

I think there's gonna be certain workflows and certain personas that reach for AI first. You know, when I think about our dog food instance, again, I'm embarrassed to admit, um, I don't know all of the fields and the schemas and the, the perfect questions to ask anymore. But at the MCP server, that translation layer makes it effortless.

I'm probably gonna be an AI first honeycomb user from now on, which is, again, very scary. And God, very exciting to think about. For engineers who are really deep in the flow, who know exactly what they're asking for, who kind of don't wanna deal with potential misinterpretation, I think that there's gonna be paths that are not AI first, and that's okay.

And the exciting part for vendors is now how do you find that balance? Who, who falls in which buckets? How do you, uh, you know, innovate in each one and then figure out the right pieces to cross port into the other?

We'll, see, I, it still blows my mind that four months ago, MCP was like a brand new thing that I know. And now it's like, it's of course, it's the standard. Everyone's doing it.

Yeah. It's wild. It's a wild world.

We live in, you know, another article by the time this interview is, is up, but it's another article I just finished. It's called Daddy, where does Software Developers come from? And, um, because, you know, at the same time we're doing this, what we're saying is what we're kind of saying is, Hey, if AI could do that job better, let AI do it.

'cause that's gonna, you know, for our senior engineers, our senior developers, our senior people, they're gonna leverage ai. But then you read about, you know, Microsoft and sales. Microsoft laid off all these people.

Salesforce isn't hiring any more quote unquote junior developers if we don't train these junior developers to do the kinds of things you were just talking about, Christine, what's gonna happen when these senior developers go away is just ai do it all. And then are we at the mercy of, of what the ever the AI tells? I mean, where, where's tomorrow's developers and, and observability engineers and DevOps engineers and platform engine?

Where are they coming from? I think we as an industry have a, have some thinking to do about what fundamental skills are actually necessary to work with these ais and which skills we considered were fundamental, or actually not anymore. I think it is not hard today to find someone to be like, oh, well, writing code was never the hard part of being a software engineer.

Fine. Uh, but I am hearing stories about baby startups that are hiring folks who weren't software engineers and just expect them to take LLMs and go run vibe. Yeah.

They're vibing and, And, and it, it's always the day two, day three debugging judgment pieces that are, um, that, that start to hit some road bumps. And, uh, I think it's kind of shortsighted to do really blanket. We're not hiring, um, junior engineers anymore, but I, those companies, every company should be thinking about what are our expectations for junior engineers, for someone new coming in, who has these AI agents who's sort of grown up with them over the last two years?

What, Or four months? Four or Four months. Like where, where do we, where do we expect them to lean in?

Um, charity's got a post that she's drafting right now. Maybe by the time this video is up, it'll be published about, uh, durable code versus disposable code. Mm-hmm.

And sort of sort of playing with this idea that, uh, not only are there these two buckets of code that in the past we sort of used the same work, uh, skillset to produce. She's like, maybe there will be a world where disposable code and durable code have entirely different workflows, entirely different skill sets, vibe, coding for disposable code for like little toy apps or things that you're not expecting to, that you kinda just want to get up and running. Awesome.

Vibe code for that, for the, I don't know, airline booking systems, or, this is actually a terrible example, but I'm gonna say it anyway. Air traffic control software that you need to be durable. I know.

Mission critical. Yeah, mission critical stuff. Like, it's, it's, it is a different set of expectations, and we have not been, we as an industry have not been incredibly introspective about what No.

What to look for in each one. Well, Not only that, we're repeating the same mistake. So I'm a security person.

I've been in security 30 years long before I got involved as a a media person, right? And, um, this is the same mistake we always make when it comes to security. We, we'll, we'll get to security.

We'll bolt it on. Don't worry. You know, I, I wrote in another article I wrote last week, the s in Vibe coding stands for security.

There is no s in Vibe coding. And, and you know, we, and we're gonna pay the price for it. Make no mistake, you know, we will all kind of pay the price until someone says, oh yeah, we should really do something about that.

Um, but, you know, fools rush in where Wise Men Dad or Tread or whatever the saying goes, maybe there's a little of that going on here. I just a little, I am, I feel like there's so much to keep up on the, the, the, the part of, um, the set of changes we're all experiencing right now that I, I seem to not be able to get enough of is the, how is it changing our teams? How is it changing how we think about this craft?

How is it thinking about the skills that we value? Um, and yeah, I can't, I can't wait to see sort of how things shake out. Absolutely.

It's, you know, the old Irish proverb, right? May you live in interesting times. Indeed.

It certainly is interesting times, and it's a great time to be a CEO, right? I mean, to really, 'cause you gotta think about, you know, if you were just a developer, old Christine, little Christine, whatever you wanna call it, you would just be thinking about how does this affect code? Or, or stuff like that.

But you gotta, you, you know, and today's Christine needs to think about all of these kind of macro things and, um, You know, a secret about me, Interesting stuff. I really like to get excited. I just, it's just, it's fun to have like new things to get excited about.

And it feels like right now, everywhere I turn, there's like something cool. There's tons of hype over here. And then holy cow, this tool actually lives up to the hype.

Uh, it's, it's, it's a fun time. It's a fun time to be where we are. Me.

Oh, I'm, I'm, yeah, I hear me too. Me too. No doubt about it.

But some days I wake up and just say, wow, I can't believe, I can't believe that what I live through right now. Like, you know, I mean, the things I'm able to do that I never dreamed Yeah, right. That I'd be able to do kind of stuff.

But like, what are the implications? And I, you know, I, I get this like, are we all gonna end up like Wally, you know, Les running around on, on server chairs with remotes, but we'll see how it turns out. But honeycombs on top of it, that's the important thing.

Um, hey, this MCP server, we mentioned it's available in the AWS marketplace in the AI agents and tools category. Where else? I assume it's available anywhere else or No, Absolutely.

It's open source. Folks can, you know, you can host your own. We also have the hosted version, so you don't need to deal with setting up bot and all that on your own.

Um, you know, our goal with Honeycomb is always, we wanna work, work where you work. Um, and so we had the easy buttons and then we have the, like, set it up yourself. Cool.

And it's Honey k Mayo. Christine, it's great seeing you. If nothing else, maybe I'll see you at CubeCon in Atlanta.

I guess that's November or something. That's a, maybe that's a maybe. Uh, but May a solid.

Maybe solid. Maybe you might be sending your AI there. Who knows?

Who knows what'll change by then? It's true. Yeah.

Well, let's make sure we talk before then. Anyway, continued success at Honeycomb. You know, I forgot to an ask.

This MCP service is out now. Indeed. All plans, including our free one.

Great. Thanks for coming on. Say hello to Charity and everyone else.

Keep up the great work. Christine, always a pleasure to have you. Likewise.

Thanks so much. Alrighty. Christine Yen, co-founder, CEO at Honeycomb here on Tech Drunk tv.

We're gonna take a break. We'll be back. Hey guys, thanks for the throw.

We're here with Emmy Linder, who's the newly appointed CEO for stairwell, and they specialize in analyzing data for cybersecurity purposes, hopefully using something that feels familiar to folks like a search tool. But, uh, there are new challenges and new issues, and we're maybe on the cusp of the age of ai, so who knows what's gonna happen next. But Emmy, welcome to the show.

Thank you. Happy to be here. Alright, So you are taking on this role, the current CEO and founder is moving over to be, uh, the CTO, I believe his name is Mike Wak.

But what attracted you to Stairwell and, and what are you gonna try to achieve? Yeah, so Stairwell, um, you know, when I first, uh, started talking with Mike a few months ago, um, the things that I really found compelling about Stairwell is that there's a lot of really strong components within the company. First and foremost, Mike himself is an amazing founder, huge, uh, incredible cybersecurity background.

NSA Google, um, Chronicle. But that was one of it. The others that I really was excited for him to stay on and wanted to be, remain very relevant and very involved in how the company continues on its technical vision.

Um, top tier investors, solid runway, obviously security is always a good space to be in. And with ai, it's even worse depending on your perspective, right? But from the CISO perspective, defending and protecting companies has become that much harder.

So they need more innovation and more technology. And, um, and I think where the company is and where what I can bring to the table is a really good fit in terms of Mike staying on with technical vision. And I'm coming in with operational excellence strategy and the ability to kind of match the product and the company to what the market needs and the pain that we're trying to solve to, for CISOs, right?

A lot of people have long said that security is really a data management problem in that sense that we're looking for anomalies, AKA needles and haystacks and, but we don't really have the tools to go after. And, um, is that changing? And, and where does stairwell fit in that conversation?

So what Stairwell does, which no one really does, is it retains, um, files, executable files forever. And, um, it's not, that is normally no one can actually do it. Sims or eds or other security tools kind of have moment in time windows, and then they kind of vanish and they move on further into the future.

What Stairwell does is once you've, once you have it, it will forever look back and will continue to collect forward so that you can actually search and see whether you have, um, anything bad or were ever impacted by something bad. Moreover, it can, it can find variances variance, which is really, um, that is something that is very unique because you can have a known bad or a known hash or a known, you know, some sort of indicator signature, et cetera. But to be able to say, okay, based on these parameters, this is going to be, you know, the difference between, uh, what is coming and going.

Um, that variant analysis is something that no one does. So those two things combined are actually one of stairwells, um, key differentiators, data forever, and ability to, to look and understand variants across the board. And that's critical because in my mind, we always have these moments where somebody says, we've discovered a breach.

And then the next question is, well, how long has this been going on for? Were you attacked? Are you, do you have it on your environment?

Exactly. And if you, even if you, if you had it, great. So that's an answer.

And you, you kind of know, but you really wanna be sure. And many times, if you, if you go to your environment, if you go to your sins, if you go and search your a DR tools, you might get it not right now. And the not right now is not helpful because depending on what it is that you're protecting and defending, sometimes you need a definitive answer and then you need to go back and fix it.

And many tools don't have that ability because they don't look so far in the past. And variants and, and executables, they know how to linger, right? They know how to kind of simmer and pop up when needed.

And so no one does that. And it's, uh, it's, it's a really key differentiator. And has the tenure of the conversation changed among the regulators?

'cause it seems to me increasingly they want a definitive answer. So it depends. I think it depends on the, on the industry, it depends on the appetite, both budget and, uh, and security know-how and sophistication.

But for the most part, um, those that have, uh, critical infrastructure, critical data that they need to protect, always wanna know the answer. Regulators come in, and it depends on their answer. It's not necessarily the case.

But if you really wanna know a hundred percent, there's no other way. Mm-hmm. So as you look into the future, you know, what's on the agenda for you guys?

Where do you go from here? What's top of mind? So what we're, what we are focused on right now is making sure that we clarify what it is that stairwell does and what problem it solves for CISOs, especially in the AI world where attacks are, you know, we used to say attacks had, you know, they had to be low and slow, taking months and weeks in order to be able to not spark any kind of alarm or alert for, for socks to find in the world of ai, you're in the fast and furious.

They're very good, they're very frequent, they're very believable, and companies and enterprises are constantly under attack in that world. You need more defense, not less. And you need to be, um, you need to as usual, right?

There's always a resource problem with security. You wanna be able to get to the answer fast. You wanna know if you have to do anything, yes or no.

And if so, what is the best course of action? And that's what we wanna make sure that companies are aware of and what stairwell brings to the table, because no one does what we do. Mm-hmm.

Um, you mentioned ai and it occurs to me, I Mean, how could I not, how could I not? There you go. But we got all the way into about 10 minutes or so without mentioning, yeah, we go, um, but there will be AI agents going forward, and I think it'll work out this way.

Won't I, as an security analyst ask the AI agent to go find something and search for something, and more than likely that AI agent's gonna call stairwell? Is that how that's gonna work in your mind? So if, if you have stairwell, the AI agent will be able to go into the data and understand whether or not you've ever been touched by that bad thing.

If you don't have stairwell, you won't be able to have that answer, even if you have a thousand AI agents there, and that's where the uniqueness of stairwell shows it. You, you first and foremost need the data, then you can do a bunch of different things. We have the data, we store the data, and we supplement it with a lot of additional information, not only from your environment, but from other, other feeds, Right?

Of course, when any, lemme try that again. Of course, whenever somebody says, we store something forever, everybody's mind immediately goes to, well, how much does that cost? 'cause storing stuff for a long time is expensive, so how do we kinda store everything forever without breaking the bank?

So this is where, this is why I love Mike, because the way that he, that they built it and that he and the team built it, I, I just walked in the door, right? But they built it in a very efficient way using a lot of kind of, you know, they're all Google background, right? So everyone's got that search and the, and the billions of people that need to use it and access it type of mentality, um, they found a pretty efficient way of doing it.

So once you get the data in, the way that we ingest it, the way that we store it, the way that we can call on it, is really efficient. And the storage costs the cogs. And believe me, that was one of my first questions, pretty good.

When you kinda engage with these customers and you talk to them, what are they telling you their number one pain point is? So the number one pain point is, has been and continues to be time, resources, and, um, too much information that you, that you can't actually sift through. And that, that's been the case for a while in the security world with, with, uh, defenders specifically.

And that's still the case. And again, AI is making everything worse. And so anything that makes it easier, less time intensive, uh, gets you to the bottom line, answer faster is a positive.

And again, within the noisy space of security, you need to make sure that you provide an end-to-end type of value prop. So that is why partnerships is something that we are definitely looking into because it's a way for us to integrate into workflows that already exist. Alright.

Of course, this is not your first cybersecurity rodeo, but what were you doing before stairwell? Uh, so, uh, my previous role was COO of Cybereason, and, uh, last couple of years I've been doing a lot of advisory work for startups and, uh, and operational coaching as I call it, for CEOs and COOs. All right.

Um, so what is that one thing that so far in your experience when you talk to customers that you see them doing over and over again, that just makes you shake your head a little bit and go, folks, maybe we should be a little bit smarter than that. Yeah. Um, I think what, there's a couple of things, of course, but generally speaking, um, some, some CISOs and some companies want to do the bare minimum just to basically kind of be in compliance and, and be, you know, do just check the box type of security.

It's just not enough. Not in the world that we live in and not in the amount and level of attacks that are out there. And so we need to be, you don't need to overdo it.

You're not protecting necessarily, you know, government affairs, but you definitely need to be able to protect your business with tools that give you all the information that you can have in an easy way. All right, well, folks, you're hearing it here. Anytime there's an attack, regardless of how successful is, there are five questions that get asked, and they're generally who, what, when, where, and why.

I think you need some tools to go answer those questions, and that's not gonna happen if you can't actually search anything. Hey, Emmy, thanks for being on the show. All right, thanks a lot.

All right, and back to you guys in studio. Hi everyone, and welcome to the six five Summit AI Unleashed. I'm joined today by nicolo Desi, CEO at Ion Q for the Quantum Track opening keynote on AI and the road to quantum commercialization.

Nikola, it's great to have you back, Honor and pleasure. Always. Yeah.

So, you know, the, the event is all about AI unleashed, but you can't really say AI and look at the future without talking a little bit about Quantum. So, you know, we've had a long relationship. Um, you know, we've, uh, we've known each other for a long time.

You have been watching the commercialization of quantum take place. You're doing deals, you're expanding the company. Let's just start off, you know, give us a little bit of the kind of quantum state of the union, because as AI continues to proliferate, we are seeing quantum interest growing bounds and leaps and bounds as well.

It's understandable. I mean, I, INQ has really led the market, uh, in terms of quantum ai. Uh, on one World Quantum Day, we had an event to the NYSE, uh, back in, in March, April, uh, where we, we put, put out a couple papers on in what we call industrial ai.

So quantum, uh, machines that we already operate, partnered with, uh, a couple, let's just say manufacturing industrial companies. Um, we showcased, uh, what we call narrow quantum commercial advantage, where we're able to improve a classical LLM generative adversarial network, um, by, you know, a, a meaningful, uh, albeit single digit percent, uh, on our current machines. We also showcased, uh, at the same event, uh, an ability to provide training data for industrial AI applications that have a lack of such training data, right?

So already Quantum is a partner for classical LLMs, and I think that what we're gonna see in the coming year or two is the energy advantages of Quantum, as well as the unique insights that Quantum provide, uh, are gonna be more and more valued. Yeah. And, and you have been very focused by the way, on expanding your footprint at IQI think over the last month, probably from the time that, uh, the event kicks off here.

You've made a couple of deals, I think, well north of a billion dollars in, in, in m and a, um, also some strategic partnerships kind of around the world. Talk a little bit about kind of how you're thinking about strategic deal making and kind of how you're thinking. 'cause a lot of your partnerships have been with governments, uh, you know, who are very interested in, in making big investments in Quantum.

So give us a little bit on the m and a and then a little bit of these sort of strategic partnerships. You're, you're, you're signing around the world. Yeah, sure.

Now, we, we've announced, uh, three acquisitions, uh, since I took over as CEO on February 26th. Uh, the first one was a business called, uh, Lightsy. Uh, and that is a quantum memory company, uh, spun off from Harvard University.

And Lightsy actually does two things for INQ it, it supports not only the scaling of our quantum computers, which is the core business, um, but it also supports the extinction of our quantum networking business to longer distances, uh, which supports our quantum networking business. So all of a sudden we can go from 10 kilometer, 20, 30 kilometer, uh, you know, kind of citywide quantum networks, uh, to hundreds of kilometers with their technology. Um, on the compute side, it accelerated IQs roadmap on its own.

Uh, and we closed Lightsy, uh, just a few weeks ago actually. Um, we also announced the acquisition of Capella, which is a satellite signals intelligence, and, uh, we will turn it into a QKD satellite business. Um, and the reason that's important is that it, it impacts the ability of communications to remain secure as we go from ground to satellite, satellite to satellite and satellite to ground, which the most vulnerable parts of the communication roadmap, uh, that we all rely upon for civilization as we know it.

Um, the third acquisition we just announced, uh, very recently, um, and that was, uh, Oxford Ionics, uh, on June 9th. Oxford Ionics is a biggest transaction that, uh, INQ has ever, uh, announced in its history. It is a transformative deal for both the US and the uk.

Um, and the reason it's transformative is we're actually getting the combination of INQ standalone plus our light syn friends in Harvard, uh, outside Harvard and Boston, plus our, our new friends at Oxford. And, and one plus one plus one is, is not three, it's like 30 or 300 or 3000, uh, at the rate that we're scaling. Exponential.

Yeah, it, it really is probably doubly exponential because as you know, Daniel, every time we add a Q cubit, we double the compute space. So it's not just Moore Moore's law where you double every 18 months. It's like we could be 200 million times bigger, uh, in 18 months.

Um, and I think that's one of the things that, you know, you, you, you may well understand, uh, you know, in your gut a lot better than industry observers typically do, which is that humans don't have brains that understand exponentials very well. They really don't understand double exponentials. And so the difference in compute power, every generation is actually gonna steepen in quantum computing and steepen at a, a, a pace that is, you know, very difficult to fathom as we go from 36 qubits to 256 to 10,000, 20,000, 200,000, and we announced a roadmap, uh, on June 9th that takes us out to 2 million qubits in 2030 and 80,000 logical qubits.

And we think that puts us, you know, 50 times better in logical cubits than anybody else. We think it puts us potentially billions of times faster in gate operations, uh, per clock tick than anybody else. Um, and it really is showing that our inorganic strategy is driving synergies around the technical roadmap and attracting talent and enabling us to put together the pieces to scale much the way Nvidia did when it bought Mellanox in 2020.

Right. So light sync is our Mellanox deal for quantum computing. It's just that it's about fo photon memory and photon entanglement quantum memory as opposed to just data center expansion.

Yeah, Well, the talent in this particular space, nicolo, is, is exponential in itself. You know, we've seen across the AI space, the race for talent. You hear stories about meta tracing, $2 million research engineers and losing because they want to go work for something.

You know, I think there's gonna be a very similar, but actually maybe even more challenging talent race in quantum because the number of the quantum physicists and engineers that are going to be available to help solve these problems are gonna be pretty limited. I do think it's also worth pointing out, um, I think the billion dollar deal you did, the just over a billion dollar is the largest quantum deal in the space to date. There has not been a deal of this size yet.

So you also broke a, uh, you broke ground in terms of crossing the, the, uh, 10 zero club. Um, so congratulations on that as well. So talk a little bit about the relationship with Quantum and ai.

You've made a few parallels throughout this conversation. I like the one about the Nvidia network, you know, you know, quantum network, but I think sometimes it gets a little bit lost because people wanna compare these things somewhat discreetly, like is it quantum or is it ai? And just like we found with classical computing, um, and quantum, the relationship is more symbiotic in the AI compute era, there's gotta be a almost an exponential double, triple exponential impact that could be expected over time when the two things actually work synergistically.

Yeah, no, well said. I mean, look, yesterday, you know, on June 9th, we also announced a 20 x speed up in partnership with Nvidia and AstraZeneca and Amazon AWS, uh, around computational drug designed. Um, and so we're turning months into days, right?

I mean, I can't stress that enough. It's, you know, a combination of a quantum computer on the loop with classical workflow in combination have, have allowed us to effectively do something in days that normally takes months. And I would say that that's gonna have a huge impact on, you know, humanity.

Um, you know, your ability to drive more early discovery and early, let's just say investigative work around potential clinical paths and potential candidates to be drugs, um, is going to expand the pipeline, expand the top of funnel, if you will. Um, and we're just getting started, right? It's an AQ 36 system.

Uh, it will be much more powerful every generation, hence fourth, um, hundreds of millions, even billions of times more powerful, you know, in the coming years as we move from early commercial advantage today through to early fault tolerance, through to full fault tolerance. And so, um, I think you're spot on there. Um, we look at this as quantum expanding the overall compute market, and they were expanding the pie and addressing problems that can't be solved otherwise, as well as making, you know, all of this happen with a much lower ratio of compute success to energy, right?

And we, we are not, I mean, Microsoft is, you know, recommissioning three mile island to power their data centers, and that is kind of a sad day for humanity, right? Um, we are not in need of that. Our systems plug into a wall socket.

Um, I like to remind people, uh, Daniel, that, you know, the human brain, at least in my estimate, is probably a quantum computer of sorts for two reasons. One, it only uses five amps of power. It doesn't need a nuclear power plant, right?

Two, it learns much faster on less training data. One of the things I'm always amazed by, you have, you have kids, and so like, you remember your kids when they were six months old, a year old, 18 months old. Um, I was always marveling at my own daughters who could look at a new object in a book, learn the name of it, you know, car, train, plane, whatever.

And then you'd, you'd go outside and they could point at that object if it happened to come across their worldview. And that's amazing on one piece of data. I mean, we have to show cha LLMs, you know, millions and gazillions of pieces of data to train 'em on what an airplane looks like, 2D to 3D real life, right?

Um, and so look, at the end of the day, this is the biggest revolution in computing in 80 years, right? We're moving from classical to quantum. They're gonna run alongside each other for a long time.

You know, we're gonna do pieces of problems that are just not solvable by classical computers, either ever or in a time horizon that is relevant to create value. And over time, the energy advantages will allow us to nibble on, you know, more of the GPU space. Um, but I, I think right now, the next few years is all about, you know, exponential growth and what can do and exponential growth in, in the partnerships between us, you know, and LLMs to make LLLM AI quantum AI breakthroughs happened that no one's even thought of.

And I think the one thing for our audience that's really important to understand is each type of computing does different things really well. That's right. So, you know, the CPU era of computing, uh, you know, running traditional applications, you know, running your ERP system like that is not something you'd wanna do with a quantum computer, by the way.

It's not something you'd wanna do with a GPU AI system either. That's right. AI computers are done specifically developed for certain things.

You know, you were hearing a lot of at this event about, you know, there's the pre-training error, you know, certain types of compute parallel processing yet. 'cause you need those trillions of, of if tokens, for instance, on the inference side or on the training side, you need to see tons and tons of data for it to learn and become useful. Quantum, like you said, needs less, potentially data can do more with it, but it's for specific kind of applications.

So it sounds like everything's very symbiotic now. You kind of, you kind of started going down this path, nicolo, and I think it's really a great way to end this conversation is the debate is always commercialization. Mm-hmm.

The debate is always, you know, when does it become less academic? And a big part of your recent talk track has been, that's now, um, but then if that is now, how fast does it, like, are our, if our kids do go to college, if college is a thing, when your kids, my kids, my younger kids, my older ones are already there, um, when they go, are they gonna be playing with, like, is the average student gonna be playing with a, with a quantum computer? I mean, like, what is the kind of pace that we start to see with these next wave Breakthroughs?

Yeah, I mean, look, you, you started this conversation on, on partnerships and I'll, and I'll bring it, bring us right back to that, which is governments and Fortune 500 companies want the INQ ecosystem. And the reason they want it is they want a futureproof job creation. They wanna future proof universities learning, they wanna future proof research and entrepreneurs starting companies.

And we have, uh, I would argue that, you know, the leading ecosystem in quantum networking and quantum computing working together, uh, at INQ, not only symbiotically between the quantum businesses, but also branching out into, you know, working with the G-P-U-L-L-M and CPU ecosystem. And so I, I absolutely think, look, I mean, Mellanox was bought by NVIDIA in 2020. So, you know, people said something similar in 2020 that, you know, NVIDIA's an interesting company.

It's kind of a GPU graph, you know, graphic frost game company, right? And then chat, you know, chat, GBT went from curiosity to hobby to, oh, goodness gracious, like there's something happening here, right? And so I think we're on a similar time horizon, right?

You're gonna see the inflection in quantum computing happen in quarters and very low single digit years. Um, and you're gonna see it open up, you know, a world whereby, yes, I think five years from today, three years from today, certainly when you're a younger one, you're in college. Certainly quantum computing is on the curriculum.

I mean, look, I'm a physicist originally, Daniel, and when I look at old university entrance exams at Cambridge University where I went to school, they didn't have any quantum on it 40 years ago, 30 years, but 30 years before I went to school, there was no quantum on the roadmap for, for getting to Cambridge. I couldn't do those exam papers, by the way, when I took them, because we learned a bunch of quantum stuff that the guys in the sixties didn't 'cause it was too new a theory or another 30 years on from that, right? So, you know, you're gonna school in the 2020s and 2030s.

Quantum is most of what matters. If you're a physicist, not a small part of it. It's most of what matters.

Um, and I, I'm really excited about the proliferation of technology. I mean, when we iPod four and a half, five years ago, quantum was very much a frontier of science Today, it's an early adopter business that every Fortune 500 CEO has a strategy around. And frankly, every large government agency has a strategy around as well.

And that's big progress in, you know, four years, four and a half years. Imagine where the next 2, 3, 4, 4 and a half years is gonna bring us. I mean, exponential, you know, deepening of that ecosystem and exponential inclusion of more industries, consumers, researchers, students in our ecosystem as well.

Yeah, I feel Like the title is, uh, this opening session here and our, our quantum track is gonna need to be something about double triple exponential. But, you know, your, your final question on the, to go on the way out, nicolo, is can a quantum computer help a guy like me grow his hair back? You know, it might, it might Daniel.

Uh, it actually might, uh, you know, we're, we're really heartened by how successful we are at modeling molecules, uh, you know, for both chemistry, material science and pharmaceutical purposes. Um, and so, you know, as we go and, and we're also good at topological data analysis. So when I, when I, when I look at Alzheimer's, uh, you know, cures and kind of cocktail approaches to solving something, I think we can help with a cocktail approach to that.

Daniel top watch with data analysis. I think we can help with molecular modeling early on. We can explore more pathways, uh, that can individualize perhaps, uh, you know, treatment of, uh, of what, of, of making your head look more like mine.

Um, Right. And, and, uh, and the reality is, you know, we're just getting started, right? So as quantum companies become bigger, we'll be able to mo model larger molecules and more pathways faster.

Because end of the day, molecules are a quantum process. They're not a classical process. They're like the very embodiment of a quantum process.

So a lot, lot of lot come in here in the chemistry and pharmaceutical space. I'd love to see us, uh, us tackle Alzheimer's and, and Parkinson's and some much more complicated, much, much more degenerative type of diseases, uh, immediately and more quickly. I, I tell you what, I think I pull this ball look off pretty well, Nicolo Dessi, C-E-O-I-N-Q, thank you so much for helping us open up our quantum track here at the six five Summit.

Look forward to chatting with you again soon. My pleasure. Thanks so much, everybody.

com slash summit. Stay tuned for more in-depth conversations where we bring you the next generation insights. Isam.

So great to have you back for this year's six five summit. This is become a bit of a, of a, of an annual thing for you. And I, It's, I love this.

And, uh, I do, I, by the way, I genuinely love these conversations, I talking to you. Um, let's, uh, let's jump in because you know, the audience here, I mean, there's so much happening here at the summit. Um, you know, lattice is a company that of course has really grown its presence over the years known for being a leader in FPGAs, but is also becoming a leader in ai.

I mean, you're doing many things in ai, you're involved in data center, ai, edge ai, just give, give everybody out there a bit of the background of what you are focusing on across the AI space right now. Yeah, good question. So, like you said, there's a variety of places that we play when it comes to ai.

So when you look at our AI related revenue, it falls into, let's say three buckets. The first bucket being, um, we enable a lot of the AI specific servers, uh, that are in the data center. So when you think about the AI racks that are there, we play a critical role in control management, uh, and also security type application.

So, um, if you take a, a rack, for example, you're gonna find a number of lattice FBJ devices there. And it's not just enabling the AI specific servers, but we're also in the routers and switch cards, uh, NIC cards, et cetera. So we enable a lot of what's happening in the data center with Lattice FPGAs.

Now, the other bucket of AI stuff that we work on is, think about all the stuff that's happening in the automotive market or industrial market. I think it was Mazda actually that, uh, released a press release about how Lattice is enabling their ADAS systems. And what we're doing is we're taking all these sensors that are being deployed across the world, just more and more sensors being deployed.

And the, the problem statement is that the typical compute accelerator wasn't designed to take in all these different types of sensors. There's so many. It could be an image sensor, a gyro, it could be a pressure sensor temperature.

So what we do really well is we take that sensor into our FPGA, and then we aggregate it. We do like sensor fusion, and then we pre-process it, and then we do some of the secret sauce that's needed to go to the GPU, for example. We've got a partnership, a formal partnership with NVIDIA that we can take that sensor data.

And some of that secret sauce we do is kuda secret sauce, and then we send it into an Nvidia compute GPU, and then we can run the AI on the edge. So we do a lot of that, not just in automotive, but in industrial equipment, as you can imagine as well. So we're in the data path of an AI application.

Now the third bucket is, hey, why not do the actual AI on a lattice FPGA? And there are numerous applications today that are actually putting the inferencing on the FPGA and that has value as well. FPGAs are reprogrammable, you put your secret sauce models, change, update your models, update your use cases.

So where do you find those type of things? Well, if you look at a client device, we talked last year about how, uh, Dell is ramping up with XPS, uh, models and latitude models and the lattice FPGAs right by the camera there, doing a lot of the AI use cases, uh, in, in a client device. We're also in factories.

We do sorting, so object detection for defects. Uh, we're in smart, uh, control panels and factories that recognize the operator coming, approaching it, but also making sure he's hitting the right button that he's not looking somewhere else and hitting buttons randomly. So there's lots of applications where you can actually put the inferencing on the FPGA and the beauty, again, reprogrammable, update your models, your secret sauce, but very small, very low power.

Yeah, I think it's really important too to, to double click there because people might interpret that as you saying, Hey, we are, uh, you know, a accelerator or a GPU, and really you're compliment, right? In so many cases where some, it might be a, Hey, you've got this one or two specific tasks that we can just do better, more efficiently than, you know. And then again, kind of feeding, you know, it's never bad to announce or discuss partnerships that you have with Nvidia.

It seems to be a very popular thing these days. But that you have, and by the way, I remember for a few years now, 'cause I've come to your big, uh, annual devcon and you've been showing this stuff off, you know, sensing technologies related to adas because you kind of said, oh, we're helping with adas. People go, oh, you're doing a dash well, we're helping with adas.

Like, enable, You enable it. That's right. Yeah.

You're an enabler for it and without what you do. And I also have, I've seen over your, you know, for any, anyone out who's not really super familiar, but like everything that's sort of sensing, you know, a lot of people, I don't know if you've noticed, but I know I, I'm on a new laptop here, um, that's why it looks so great. I won't say whose laptop it's, but like, you know, you walk up to a laptop and you don't touch it and you just get in front of it and it pops up, powers on and does the facial recognition and loads on it.

Yeah, I think there's technology from a company maybe like yours that actually helps with that. Or when I'm sitting here working and maybe someone walks up behind me, me, yeah, shoulder Surfing, it's called shoulder surfing. We, we detect that by Way, very high vulnerability.

Uh, you know, a lot of people think every hack is this really crazy engineering. A lot of it's just, uh, user, user, uh, being lax in their behaviors. You know, I say it's the stickers you leave on your laptop, it's also just literally having something like a password sitting up on your screen or someone watches you log in and then makes a note.

And you guys have built a lot of technology for, for that. So you're really also always complimenting, enabling and thinking about things. And that's both ai, but also the power of a specialty FPGA that can sit inside low cost, low power that can handle a workload like that without putting extra, uh, cycles on the, you know, the SOC that's being used For Traditional and core applications.

And, and you bring up a really good point. Um, when people think about taking AI on the edge, you really have to understand what is that AI application you do you wanna do on the edge? Are you really after the fastest performance, which comes at a very high cost of power?

Or is this application doesn't require that hybrid doesn't require A GPU if it doesn't require a GPU? Why pay the money for A GPU? Why take the space of A GPU?

Why consume all the power of that gp, especially for handheld devices or battery operate devices? So there is a spectrum of applications, some that yes, may need the highest power and performance, but the vast majority that we see working with our customers don't require that GPU type, uh, performance. In fact, if you think about the architectures of some of these GPUs, they were mainly architected for more machine learning than they were for inferencing.

So if you really wanna do inferencing and you wanna do it efficiently at a much lower, which a much better value proposition, which is size and power and cost, then adopt a piece of silicon that's really tuned for inferencing. And that's why our FDS on the edge have a, a really good value proposition. Yeah, I love that.

And um, I also have always been a big fan of some of the security applications you focus on. You know, things obviously where vulnerabilities are high, like in boot, you know, when, uh, you know, that stuff that FPGAs have been used for Lattice specifically, another great example you've shown at your, uh, developer events, uh, you know, that I think have been very successful. And I think this really all speaks to the company and its agility, its ability to sort of adapt.

When I first, uh, was getting to know you, it was very much kind of in the low end, um, you know, since you've been there and under your leadership and, and the leadership of, of of Lattice, you've definitely come up market. You've seen a, a gap in sort of the mid-market. Um, you know, talk a little bit about kind of how customers, partners, uh, and, and, and sort of driving this agile approach and sort of been shaping your strategy because the strategy has evolved pretty quickly, um, in your space.

Yeah, I'm, I'm gonna give credit also to the entire organization. We've got a lot of talented folks within Lattice that are really leading us into this new era of, uh, adapting programmability into multiple types of application. I think one of the good things about what we do is, like you said, we're very focused.

If you look at the FBJ market, they're small, mid-size and large, and we're really focused on small and mid-range FPGAs, and that's where we believe the core mainstream of FPGA usages in, in our industry and in the markets. The other beauty of an FPGA is because it's programmable, it's actually a horizontal solution. Like you hear us talking about, Hey, we're solving stuff in compute.

We're solving stuff in communications, we're solving stuff in industrial aerospace, and defense, automotive, even in consumer FPGAs, can actually do multitude of functions. And that goes across multiple end markets. And so whether it be, um, bridging, interfacing, motor control, ai, uh, security, like you said, vision type applications, what it allows us to do, which I think is unique to Lattice, is see what's happening across the industry across all of these end markets.

So we get good telemetry on, hey, what are these emerging trends that are coming out? You know, some, somebody wanna do something that's new, an emerging trend, and they don't have the right solution, we'll use an FPGA. And so working with our customers and working with the system architects, whether it be in compute or comms or industrial automotive, et cetera, we get really good telemetry of what their challenges are, what are they trying to do, what's the emerging trends, and then we build roadmaps and software solutions that help our customers solve those, those, uh, emerging trends and the applications that they're trying to address for the future products that they're building.

And then we tie that into our roadmap, and the team does a great job on executing. Yeah. And, and, and I've been witnessed to that.

I've seen the first of all, the expansion. I mean, how many, how much has the developer ecosystem expanded with your software? I mean, this was probably one of the best indicators in any, any case of a company, especially when you're kind of trying to make that just being a raw hardware company to kind of a hardware, software, uh, developer stack.

And I, I, gosh, I remember, like I said, from the first developer conference, I went to say like the third or fourth, like, is it, do have you done three or four? Uh, we've done three now. Okay.

The third. Yeah. I mean, it was like the first one.

I mean, you know, this is a compliment and it was like a small, it was small number, but it was like a change of the guard by the third one. It was like, oh my gosh, you've got some of the biggest companies, biggest partners showing up here because they're really seeing the value. I mean, I imagine the developer ecosystem growth has been a pretty strong indicator of everything you just talked about, but also is a strong indicator of kind of how much this platform has room for growth.

Yeah, and our last developers conference, uh, we had a record over 6,000 registrants. And if you look at our developer's ecosystem since we started this, uh, it's grown by six x and it continues to grow as well. More and more companies and developers are adopting and building solutions for lattice FPGAs.

And, uh, you saw that as well, that ecosystem of developers are not just people building soft ips that can be programmed to our FPGAs, but it's also other semiconductor companies like the partnerships we talked about with Nvidia and others where they're building reference designs and solutions now leveraging Lattice FPGAs as well. And then if you need a design house to help you build your system, well, they're trained on lattice devices as well. So that ecosystem has grown tremendously.

And that's something that we're gonna actually continue to invest in because as we go into these new emerging trends around vision, around AI and security, we're finding more and more people needing assistance. 'cause these are complex teams, these are emerging people are learning. And so we're investing more in our ecosystem to help our customers get these new applications into market as well.

Well, Asan, we've got just a couple minutes left here. We'd love to sort of hear where you see this going next. Uh, how much does AI continue to shift your business?

Um, give us a little taste of what we can expect. Well, I tell you what, what I tell, uh, my employees and colleagues here is I truly believe the best of times are to come. When you look at what's happening in our industry, uh, what we're doing in lattice, around security, around ai, around vision, our new products that we're introducing, all the solutions that we're bringing to market with our software solutions stacks, what we're enabling, I think the most exciting times are in the come and, uh, we're gonna continue to work closely with our customers, with our ecosystem partners, and continue to make a difference in our industry.

Well, ASAM, you know, every, uh, time we connect, I become a little bit more enlightened in what you're doing. And I continue to see, you know, you literally connecting the dots across, uh, so many different areas. It's been great to watch the success.

It's been a, you know, an interesting couple of years, the market as a whole. And, uh, you know, with, uh, you know, I, I like that you kind of gave shout to all the team. Of course, it's been good to get to know Ford, uh, Tamara, your new CEO.

Um, it looks like, uh, the market's sort of finding shape. AI has sort of been an accelerator. I think we're also gonna start to see some of those markets that were slower coming back.

And I think, uh, even just the last quarter, uh, that we talked, some of the results were starting to show that, um, semis is, is back, um, beyond just the AI chips. Like there is actually a lot going on and you know, everything and, and f PGA as you continue to be one of the companies. It's one of the only actually at this point that's really focused entirely on this particular space.

Um, and I expect to continue to see good results. So, Samm, thanks so much for joining us at this six summit. Let's, uh, we do this again.

We should, we should. I really enjoy what you guys do around this summit. It is informative to me when I listen to the other guests as well.

And so I really appreciate you having us here and look forward to next year as well. We look forward to having you back. Thanks for joining us.

We're gonna head off for a little break here. Uh, stick with us here at the six Box Summit, GPU Hammer time Cloud Native Security, slow AI coating the FCC resale shenanigans. Continue.

AKA's a ai, somebody got swept up in a typhoon, Google goes windsurfing. And we have some new reports from the Future Room Group in this week's episode of the Tech Field Day Rundown. Hello everyone.

Welcome to this episode of The Tech Field Day rundown. It is July 16th. Are the odds of July a thing?

I don't think they are. Uh, we could ask Caesar, but of course, you know, uh, he's too busy making salads and, and whatever else that he does. What you should be doing though, is checking out this wonderful edition of The Rundown, because I wish I was kidding.

It really is National AI Day. So how's that different from the other 364 days a year? Well, this time you actually have to get AI a gift.

Um, I recommend something in the electrical variety. Um, but, uh, one thing that is never shocking is the intelligence and charisma of my amazing co-host, Mr. Alistair Cook.

Al, welcome to the show. It's truly hard to be shocked by the completely absent. Thank you, Tom.

It's a delight to be here with you on national personal shift day, and hopefully your personal chef is going to make some nice dish for your AI in order that out. Future AI overlords look kindly upon us. Yes, yes, they should, because we always talk so highly of them here on the rundown and would never hope that they would troll the internet, uh, video archives to punish us in the future.

With that being said, let's jump into the episode because things may not be looking so rosy for AI right now. That's because Ro Hammer is back. Yeah, that's not the one that Thor uses.

Instead, this is a huge CPU exploit. Well, guess what? It's not just a CPU exploit anymore because it has been demonstrated to be effective on the Nvidia RTXA 6,000 GPU.

That means that you could potentially lead AI models on that JPU to produce inaccurate results. The research team that found this little jewel, uh, said that the effect of the attack was like giving a GPU catastrophic brain damage, which is kind of how I feel when I have to read another AI story. Uh, this thing though, it's only possible in a research lab, right?

Al, like there's no way that someone could potentially give my, uh, AI application brand damage, right? Sure. We'll tell ourselves that.

And that's the thing. This has only been done in in a research lab. Uh, however, RO Hammer is really well known.

I mean, CPU Ro Hammer has been a known thing for some time. The ability to attack one section of the memory within a physical server and have that cause a a bit flip a corruption in the data. And another section of memory, the mitigation for this has always been error correction and preferably multi bit error correction.

So that, uh, if a single bit or a couple of bits get flipped, you can correct that because you've stored additional data. Now, the problem with storing that additional data, that error correction data, is it takes up more space and then computing the actual error correction codes on a regular basis also adds more load to the CPU to do these functions. So there's protection against it and it's gonna come at an expense.

So who needs that protection? RO hammer and now GPU Hammer, which is the, the name of the script that they used here. Uh, these are most dangerous on places of mixed trust.

So places where you have some high trust applications using exactly the same physical hardware as applications you don't trust, I would ca categorize that as public cloud. Another tenant could be right alongside your workload. Now, their ability to know that you're alongside their workload is pretty limited.

Uh, but their ability to damage your workload is, is less limited than possibly we thought it was before. So I think we'll see some moves towards more dedicated hardware for this. If I am controlling all of the workload that's on a specific GPU as well as its associated CPU, I'm much less exposed to some sort of behavior that another tenant is doing.

So this doesn't necessarily mean that everything has to come back to on on-premises to mitigate this, uh, couple of techniques, if you're staying with multi-tenancy, then use a multi-tenancy that uses error correction code in the memory. And then the other mitigation is to use dedicated hardware actually on the cloud rather than using multi-tenant hardware. Is this something that an, an attacker could use as a way of doing a distributed denial of, or a denial of service against your application, right.

Giving your AI application brain damage because it's running right alongside it, yes, technically could happen. In fact, it's more likely to happen on premises because then the attacker knows what applications residing on the same GPUs and cpu. Uh, it's less likely that you'll be the target of an attack on public cloud since there's less of that awareness.

But you could just be unlucky that some person of questionable character, uh, is experimenting with this on an adjacent GPU or is just ha has been compromised by somebody who just means the world ill who just wants to see the world burn. Should you be worried about it? Yes, you should, should be aware that this is a possibility and that it really does tank the results of your AI application.

One of the things they said is they benchmarked the accuracy of the results of an AI model that is running adjacent to this GPU hammer. 1% accuracy. It really is that sort of catastrophic brain damage that we definitely don't want in our application.

So yeah, the, uh, the impacts are pretty severe. The probability of being affected by it, it's fairly low to protect it itself against all kinds of, uh, unpleasantness. Some sort of cloud native security fabric might be very useful.

Maybe is building just this to simplify and cut clouds, either security costs by enabling real-time policy enforcement as well as traffic encryption. This approach aims to fix current challenges, outdated tools and security gaps across multiple clouds. How important do you think it is to have these cloud native solutions across multiple clouds?

Tom? So lemme ask you this question. How would you configure an access list on a Palo Alto firewall to restrict all inbound traffic from a certain geolocation ip?

Now I need you to translate that into AWS Azure, Oracle and Google too late if you didn't already immediately know how to do that. You've been violated and breached and you're gonna be in the news. Sorry, that sucks.

One of the reasons why we have such a hard time securing workloads in the cloud is because the way that our tooling operates is very much stuck in a bastion host on premises methodology. And that doesn't work in the cloud, folks, you, you cannot secure things VPC by VPC. You.

Well, it's like, remember the old days of managing things server by server by server? Some of you probably are wistfully looking off into the distance. Oh yeah, the good old days as you, I don't know, puff on an old pipe and, uh, complain that your joints are getting creaky.

'cause the weather's changing outside. Anyone under the age of 50 is gonna go, no. It's horrifying to think about managing things server by server.

That's why we invented virtualization containers. So many other things. The same thing applies to security and even more so because security rapidly evolves.

So you've gotta be on top of those things. And I think what aviatrix is building here, a cloud native fabric that spans multiple clouds is absolutely critical because it means that you only have to build your policies once. And then when you tell the system to deploy those policies across whatever clouds you're using, that means that you don't have to worry about all those translation mechanisms.

What did Amazon call it? No crap. What does Microsoft call it?

Is it the same direction? Does it apply at the same location? It's all taken care of for you.

And when your executive comes back from their latest meeting and gets told, oh, we, uh, we need to be in this other cloud now because they're great for insert inane reason, that salesperson told them here, now you just click a couple of buttons in the aviatrix Cloud Native security fabric and you can be secure by default over there instead of scratching your head for a few hours trying to figure out how they do an contract policy model remediation. Yeah, I'm already lost. So, bravo to the folks from aviatrix for, for taking a look at this.

You may have seen them previously at, uh, field day events in the past Cloud Field Day Networking Field Day. We'd love to see 'em at Security Field Day to talk about this. So, you know, aviatrix, we'd love to get an opportunity to check that out.

Uh, we do offer security field day events throughout the year. You've probably heard about AI helping you code things. Well, someone actually did the math and they did a study of 16 experienced developers, uh, that have completed hundreds of coding tasks.

And they found that having an AI coding assistant is making these developers less productive. The developers believed they were nearly 20% more productive with assistance like Cursor Pro or Ros Claude. However, the study found the exact opposite.

The tasks took 20% longer to complete using the ai. I guess the question that I have to ask Al is AI the new clippy? Well, that's a an interesting parallel because you know, Cliffy didn't do a heck of a lot for us, but it certainly took a lot of our attention.

And that's, uh, what we're seeing here is that these AI coding assistance are more useful than Clipper, but they are taking a lot of attention. So this specifically was some senior experienced developers working on open source projects working through about 260 different tasks. And the developers themselves, as I said, they've, they've reported somewhere around 19 to 20% more productivity felt productivity than without using the AI tool yet it took them 20% longer to complete that productivity, which doesn't sound like greater productivity to me.

Uh, the study looked a little bit deeper. One of the things was that, uh, these developers spent a lot of time reviewing what AI had provided them as code, as well as spending a lot of time crafting the prompts to get AI to give them what they wanted. And so it feels like there's still an awful lot of work that needs to be done to get assistance from your system.

But I should caveat though, that these were some experienced developers and so they knew what they were working with. They knew the environment, they knew the, the tools that they were using. So one of the use cases for these AI assistance is to speed up the training of less senior developers, uh, make it easier to work with things you are less familiar with.

And I suspect the results would've been better in that use case. So making software development more accessible for junior developers and people who are just starting out, I think is the place where these coding assistant are very useful at the moment. Uh, I think we're a little way away from these coding assistance being an expert programmer and making expert.

Uh, that's definitely the direction they're heading for, but the study suggests that we're not anywhere near there yet. They also tell us that we're not very good at judging our own productivity. A new law restores the fccs authority to auction spectrum and requires at least 800 megahertz be sold, potentially pulling it from the six gigahertz and CBRS bands currently used by wifi and rural broadband providers.

While mobile carriers like at and t and Verizon support the move to give them more bandwidth, uh, for 5G expansion, critics warn it could slow your wifi and harm small ISPs that rely on these bands. The law reverses earlier efforts to protect six gigahertz for unlicensed use and reflects growing pressure from the wireless industry now backed by former FCC who leads a major telecom lobby, Tom, 800 megahertz. Is that a lot of bandwidth to pull out of these bands or is it just around the era?

It's only a flesh wound? How bad could it be? It's not like the six gigahertz spectrum is 1200 megahertz in whip.

Oh wait, that's exactly how big it is. Yeah, my political soap box for this little story is probably gonna be more like one of those CrossFit box jumps that's like five feet tall. So allow me to elaborate on that.

We opened up these two bands and, uh, we opened them up under extreme pressure from lobbying groups. The first band, of course, was CBRS, where, uh, you've probably heard of this for things like IOT or for warehouse locations that are not well suited to wifi that need large coverage for devices. And, uh, we had to play with some certain rules.

Uh, the CBRS band is lightly licensed, which means that there are people who can buy a license in that band and then they can operate in that band. And then if you are trampling all over their license, you have to move. Uh, so there's, there's the ability to operate unlicensed in that band.

Then we have the six gigahertz band. And, uh, this is an interesting story. Uh, the short version is, is that, uh, when they were doing all the original surveys, they got a lot of feedback from saddle leak ally providers and from telecom companies who said, oh, we need way more studies before we can let you use this.

And the incumbent wifi companies like Aruba and Cisco said, no, you don't. Uh, we've proven that it doesn't actually interfere with anything and there's no problems and you should open it up. And they did under AGI PI when he was the FCC chairperson.

And then we come to this year when someone snuck a little rider into the bill that was passed, it was neither big nor beautiful. It was kind of a pain in the neck. Uh, among other things, you know, like trying to repurpose the space shuttle discovery to Houston, they also said, well, remember those things that we set aside for everyone to be able to use wifi?

Yeah, we would like to sell those to somebody. Now, well, who was the first group of people lining up at the door cup in hand ready to pay for the thing that they claimed they needed more time to sort out? That's right.

It's the telecom providers. And who's the lobbyists for the lobbying group that's working on behalf of them? Oh, that would be the same person who opened that spectrum up a few years ago who got fired or was departed from the FCC and got himself a nice Cushing lobbying job.

Hopefully he's drinking out of a 55 gallon drum instead of that comically oversized coffee cup that he loved to put on TV so much. This isn't about coffee, this is about going back on a promise, a promise that said that we're gonna give you this band to expand. 4 gigahertz recently because that's a trash band if I've ever saw it.

And two, it's funny that that's what they're telling the government because what they're telling the investors is, oh, we got plenty of room to expand. This won't be a problem. Please don't let the stock price dip because we lied to you.

Huh, that's really weird. It's almost like they're talking out of both sides of their mouth, which honestly is not something I should be surprised about because it's a political discussion. If you wanna know why Agit Pi was so, uh, ready to say that six Gigaherz isn't being well utilized, maybe you could go back to the FCC and ask them about why they had so many crazy rules in place that wouldn't allow six gigahertz operations outdoors without, uh, the, uh, a FC databases being brought online and having tons of studies done before you would even light it up.

And oh my God, if you ever get a radar hit in an airport, we gotta shut this thing down for six months to figure out why the airport radar systems are so comically backwards that they don't operate well with anything. Because that was another group of people that kept saying that they can't operate in those bands. And guess who was the other group behind the airports trying to get six gigahertz shut down.

That's right. It was at t and Verizon. So you'll have to excuse me if I think that maybe the motto of the US government should not be, why do we give away things we could sell to people?

Do better guys and quit trying to sneak these things into must pass bills because if you don't, well, who knows what might happen the next time somebody is in office who actually understands how technology works there. I'm gonna jump down off my soapbox. Now.

Let's talk about something more fun because ACA has unveiled a new suite of integrated tools that they're calling ACA orchestration, ACA agents, ACA memory and ACA streaming, which is designed to speed up the creation of agentic AI systems. The platform promises three times the performance in just one third of the compute cost with the support of wide range of applications from autonomous and adaptive systems to real-time and edge use cases developed in close collaboration with customers who are managing massive AI workloads. Uh, some of those are exceeding over a billion, that's with a b tokens per second.

The new offerings are now available to all existing ACA license holders. A live demo and deep dive are gonna be scheduled for a webinar tomorrow. Al, what has you excited about acas new lineup?

Well, for a start, you've gotta put this in a, a context of what ACA is. Uh, ACA is, well, it's a company, but it's also a framework for building distributed applications. And it's that basis of we've gotta distribute application framework and what feature set the people want to add to their distributed applications.

Ai and particularly in a distributed architecture agent ai, this is where the AI takes some input, does some action, and then produces some next step in the process. And often we we're expecting to be seeing one agent talking to another agent. And these, these agentic AI is in some ways is like the microservices version of ai, where the microservices talk to one another while these AI agents talk to one another, each one achieve some small task, but together they achieve a much larger task.

So having this capability as a platform that you don't have to build yourself all the way from the ground up as here's a, a set of capabilities, a set of libraries that you can use to build an agentic AI application. And they're using all of the right acronyms in here like MCP and having, um, uh, really nice API integrations for this. The ability to tie these things together in in large scale.

Um, this looks really cool. So having orchestration of an entire workflow that might go through multiple different AI agents running those agents themselves, uh, through HGDP APIs for the agents themselves. Uh, one of the nice things is separating out a, a section called ACA memory.

This is the ability to persist all of the context information that's associated with a particular query or a particular sequence of queries or discussions that are happening between these agents. Transferring that, that context information from agent to agent and persisting it over a long time is gonna be absolutely vital in here. Uh, some nice interesting things about, uh, streaming in here.

So, streaming data rather than just transactional data, uh, streaming data methodologies for driving in, having this stream of data coming in triggering the a triggering the agents to run and get some insights from these individual units of data streaming in. Again, we've got this, uh, ACA platform that allows you to build large complex applications that to say billions of tokens per second, sorry, billions of tokens per second. Uh, as as we're getting the, these big applications running through.

I think this is really cool. I think this is gonna be incredibly useful for building out these large applications or more likely adding a gen AI features to the large applications you are already building with aca. Alright, you ready for your soapbox?

Again, I don't, I don't think this one's gonna be the soapbox. I think you're just gonna be, uh, frothing it in the mouth with joy on this one. US officials want to extradite Jews' way a Chinese man arrested in Italy working for the silk typhoon, a Chinese government-backed hacking group.

He's accused of helping steal COVID-19 research from US universities and joining the 2021 Microsoft Exchange server attack. Uh, prosecutors say he was following his orders from Chinese intelligence to serial sensitive data. His family and lawyer say he's innocent and may have been hacked himself.

Experts say the arrest is a rare win that silk typhoons tie, the cyber spying is still likely to continue, isn't it, Tom? It's, and the reason why is because as we all surmise, the people backing the typhoon groups are well connected and are basically training people to kind of accomplish things that they could get in trouble for otherwise. And unlike Elmo this week, I don't think our friend here was hacked.

I think that it was quite the opposite that he was on the, the receipt on the producing sign of that. And again, this is all alleged, this is, you know, what's happening. But what's important here is the fact that this was an arrest made in Italy on behalf of the US government that immediately filed for extradition.

They want somebody to hang this on. Yes, stealing COVID research data five years ago, bad Microsoft really wants to hang the exchange hack on 'em. But here's why it's important because if you remember the other Typhoon group, salt Typhoon was responsible for a massive hacking campaign just last year that hit a lot of people that are currently in power, and they have the potential to be embarrassed at the comically inappropriate time, like during a negotiation with a foreign power over trade deals, for example.

And I think that this is trying to send a message to a group that allegedly is state sponsored, knock it off with hacking us or we'll start arresting your people in every country in the world, at least the ones that we have extradition treaties for. Now, the question is, does this cause the typhoon groups to back off of certain campaigns or does it cause them to escalate? Well, if you're gonna threaten me, I'm gonna get enough blackmail to threaten you right back.

I don't have a good answer for that because we know we have never seen hacking on this scale with this many involved people. And honestly, I don't think one person is gonna turn that tide, especially if they can't get 'em out of Italy, because as we've seen, the Chinese government can be very, very convincing when they want to keep one of their nationals out of an unfriendly country where they could potentially be prosecuted for something that quite honestly, the Chinese government doesn't think is a crime, or at least the way that it was directed. Or maybe they cut 'em loose.

I don't know, maybe they're like, yeah, you can get your sacrificial lamb and while you're busy putting 'em on trial, we're gonna steal a whole bunch of other stuff, folks, the typhoon thing isn't going away anytime soon. If anything, I expect to see four more typhoon groups come up out of this, and they're all gonna be named some other kind of element that will eventually form the Typhoon megazord, and then we're gonna have a kaiju battle on our hands and it's gonna get real entertaining. So make sure you stay tuned, because this one's about to get juicy folks.

OpenAI has been buying everything that's not nailed down, but they missed out on something because the $3 billion deal they had to buy, AI startup Windsurf collapsed. So guess what happened? Google quickly hired Windsurf, CEO co-founder and a whole bunch of other staff to its DeepMind team focused on agentic coating.

Google also got themselves a nice little license to use some of Windsurf technology. However, it doesn't own the company, it just owns most of the people who work there. Failed deal comes among growing tensions between OpenAI and their daddy and Microsoft, who, uh, it's highlighting the fact that there's fierce competitions among tech giants to be the leader in ai coding experts say that OpenAI is facing challenges while rivals like Google and Anthropic are able to move faster in this space and potentially challenge to take the Crown.

Al, did Google get themselves a real win by effectively buying everything from Windsurf, but the name, It feels like that, doesn't it? And you, you take the CEO, you take, uh, co-founder, you grab some of the, well, apparently Windorf had 250 staff and most of them are still there. But if you pulled the guys out at the top, the founders, the people who had the foundational knowledge for this, how much is Lyft?

It's an interesting thought. Uh, Google is also not saying how much this cost to pull these people in can imagine that when the current, uh, massive demand for skills around building AI applications that these people are, are demanding pretty serious amounts of compensation. Uh, that $3 billion is probably a little more.

The, the, again, uh, the $3 billion OpenAI was gonna spend is probably a little more than the salaries that these people are are getting from Google. But maybe just the, the joining bonus will be up there. So, um, it'll be interesting to see what's left of Windsurf, how long and, and how far Windsurf continues.

They, uh, formally known as Codem and are a provider of AI coding assistance. You may recall my earlier story on AI coding assistance. Uh, definitely good to see the Google team making commitments to, to getting hold of some more great skills to add to their existing amazing skills.

And in the DeepMind team building all of the AI engines and, um, building some pretty cool technology within Google there. Um, does this mean that Open AI is, is gonna sort of fall off a little bit and not succeed so well, a little too early to say? Uh, it is definitely a, a fractious relationship between the various players here, because whoever builds the best generative AI foundation model is gonna be in a position to license those models out to the commercial organizations that are going, going to build AI applications, uh, commercial organizations, Aren gonna build foundation models.

They're gonna fine tune those foundation models that they've licensed or they're gonna use technique, slight retrieval, augmented generation to get some business awareness to those foundation models. But it is ridiculously expensive to build a foundation model. And so it's high stakes fights between the various people who are building these, uh, foundation models.

It'd be interesting to watch this from the sidelines with, uh, a, a big bucket of popcorn and a nice cold drink. New report from the Futurum group. Of course, our great friends, and, uh, lemme try that again, again.

A new report from the Futurum Group highlights some big news in the data intelligence space. The global data intelligence, analytics and infrastructure di AI market is projected to grow significantly more than doubling from 44 9 billion to 876 billion going from 2024 to 2029. 5% compound annual growth rate according to our Futureum group study.

Uh, Futureum intelligence is an awesome part of the wider future group, which of course includes Tech Field Day. The surge in in the, this surge in the DI AI is primarily driven by the accelerated integration of AI into enterprise data infrastructure, making data, a crucial investment for future-proofing businesses through AI driven automation and real time decision making. Tom, growth in storage is unsurprising.

Is this something slightly different? Uh, it is a little bit different. And normally I'd say, you know, more data people are munging through it.

That's, that's one thing. But when you look at the fact that this is literally what AI runs off of, like you, you'd be crazy not to think about that. Look, the data intelligence that is the raw materials that your algorithms, your models are running on and running AI without good data intelligence is a lot like buying a Ferrari and then parking it in your driveway.

If you don't have the place to run it, why have it? And look at all the things that go around this. I mean, we've been talking about this for over 10 years.

Remember we used to do data field day and now that we've kind of come to where we're at with ai, it's like the new data field that, and yeah, I, I get it. Everything old is new again. But more importantly, what people need to understand is that everything that surrounds AI is growing at a rapid rate.

I mean, look at the number of news stories that we've done already this year about power consumption. Like the electricity market is doing gangbuster business. We can figure out a way to do cold fusion.

I mean, AI will eat it up. But more importantly, what is the other thing you gotta think about is like the, the data inputs are gonna be growing and how do you store those? Because I promise you AI hates unstructured data.

You've gotta be able to look it up in a table. You gotta be able to index it. You gotta be able to do all of these things.

And that means that organizations who have historically done a terrible job of holding onto that data are gonna try to find a better way to do it. And that means they're gonna be investing in all of these things. It's like cleaning up your house before you buy new furniture, right?

We want it to be good. We want everything to be ready so that when we get the new stuff in here, it, we don't need to clean around it. And one of the things that all of the great people over at the Futurum group have done is they have put together the numbers, you know, one of those amazing reports that, uh, you know, is easy to read, easy to consume, and can give you an idea of where the growth areas are and what people are using it for.

And don't forget that you can head over to the Futurum intelligence platform. And if you're already a customer, you can download this. If you're not, you can sign up and get access to it and we'll, uh, link to the blog post with the announcement there where you can get all the information that you need.

So congratulations to that team for putting out an amazing report. I think that it's really gonna help people understand not just where a growth is in ai, but all of those satellite industries, right? Like when you see like a neighborhood being built, what's your first thought?

I need to put a Walgreens here somewhere because somebody is gonna need to go to the drugstore sooner or later. That's what we're talking about here, folks. We didn't have time for a closer look this week 'cause we had so many stories we had to cover that, uh, you know, we just wanted to make sure that you knew everything that was going on.

Just like we wanna let you know everything that's going on in the world of Tech Field Day, because that's what's near and dear to our hearts. The next thing coming up is actually in just about a month, we're gonna be Tech Field Day Extra at Cher Cleveland. That's right.

Stephen Foskett is back and he's gonna be having some great presentations from an amazing lineup of presenters. You can find more information about them on the Tech Field Day website, and I'm sure that Steven's gonna have, uh, a little promo coming out pretty soon with a list of who's gonna be presenting, as well as a fun little podcast that you're not gonna wanna miss. Uh, but coming up after that is Mr.

Alistair Cook. You're back in the US for some fun, aren't you? I am.

I'll have a little bit further of a commute than Steven does, getting up to Cleveland. I'll be back for AI infrastructure field in the, uh, San Francisco area. Uh, it's gonna be fun.

We're gonna have Broadcom and Ter and Hammer Space and Martis all turning up for that one plus a few more that will be added. If it's gonna be anything like the last infrastructure field day. There'll be a lot at this event.

Of course, couple of weeks after AI Infrastructure. Field Day is Security Field Day. That's a time event.

It is a Tom event. And, you know, we talked a lot about security in this presentation lineup. Um, we had a lot of great stories about security.

Think about all the ones that we've had so far. Uh, we got some cool stuff coming up that you're gonna wanna talk about. And I actually wrote a blog post about the one of the companies we've got presenting Secure X.

Uh, but more importantly, we'd love to see more great companies. I talked about Aviatrix here, but we'd love to see some of the data protection companies out there. We'd love to see some of the companies doing endpoint security, XDR, whatever.

We, we wanna open it up to as many people as possible because we want you to get a full understanding of what goes on in security. There's more to it than just changing your password every 30, 16, 90 days, whatever it is. Um, so make sure you tune in September 24th and 25th.

Consider it a birthday present to me because my birthday is that week and I would love to be able to celebrate with a very full event and cake, mostly a full event. 'cause I don't need anymore cake. But what you need is more of us on the rundown every week.

You know where to find us. We're here on Wednesdays, we publish our videos in the afternoon so you can, uh, drink that afternoon cup of coffee, stretch out and, uh, hear us wax intellectual about the news. But more importantly, you can listen to us in audio podcast format.

Maybe you're, uh, stuck on a long commute and you need something to pass the time and you wanna listen to the dulce tones of the Kiwi and the Nerd talk all about all of the stuff that happened this week. Well, if you do that, just search for Tech Field Day rundown. I actually told someone who's not in tech about that yesterday.

So, Kristen, if you're listening, I hope that you enjoyed this little shout out, and I hope that this didn't bore you to death, because I know that tech isn't your primary job. You're just doing this because you think I'm cool. Let's be fair.

Nobody thinks I'm cool. Uh, but we think you're cool and we love that you are part of what we do here at Tech Field Day. And on the rundown, we hope that you tune in for all of our events.

But more importantly, we hope that you're back next week for the rundown. Like, share, leave a comment, let people know what we're all about because that's how we get more listeners and we want to hear from them, from you, from everybody. Thanks for tuning in.

I know it's the summer, at least up north. It's the summer down south, it's the winter. Stay warm or cold hours appropriate.

And we'll see you next week for another great episode of The Run Back. Hello everyone. Welcome to the next of our breakout sessions as part of our Cloud Fridays event.

With this session, we're going to talk about the power of Ansible automation platform and how you can now buy it through AWS and take advantage of some even more, um, powerful solutions Red Hat and AWS are bringing to market around it. I'm Simon Briggs. I work as a Red Hat ecosystem solutions architect, focused on AWS, and I'm here now with my colleague Farley, who will, uh, introduce himself and then explain about the power of Ansible Ansible automation platform.

Thank you Simon. And I am a specialized social architect, specialized into Ansible, uh, mostly working with the ecosystem, but also with direct customers as well. And one of the thing that we have seen is that the kind of issues, the challenges that are, um, seen in organizations are fairly, uh, the same, um, across all organizations.

And the, the issues start with the fact that there are a lot of different people into the organization, different skills, different roles and responsibilities, and that have to handle a lot of different use cases. But the reality is that all of these, um, processes are, uh, on top of the same base, uh, concept, same base, uh, constructing blocks of it, which are compute, networking, storage, and security. And those can be also, uh, physical virtual in cloud on edge, and, and they can be slightly different.

But the reality is that those same components are the one that are, uh, the basis for all, um, architectures within it. And the complexity with this is that if we have automation that is completely different, uh, from one team to another team, uh, the, the result will be that we'll have a lot of duplicated effort as well as completing automation that will create problems over the course of time. So the solution to this is to have a unified approach.

Having a unified approach means that we can break the silos, um, across those different, uh, use cases and teams and domains, and have a unified platform that allows us to automate every single use case on every single architecture and using every single component, um, to have that same behavior, um, across the whole organization, which also allows us, uh, to have governance around all this automation and therefore the under, uh, underlying it, um, that is, uh, co um, cons coherent, uh, across the whole organization, which means, uh, that we can have that, um, consistency, uh, across the organization. Now, what Ansible provides is exactly this kind of automation. So what hassal provides is an increased speed, uh, to delivery because, um, a lot of operations in IT are usually done by clicking on, uh, user interfaces or maybe providing some comments, those kind of things.

What Ansible can provide you is, um, having automated those processes, that means way less clicking and less clicking also means reduce human error because, um, once you have an automation that has been scripted, um, in, in the unsold mission platform, you can simply rerun the automation multiple times and every time you will get exactly the same result, which is not what you usually have if you have people doing the process themselves manually. So, um, the result of this is also a higher, a higher level of consistency, uh, because of that getting exactly the same result without errors, um, every single time, which allow, allows us to have a more coherent, um, environment, uh, in our it. This also allows us to evaluate, uh, the whole lifecycle of an application and automate the whole lifecycle and of an application.

And you can think about the lifecycle of an application roughly dividing in three different parts. Um, the first part is about the provisioning and the setting it up or day zero as is, uh, often called. Then there is a second part, uh, that is about, uh, the, the operation part of this, uh, the, the visibility and so on.

And then the third part is about the governance. Uh, so as you can see, we uh, tend to graph it in this kind of way because ssi, it's a life cycle, and usually as soon as it, it hands it start again. Um, and Ansible is, um, able to provide you help and support for every single one of those, uh, aspects.

And one of the critical aspect about Ansible is that it allows you to automate, uh, your IT processes in the pub public cloud, uh, in cloud native ways, but also in the private cloud or, uh, on data center on the edge and so on, which means that it becomes kind of, um, lingua franca, uh, across your whole it, uh, so that every item in your IT is configured exactly in the same way, which also means that it becomes easier to cross pollinate, uh, across, uh, the organization best practices, um, standards, uh, guidelines, as well as, uh, for people to move from one side of the organization to another because they, uh, already use, uh, the same tooling at least, uh, for the automation of the processes. One of the critical aspect, uh, about Ansible is that it's not just Ansible itself. The value of Ansible is, uh, all the integration it has with the huge amount of, um, it, uh, partners that we have, uh, which means that you can automate, uh, not only, uh, for instance Linux boxes or Windows boxes, but all the networking, um, parts as well on the security across it, um, as well as, um, items on public cloud, private clouds edge and so on.

But also it can be integrated with ITSM systems such as ServiceNow and many others. And all of this can be done through, uh, collections, um, which are basically bags of tools, uh, that, that you can use to integrate the automation with, um, that specific IT technology. Um, and we have the concept of certified collections and verified collections, which basically our collections that get provided either by Red Hat or third parties, but are validated by Red Hat at least.

Um, and you can use them, uh, with the security of, um, getting, uh, this, uh, bits from a trusted source such as Fred Hat. The, um, red Hat Solution, uh, red Responsible Solution is a strategic solution, uh, because it encompass all the possible, uh, use cases, uh, that are usually found in IT departments. Uh, it's basically first multiplier, uh, for the operation side of it.

And it's critical nowadays, even more probably than, uh, in the past because now everyone talks about ai, AI is great, but AI can be built only if you have already an organized IT department. If you have an IT department that spends all their time, um, around fires and issues, it's going to be very, very hard to, uh, have the, the time and focus, uh, to then work on, um, AI or whatever. Uh, next, uh, big technology, uh, will come out, uh, in the IT space.

Ansible automation platform has been already selected by many, many customers and also, uh, a lot of analysts such as, um, Forrester, Gartner and many others in this case. Uh, this is the Forester wave, um, about infrastructure automation platforms, uh, that, um, place, uh, Redde solution, which is a, a p, um, as a leader into the space. And this really, um, is a testament, uh, to all, uh, the part that we discussed so far, uh, but also, uh, the, the integrations, uh, that, uh, are, are very, very useful, uh, to all our customers.

So thank you Farley. That's a brief description of the power of Ansible automation platform, the tool itself. And what I'd like to do now is concentrate on how, um, red Hat works in tandem with AWS in our long established partnership to allow our customers to get the benefits of both providers capabilities when consuming Ansible automation platform.

And firstly, um, I'll call out something which fate just talked about. So we actually have an AWS centric collection available to our customers as far explained needs, a supported, um, collections of blobs and add-ons, et cetera, plugins, which customers can take to help them augment their playbooks and their automation scripts around this technology. And it allows them to understand the validated framework around which they can most quickly deliver the value of the sophisticated automation that Ally talked about to their organizations.

And actually, if you think about cloud, um, infrastructure and engineering, we'll also, uh, often be working with engineering teams who are working with Ansible, but are also using other automation technologies. Often organizations in the cloud use Terraform, for instance, for infrastructure building and con, um, setup and some organizations in the AWS context use, um, cloud formation. These are both really powerful tools, but they do slightly different things to Ansible automation platform.

And because of that, we are able to integrate with those toolings. We do have plugins to be able to work seamlessly with them from our tooling or to call out to, um, Ansible or if customers, um, are completely, uh, new to using automation. Um, within the, um, virtual environment of hyperscalers, Ansible has the capability to do everything that a customer would need to be able to deliver.

But recently Red Hat announced some new capabilities from our AWS platform. And what that is, is, um, described in this visual. So as you can see, if a customer wants to use Ansible automation platform on AWS today, they can go down the left hand channel here, where they would deploy directly onto AWS using Red Hat Enterprise Linux or Red Hat OpenShift tooling to be able to deploy Ansible that they purchased directly from Red Hat already.

But importantly, in December this last year, uh, red Hat announced AWS reinvent, um, their big global summit each year. But Red Hat is now making Ansible automation platform available in two forms to buy directly from the AWS marketplace. The AWS marketplace is a very powerful, um, environment that allows independent software vendors such as Red Hat to sell its software like Ansible to its customers, but within the partnership of AWS, this then unlocks a lot of value add procurement capabilities for the customer.

It streamlines their procurement. They already probably have extensive, extensive relationships with the hyperscaler, so that streamlines under that process. And it also possibly unlocks extensive commercial agreements that a customer might have in place already with AWS allowing this, um, purchase to be, uh, recognized within an EDP, for example, or a private purchasing agreement, um, that allows customers to buy in two different ways.

Firstly, a customer can buy and deploy themselves in very much the same way that they would if buying directly from Red Hat. So they would just buy the subscriptions from Red Hat, but via the marketplace and then deploy them themselves. Or they can use a new technology that Red Hat is very proud of, called Ansible Automation Platform Service on AWS that capability is, um, innovative.

It is a solution where if the customer chooses to subscribe to the technology through the AWS um, uh, marketplace, they can very rapidly commission an Ansible control plane, which is fully managed by Red Hat delivered on a W S's back plane, um, technology. And that technology then is available for customers to start deploying execution planes from Ansible anywhere they would like. This allows customers to use a, um, AWS based Ansible automation platform, um, control environment to run automation across the many different environments that Farley talked about, be it other cloud vendors, clouds, be it on premises or even in colo locations.

The use of an execution plane allows the customer to do it from, um, the centralized AWS deployed Ansible automation platform. Either way, the customer gets great benefits from that Ansible, they're able to procure it through the AWS marketplace, which allows customers to save, um, commercially if it's available to them. And they're also able to rapidly take that technology and start deploying it because time to value within any of the technologies that Red Hat delivers is a very important facet of the services that we provide.

And of course, if customers do have buying commitments such as an enterprise discount program or um, a, uh, um, a purchasing agreement with AWS, then you will get further benefit from that approach. I mentioned earlier that you can deploy your, um, managed Ansible automation, um, to AWS, but still be able to deliver automation across any, um, environment you choose to work on. And this diagram helps, um, explain from a very high level how that is achieved.

So on the left hand side, on AWS we deploy in the managed service, we deploy a control plane of Ansible automation platform and we, um, deploy what are called hot nodes that allow the customer to then manage their execution planes, which are represented to the right there. They have the ability to set up an execution node, which then allows the individual manage nodes within that environment to be, um, managed, however the customer wants to secure that environment. Um, importantly, this is available today.

We have many customers using it, but I want to call out. You will see there on the right hand side on the bullet notes that, um, event driven architecture is the one large feature from the solution set on A A p, which isn't available today on release. We are working hard to make sure that technology is, um, consumable, but it, it isn't there as of today.

Um, I'd also raise that, um, at the moment, the control plane that customers commission can be deployed to three different regions within emea. That's, um, EU, west one and two, so that's Dublin, London, and EU Central one, which is Frankfurt. Um, those regions deliver the control plane.

So we actually have many customers who use an ex execution plane in a different AWS region using it today across emea. Um, but be aware that we are looking to roll out the control plane capability to all the regions that our solutions are available in for other products today. So that would be most of the regions across the whole of amea, um, very soon.

What I'd like to do then is talk about one of those customers who's using the technology today. So I've already said this technology is quite new. It's only just, um, been announced, um, to the market and we already have customers who are using it.

The reason being that ultimately many customers are already very familiar and very happy with the value that Fall was talking about Ansible as an open source project and Ansible automation platform as a supported product from Red Hat is extensively used within, um, the IT industry today. And as such, there was many customers who were very keen on utilizing a managed version of our Ansible automation platform to further extend their ability to concentrate on building out automation from Ansible rather than concentrate on managing the control elements of an Ansible automation platform deployment. Um, one of those organizations is a department of work and pensions in the uk.

For anyone who doesn't know that, um, department, they're a very large government body within the uk, one of the largest we have. Um, I say we, because you can probably tell I've got a UK accent and I am resident in the uk, and, um, they have found that through Ansible and then a recent investment in taking on Ansible as a managed service through AWS, they've been able to, um, very, very drastically drive down their ability to, um, deploy technology to the right place in the standardized form that Fally was talking about. So avoiding, um, the challenges that come about with multiple different individuals being involved in a delivery pipeline, et cetera, they've been able to move away from that and have seen drastic reductions in time for deployment.

Um, and they talk about 50 minutes breaking down towards 10, um, for particular virtual assets. They've, um, also talk very, um, strongly about the fact that they've been able to create a consistent deployment environment across different parts of their IT infrastructure join, bringing together that hybrid cloud capability that Red Hat has for a long time in messaging the industry about being able to do the Department of work and pensions will always have an on-premises, um, delivery capability and what they, uh, do for the uk, they won't move away from that very quickly. So they were looking to be able to, um, get consistency across not only their on-premises assets today, the Ansible usage they've had in those environments, and then extend that capability out for consistency across their very extensive now AWS commitments.

Um, and have also talked about in their referenceability on this, um, project, the fact that they found the move towards Ansible automation platform a, um, standardizing effect from their troubleshooting and management point of view, they're able to use what are, um, essentially very, um, human readable, understandable scripting approaches. They've been able to utilize that capability to take the standardization across the environment to make sure they, they can remove those silos or pockets of infrastructural management that they had previously that would be very difficult for other members of the organization to be able to utilize without that standardization. Now we've added a QR code onto this slide and I've talked to it for some time.

So if you are interested in the detail around that, um, that customer's case study, you can follow that QR code, it will take you to our website where we've got it, um, written up in a lot more detail than I can do justice to it. And you can hear from, um, the DD wps own people about how they found the investment that they've made on this fantastic technology. What are the next steps I think you'll be asking yourself at this point, what, what is Red Hat asking you to do?

Well, if you are thinking about extending automation within your organization, and if you like what you see about the, um, Ansible automation platform becoming a powerful tool to standardize, um, and take your technology automation to the next level and also use AWS, then we've got a few assets here that might help you. Again, we've added some qrs so you'll be able to, um, watch this video about maybe, um, take a, a quick read of those links and that will take you to several assets we have within the, um, AWS um, web presence, um, to help organizations get more understanding of this technology stack. Um, the, um, they, the different QRS take you to different, um, documents essentially, um, from explaining the technology itself with the Ansible automation platform on AWS, um, to some labs where you can actually get hands-on experience of driving through the technology if you've never had experience of using Ansible automation platform in your environment.

Um, we've actually got a very detailed ebook which helps organizations understand how to run automation in a hybrid cloud at scale, which obviously is the next next challenge. Being able to understand how to pull and twist the levers of a technology is very useful, but our guidance there helps organizations understand the different challenges that they will face in taking that technology and running it out into very large production environments, particularly ones that AWS scale will allow organizations to achieve. And there's another document there that goes into much more detail about specifically how Ansible works with public cloud.

You're very welcome to use our assets. You probably also noticed at the start, um, that Bally and I didn't hide our email addresses. The reason is we're very happy to talk to our customers about our technologies.

So if you would like to speak to us as well, please reach out to us at any time to ask your questions. Just a quick note to tell you about upcoming sessions. Um, this session is obviously part of our fantastic Cloud Friday initiative.

Um, if you stay, um, around, we are going to talk about, um, different things. We are gonna have a, a summary session where we're gonna deal with some of the q and a that arises during all the sessions that we've run during this. Um, this Cloud Friday event.

And then afterwards we're going to host something called the Networking Lounge. Now, uh, this is totally optional. If you feel like joining us to have a much more informal personal chat with presenters like fia, myself and the other members of our team that have helped today, please use us.

And with that, all I need to do is thank you. Thank you for staying on, listening to Fian and myself. We've really enjoyed being part of this session.

And also thank you for being customers of Red Hat and AWS and we look forward to seeing you more often. Goodbye.