Latest / Logixly Live Unsupervised Learning Podcast / Anthropic IPO Reality Check

Transcript

Anthropic IPO Reality Check · Logixly Live Unsupervised Learning Podcast

• 0:00
  SPEAKER_04: Welcome, everybody.
• 0:01
  SPEAKER_04: Another episode of Logics Lee Live, where we're going to cover some of the big stories in AI.
• 0:05
  SPEAKER_04: I feel like this week there's lots of options to choose from.
• 0:08
  SPEAKER_04: We were just talking about one right there with Muse or the open AI announcements today, a dot.
• 0:14
  SPEAKER_04: But we're actually going to go a little bit different direction.
• 0:17
  SPEAKER_04: Uh, you know, as always, we got Ted joining us and Katie back.
• 0:22
  SPEAKER_04: Ben was um sitting in last week, had a great time with Ben, but welcome back.
• 0:26
  SPEAKER_04: You're at our nation's capital.
• 0:27
  SPEAKER_00: I was getting rained on.
• 0:30
  SPEAKER_04: Ah, wonderful.
• 0:31
  SPEAKER_04: Well, a lot going on in the nation's capital as it relates to AI as well.
• 0:34
  SPEAKER_04: But the topic we're gonna hit today is anthropics upcoming IPO.
• 0:40
  SPEAKER_04: Okay.
• 0:41
  SPEAKER_04: And I feel like kind of like what's changed, what's newsworthy.
• 0:44
  SPEAKER_04: So uh they have filed a um an S1, I believe, with uh the SEC.
• 0:51
  SPEAKER_04: It is private.
• 0:52
  SPEAKER_04: This is not public, but Reuters reviewed a copy of the S1.
• 0:57
  SPEAKER_04: So they they did not get a digital copy, as I understand, but they got to review it and read it.
• 1:02
  SPEAKER_04: And so I'd love to dig into kind of some of the highlights of the reporting there.
• 1:07
  SPEAKER_04: But before we jump to that part, I wanna I actually want to start with some quotes from the all in podcast from last week.
• 1:16
  SPEAKER_04: Because I again I feel like we're gonna we're gonna go into the anthropic S1 and it's gonna be all about frontier intelligence, big build, big spend, and frontier models and and everything.
• 1:28
  SPEAKER_04: And I just, you know, Chamath in the All-In Podcast last week, he had a great quote.
• 1:32
  SPEAKER_04: I thought he had a great quote.
• 1:34
  SPEAKER_04: He said he was talking specifically about frontier models, and his he said something I just I found it so funny.
• 1:40
  SPEAKER_04: I literally, it was one of those moments when I um I pulled the car over, I wrote myself an email so I did not forget this, right?
• 1:48
  SPEAKER_04: Um, but it was he's making the point that the premium stuff is the best, and you need the best for things like advanced engineering, mathematics, drug discovery, and the premium being the premium models like the edge of the frontier.
• 2:02
  SPEAKER_04: And then he said, but but you don't need it for narrow use cases.
• 2:08
  SPEAKER_04: And I just like the dichotomy to me is like, you know, we're in the window of AGI, the frontier models are needed.
• 2:15
  SPEAKER_04: We need to have this artificial general intelligence, and then to follow it up by saying, Well, it's actually really for narrow use cases we want to have this general intelligence, which uh to me is um just incredibly fascinating.
• 2:30
  SPEAKER_00: Um I we have a client that I love their use case because we tested against LLMs.
• 2:37
  SPEAKER_00: They have an SLM, right?
• 2:38
  SPEAKER_00: It's a small language model and it is narrow, it is just to their use case, and it outperforms what we're getting if we use the other embedding models, right?
• 2:49
  SPEAKER_00: Like it does a great job, yeah, but it also I think the actual performance time difference was 1200%.
• 2:57
  SPEAKER_00: Yeah, like it is shockingly faster when you're trying to do things at scale.
• 3:01
  SPEAKER_00: Like the 100% you don't need this perfect if it is for a very narrow use case.
• 3:07
  SPEAKER_02: Like yeah.
• 3:08
  SPEAKER_00: So we like I see narrow in the business sense of like there there are great smaller models for narrow in the use in that.
• 3:15
  SPEAKER_00: And I honestly, like from a general standpoint, I don't think it's gonna get a whole lot better at rewriting my emails.
• 3:20
  SPEAKER_04: Well, I and that's what I feel like it comes, you know, we talked a lot about Jev last week and the unbundling of AI, if that's gonna be a theme.
• 3:27
  SPEAKER_04: And lo and behold, today OpenAI.
• 3:31
  SPEAKER_04: I I think we put it at a couple of weeks.
• 3:33
  SPEAKER_04: You said a couple weeks.
• 3:34
  SPEAKER_04: Yeah, it was a couple of days.
• 3:36
  SPEAKER_04: OpenAI has their decision decisions API, which looks more or less to be functionally equivalent to what Jev had produced last week.
• 3:45
  SPEAKER_04: Um, but but in that mode of like this unbundling, um, you know, to take some of the the thoughts from Chmoth is like, hey, you know, if you unbundled coding, like like you're gonna get a point to where this is state of the art, more or less on agentic coding, and it doesn't need to be wrapped up in fable, right?
• 4:07
  SPEAKER_04: You can you can pull that out.
• 4:08
  SPEAKER_04: Katie, we do so much document capture, extraction, classification, summarization.
• 4:13
  SPEAKER_04: You just said writing emails.
• 4:15
  SPEAKER_04: Yeah, like those are great examples of where you can unbundle pieces 100%.
• 4:21
  SPEAKER_04: And we and I know we've done that, but but but I I I just feel like the thinking that you have to go and get the biggest and best off the shelf and apply it everywhere, that's just not the case, right?
• 4:32
  SPEAKER_00: Yeah, it's not, but nobody grew up going, man, I can't wait until I have a Toyota Corolla.
• 4:37
  SPEAKER_00: That wasn't anybody's like dream car as a kid, right?
• 4:40
  SPEAKER_00: Yeah, you look at a model and it's like here's this really big, great one.
• 4:43
  SPEAKER_00: Everybody, you want the new shiny thing because you think it's gonna be better, even if it's not.
• 4:48
  SPEAKER_01: That's exactly right.
• 4:49
  SPEAKER_01: That's a great point.
• 4:50
  SPEAKER_01: Yeah, and if just just the general models are are gonna solve at least 90% of the use cases for businesses, the differentiator differentiator comes down to your change management, you know, how you lead your teams.
• 5:05
  SPEAKER_01: It it becomes like a utility, like a like the electric company to where everyone has it, you know.
• 5:11
  SPEAKER_01: They all have the same knowledge, but now it comes down to execution.
• 5:14
  SPEAKER_00: Yeah, we'll drill in on that later.
• 5:16
  SPEAKER_00: Yeah, it's a I like that.
• 5:16
  SPEAKER_04: Yeah, that's actually a thank you.
• 5:18
  SPEAKER_04: That's an incredible segue.
• 5:19
  SPEAKER_04: Is that a segue?
• 5:20
  SPEAKER_04: It is a segue because so we'll we'll hit you know going into the anthropic S1, the um we'll get to the utility conversation, but I want to go through it first from the revenue standpoint.
• 5:32
  SPEAKER_04: So, in case you are not familiar with the revenue ramp from anthropic, and then I'm gonna quiz uh Katie and Ted here real quick, but the um revenue ramp for anthropic is 2023 uh essentially zero in revenue.
• 5:47
  SPEAKER_04: 2024 is 400 million in revenue.
• 5:52
  SPEAKER_04: 2025 is 4.5 billion in revenue.
• 5:58
  SPEAKER_04: 2026 run rate.
• 6:01
  SPEAKER_04: So you take, I think they have July's revenue, and this is actually not from the S1, this is from another source, probably an estimate, but uh 65 billion run rate as of July for 2026.
• 6:16
  SPEAKER_04: Okay.
• 6:16
  SPEAKER_04: Pretty massive, right?
• 6:18
  SPEAKER_04: Like that literally is I believe that is the the quickest scaling of any company ever.
• 6:23
  SPEAKER_04: Um okay, so 65 billion in annual run rate.
• 6:28
  SPEAKER_04: I I'm gonna I'm gonna guess, I think most people are, if you took the December run rate, they that could be at 70 or 80 billion, but let's just let's say 65 billion is what they do in 2026, okay?
• 6:44
  SPEAKER_04: Do you um well one what is the largest revenue company, US-based company, uh, right now?
• 6:56
  SPEAKER_04: What is the largest revenue, dollar revenue, top line revenue?
• 7:00
  SPEAKER_04: Take a guess.
• 7:02
  SPEAKER_01: I'm gonna say Walmart, but they're the largest employer, but I don't know if it's a lot of people.
• 7:06
  SPEAKER_04: Walmart's number two, and revenue wise, they're very close to number one.
• 7:09
  SPEAKER_04: So do you know number one?
• 7:10
  SPEAKER_04: I don't.
• 7:11
  SPEAKER_04: Okay, so Amazon, Walmart, both at right around 700 billion in annual revenue.
• 7:16
  SPEAKER_02: Yeah.
• 7:17
  SPEAKER_04: Okay.
• 7:18
  SPEAKER_04: Apple comes in at around 400 billion, would be number four.
• 7:21
  SPEAKER_04: So again, the data point, just to put it in perspective, in your basically four of revenue, Anthropic has scaled it to 65 billion.
• 7:31
  SPEAKER_04: Yeah, and that's um, they're growing 10x.
• 7:34
  SPEAKER_04: If they do another 10x, they're really right, right?
• 7:37
  SPEAKER_04: If they do what they've done the last few years, they're literally on pace to become the one of the larger.
• 7:43
  SPEAKER_04: Okay, so now companies that they do more revenue than right now, uh Uber, they do more revenue than Uber, Uber comes in at 58 billion.
• 7:54
  SPEAKER_04: They do more revenue than Netflix, Coca-Cola, salesman.
• 8:01
  SPEAKER_04: Nike, Visa, Starbucks, McDonald's, Adobe.
• 8:06
  SPEAKER_04: Isn't that wild?
• 8:07
  SPEAKER_04: Yeah, yeah, it's crazy.
• 8:09
  SPEAKER_04: Yeah, yeah, it's crazy.
• 8:11
  SPEAKER_04: I don't know.
• 8:11
  SPEAKER_04: You put some of that stuff in perspective.
• 8:13
  SPEAKER_04: I just think it's fun to like uh good point.
• 8:16
  SPEAKER_04: It doesn't right, it speaks to um there is clearly leverage in AI and tokens.
• 8:22
  SPEAKER_04: Now we don't have it all figured out and what it's gonna look like, but I mean the the revenue ramp is absolutely amazing.
• 8:28
  SPEAKER_01: But just from a standpoint of we've all grown a business, right?
• 8:32
  SPEAKER_01: And so when you think about that scale and how fast, I mean you gotta tip your hat, like you know 100%.
• 8:39
  SPEAKER_01: Yeah, there's a lot of craziness out there, but my God, if you can do that, that's an incredible accomplishment.
• 8:44
  SPEAKER_04: Uh and I think again, I guess, and I'll just acknowledge it, I am kind of an all-in podcast fanboy.
• 8:50
  SPEAKER_04: That's what we were talking about to get in here.
• 8:52
  SPEAKER_04: And Chamath had made the point that um the this you know, the CEO of Anthropic clearly knows how to do something right.
• 9:03
  SPEAKER_04: Yeah, yeah.
• 9:04
  SPEAKER_04: Because a a year ago, well, I mean, I gotta think now.
• 9:07
  SPEAKER_04: A year ago, yes, a year ago, we had not been clod pilled, and it was Sam Altman and only Sam Altman, right?
• 9:14
  SPEAKER_04: And how quickly that has 12 months later, right?
• 9:18
  SPEAKER_04: It is it six months later.
• 9:20
  SPEAKER_04: I mean, January was basically when every everything started to flip, right?
• 9:24
  SPEAKER_04: And it is kind of amazing.
• 9:26
  SPEAKER_01: Yeah.
• 9:26
  SPEAKER_01: You just gotta wonder too, is there a boom and bust?
• 9:30
  SPEAKER_01: Is you know, is what think what goes up, you know, usually must come down.
• 9:35
  SPEAKER_01: And so you you have to wonder what pressures and with technology moving that fast, how fast could that turn?
• 9:41
  SPEAKER_01: But hey, they're at the top of the heap right now, right?
• 9:43
  SPEAKER_01: Enjoy the view.
• 9:44
  SPEAKER_00: I feel like Claude, so everybody sort of went on their own path of like, I'm gonna specialize in this thing, right?
• 9:50
  SPEAKER_00: And Claude really doubled down on code.
• 9:53
  SPEAKER_04: Uh I Katie, I would love to know percent of all tokens that Anthropic sells, yes, which percent are coding tokens?
• 10:02
  SPEAKER_00: 100%, right?
• 10:03
  SPEAKER_00: I it it well but when you look at it like okay, you are an AI expert, Matt.
• 10:10
  SPEAKER_00: How many times a week does somebody come up to you and go, what's the best model to use?
• 10:15
  SPEAKER_00: Like I want to do some AI stuff at home.
• 10:18
  SPEAKER_00: Anyone who's a developer right now, you are getting that question a couple times a week.
• 10:23
  SPEAKER_00: And if you're using it for code, you're gonna go, man, Claude is the best thing ever.
• 10:28
  SPEAKER_00: I don't necessarily think Claude is the best tool for like, I'm gonna answer my home questions.
• 10:34
  SPEAKER_00: Right?
• 10:34
  SPEAKER_00: Like, I it's not the one that I default to for questions like that.
• 10:38
  SPEAKER_00: It's fantastic for code.
• 10:40
  unknown: Right.
• 10:40
  SPEAKER_00: 100%.
• 10:41
  SPEAKER_00: But like I feel like they sort of bought the developers, and the developers are the like influencers of the AI generation.
• 10:49
  SPEAKER_00: And like that was a way to go.
• 10:50
  SPEAKER_00: That's an interesting one.
• 10:51
  SPEAKER_04: Oh, yeah, I can totally see that.
• 10:53
  SPEAKER_04: And I'm curious, I think they, you know, as part of the S1, I think they have to, and I I know they spend a lot of time outlining risks, but I I feel like the, you know, so last year it's open AI, this year it's anthropic.
• 11:05
  SPEAKER_04: The switching costs, and I think they they I don't think they described it as switching costs, but I think they said, hey, their their enterprise contracts are not all that long, right?
• 11:14
  SPEAKER_04: And but I think we all know the sw and actually I'll give you an example, you know, the the couple of organizations that we've spoken with, and we've been looking that at this ourselves, are you know, it was GitHub Copilot, and then you know, Claude seemed to take a full step step ahead, but now like it it it's also going back and and talking about, hey, what about other harnesses and other coding models, right?
• 11:39
  SPEAKER_00: Or you know, they're they're losing that edge, they gained that edge by digging in on a place where other people were digging in on I'm gonna generate cooler images.
• 11:49
  SPEAKER_00: Which like that's great if you're looking for like, you know, you go to the T you know Jags game and they've got like, I'm gonna make you look like Trevor Lawrence, like great, that you know, you've you've made a model there, but like that's not what Claude put their attention on.
• 12:01
  SPEAKER_02: Right.
• 12:01
  SPEAKER_00: And they made a great move there, but yeah, code harnesses like a I I would default to GitHub Copilot right now.
• 12:08
  SPEAKER_00: Yeah.
• 12:09
  SPEAKER_00: You know, it's everybody's got their thing that they default to, but you can switch.
• 12:13
  SPEAKER_00: And what I think is crazy from their uh S1 filing is if you look at the amount of committed dollars they have in spend, yes, right, which is 500 billion.
• 12:25
  SPEAKER_00: 500 billion, 400 billion of those are non-cancelable.
• 12:29
  SPEAKER_00: Like they have said we will spend these dollars even if we don't utilize them, we're still gonna pay you for this.
• 12:36
  SPEAKER_01: So I gotta ask you guys a question.
• 12:38
  SPEAKER_00: Yeah.
• 12:39
  SPEAKER_01: What would you compare?
• 12:40
  SPEAKER_01: Because you guys are the coders, you guys are the technologists, switching from one LLM to another, how painful is that compared to switching your internet provider at home?
• 12:52
  SPEAKER_04: Great question.
• 12:53
  SPEAKER_01: So, like ATT, your Matt, I remember you got very upset with ATT.
• 12:57
  SPEAKER_01: We heard a lot of four-letter words.
• 12:59
  SPEAKER_01: Yes.
• 12:59
  SPEAKER_01: And then so yeah, I'm going to Comcast.
• 13:01
  SPEAKER_01: Yeah.
• 13:01
  SPEAKER_01: And Comcast says we'll show up between 6 a.m.
• 13:04
  SPEAKER_01: Right and midnight.
• 13:05
  SPEAKER_01: So give you a little window.
• 13:06
  SPEAKER_01: Yeah.
• 13:07
  SPEAKER_01: But what would you compare?
• 13:08
  SPEAKER_01: Like if we as a company, if we're switching, or you as a person, like how is that, how is that, what's that pain?
• 13:14
  SPEAKER_00: Have you ever gotten a new car?
• 13:17
  SPEAKER_00: It's been a while, but yes, I think.
• 13:23
  SPEAKER_00: Like, sorry, sure, you have to go and you have to go buy it.
• 13:26
  SPEAKER_00: And like, it's not not a huge amount of time, right?
• 13:29
  SPEAKER_00: You might be someone that has to test drive everything, or you could just be like, nope, I test drive that one, love it.
• 13:34
  SPEAKER_00: I'm gonna switch.
• 13:35
  SPEAKER_00: Easy to switch on, and then like, yeah, there's some growing pains, right?
• 13:38
  SPEAKER_00: Like, oh, the the blinker, I have to reach an inch more further forward.
• 13:43
  SPEAKER_00: But like, I'd be I would be more willing to switch LLM providers right now than I would to use a Mac keyboard.
• 13:50
  SPEAKER_00: Okay.
• 13:51
  SPEAKER_00: Like, because the short keys and things, like there, there's so little that like really ties you to it, especially if you're using it from like an API standpoint.
• 13:58
  SPEAKER_00: Like, I just don't feel like there's a ton tying you to it, other than the like, I'm kind of a fanboy of this model right now.
• 14:05
  SPEAKER_04: Yeah, I'm I'm I cannot think of a and I can't think of a right example of an application where from a technology standpoint, like like I'm trying to think of um something where you feel like you always have this FOMO because that's what I that's to me that's what drives providers, the TV providers, Netflix versus Hulu versus like at one point I've got like 18 different services because like I don't want to miss out on Ted Lasso.
• 14:32
  SPEAKER_00: I'm like, yeah, but switching.
• 14:34
  SPEAKER_00: If you were to go right, Netflix is not what I'm gonna use anymore.
• 14:37
  SPEAKER_00: I'm switching to HBO.
• 14:39
  SPEAKER_00: Like, yeah, you might miss out on a show or two.
• 14:41
  SPEAKER_00: Like it's gonna take you a couple minutes, you gotta fill in a couple forms to get the thing set up.
• 14:45
  SPEAKER_00: It's not really that big of a deal.
• 14:46
  SPEAKER_04: That's a good I like that because the streaming subscription.
• 14:49
  SPEAKER_04: Now, the internet provider to me is a pain in the butt because I have to call, cancel, be at home at a certain time to your point.
• 14:56
  SPEAKER_04: Yep.
• 14:57
  SPEAKER_04: Um, you know, all the it get new route, like get passwords.
• 15:00
  SPEAKER_04: Ugh but um uh yeah, the streaming that's a great example because my now I'll give you an example.
• 15:06
  SPEAKER_04: Um Lioness.
• 15:08
  SPEAKER_04: Do you guys watch Lioness?
• 15:09
  SPEAKER_04: Have you heard of?
• 15:10
  SPEAKER_04: I don't know.
• 15:11
  SPEAKER_04: Okay, Sarah has getting some TV regs.
• 15:13
  SPEAKER_00: Yeah, I don't know.
• 15:13
  SPEAKER_04: Katie comes home and she's like, everyone's talking about this, and of course, it is the one streaming service that we don't have.
• 15:20
  SPEAKER_04: Yep.
• 15:20
  SPEAKER_04: So what do we do?
• 15:21
  SPEAKER_04: We just you know dial one down, dial one up.
• 15:23
  SPEAKER_04: It's pretty much painless, and then um that's a great example.
• 15:27
  SPEAKER_00: Yeah, it's crazy, but it's not it's not huge amounts.
• 15:30
  SPEAKER_00: So you have not a ton of friction companies that are not tight.
• 15:34
  SPEAKER_04: Look like no moat.
• 15:36
  SPEAKER_04: Well, that's what it is.
• 15:37
  SPEAKER_04: It's exactly what it feels like is the moat is what?
• 15:41
  SPEAKER_04: And then and then I this is what you know.
• 15:43
  SPEAKER_04: I think I said this on previous one.
• 15:44
  SPEAKER_04: This is where I like if you're Altman or or Dario, it's like, man, we are on this treadmill of the frontier.
• 15:50
  SPEAKER_04: Can we get off this treadmill?
• 15:52
  SPEAKER_03: Because it is really expensive to every three months wow the world constantly.
• 15:56
  SPEAKER_03: Yeah, but you have to.
• 15:58
  SPEAKER_03: But that's that's what that's what Altman's doing today.
• 16:00
  SPEAKER_03: He has to wow to beat last week's news.
• 16:03
  SPEAKER_00: But if you go What a terrible place.
• 16:05
  SPEAKER_00: Well, I mean, he also makes some news that he maybe they make bad decisions.
• 16:09
  SPEAKER_00: We've talked about that.
• 16:10
  SPEAKER_00: Yeah.
• 16:11
  SPEAKER_00: You know, their marketing team, I don't know what's going on there.
• 16:13
  SPEAKER_00: But like if we talk about the no-moat piece, that's where you're talking about a utility, right?
• 16:18
  SPEAKER_00: Like, that's where you're talking about I'm gonna use JEA for my power.
• 16:21
  SPEAKER_00: And I'm not using JEA for my power because I'm like, man, they've got the best electrons, right?
• 16:25
  SPEAKER_00: I'm using JEA for my power because that's who I got.
• 16:28
  SPEAKER_00: And like, thank you, JEA.
• 16:30
  SPEAKER_00: You do a great job for us, right?
• 16:32
  SPEAKER_00: But like it's not a choice when you go and like they're saying this is a utility, you know, this isn't a service, it's not a software anymore.
• 16:42
  SPEAKER_00: Like, this should be treated like a utility.
• 16:44
  SPEAKER_00: It's building that no-moat perspective, right?
• 16:47
  SPEAKER_00: But it's also building in regulation, it's stopping the competition side of it.
• 16:51
  SPEAKER_00: You're sort of building your own little monopoly.
• 16:54
  SPEAKER_00: And it's how that double-edged sword and trying to go the utility route, like very, very strange take for me at least.
• 17:02
  SPEAKER_04: Yeah, I mean, I guess I I like that too.
• 17:04
  SPEAKER_04: So thinking about a utility, which to me, I feel like Comcast ATT is not terribly different because they have to get out and physically dig.
• 17:12
  SPEAKER_04: And there's there's this thing of I can't even buy the thing I want because they're not in my neighborhood.
• 17:19
  SPEAKER_04: So I my options are one of two, and you know, neither are great, as opposed to once you have the connectivity, you go online to streaming and you get all these options, right?
• 17:31
  SPEAKER_04: You can dial them up and down, which is where I think the consumer wants to be, is I want to have and the enterprise, I want to choose the right coding agent for the right task at the right quality, cost, speed point.
• 17:45
  SPEAKER_04: I don't want to only have to go to two providers, but there's room for improvement, right?
• 17:50
  SPEAKER_00: It's we want that flexibility because there's this chance that next week Gemini comes out with something incredible, and I want to use Gemini.
• 17:59
  SPEAKER_02: Right.
• 17:59
  SPEAKER_00: Or open AI, you know, every three months we're wow in the world, and maybe you want to shift things around based on these, you know, new wows.
• 18:06
  SPEAKER_00: That's not happening in utilities.
• 18:08
  SPEAKER_00: Like we're not seeing electricity change.
• 18:12
  SPEAKER_00: Like, yeah, there's some price differentiation, but it is one consistent product.
• 18:16
  SPEAKER_00: We don't have one consistent product here.
• 18:18
  SPEAKER_00: We have a thing where you've got this constantly sort of steeple chasing of what are we doing from a quality standpoint?
• 18:24
  SPEAKER_00: So nobody's gonna tie themselves to you forever because you're good today, that doesn't mean you're the best tomorrow.
• 18:30
  SPEAKER_01: So low barrier to change.
• 18:32
  SPEAKER_00: Yeah.
• 18:33
  SPEAKER_01: You have client concentration.
• 18:34
  SPEAKER_01: We talked about that before the show.
• 18:38
  SPEAKER_01: You've got um obviously from a competitive standpoint, and you've already publicly announced that your product could destroy the world.
• 18:45
  SPEAKER_01: Yeah, are you gonna invest?
• 18:48
  SPEAKER_00: The destroy the world, it's really like such a weird stance to take.
• 18:53
  SPEAKER_01: Like, yeah, it that's a risk factor.
• 18:55
  SPEAKER_00: I I get it, and I get that you have to identify these things, but like when we when we're gonna release a new car, do we come out and go, it could run people over?
• 19:05
  SPEAKER_00: Like, I understand it can do bad things.
• 19:08
  SPEAKER_00: We are there are guardrails.
• 19:11
  SPEAKER_00: You know, I I think you and I talked the other day about like, oh, a release didn't go out because it wasn't safe, and we're all supposed to clap that you didn't do a bad thing.
• 19:19
  SPEAKER_01: Yeah, congratulations.
• 19:21
  SPEAKER_01: You didn't release a faulty product.
• 19:22
  SPEAKER_00: Hey, that's right.
• 19:24
  SPEAKER_04: Here's a press release.
• 19:24
  SPEAKER_00: Here's a press release.
• 19:26
  SPEAKER_00: Yeah, it's a wild place to be.
• 19:29
  SPEAKER_00: Where and honestly, like this is it's unprecedented for a business to be able to do this, right?
• 19:34
  SPEAKER_00: To go from zero to sixty-five billion, that's that's a gargantuan change in four years.
• 19:40
  SPEAKER_02: Yeah.
• 19:40
  SPEAKER_00: And so, you know, it's really easy to go, like, oh well, they're not doing all of this right, because well, nobody knows how to do this right because it's never been done before.
• 19:48
  SPEAKER_02: Right.
• 19:48
  SPEAKER_00: But it is some of the ways that they're coming out with stuff.
• 19:51
  SPEAKER_00: I'm just like, eh, I kind of rolling my eyes with that guy.
• 19:54
  SPEAKER_04: Yeah.
• 19:54
  SPEAKER_04: Yeah.
• 19:55
  SPEAKER_04: Yeah, I agree 100%.
• 19:56
  SPEAKER_04: Um just scanning through here other other the the headlines from the um from the uh S1.
• 20:05
  SPEAKER_04: So you already hit the infrastructure piece and any other thoughts on I think the only other one is it well let me ask you this.
• 20:12
  SPEAKER_04: Is it is is intelligence more important to the global economy than industrialization, electricity, and the internet?
• 20:25
  SPEAKER_04: And I won't say like this year, next year.
• 20:26
  SPEAKER_04: I'm saying like when we look back in like five or ten years, are we really gonna say, yeah, that that will because he's not the only one to say it.
• 20:35
  SPEAKER_00: Sundar Pachaya said it years ago.
• 20:36
  SPEAKER_00: I use it in almost every talk I give publicly as a quote that I hate, right?
• 20:40
  SPEAKER_00: Which is AI is gonna be more important than electricity, because I don't know if you guys realize, but the AI actually takes a a lot of electricity.
• 20:48
  SPEAKER_02: Right.
• 20:48
  SPEAKER_00: It's kind of the foundation of the whole AI thing in the first place.
• 20:51
  SPEAKER_02: Yeah.
• 20:51
  SPEAKER_00: So like it does it have massive changes like having electricity did, yeah, 100%.
• 20:58
  SPEAKER_00: But like electricity, I that's a game changer.
• 21:01
  SPEAKER_00: Like it's so it is so much bigger.
• 21:03
  SPEAKER_04: Okay, so what if we said what if we just said internet?
• 21:06
  SPEAKER_04: And I know it's like it's it's really hard to answer this, right?
• 21:10
  SPEAKER_04: But um well, I mean, I so again, five, ten years from now, we we have we AI is you know kind of everywhere.
• 21:19
  SPEAKER_04: Like, will we be to a place where internet is actually infrastructure for agents and not for humans?
• 21:26
  SPEAKER_04: So so my example of hey, look, I'm I like to book my own travel.
• 21:29
  SPEAKER_04: Uh if I mess it up, at least I messed it up, which I've done all the time.
• 21:33
  SPEAKER_04: But like in a world where like maybe we are using agents to book travel, like the internet is less about an interface that that we use as like I'm not going on to travel sites, but will we look at it and say, hey, it's actually infrastructure for agents?
• 21:49
  SPEAKER_00: I mean, I possibly, but like but and I think maybe this goes back to the utility side as well.
• 21:55
  SPEAKER_00: Like, I'm not gonna set up a little power plant in my basement, right?
• 21:59
  SPEAKER_00: I'm not making In my own ISP, but I could host my own small model, right?
• 22:05
  SPEAKER_00: Like you you could do this stuff without the internet, theoretically.
• 22:09
  SPEAKER_00: Like you can still get that benefit.
• 22:11
  SPEAKER_00: Yeah.
• 22:11
  SPEAKER_00: So like, yes, there is like the internet does that.
• 22:14
  SPEAKER_00: I think that we're all I don't I this is terrible admission, but like I'm gonna sit and doom scroll on Instagram all day long.
• 22:20
  SPEAKER_00: So you know, I I think we're still gonna use it for that.
• 22:24
  SPEAKER_00: But yeah, there is a huge piece of the internet that goes on, like, you know, entertainment.
• 22:27
  SPEAKER_04: I get it.
• 22:28
  SPEAKER_04: Yeah, I yeah, yeah.
• 22:29
  SPEAKER_00: I wish I didn't.
• 22:30
  SPEAKER_00: If somebody's got a plan, you let me know.
• 22:33
  SPEAKER_00: I'll get off the Instagram.
• 22:34
  SPEAKER_04: Yeah, what about you, Ted?
• 22:35
  SPEAKER_04: Any thoughts on like it is like because I kind of feel like you have to move away from AI as an assistant and like an accelerator to like AI as like a paradigm shifter that five like business will just be totally different in five to ten years because it it like business is fundamentally different with the internet now, right?
• 22:55
  SPEAKER_04: Like you if you don't have a website, you don't really exist as a business.
• 22:59
  SPEAKER_01: Yeah, I think it is.
• 23:00
  SPEAKER_01: I think I think there's gonna be huge shifts.
• 23:02
  SPEAKER_01: We've just seen it in the 15 years that we've been in business and some of the transformations that that we've worked with our you know our customers on to see their business transform, you know, and be go from very labor intensive to to not a lot, um, or with a lot of automation.
• 23:20
  SPEAKER_01: So I think, yeah, in five to ten years, it's it's gonna be I I think we're gonna have huge, huge gains.
• 23:25
  SPEAKER_01: And I think it's gonna show like in you know the GDP and it's gonna show in in quality of life.
• 23:30
  SPEAKER_01: I think I think things will be better.
• 23:31
  SPEAKER_01: I'm going with the uh glass half full thing.
• 23:33
  SPEAKER_00: I yeah, like the quality of life thing.
• 23:35
  SPEAKER_00: I was talking to somebody about this in the medical space.
• 23:38
  SPEAKER_00: Like when you look at a doctor, go talk to any doctor and ask them what they're doing at nine o'clock at night because they're looking at your charts, they're putting in the notes, right?
• 23:47
  SPEAKER_00: Eight it's this AI is gonna take our job.
• 23:50
  SPEAKER_00: No, it is so many people are working 90 hours a week.
• 23:53
  SPEAKER_00: And you know what?
• 23:53
  SPEAKER_00: I really kind of want my doctor to focus on me and our appointment, not what they have to do in a chart, right?
• 23:58
  SPEAKER_00: Like there's so much of this stuff that is just like better quality of life.
• 24:02
  SPEAKER_00: That yeah, oh my gosh, a good example, yeah.
• 24:06
  SPEAKER_00: It's great, but like it's super interesting topic, I think.
• 24:10
  SPEAKER_00: Very much on the same team of like it's not yeah, it could destroy the world, but like a lot of things could destroy the world.
• 24:16
  SPEAKER_00: We still do them.
• 24:18
  SPEAKER_04: Yeah, I mean, um I don't even know.
• 24:22
  SPEAKER_04: It it's I I I I feel like if you're if you're like we've been in the industry working on it forever.
• 24:30
  SPEAKER_04: I I feel like I it's hard for well, we did the panel last night, which is about cyber.
• 24:35
  SPEAKER_04: And uh so it was a panel last night with University of North Florida and and we were talking about things, and I feel like the people inside the industry, unless you're at one of these frontier labs, right?
• 24:47
  SPEAKER_04: Are but but people who are in the industry working with it, I feel like we're I don't know, more pragmatic.
• 24:52
  SPEAKER_04: Like, like yeah, it's like every technology.
• 24:54
  SPEAKER_04: You can you can do bad bad people can do bad things probably faster with better technology, but you can put any technology in there, it's not AI, it it literally is any technology.
• 25:06
  SPEAKER_04: Right.
• 25:06
  SPEAKER_04: That's that is that is that is the way the world works.
• 25:09
  SPEAKER_00: Well, I think that you know, like the industry, yeah, we know more about this just from from doing it, and like that's why people are more pragmatic.
• 25:16
  SPEAKER_00: But if you go on Google, like AI news, I do this once in a while for like I gotta go do a presentation, I want to get some headlines that are from the last couple days.
• 25:26
  SPEAKER_00: The only headlines that came up yesterday around were anthropic says AI will end the world.
• 25:32
  SPEAKER_00: This thing is the world's biggest disaster.
• 25:34
  SPEAKER_00: Oh my god, everything's awful.
• 25:35
  SPEAKER_00: Like, it is journalism, like journalism is catastrophizing a lot of it.
• 25:41
  SPEAKER_00: And like part of me wonders how much of this is AI creating the headline about how bad AI, like it's so much.
• 25:49
  SPEAKER_01: I mean a tremendous amount, right?
• 25:51
  SPEAKER_01: Just is that a real word?
• 25:52
  SPEAKER_01: Catastrophizing?
• 25:53
  SPEAKER_01: Is that like a clay county thing?
• 25:55
  SPEAKER_00: It might be a clay county thing, but you know what I'm keeping.
• 25:58
  SPEAKER_01: Let me spell it for you, Ted.
• 25:59
  SPEAKER_00: Okay, thank you.
• 26:02
  SPEAKER_00: You know what we are here for clay high?
• 26:06
  SPEAKER_00: Used to be C-L-A-Y, Clay High.
• 26:10
  SPEAKER_00: We can spell a four-letter word.
• 26:15
  SPEAKER_04: That's great.
• 26:15
  SPEAKER_04: That's good stuff.
• 26:16
  SPEAKER_04: Well, uh, we're gonna leave it for there.
• 26:19
  SPEAKER_04: Um, and uh, and you know, obviously, as we always say, if anything comes up and we've had some come up, uh feel free to reach out.
• 26:27
  SPEAKER_04: We love to hop on outside of the show or give us ideas for the next show.
• 26:31
  SPEAKER_04: Again, there's absolutely no shortage of of stuff to talk through and um love to have those conversations one-on-one if you want to just just connect with us.
• 26:40
  SPEAKER_01: Are we having a guest?
• 26:42
  SPEAKER_04: Uh we oh yeah, so let's let's promo that real quick.
• 26:45
  SPEAKER_04: So, yeah, so this is the unsupervised learning kind of brand that that you get the three of us jokers um kicking it around a little bit.
• 26:54
  SPEAKER_04: But we're also having a Logicsly and Friends podcast where we're gonna bring in some business leaders and I and we're gonna do um kind of, you know, this clearly is not formal, so that's not us.
• 27:06
  SPEAKER_04: We're gonna do kind of an informal conversation where we're gonna talk about businesses and business leaders and go through origin story, industry, and then business, and specifically hone in on how AI is changing business and what businesses are doing to adopt.
• 27:23
  SPEAKER_04: So, yes, so we're gonna film that next week with Gray Mabry from iVenture Solutions.
• 27:28
  SPEAKER_04: He will be joining us right here in this room.
• 27:31
  SPEAKER_04: And we've got a whole, I'm not gonna give it all away, but we've got a whole backlog of uh really interesting um businesses that we're gonna dive into.
• 27:41
  SPEAKER_04: So we'll we'll still be here at this time every week, and then kind of out of band, we will have some some other conversations.
• 27:48
  SPEAKER_04: So yeah, excited for that.
• 27:50
  SPEAKER_04: So thanks for thanks for that reminder.
• 27:53
  SPEAKER_04: And uh be sure to join us uh next time.
• 27:55
  SPEAKER_04: And until then, keep learning, even when it's unsupervised.