AI Agents and Beyond: Infrastructure, Super Intelligence, and the Elmo Hack Fallout | TSG Ep. 883

Looking for super intelligence. I'd be satisfied with anyone who's smart out there. You're watching Textron Gang.

Hey everyone. Happy Friday. Wow.

Friday. I, I just felt like we were talking about AI on Monday, you know, and then maybe it was Tuesday or Wednesday. Artificial general intelligence, a GI today is super int intelligence.

Are we all dumb? I don't know, but we've got a lot to talk about here at Textron Gang. We've got great people to talk about.

We've got a whole Seattle crew going on today. He's always in Seattle, usually, unless he's traveling. Our friend Fred Wilmont, joining him in Seattle though, and he is traveling, is Ira Winkler staying on the West Coast man with in Silicon Valley, John Schwartz, and still up.

I don't know if you're in Lake Placid still, but the dean, Mike Vard. Hey Mike, how are you? Still am and boy, it's hot and sweltering in Lake Placid.

There's no Such local as it is in Seattle, as it is in Seattle, we hear. Not as hot as it is in Dubai probably. But anyway, that was the green room discussion.

Let's jump into our real discussions for today. We're gonna start off, you know, AWS had a bit of a shindi this week, week, and, uh, they made a whole bunch of announcements as you would imagine around ai, AI agents, Q bedrock, all, all the AWS buzzwords, as well as opening up registrations for reinvent in December. But Mike, what's the deal with AI agent builders?

It looks like AWS wants to be king of the hill. Yeah, oddly enough, AWS and all those folks in Seattle descended upon New York this past week and launched a thing called Amazon Bedrock Agent Core Services. And it's basically a, the frameworks that you need to go build and deploy AI agents at scale.

And they're pointing out that all these things right now are missing for all our chitchats about AI agents taking over the world. They're like saying, you can't really do that unless you have things for securing their identities and giving them gateways to access APIs. And then the tools we're building and deploying them.

And admittedly, this is a preview of something that they're rolling out, so they're a little bit early. But one of the things that seems to be coming through loud and clear is that, uh, enterprises are now looking at all this stuff and starting to put the pieces together. And it's not sure to my way of thinking, but maybe more of them are gonna rely on the cloud than ever because they don't have these capabilities in an on-premise environment.

At least that's what AWS is betting on. And it was interesting 'cause they had the head of development for Intuit was there talking about how they built a whole operating system for AI agents on top of the AWS service. And Alan, this will warm your heart, but the, uh, head of platform engineering enablement for Reuters was there also talking about what they're up to and their plans for building out AI agents.

And it's still early days as far as the enterprise is concerned, but it looks like I'm betting that AWS all these services will be copied by every other cloud service provider shortly. So Alan, or it looks like we're in a race for AI agent infrastructure. Yeah, well, there's a lot of, as in that AI agent infrastructure thing.

But look, this is not a new story though. You know, for me, the agent Agentic AI thing first kind of burst on the scene with Benny, often Salesforce agent force and building thousands and tens of thousands of agents. And ever since that day, every vendor under the sun has said, we wanna be your agent, man.

I want to be your backdoor agent, man. And why wouldn't AWS want to be that too? My, my issue here though is I'm sure they'll build great agents.

B Bedrock is a fantastic platform. Yeah, but let me, oh, sorry Alan. One second.

Let me just put the cherry on. The problem is the world is moving away from single cloud platforms to multi-cloud. Everybody wants to put, you know, a little piece here, a little piece there.

They don't wanna be locked in any one platform. Do do the a, the AWS agents work outside AWS ira. Go ahead.

Well, that was actually kind of the point I was gonna bring up because really, I, I hate the whole broad term ai, if you haven't known that before. Mm-hmm. And because AI is really just mathematical, um, library programs, it's algorithms.

And so what AWS is essentially doing is they're putting together and packaging and making it easier to implement the algorithms when you're using it within AWS. But just because you have these algorithms in place and something easier to make them possible, the problem is you need the data to put into it because it, you know, you have to train the ai, the AI has to be able to make good decisions in order to make good decisions. It has to be able to have access to the appropriate data and so on.

And having the program itself is one thing, but having all the data being able to pull into it is another thing. And to your point, Alan, if it's sitting there looking at, if all your data is on AWS in buckets that are accessible to the agent, maybe you have access to the data, but it still has to be in the right formats and stuff. And that's one of the problems, however.

But if you're in a multi-cloud environment and all this stuff, it's, it's like any other program. The program is only as good as the data that you put into it. But the problem is most people aren't sure what data to put into it first.

And they're guessing, and this is why, for example, a chief data officer is critical in most large organizations now, because they have to be able to architect all the data in the organization to be able to feed these a God, I hate the word AI programs, and this is where AWS it's, it's kind of nice to have, you know, but at the same time, you know, it's kind of like going to one of these little, um, maybe I'm dating myself, where they used to have these little art studios where you could go in and make your own ceramics. They still don't have those places, the Pottery places. Oh, that's good to know.

But you could go in there and get one of those, you could get one of those turn tables. But the problem is you still have to get the right clay, you still have to have the skills to make the clay in the right area, even though you have the turning tables. So AWS is giving everybody those little spinny tables, but they don't necessarily have the right clay.

They don't know how to architect the programs to do it. It just makes it easier for people who know what they're doing to have a set of library programs to implement. Yeah, I would agree with you in the sense that, um, there's a handful of organizations like maybe Reuters and, uh, Intuit that have the tech resources and have been early on this whole journey.

But I'm, I'm betting that 95% of the companies out there are still trying to figure out, you know, how do we make the actual AI agent never in mind actually give you data or anything else, they're just struggling with the little piece of code that they're gonna use as the baseline. So we live in an AI bubble. I'm sorry, John or Fred, go ahead.

Yeah, so I, there's a couple things that I think are interesting here. Um, both of you guys are right, but the challenge is where AI workload's gonna get executed. And so who owns the agent, right?

Owns the workload regardless of where the data is. And the challenge is, it's inaccessible to a lot of folks, right? And, and, and here until now, there wasn't an SDK, there wasn't, you know, sort of this reach into the browser.

There wasn't some of the buildable tools. And then also the thing that everybody fights about is which service during what time of day gives you enough course to return your AI workload at the speed you needed with the amount of tokens required to generate it. And so, you know, doubling down on the amount of infrastructure they're willing to put behind it, I think is a statement.

And I would argue, you know, the rationale behind choosing between, you know, using Gemini for this particular workload or, you know, the bedrock, bedrock framework that allows you to get access to all of these other models. You know, AWS says, Hey, look, I have the highest percentage of the infrastructure in the cloud. I want everybody running their AI workloads here, but adherence on now they haven't given us any tools aside from like, let me weaponize bedrock specifically in order to make, you know, the entire user flow easier for people that don't do this organically and aren't, uh, data scientists and, and, uh, platform wizards as well.

You know, that's an interesting take, Fred. It's, it's a discussion I had not on the gang, on a Textron interview text strong TV interview I did yesterday or the other day. And that is, we are at a funny juncture in this AI journey right now.

I think you're a hundred percent correct. Initially, the people who were, they were building this for, the people who were using it, the people who were pushing the envelope were very much, let's call them AI engineers, data scientists, you know, tho these kinds of people that they were the initial audience, if you will. But now I think it's transitioning to, we want platform engineers, we want developers, we want security people, we want DevOps engineers, we want SREs, we want that sort of traditional IT stack from the developer on down to start using ai.

And that that means what we've gotta change the interfaces, the use cases, the nomenclature to make it fit their world, not the world of the data scientist or the AI geek. And it's gonna take time, right? There's, there's a transition that's going on right before our eyes, but you're right, he who controls, you know, the platform that these agents are being built on and going forth is in the cat bird seat.

And the way I see it is everybody and their mother recognizes that in, in the tech world, and they're all vying for it. And part of the problem with ai, we discussed it on yesterday's show, is, you know, it's the ultimate billionaires club. Yeah.

You gotta have serious cash to be able to play in this. So just say, I think DevOps engineers should just rebrand themselves as ai, DevOps engineers get themselves a 30% raise. And who's to say they're not an expert?

'cause nobody else knows what they're doing either. So Maybe, yeah, You know, there was an interesting, was it, I think part of the announcement with that this, uh, new addition of market marketplace, um, kind of as a central hub to, for businesses to deploy and manage specialized AI agents. And again, you talk about the billionaires club.

So there's AWS and I think Microsoft has agent store Google Cloud's, AI agent marketplace. I mean, there's kind of paring Well does agent force, right? They want it.

Cisco's doing its service now is doing it Again. But these are the biggest of the biggest companies, Right? It's, it's the billionaires clubs.

Well, yeah, I still think I'll use the example and what people need to understand. To me, it's, I like analogies. It's like a spreadsheet right now.

The problem is that if you go to a spreadsheet, like most people do a spreadsheet to create a very, very simple table because, and, and it's like you're paying a lot of money for something that could be done in a Word document, frankly, at the end of the day for most people. But if you happen to really, really know how to work a spreadsheet, you could work magic. You can run your company off of Microsoft Excel.

Not that I'm saying you should, but you could because the capability is there. The problem is what AWS is doing sounds incredible at the moment for a variety of reasons. You know, much like Fred said, it's enabling a lot of what companies can do.

The problem though is most companies don't have to Fred's and everybody else's discussion, they don't have the expertise to take all these library of agents. They don't have the data, they don't have all this other thing to actually make it useful. They, to Alan's points like a mil, you know, billionaires club, where a few large companies will have the people, it will simplify their way of ramping up.

But it's a people who are few and far between who have the ability to staff it up and like what we're talking about in a later segment today, you know, you have to throw lots of smart people to make it useful at it with lots of money. Absolutely. So we're gonna wrap this one up and take a break, but, you know, we're creating ai, have and have nots, and maybe that's what we're destined to, you know, You know, I was gonna say before you, before you end this segment, um, we, we can maybe talk about it later in another segment, but Oxford University has been doing a lot of research into the haves and the have nots, and it's a very small pool of the haves and the vast majority of the rest of the world are the have nots.

Oh, no, no doubt about it. But those haves will accelerate, right? Become super intelligent.

Maybe that's what we're gonna talk about next. Nice segue, John. You're watching Textron Gang, Discover Textron Group, the epicenter of tech innovation.

We are your go-to for reaching IT leaders and practitioners worldwide. Our secret impactful content that sparks awareness, engagement, and top quality leads with us. You'll access editorial websites, streaming videos, virtual events, custom content analyst research, and more.

Join our satisfied clients. Let's revolutionize your tech journey. Contact us today and tell your story to the world in the most powerful way with Textron Group.

Hey folks, we're back and we're talking as Alan John. We're talking about super intelligence, which I'm not quite clear is any different than what we call, uh, a is it A-G-I-A-G-I? It's artificial general intelligence.

But there's a report now that meta's out trying to train machines to think like humans by starting over and treating them like babies and trying to teach them from the very beginning everything you need to know to be a human. And of course, they have been spending in mountains of money just trying to attract talent. John, I know you've been covering this space a little while, but it seems like especially out in the Valley, everybody in his brother who can have AI on their resume is now worth a billion dollars plus.

But how real is superin intelligence to you? Uh, I think it is real. I mean, we, you alluded to Mike the, uh, this teaching study.

Well, what Mike Metta is doing is their researchers are studying how babies learn. So one of the examples or a simple action like throwing a tennis ball under the air and predicting what it will do to require to catch it. And they're, they're looking at things like the ability to form full sentences.

Uh, there's a, uh, chief AI scientist at Meta who said they're building a world model akin to embedding an AI agent with common sense. But it kind of all goes back to, we talked about super intelligence and a GI, and again, they kind of blur together for me at times. But simply Meta is trying to create this AI lab that would surpass human intelligence with super intelligence.

So there's this research lab under development. This is the, the, the news where Meta had acquired essentially scale ai, CEO. Uh, they made a, an investment of about $15 billion of scale ai or 49% of the company.

And Zuckerberg is very heavily engaged in this, evidently, reportedly spearheading this effort. And he's trying to put together this super intelligence group of 50 top AI researchers and engineers. And since he's not happy with AI's, that meta's AI efforts.

And so he has kind of taken, taken the lead in putting together this, this team in which they're offering these incredible pay packages and are trying to poach people from open AI with packages worth up to a hundred million dollars. And the idea is to move beyond AI research to a more specialized focus on a GI and ultimately super intelligence. And the idea is they're gonna weave a GI and let's just say, let's just stick with a GI that you wanna weave it across their suite or chat bots, augmented reality glasses and other projects that we don't know about.

So in a sense, this is where they're going. And it all kind of ties in what we talked about yesterday with this incredible amount of money that these companies are spending. I think in the ca uh, on data centers, in the case of meta, I think it's 60 to $65 billion they claim this year alone.

Um, it's gonna be interesting to see this race and this concept of a super intelligence lab. Um, it it, it also shows Zuckerberg's ultimate frustration with LAMA four, um, and, and kind of trying to move forward and trying to gain some sort of competitive edge. But again, as Alan has said, and we all have agreed upon this the last couple of days, it's all about the companies with the resources sitting on piles of cash who are gonna be able to do this and try to gain some sort of advantage over everyone else and then cook the rest of us in.

So, John, I'm so confused. I I, before doing the gang today, I was texting with Daniel Newman from Future. Yeah.

Oh, yes, yes. He, he's, he's been very keen about this. About, Yeah.

So, but we were talking from the way Daniel and I were talking, and the way Daniel's explaining it, super intelligence is sort of a way station on the way to a GI it sounds like you are saying super intelligence is beyond a GI Yes, that's, that's based on what has been reported out of meta. They consider it's the next threshold or next level. But then again, they use these phrases so, so often interchangeably that I almost would prefer just to make, to simplify things, to just continue with the a GI reference.

Super Intelligence is probably new marketing spin or a marketing, uh, branding of its, uh, maybe they've dec decided that it has more sex appeal, but in a sense, the, the project within Meta is this super intelligence lab that they're working on. Yeah, Well the, the data center, the building, two of these data centers, one's in Ohio, and that's the big one that's, uh, uh, I forgot Hyperion, which is The side, no, Prometheus. Prometheus is the one in Ohio.

Yeah, Louisiana. Hyperion is in Louisiana. It's gonna be bigger than Manhattan ira, right?

That No, I'm, I'm not kidding you. Seriously, they said it's gonna be bigger than Manhattan one day, a data center complex. You Know what also OpenAI is working on something in Texas that they claim is a data center that can be the size of, uh, central Park.

Yeah, no, that's, well, but that's, that's small potatoes. That's a park that's smaller than Manhattan of Manhattan. Yeah.

But it's just, again, a scale. You think about it, it's like near, it's crazy. Let's bring, let's bring Lisa in here for a minute.

Um, it seems to me, Lisa, that these definition of a GI keeps changing and it is becoming more of a marketing term and the goalpost are shifting and, you know, for what's supposed to be really just research and development, there sure is a lot of marketing spin here. What's going on from your perspective? There is, I actually, um, great to see you guys.

I, I agree with John that super intelligence is a sexier term and folks are really wrapping their heads around it. People can, I don't wanna say people can wrap their heads around it, but I was just on stage at a big conference a couple months ago and interacting with a robot, talking about super intelligence and what it's going to be enabling everyone to do from marketing perspective. Marketers I talked to every week, everyone's, and I've saw this on this program before, heavily leaning into AI powered marketing tools.

What they're gen ai, they're using it for content creation, for videos, et cetera. Now they're leaning into how to deploy AI agents. So that's where marketers are in really understanding, well, how can AI agents actually be on our org chart, help marketers leverage these tools to enhance campaigns, improve customer engagement, how can agents help marketers create more personalized, um, more engaging customer experiences?

But I think from a terminology perspective and is a little bit blurry, and I think I like John's point of, of, of super intelligence being, it's kinda like a sexier, well, you, You know, and, and a weird, I'm, I'm glad you mentioned that, Lisa, because in a weird way, think back to when, uh, Facebook or meta, whatever you wanna call it, well, Facebook, when they were throwing billions of dollars into the metaverse, remember that, remember that big push a metaverse then became, and, and I went back and forth with Facebook about this, they renamed the company and I said, so you basically squandered billions of dollars in this metaverse push? And they're like, no, no, no, no, you, you're wrong. It's not a metaverse, it's it's AI now.

So they it is like a moving goalpost. Yeah. It kind of redefine Well, they're just right.

That's pure spin Again, actually. Well, but guys, let me, I'm sorry. Go di Well, could I just point out what this company is essentially doing from, I like to simplify things.

This company is essentially training AI data right now because this whole concept of super intelligence, again, I look at the fundamentals fundamentally, when you have artificial intelligence, using my Dr. Evil quotes, artificial intelligence requires models, it requires data, and it requires the tuning of these models with the data at hand. So when they're talking about, like, you know, when I read the articles, it's like, oh, they're talking about a child, how a child slowly learns No, what they're doing is they're talking about taking data, forcing it into their programs, and then figuring out how to tune and train their AI programs to anticipate the right data.

Like you said, a ball being thrown up in the air. What they're really doing is they're saying, okay, here's the trajectory of a ball. A person watches the ball intuitively and figures out where it's gonna land.

What they're essentially doing is tuning their mathematical models to calculate where it's gonna land. Now, to Alan's point again and again, this is a case where the have, where the haves are gonna beat out the have nots. And the reason that this is important is that Zuckerberg and everyone else are investing billions of dollars into pre-training these AI models so that they can better think.

And the fact that they have these very well trained AI models, possibly definitely years, if not decades ahead of everyone else, is gonna make their models that much better in the future. And everybody's going to either have to license them or catch up. Now I see Fred shaking his head, I don't know if I he agrees or disagrees.

Yeah. What I think is interesting is the way that we talk about the, the storyboard of what the mission is in these particular cases. Mm-hmm.

And so if you look at the mission for like anthropic, right? Which is to, you know, have a steerable interpretable ai, and you look at X AI is to understand the universe, you know, what meta wants to do basically is have a, you know, a a, an AI assistant for every person on the planet. They're very different sort of goals.

And when you look at the overall outcomes, you know, IRA, a hundred percent, if you look at each of those organizations, their mission's based off of the data they have, for sure. But the in the intention behind that is the question mark that I have, right? These labs are, are, you know, uh, basically a supercomputing, uh, cold war that's happening right now.

And the characteristics of those things are how fast, which models and then what's the value for the end user? I think the curious part there is going to be what happens after all of that data analysis has been, has been run through, and what does the democratization look like? Hmm.

So on yesterday's show, guy Coer, our friend Guy Coer said something I re really made me think, and I thought about it a bunch yesterday and last night. You talk about their mission. Fred, let me tell you what could be another mission here that is not said out loud.

When you look at Zuck and you look at Elon Musk and you look at, at the, you know, Larry and Serge from Google and Sam Altman and all these AI tech bros, Peter, Theo, the whole bunch of them, this whole cabal of tech bros, multi-billionaire, AI tech bros. Are they really doing this for the money? Are they doing it for the ego to say they were first?

Or are they doing it for immortality? Are they looking to set up systems that they could download into and live forever? And it's gonna take, that is the ultimate billionaires club, right?

Is there, is there anybody I think there who thinks that, you know, training AI for intelligence based on humanity is kind of a fool's errand? Because I'm like, Well, no, because Facebook is trying to give it a human, a human touch because it, for them, it's about robot they're going at, they're not looking to get developers developing more code or security, people making better security. They're looking to do humanoid robots, very humanistic sort of interface.

But make no mistake that that immortality is, is a thing. Yeah. And you're absolutely right.

Agree. They're, they're motivated by all, all of the above. Agree.

But especially the brutality angle. Sure, everybody has to live for others, but they have the money to actually try to, and they can Do it. Yeah.

To try to do it. And is that of a Hollywood kinda script, right? Anyway, let's take a break.

We're gonna be back here on Textron gang. We've got one more poor Elmo. What happened to our little Elmo?

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 in Poor Elmo. Nobody's safe these days. He got hacked and was made to say all kinds of mean and nasty things.

But Lisa, let's jump to you right away. This feels like to me, like every marketer's nightmare. Oh, is this gonna become more commonplace?

Yes. You know, when you think about Elmo, he's lovable. He's furry, he's eternally three and a half years old.

The, the picture, the face of innocence in the cyberspace. He's been a force for good for Sesame Street since the 19 hundreds in recent years, he's become more cyber obviously on X. But you know, what happened was really shocking.

Many people I, I thinks got about 650,000 followers on X and the Sesame Street workshop that runs it was absolutely shocked. The Anti-Defamation League was shocked. Emma Kel was hacked over the weekend and it included a string of posts.

They were racist, they were anti-Semitic messages criticizing the president calling for the release of the Epstein files. I mean, it was really, really hot. And so now, now, while Sesame Street Workshop has addressed these quote disgusting anti-Semitic posts that were hacked, there's still screenshots out there.

That's the challenge with things on the internet, is they can live forever. There's about six screenshots in total that are going around. Sesame Street says the account is now secure.

But what it shows is no one is safe. You know, they, they went to something that is probably the essence of innocence on the internet and have it now. That also kind of reminded me of some of the grok troubling posts that we've seen just a couple weeks after XAI put a hold on the X account because Grok went crazy posting similar messages.

It posted an apology, but still, um, yeah, but, But Grok wasn't hacked. There's a difference. That's true.

That's a good point. No, Elmo was, was purely innocent and purely hacked. Right?

And it just goes to show that parents need to know cyber and kids what parents need to do. Um, and it's some of the basic things like multi-factor authentication. I don't know the specifics of what happened with Elmo's account, but it just goes to show that nobody's safe.

Well, Could I jump in as the only person who's probably been personally targeted by the New York Wa, Washington Wa what was it? The Wall Street Journal? Um, if you go ahead, this is, there's zero new to this in the first place.

Kids who will follow Elmo shouldn't be on Twitter anyway. That's they, you know, like 13 years old. This is like, oh my God, Elmo, I, I don't think Elmo's the icon of innocence.

I mean, Elmo was manhandled by a sexual predator for a while. This is not shocking in any way, shape, or form, but let's look at what happened. We're talking about the takeover of Twitter accounts, which is not new at all when you go back and search for ever all the listeners, uh, while you're listening, Google Iber, Winkler Syrian electronic Army and Cockroach.

And you will see that the Syrian electronic army targeted me personally. I was told by the FBI, they hated me more than Obama, by the way. And they hacked the Wall Street Journal Twitter feeds to call me a cockroach.

They hacked Buzzfeeds Twitter feeds to call me a cockroach accounts, which are much larger. They hacked the Associated Press Twitter feed at one point in time saying that there was an attack on the White House, which moved the stock market. So what we are talking about fundamentally are attacks against personal accounts.

And even the DOD accounts were hacked in the past. So what we are talking about is an identity compromise because of a lack of multifactor authentication, sharing of accounts and a bunch of other stuff. How Elmo was hacked.

It's probably the same thing. They guessed the password or some or phish to password more likely and were able to go ahead and hack. But the bigger lesson is not poor Elmo.

I mean, anybo, anybody on Twitter who's looking at Elmo, you're not looking at him for serious motivational, good warm, fuzzy feelings. In general, if you ask me, it's kind of a get a joke account, but it's just a sign that there are much more critical accounts. I mean, Trump is using Twitter as public policy, which to me is horrifying.

But at the same time, his account was hacked in the past. You know, you had Biden's account hacked in the past. You had all these other people's accounts who were hacked.

And we have to learn, yes, we need security, but we also need to learn, wait a second, what it goes on on these accounts might not really be what is going on on these accounts. And it requires public perception, public awareness that Twitter is not an authoritative source on that identity, even though you think it should be. You mean if I don't pay Elon the money to get the check mark in everything, I'm, I thought if you did that, you are secure.

I can probably pay for a check mark and start an Alan Shimmer account for you if you want a check mark. And I'll be happy to, to tweet out. Should I would disagree with Ira on one point.

I think it might be helpful for adults to check in with Elmo every morning. 'cause you know, it just levels you out and maybe makes you a better person. But let me ask a question.

So Elma was a victim of abuse and manhandled. I I thought it was Cookie Monster. No, it was Not.

It might have been the same one. Dev guy has two hands. Remember that?

No, I mean, look, this is, you know, I filed this under why we can't have nice things. Mm-hmm. Right?

Whether you think Elmo should have an account and how much you believe in Elmo Santa Clauses and the Tooth Fairy, the fact that, you know, people will use these accounts to spout this kind of garbage. And like it or not, there are miners on the internet. They are on X and they're gonna see it and eh, you know, this is very nice people.

Hey, you're, I guess that means you're not anybody till you get your account hack is that I was hacked Black Hack 2005, also anti-Semitic stuff and it was pornography, it was all kinds of nonsense. And I had to get all my emails back. I know Fred, you knew me then.

You might probably remember. Yeah. Um, what a pain in the ass outwards.

But anyway, it, it, yeah. It's not fun. It's not fun.

And you know, I found out who hacked me and he's in jail now, not for hacking me, but he deserves to be in jail. Wow. Um, anyway, poor Elmo, you know, but look, shock and all, as Iris says, Elmo's really kind of the tip of the iceberg here.

Yeah. Somebody better protect big bird. That's all I gotta say.

And Grover, let's not forget. Grover. Grover.

Grover. Grover's. Awesome.

Poor Grover. I don't, guys, There's more lessons from Oscar the Grouch, if you ask me Oscar, the Oscar was definitely a person. Guys, we, we gotta, we gotta pull the plug on this Great Friday.

Lisa, thanks for joining in. I know you had some technical difficulties and we appreciate you fighting through those rest of my gang members. Enjoy the, the cool Northwestern weather up in Seattle, out in the West Coast or Lake Placid or whatever you are.

Mm-hmm. Of course we're roasting here in, uh, south Florida, but we do have a full tech, strong TV lineup immediately following the gang today. Check that out.

We'll be back Monday with lots more new good stuff. Enjoy the weekend. We're out.