Episode Transcript
[00:00:00] Speaker A: Hey, welcome back to this special live weekend edition of the Gregory and Paul Show. I'm Gregory.
[00:00:07] Speaker B: Hey, everyone. Paul.
[00:00:09] Speaker A: And we always break down the latest in SaaS, startups, AI. We have a bunch of topics that we want to cover today. We're obviously going to start with the Technology Brothers podcast acquisition by OpenAI, which Paul's very excited about and I'm excited about too. We're also going to talk about.
Yeah, yeah, I know you're inspired.
We'll talk about OpenAI and their experiments with advertising. Paul's got a bunch of cybersecurity topics he wants to dive into. Do you want to share anything?
[00:00:39] Speaker B: Yeah. Lots of, lots of exciting cybersecurity news this week related to Axios NPM library with a hundred million downloads a week getting hacked and a bunch of other crazy shit shenanigans happening.
[00:00:54] Speaker A: So cybersecurity. We'll talk about Waymo hitting half a million view. Half a million rides. Not views. Half a million rides. We'll talk about the first billion dollar person to two person startup.
[00:01:04] Speaker B: We think maybe allegedly.
[00:01:06] Speaker A: We'll talk more. Allegedly.
We want to hit on layoffs. Fortunately, Oracle announced a big layoff and there's been some other layoffs have been announced. So we'll dive into that. We'll talk about Artemis going to the moon and of course we have a really fun meme to finish off the show with.
[00:01:24] Speaker B: That's right.
[00:01:25] Speaker A: All right, Paul, should we. Do you want to kick off the Technology Brothers thing and then I'll chime in?
[00:01:29] Speaker B: Yeah. Well, I mean, for anyone that is not aware, G Brothers is a super popular tech niched podcast, YouTube show, live stream that's been around for just a little bit, I would say. What, more than a year? A little bit more than a year. Recently it got acquired for the low hundreds of millions by OpenAI. Incredible. Absolutely incredible. So I think I actually recently just saw a tweet. Somebody said their combined stream per episode is about 70,000 views.
[00:02:04] Speaker A: I saw that too. It's not that big.
[00:02:06] Speaker B: Yeah, it's not that big. So niching down. Oh, I mean, okay, the comparison was against them, against Joe Rogan. Right. Joe Rogan's podcast gets hundreds of thousands, millions of views.
[00:02:19] Speaker A: Hundreds of millions. Right. It's a. It's 100. It's huge. 100 million.
[00:02:21] Speaker B: Huge. Widely distributed, widely popular, not very niche. I think he was acquired for $250 million by. Was it Spotify?
[00:02:30] Speaker A: Yeah. So maybe maybe a similar number. Look, I think people maybe are exaggerating, but I think it's a Hundred million that got acquired for which is still an enormous sum.
[00:02:37] Speaker B: Yeah, that's still enormous sum for I have a really. I.
[00:02:41] Speaker A: So I've looked into this before on different occasions regarding like let's say media valuations and audiences. So it's actually not that uncommon or if you look across the spectrum of like what the valuations are versus the audience. CNBC only has like a hundred million, a hundred thousand concurrent viewers. But everyone knows that every CEO watches it. Jamie diamond watches it. Like really influential people watch it. We're on the consumer side. All the research I've ever done and I don't do a lot of consumer.
Millions and millions of followers, millions and millions of views is not that valuable. So like an Instagrammer that has a million followers, that's just consumer fashion, beauty, whatever.
It's not even enough to monetize. You need to do like 5, 10 million in order to like make what I would call like real money doing it. And I think this same thing applies to like streaming to media properties. So a show like Joe Rogan which has enormous amount of downloads and views. Right. That thing is giant, hundreds of millions of people. I mean it's, I can't even imagine how big it is. Right. But it's, it's about the same value ultimately as an audience like the Technology Brothers podcast is able to.
[00:03:53] Speaker B: Yeah.
[00:03:54] Speaker A: Command. So I don't think it's unusual but that's difference you like a B2B finance audience. And the last thing I'll say here is like even if you dive into the CPMs and the payouts on YouTube, the same dynamic exists. B2B and finance categories pay way higher, like 10 times higher and the audiences tend to be 10 times smaller or even a fraction of what like a large audience is on a consumer or property.
[00:04:19] Speaker B: Yeah. So niche, niche does is actually much better for certain, certain niches.
[00:04:25] Speaker A: Business, finance, technology.
The, the niches that like there's a high value audience. I mean the people that I would imagine watch us and watch Technology Brothers and pay attention to technology media, they're all people that work in high tech. They are generally high paid in relationship to the average, the average consumer.
[00:04:43] Speaker B: Do, do you think it's also a bit on the perception side perception that technology or finance is worth more than you know, other stuff like consumer. And also it really depends on the company that's acquiring.
[00:04:57] Speaker A: Well yeah, I mean I, I, that's what I thought you're going to ask is like the perception of technology brothers in particular, like I don't know, do they, those guys just, just knocked out of the park.
[00:05:05] Speaker B: That too, right?
[00:05:06] Speaker A: Actually they're so dreamy and good looking and they have gray hair which they're good looking.
[00:05:09] Speaker B: That's fair.
[00:05:09] Speaker A: I'm a little, I appreciate, I like, I like the hair. I like the hair.
[00:05:13] Speaker B: I love the hair.
[00:05:14] Speaker A: I, I, I would give, I would give him a premium valuation just for that.
[00:05:17] Speaker B: I love the hair, I love the suits, you know, the production. I'm a little bit jealous. Gotta say, Gregory. I'm a little bit jealous. I am a little bit jealous.
[00:05:25] Speaker A: I mean it, it's, it commands a premium valuation for sure.
[00:05:29] Speaker B: Sure.
[00:05:30] Speaker A: But more seriously, they did knock out of the park. They somehow captured the attention of like the high tech audience in a way that like no one has.
[00:05:40] Speaker B: It does not mean that we're going to start wearing suits though. Hopefully that, that's funny.
[00:05:46] Speaker A: Anything else? Do you have any other comments on this?
[00:05:48] Speaker B: Oh, let's move on.
[00:05:49] Speaker A: Okay. So we should talk about the OpenAI advertising product went live a couple months ago. They've been signing up advertisers and they claim to have hit a hundred million dollar what I would call run rate and re new. A lot of people explain what.
Yeah, yeah. This is super for me it's super important. I'm reading like professional journalists calling it ARR annually recurring revenue which is in my opinion a very specific measurement of revenue for SaaS. Companies that have long term contracts that are almost guaranteed to get that every year. And advertising, it doesn't work like that where you have to sell the customer every, every IO, every campaign has to be resold so it's not annually recurring revenue. I was really surprised to see so many people talking about it in that way to the point where like you could even make a very dramatic statement that it's fraud. But like this is a really important thing when it comes to understanding the valuation of a company and claiming that revenue is annually recurring in a contractual way that it's not, it's not, it's, it could be, it could be considered fraud.
[00:07:01] Speaker B: Yeah, I don't know.
[00:07:02] Speaker A: I, I have to be honest. I don't know what their contracts are. Maybe they have an annually recurring contract for advertising revenue. I've never seen that. It's not a common practice in the media business, but it is possible. So I, I'll leave it, I'll leave that out there. Just that like I don't have the, I don't know all the details but that's my perspective.
[00:07:19] Speaker B: I wonder if they have like a minimum spanned locked in for a minimum span also. Yeah, they don't.
[00:07:24] Speaker A: I knew. I know, sorry. I know that minimum cents are really high. They were like they wanted. I forget what it was like millions. It was very expensive and they didn't have a lot of the targeting or other features that people would expect.
[00:07:38] Speaker B: Yeah.
[00:07:38] Speaker A: And I've heard, I've heard like a mixed bag on the results. Some people seem to be happy with them, other people are not perhaps as happy with the early results.
[00:07:48] Speaker B: Ditto idea of who's happy, who's not happy particularly.
[00:07:55] Speaker A: I don't have a list or anything.
[00:07:56] Speaker B: Okay.
[00:07:57] Speaker A: I did read on Twitter that like I, I forgot I thought there was Walmart. Right. Came out and said that the click through rates were much lower than what they experienced other media. So that's I think was one that was on the record.
[00:08:06] Speaker B: Much lower than ex. What they expected. I think.
[00:08:09] Speaker A: Yeah. And I think lower than like what they typically see with search is I think how they, they.
[00:08:15] Speaker B: Yeah. So it's interesting, right? Like all of these AI companies are trying to become more profitable, I suppose. Claude just raised.
[00:08:24] Speaker A: I don't know if they're going to be profitable. They're definitely trying to get more revenue.
[00:08:26] Speaker B: They're trying to borrow. Right. Get more revenue. They just raised what, a hundred some odd billion once again. And the story that I hear, yeah, the story that I hear is even with all of that money raised, they're still about 18 months of Runway away from needing to raise again, which is incredible. Incredible or insane.
[00:08:47] Speaker A: So. So dude, what I, my understanding is that the compute costs are that high.
[00:08:53] Speaker B: Yeah, that's right. Yeah, that's right.
[00:08:54] Speaker A: So and that, that's why they. Go ahead, you finish.
[00:08:57] Speaker B: Exactly. No, no, no, exactly. It's, it's basically they're being pressured to invest into compute, right? Into GPUs, TPUs, the infrastructure. And if you listen to Dario, the CEO of Anthropic Talk about all of this, he is basically saying he's being pressured. But if he gets the timing wrong. The timing as in if he invests 18 months into the hardware and he doesn't get the incremental benefit of all of that hardware, the company goes bankrupt. And this is Anthropic, probably one of the largest competitor of OpenAI. So all of these companies are having go having the same pressure.
[00:09:36] Speaker A: Yeah. I read that the reason they shut down Sora was not that no one used it, it was that they needed the compute.
[00:09:44] Speaker B: Yeah, they were burning something like $50 million a day on compute and they
[00:09:48] Speaker A: wanted it to train a New model was like a rumor I heard, which is like, really logical. Like, there's totally logical. Yeah. The compute constraints are that acute that they would need to like turn off that product.
[00:09:58] Speaker B: Which brings back to the last point is there's only a couple company in the world that's doing AI, that has compute, that has the time horizon, can actually outlast everybody, which is probably Google. So I coming back is saying I think Google's just gonna eventually own the distribution for B2B side. At least I do.
[00:10:21] Speaker A: I agree a hundred percent. I mean, I'm a Google shareholder. Yeah. I identified this at the bottom of like, let's say that the trough of when they just appeared to be weak.
[00:10:29] Speaker B: Yeah, that's right.
[00:10:30] Speaker A: Against OpenAI. And I told everyone I knew I was a republic. I'm like, look, I think counting Google out is foolish. They have all their resources in the world. In fact, like, I even think this is their strategy. Let OpenAI lead the way and then they've got everything they need to swoop in. I still think that that's their strategy when it comes to advertising. I don't know if this is actually how it'll play out, but my guess is like, or at least if I was Google, I would let OpenAI break ground with the advertising, which still be backlash from users. There'll be growing pains with advertisers. So let all of that play out. And then if I was Google once, that's all, you know, happened, and some of it will be challenging for OpenAI no matter what, then you can kind of swoop in and either make, make a determination, like, do you want to undercut them and go for lower prices? Like, there's a lot of angles that they can take. So they let OpenAI shoot the first shot, and now they can, like, they can look at that and make a strategy around it. I think it's quite smart. And so I agree with you that I think Google's in just a really strong position.
[00:11:32] Speaker B: All right, all right.
[00:11:33] Speaker A: Cybersecurity.
[00:11:34] Speaker B: Yeah, this is good.
[00:11:36] Speaker A: This is good. Okay.
[00:11:37] Speaker B: This is my favorite topic of the week, only because of a couple stories on how exactly the hack happened. Okay, so background story. There's this library or this package that gets a hundred million downloads a week, that software all across the world.
[00:11:53] Speaker A: This is the Axios thing.
[00:11:54] Speaker B: This is Axios.
[00:11:55] Speaker A: Yeah. I had no idea where this thing was.
[00:11:57] Speaker B: Okay, so just like very, very short bridge version. This library helps APIs talk to each other. So if you're building an API that needs to call out to some other API.
You basically use this library to do the HTTP communication. This library got hacked. The story, or the postmortem of the story is one of the core maintainers got social engineer to install a virus on their computer. Essentially the end to end story was they got invited to join the Slack room.
[00:12:31] Speaker A: Yeah, this, this is good. This is good. Okay.
[00:12:33] Speaker B: It's crazy. They got invited to join a real Slack workspace and this workspace was branded as a famous company.
[00:12:41] Speaker A: So hold on, it was branded as like a well known company.
[00:12:45] Speaker B: That's right. Yeah. It's brand.
[00:12:46] Speaker A: So like a CrowdStrike or, or Salesforce or something.
[00:12:51] Speaker B: Somebody that could potentially be like a customer.
[00:12:54] Speaker A: NetApp. I don't want to, want to like slander anyone, but yeah. Okay.
They picked a well known SaaS company. This is fascinating. And okay.
[00:13:03] Speaker B: And then basically the person who reached out cloned the founder of the company with the same likeliness. So everything was set up to mimic this company. This fake, this copy of the company.
Right. They even had multiple channels where they shared LinkedIn polls and people chimed in and like things. So there was real social activity inside of this fake Slack channel. Super convincing. And then the hack happened. They scheduled a meeting with him to connect send him a Microsoft Teams meet link. When he clicked on the link, a fake software or virus software popped up that mimicked the Microsoft Teams update install. So he, you know, in a moment of, you know, compromise, he clicked the install and lo and behold, entire 100 million packages.
[00:13:56] Speaker A: That's how the hacker software, they got access. Access. So dude, we, we talked about this like earlier like how, I mean you have to do those zoom updates and team updates, like teams updates all the time. Like all the time. I would have been completely fooled by like this, this attack vector was, was clever.
[00:14:14] Speaker B: It's super clever. It's kind of social engineering. It's not actually anything to do with the vulnerability.
It's just this is going to happen more and more because it's going to be easier and easier to do this. Right. Like you could.
We didn't even really talk about deep fake. Let's say like people introduce some procedure into not installing or catching these kind of third party.
[00:14:35] Speaker A: Well, we've seen the videos, right. I mean they could have invited them on the teams call and then done like a deep fake with you know.
[00:14:40] Speaker B: Totally.
[00:14:41] Speaker A: Yeah, you know, yeah, exactly.
[00:14:44] Speaker B: So cybersecurity is going to be massive in the future. There's a couple more stories. So AI company called Merkor also got hacked. Mercore is basically like a Training data provider for all of the big companies. Amazon, Walmart, massive companies. And their data got leaked onto the Internet. And that basically means the hacker got access to all of the training data for these big companies. And the last one was clock code where source leak was leaked.
[00:15:15] Speaker A: Again, human mistake, right, dude, the source code. And so yeah, explain this one. This was interesting.
[00:15:21] Speaker B: Super interesting. So clock code itself is partially open source. Basically what that means is it has a whole bunch of skills and like prompts. That's open source, but the actual code that runs clock code is not open source. But over the last week, one of the releases embedded a compilation file with the extension map. That file once, once uncompiled, basically revealed all of the source code. And that was introduced because of error in the workflow. Oh yeah, basically another human error.
And so, so people are of course blaming all of this and let's say AI, right? Because people don't read code anymore.
[00:16:06] Speaker A: Was that an AI mistake?
[00:16:07] Speaker B: Not directly. AI it is also, you know, all of these workflows are. Nobody reads these workflows deployments anymore. You know, like we used to do code review. Nobody has time for that. And especially AI company that, you know, is a big proponent of vibe coding. I think the key here is that they need to slow down and hey, so, so didn't.
[00:16:38] Speaker A: Didn't they find in a source code that. Or it was revealed that there are other versions of CLAUDE internally that they don't. That people didn't know about. Yes, of course, I'm jumping ahead.
[00:16:50] Speaker B: No, no, you're.
So of course once people got access to their code base, they dug through it. Right? Of course, yes. The answer is that there is a employee only verification gate which only anthropic employees has access to. And this hypothetical version basically turns on a whole bunch of features that makes the clock code harness better. So for example, the source code explicitly tells the agent to double check its work before reporting success.
[00:17:24] Speaker A: Really?
[00:17:24] Speaker B: Which a lot of people complains about because cloudcode usually comes back and says, hey, my job here is done. And then you test it. There's a whole bunch of bugs internally. The user of Clock internally, the anthropic users has a special check to tell it to verify. It's worse.
[00:17:42] Speaker A: Is that, Is that not. Would that like double compute costs or something though? Is that why. That might be one reason why they.
[00:17:48] Speaker B: No, I think it's just bad, you know, it's. It's. If they, if they had this instruction
[00:17:53] Speaker A: publicly, it basically it would reduce confidence in their model.
[00:17:57] Speaker B: Right?
[00:17:57] Speaker A: Yeah.
[00:17:58] Speaker B: Right. Like your AI. You're telling your own AI to double check its work all the time, Right.
[00:18:04] Speaker A: Against that, I think it's still good. Anyway, I hear you.
[00:18:08] Speaker B: It's the non. You know, it's the. It's the responsible thing to do, but might not be the best pr.
[00:18:16] Speaker A: My God, what a week in cyber hacks. Amazing. You know, I'm a huge fan of Darknet Diaries, so shout out to Darknet Diaries.
[00:18:23] Speaker B: That.
[00:18:23] Speaker A: That is a fantastic part. If you haven't heard of it, check it out. It's like a cybersecurity podcast that just talks about all kinds of hacks and they cover everything. It's fascinating. There's so much going on there. Like, I always joke, Paul and I, we should like security company. I don't. I don't even know where to begin. There's so many things that need to be.
[00:18:38] Speaker B: We totally should. Yeah. So at big companies, there's red teams. Right. So red team are literally experts or internal people who sends you phishing emails, I'm sure. Yeah.
[00:18:49] Speaker A: Professional hackers.
[00:18:50] Speaker B: Professional hackers or professional DevSecOps who internally will send you these phishing emails. Hey. And they try to get you to click on the thing. If you do, then you have. You get another email saying that you have to go through security training. That's. That's a real thing.
[00:19:04] Speaker A: And I. Yeah, last time I worked at a big company, I would always flag everything I got that I wasn't sure about that asked me to sign up for something, particularly if it came internal, because that's the big trick is like it comes from the head of HR or something. Like, go here and sign up for this thing. And you're like, dude, is this legit? Like, can you punch in your Social
[00:19:23] Speaker B: Security number or like, download this file? Yeah. And then you get a talk from hr, give me all your personal information.
[00:19:29] Speaker A: I'm like, don't you guys have this? And I don't know if I'm going to give it to you.
[00:19:33] Speaker B: Right.
Yeah. And then there's outside concerns, right? Like that helps governments, helps critical infrastructure companies, and even big companies, big tech companies to red team their employees. It's a total legit business.
[00:19:46] Speaker A: So, dude, I have to share this though, because like, I was. I on my run, I was listening to this Darknet Diaries episode where the guy talked about Spotify hacking, and it was fascinating. So if you're out there, you should definitely check it out. But the thing I didn't realize or what was super interesting about it was like, I was originally trying to do auditing for record labels because it's very difficult to audit the streams. Yeah, it's logical. You're like, how do we audit this? Like, and this is a problem that exists all across the board. Right. How do you audit advertising? Like, it's very difficult to do. So he, he, he asked for the data and he was working for the record labels. And so when he started doing the audits for the music industry, what he actually uncovered was like, massive amounts of fraud that people were like, faking the streams. And the reason they're faking the streams is because they've. Yeah, yeah, They've signed up as, like, indie artists or somehow they're collecting checks for the payouts.
[00:20:37] Speaker B: Yeah, yeah.
[00:20:37] Speaker A: Which is, like, sounds like a rather convoluted way to make money. Right? You sign up for a free account or you sign up for a bunch of fake artists, or there's ways to, like, steal artist accounts and then you get the checks. And it's like, at a massive scale. It was fascinating. And so all the things that they need to do to, to manage that. I, I read that and I was like, there's so many niche cybersecurity businesses out there. Like, and it's a, It's a great business.
[00:21:03] Speaker B: Yeah. Cyber security, anti fraud. We had an issue last week on the podcast or on the show to talk about fraud in the medical. But we realized that it's, it's basically everywhere.
[00:21:15] Speaker A: So that was inspiring too. Dude, we should start a thread on. We should start something on indie SaaS or something on Reddit and say, you guys go make, Go make security companies stop making product. Product hunt clones.
No more product hunt clones and AI reply bots. We'll make some security companies. Yeah, that's a. I reply bots. Yeah, we don't need anymore. I don't need another automated SDR tool. Okay. Obviously, the industry needs more security tools that we even know what to do with because, like, when I look at this industry from, like, a high level, I can't even make heads or tails of what some of these companies do. And then they're, like, earning billions and billions of dollars. Like, there's just. I guess there's enough attack vectors that there's enough need for, like, specialization in every category in cybersecurity. It's. It's fascinating.
[00:22:04] Speaker B: So on the, on the Spotify side of things. Right. Like, I wonder if the record labels are on it as well.
[00:22:10] Speaker A: So, so, okay.
[00:22:11] Speaker B: Because, like, you just quickly. Yeah, go ahead.
[00:22:14] Speaker A: Well, I was just gonna say, like, he was hired by the major Labels to audit it. So my best guess is like, there's not really a lot of interest in them to.
[00:22:22] Speaker B: To.
[00:22:23] Speaker A: At least according to him and data he had.
[00:22:25] Speaker B: Yeah, they.
[00:22:25] Speaker A: They actually, they. They. It's in their interest to make the industry legitimate. He says all the indie artists, independents, that's where they have a. They have a real incentive to fake it. He even went as far to say that, like, there is no growth in independent music. That basically major labels, major artists are the ones that are popular.
[00:22:45] Speaker B: It gets all of the attention.
[00:22:48] Speaker A: He basic. Well, he basically said that, like, any indie artist that is popular is fraud.
I see more. He didn't want to, like, go that far. I'll go there. That was what he was saying is that all indie artists are fraudulent. Like, there. There is no a. Bottoms up indie artist.
[00:23:04] Speaker B: Grassroots indie.
[00:23:07] Speaker A: That's what he said.
[00:23:08] Speaker B: No, man. Like, that seems like exactly what the record label wants you to leave.
You know.
[00:23:14] Speaker A: Okay, all right, let's move on.
[00:23:17] Speaker B: We.
[00:23:17] Speaker A: We've. We beat a dead horse with cyber security.
[00:23:20] Speaker B: Way. What is happening.
[00:23:21] Speaker A: Wayo. Okay. Waymo. They recently announced that they hit half a million rides. And. And what I really like about this is that it's real autonomous. I would call robots in production doing like, real things in the real world. Right? You've heard a lot about optimists and like, some robots going to come into my house and do my dishes and make my bed and do my laundry and all the things like, I hate wasting my time doing, but, like, that stuff is so far away, I don't even believe it's going to happen in my lifetime.
[00:23:49] Speaker B: So you're. You're the kind of person who would want a bunch of robots in your house?
[00:23:55] Speaker A: 100%.
[00:23:56] Speaker B: Just doing various chores, cleaning things, folding laundry.
[00:24:01] Speaker A: Dude, I want robots.
My vision of robots is like, yeah, I want an optimist that does everything. Everything I need to do at my house. I never have to like, worry about cleaning, cooking, all these domestic chores. I also want an optimist that's like a truck pulls up and it's got 10 optimuses and it builds me a house or it does like a bunch of physical things, like labor becomes optimist. Like, you rent a truck with like a hundred of these robots on it, and you're like, we need to go and clear this forest or something. And the robot said, we need to fight a forest fire. 100 trucks show up, each each containing like a unit of 75 firefighting optimuses. And boom. We just put the fire out like it is Just done.
[00:24:41] Speaker B: We.
[00:24:42] Speaker A: The vision of this stuff is fantastic and I believe someday it'll happen. But I think that stuff is like so far in the future that I don't expect it to happen in my lifetime. But what I'm excited about with Waymo in particular is like this is really an autonomous robot doing like a real
[00:25:00] Speaker B: world specific real world thing. Yeah, that works.
[00:25:04] Speaker A: That's a hundred percent here today, like driving people around. There's no driver. It's amazing. Like it exists and I think it's fantastic. I've used it many times. I love it.
[00:25:13] Speaker B: So I used it in San Francisco. It's, it's great. It's absolutely great. You use the app, you call it, it shows up, it drops you off. Flawless. Do you think the price of taking a Waymo ride would be cheaper than a human driver?
[00:25:26] Speaker A: It has to be. I mean so, so.
[00:25:28] Speaker B: But there's no incentive for the car company to make it cheaper, right?
[00:25:31] Speaker A: Well, so, so this is, this is, this is actually a fairly big issue. So in Seattle they pass all these laws which I think are ridiculous to make Uber driving doordash all of these things very expensive because they have these really high minimum wages. Like I pay like double the price to take an Uber on Uber here then, then even in the Bay Area to the, to the extent of like I've been very vocal on Twitter about this. It costs me almost the same to take an Uber from where I live in Seattle to Bellevue to go like an event than to, to fly San Francisco. It is unbelievably stupid to the point that like I don't go to Bellevue and I just realized I could double down on going back to San Francisco. So like flying in an airplane is around the same price. It's crazy. So obviously that means that like with Waymo it should be a lot cheaper because it won't be.
[00:26:21] Speaker B: So it's, it won't, it won't fall
[00:26:23] Speaker A: under the jurisdiction of minimum wage. You won't have drivers.
[00:26:26] Speaker B: But then the only way to make it cheaper is to regulate it. Right? Like the airlines. So like airline nowadays is public util. Basically. No, no.
[00:26:37] Speaker A: Would. It wouldn't be. Well my. So maybe, maybe I'm missing something but like if you don't have, don't have to pay the driver, that cost is removed from, from the company. So they don't. They can offer it for less.
[00:26:48] Speaker B: Yes. But if they out compete Uber or real human drivers to monopolize the market and it would all be robo drivers and there's no incentive for them to lower the price once they monopolize.
[00:26:58] Speaker A: Well, that's, that's, that's a different scenario.
[00:27:02] Speaker B: That's what's going to happen though.
[00:27:03] Speaker A: Well, they'll be competing. They'll be, they'll be competing rideshare, autonomous rideshare companies and that will keep prices down.
[00:27:09] Speaker B: Right. So you, so going back to maybe,
[00:27:13] Speaker A: okay, you turn this into a free market discussion. Turn this into a complicated, complicated free market discussion. So like in theory, if it's a free market and there is multiple players in the autonomous driving space, it should help keep prices down. There is Waymo, There will be maybe a Rivian one. Right. Uber is going to partner with them.
[00:27:35] Speaker B: Yep.
[00:27:35] Speaker A: There's Chinese rideshare companies. Maybe Lyft will get their act together. So I, I do think that like there is, there is competition. You know, it is interesting how like Uber did seem to win big and like I used to take Lyft, I don't use it anymore at all. But they really have become like the, the dominant market player by a huge margin. So. Well, will like robo taxis play out that way? I don't know. I mean Tesla wants to be in the game, right? Supposedly you can take one of these robo taxis in Austin. If anyone's taking one, like comment or reach out to us, I'd love to know about the experience because I don't think it's available anywhere else.
[00:28:08] Speaker B: Yeah, they have the data, they have the training data, they have the cars. So they should be ahead of the game than everybody else.
[00:28:15] Speaker A: But in theory, if you remove the Uber driver, you take that cost out, then you can offer it the ride for less and you can still make the same money. I mean there's, there's other aspects too, like less accidents, safety and insurance. Well, insurance costs should in theory be lower.
[00:28:31] Speaker B: Basically you don't even need to own a car anymore. That I am very well.
[00:28:35] Speaker A: I, yeah, that, that's been me. Like I. My wife has a car, but I don't have a car. I take Uber everywh for long.
[00:28:40] Speaker B: She drive you everywhere.
[00:28:41] Speaker A: Because I'm crazy to get groceries and I don't even like that. I complain. I don't like going in the car.
[00:28:46] Speaker B: So robot, robot, in the near future doing every chore that Gregory has to do.
[00:28:51] Speaker A: I hope so. Please. I can't wait.
[00:28:53] Speaker B: I like it. I like it.
[00:28:55] Speaker A: All right, what's next?
[00:28:56] Speaker B: The first billion dollar two person startup question mark. Okay, so this, this com, this company, New York Times recently did a article on, on these guys. Let me Pull up the.
[00:29:09] Speaker A: What is the company even called? What is this fraud called?
Where I Stand on medv.
[00:29:14] Speaker B: Medv M E D V I. Okay, so the full story is that these two guys, one of them is called Matthew Gallagher, created a GLP1 telehealth company. So GLP1 is a drug to help people lose weight, right? Become healthier, lose weight. And New York Times did an article on them where their only about a year old, generated about $400 million in 2025, and they're projected to reach $1.8 billion in 2026. So the first company, two people, no VC investment, using AI.
[00:29:49] Speaker A: Allegedly.
[00:29:49] Speaker B: Allegedly billion dollar company. And then people on Twitter dug into the story and realized that first of all, they were doing something extremely illegal, which is using hundreds of fake Facebook profiles to run fake Facebook profiles of doctors buying ads on these profiles or for these profiles to show. People say, hey, like this doctor endorses this drug. You should buy it from us. Extremely, extremely legal. And they've actually got warning letters from FDA telling them that, hey, you should stop doing this.
[00:30:27] Speaker A: Yeah. Fake endorsements.
[00:30:29] Speaker B: Fake endorsements on social media.
[00:30:31] Speaker A: I think this. I'm. I'm. I'm not a fan of this. I think it's. I think it's terrible.
I think, I think they should be prosecuted for this, actually.
[00:30:38] Speaker B: So the question is like, can you get away with a bunch of money, right? And they just throw money at the problem. If, if FDA comes back with a, let's say, $100 million fine, right? Allegedly. They've already made 400 million and it's making 1 1.8 billion this year. Why not just pay the fine?
[00:30:56] Speaker A: I think that's exactly what'll happen. I think that like, fraud in sales and marketing is awful. I think it's gotten worse and worse over the years. I think there's like less people that seem concern.
I think it's harder to police. I think like Facebook and Google seem to wash their hands of this. They should be responsible too. There's an argument maybe even Facebook should be fined. Like they allow this stuff to happen. I don't know exactly what I'm looking at. It looks like they make fake profiles, so maybe they're. They're skirting around Facebook security and stuff. So I don't want to like totally drag them under the bus or whatever, but like, Facebook should be responsible as well. And when I started advertising in the old days, when we bought TV commercials and magazine ads, they were super duper strict about like, what you could say, what you could do the bar was extremely high. They made it very, very difficult actually. And to some degree there was a monopoly. Right. There was only a couple of magazines, a couple of TV networks are at scale and they took it really seriously. And over the years as like digital marketing has become pervasive and there's just so many publications and it's very hard to police. Everyone's kind of turned a blind eye to this. So that would be my kind of like long term perspective. I think it's really, really bad that fake doctors like didn't do it once. They made like hundreds of these things,
[00:32:03] Speaker B: they made 800 plus basically a thousand fake doctors. Right. That they also used AI to deep fake before and after photos of supposedly their customers.
[00:32:14] Speaker A: And it's a doctor too. It's not like a funny prank or an April Fool's Day joke or something. Like I'm all for, I'm all for that, but this is like not good.
[00:32:22] Speaker B: Yeah, they're basically.
[00:32:23] Speaker A: Go ahead. I'm sorry.
[00:32:24] Speaker B: Well, they're basically targeting, let's say my parents generation that uses Facebook still. Right. Especially like showing people that may not be realized that this is possible.
[00:32:34] Speaker A: Dude, you get me all crazy. I'm getting all passionate about that. I think it's bad. So, so okay. The other thing that bothers me is like all this talk about the billion dollar one person company. Like I feel like we already had that many years ago in consumer and it's called Kylie Jenner's Instagram where she sold her own makeup. What I read was like it was like a couple of contractors. It is actually breathtaking the business that she built with that. So it's all automated online, same idea. It's all done through APIs. There's a couple contracts to help them out and they do fulfillment outsourced. So this, we've already seen this. Right? And you have like something like Taylor Swift like making billions of dollars like through streaming media and concerts. And like it's not a big team. She doesn't employ 30,000 people like a big corporation. Right. So we've already seen this too. I feel like nerds just don't pay attention to anything that's not nerdy Sass
[00:33:22] Speaker B: on LinkedIn, B2B Sass or anything that has AI AI hyphenated.
[00:33:28] Speaker A: Yeah, it wasn't on LinkedIn, so I don't, I don't know if it's true or not.
[00:33:32] Speaker B: Right, that's right. If there's no.
[00:33:35] Speaker A: Oh my God.
[00:33:36] Speaker B: If the founder is not 30 under 30.
[00:33:38] Speaker A: Yeah, exactly, exactly. This thing's ridiculous. Okay, let's, let's go on tech layoffs. Not my favorite topic, but I do think it's important cover it.
[00:33:48] Speaker B: So. Yeah, this seems like a recurring.
Recurring topic, right? Yeah. You want to go through.
[00:33:53] Speaker A: Yeah, recurring for the last, like, I don't know, three years.
[00:33:56] Speaker B: Yeah. I mean, besides the headline, do you want to say anything about this?
I actually have.
[00:34:00] Speaker A: Yeah, yeah, yeah. I, I tweeted something about this because there's a, there's a, there's a larger chart that's been circulating that shows this like, run up in tech employment over last like 10, seven years and then how it's declined in last couple of years, continues to shed jobs overall in the, in the tech industry. And I've seen this trend happen over the course of my career. So you have these booms and busts, and the booms tend to be pretty extreme. Way too many people get into this industry than perhaps it needs and perhaps like, should be in it. Like, I think to be in this, as long as, like, I've done it, you gotta have the real, like, stomach for this stuff. You gotta really love it changes all the time. You have to like, relearn everything. And I love it. I think it's fantastic.
[00:34:48] Speaker B: Right.
[00:34:48] Speaker A: But I think a lot of people get into it. They have a great run. It starts to change, it starts to get hard and like, they're maybe not the right fit. And so that's okay. I think that's how this works. I'm not cheering anybody's like, job loss or anything, but I do think it's part of the natural process for this thing.
And I think there's too many people that are in, in the, in this space. And I think in particular areas, like, tech recruiting is an obvious area that balloons up. So as the companies get big, they hire lots of recruiters to hire people when they're shedding jobs. And it's a weaker job market. You just don't need as many people in that space. So it's a role that like, there's just too many people in, or so
[00:35:23] Speaker B: you think it's, it's just the natural cycle of the industry.
You go through the Zerf cycle and then.
[00:35:31] Speaker A: Natural cycle makes it sound so nice. I think it's the results. It's because these are big layoffs. Right.
[00:35:37] Speaker B: And there's a. It's all. And there's people part of the natural process.
[00:35:41] Speaker A: It's just part of the natural cycle. Right. It's like, that's like these economists that like, they talk about it in this very abstract way and like, dude, these are like real people losing their job to have like families and mortgages. So like I, yes, I want to be clear that like I'm empathetic to it. That's why I would call it the result of boom and bust. I think technology is just susceptible to this over for, for whatever reason. I don't even know if it's logical. Like I think people get excited about new technology, they get overly excited about it. Too much investment comes into the space because everybody wants a piece of it. So then that, that, that creates this dynamic where all these people want to get into it to make money in their career and build a career in it. Right? So it's a self reinforcing thing that just starts to accelerate and then when it pops or it breaks whatever term you want to use, it's painful and it, and it doesn't come back in the same way that people expect. You know, a couple of years ago when I was still in San Francisco after the, let's say the last bust of like 2022, you could just see all these people like expect everything to go back to 2019 and like, we're not there yet.
[00:36:38] Speaker B: Yeah, we're not there.
[00:36:39] Speaker A: It takes, it takes time. You know, you're at my event where we had some venture capitalists say like, hey, this is a difficult time for startups. And he put time frame on it, which I actually agree with. He said like it's going to be like this for five years, five more years going forward from today. I think his time frame is actually right. It's just an uncomfortable time frame to say like, yeah, it's going to continue
[00:36:57] Speaker B: to be, to tell people, right, to buckle and for five years. But it's also, it's kind of funny. So I remember what, exactly what he said, he said somebody asked him, is now a good time to be a founder? And he said, no, buckle for five years.
[00:37:13] Speaker A: Which I don't agree with that. But he was like, no. I was like, oh man, what a
[00:37:19] Speaker B: virtual capitalist tell you that hey, this might not be the best time to founder. That's exactly what you should be, a founder.
[00:37:24] Speaker A: I do, I agree with that. I think that, I think that the cycles tend to be decades. So we had like 29, like after 2008, the GFC, you had this real huge run up in high tech and, and I saw that boom through San Francisco and it was, it was awesome. Right. But prior to that, I remember the.com and I remember the bus and you could Argue that that like, kind of tech session went from about 2001 all the way up to maybe 2008. Like, there was some, like, things that happened and some interesting companies got formed. Like, even Twitter actually started in 2004. Right. So there were some cool companies that got, that got started, but it wasn't like an exciting space like it was in 2015.
[00:38:01] Speaker B: Yeah, I mean, a lot of big companies came out of the dot com boom, you know, like, I mean, Google was probably one of the biggest one that came out.
[00:38:10] Speaker A: Yeah, yeah, yeah.
[00:38:10] Speaker B: I mean they've been around for a long time, but that's when they hit the, the strategy.
[00:38:15] Speaker A: They went public in like 2002 or something. So I haven't done the analysis, but like, my guess is that the, I want to say the best companies are formed during the downturns. That's probably not true, but there's enough, there's enough data from the downturn that it's, it's hard to argue that there's a bets time that like there's great companies that get formed all the time. Like I looked it up the other day and I was like, 2004 is when Twitter started. I couldn't actually believe it.
[00:38:39] Speaker B: Right.
[00:38:39] Speaker A: Like, there wasn't a lot of excitement in the space. Like 2004. It was like right when I moved to San Francisco, like there was still people talking about like the decline in the population of the city. Parking was easy. You could go wherever you wanted. Like.
[00:38:54] Speaker B: Right.
[00:38:54] Speaker A: It was an exciting time.
[00:38:56] Speaker B: I, I think the real founders, when they know YouTube was founded 2005, I would say.
[00:39:04] Speaker A: Yeah, yeah, I remember that.
[00:39:06] Speaker B: That's, that's so many great companies were founded between 2001 and 2008.
Afterwards.
It was afterwards, it was a lot of social networks. Yeah, right.
[00:39:19] Speaker A: That was all of the, all of the Y Combinator companies that people talk about. So like there's like Airbnb, right? Yes, yes. They're from basically the gfc. So Airbnb stripe. There's a couple of Coinbase, maybe there's a couple of. I can't remember. And there's like, there is an argument to be made that Y Combinator picked up these really high quality companies in that era because venture capitalists were not making investment. So kind of like the conversation that we were, if we were listening to or the talk we're listening to, we're like, hey, now's a bad time to be a founder. So the larger institutional investors pull back at certain times and it creates opportunities for Y Combinator angel investors. Like they're technically altern alternatives to institutional venture capital. You could make an argument, do well with downturns.
[00:40:05] Speaker B: 100% agree with that. So, long story is, if you're looking for something to do, I think the
[00:40:11] Speaker A: summary is for today's show. If you're a founder out there, a focus on cybersecurity. And now's a great time to do it.
[00:40:16] Speaker B: That's right. Cyber security and not fraud. Anti fraud.
[00:40:20] Speaker A: Okay. Okay, we got two more things we got to cover. We got to cover Artemis.
[00:40:23] Speaker B: Okay. This one I'm super excited about because I'm a huge fan of space travel.
[00:40:28] Speaker A: Going the moon, baby.
[00:40:29] Speaker B: We're going. So Artemis was launched.
Launched two or three days ago by NASA. Essentially, this mission was just a quick slingshot around the moon, but it is the furthest that human or mankind has gone since literally the moon landing in the 60s.
[00:40:45] Speaker A: And wait, hold on. The rocket's just going to go up and go around the moon and come back.
[00:40:49] Speaker B: Go around the moon. Slingshot. Literally slingshot across around the moon.
[00:40:52] Speaker A: Just like a Sunday drive.
[00:40:54] Speaker B: Just like a Sunday drive.
[00:40:55] Speaker A: Okay.
[00:40:55] Speaker B: But still the furthest a living human being has gone since literally the moon landing.
Dude, it's so cool. There's also Canadian on board, so extra proud.
[00:41:07] Speaker A: Okay. Props to Canada. Props to cash up there. I like that.
[00:41:10] Speaker B: That's right. Next year with Artemis 3, that's when we hope to back up to the moon.
[00:41:17] Speaker A: Yeah. So I was, I was talking about this with my brother. He told me that it's the largest chemical rocket ever made.
[00:41:22] Speaker B: Yes, it's. It's huge. Compared to, I mean, compare. First of all, compared to the moon landing mission, this rocket is like futuristic.
[00:41:30] Speaker A: Incredible. Incredible. Dude, I watch all those crazy YouTube shows about like space travel and all the wild inventions that would perhaps be an alternative to chemical rockets. We've seen these, like, there's, they're actual designs for like these like nuclear powered rockets.
[00:41:47] Speaker B: Yeah, that's right.
[00:41:48] Speaker A: It's like, it's like literally. It's literally like lining up a bunch of atomic bombs behind you.
Like, like, it sounds cool until you think like, do you want to be the guy in that rocket?
[00:42:01] Speaker B: Like, yeah. Or the first person. First person writing this. Yeah.
[00:42:04] Speaker A: Wow. There's also. And now I'm getting off topic, which is fine. There's also designs for like the, like, there's other kinds of atomic powered rockets that they've, like, they've. I don't know if they've done tests on these things. I'm getting way off topic here.
[00:42:18] Speaker B: No, no, no, it's good, it's good. It's. It's. So like everything that we've built so far is really kind of like the.
Still the old style, right? Like we burn.
[00:42:27] Speaker A: Yeah. I would call it a chemical rocket. I'm chemical, obviously not a rocket scientist.
[00:42:31] Speaker B: So. That's right. We burn some kind of fuel, whether it's fossil fuel or some kind of chemical composition fuel. And then there's nuclear, which is atomic. Like atomic.
[00:42:41] Speaker A: Right.
[00:42:42] Speaker B: And then there's like even more crazy one where it kind of uses. Uses some other force.
[00:42:48] Speaker A: Those are all theoretical, right? They've never built one of those. What's that? I can't remember. I came. What's called. They're not. They're just. They're all just designs right there. They don't even have working models. Is that true?
[00:42:58] Speaker B: So like they have at a small scale, right, like something that sits on your desk, something that's in a controlled lab environment, it works.
[00:43:08] Speaker A: But I've seen there's like solar ideas too. Solar sails. There's actually a bunch of concepts. Yeah, yeah.
[00:43:14] Speaker B: So there's like two problems that you have to solve. One is you have to exit Earth gravity. So that's the biggest problem that takes a lot of force and propulsion to get out of orbit. And then once that, once you get into the orbit, then yeah, you could use solar sail. You could use a whole bunch of other ways to proportion yourself.
[00:43:31] Speaker A: But yeah, if you're going that, if you're going that fast, like 1% speed of light when you get there, you gotta slow down.
[00:43:37] Speaker B: You gotta, you gotta slow down.
[00:43:40] Speaker A: No one thinks about this problem.
[00:43:41] Speaker B: I'm serious.
[00:43:42] Speaker A: I'm totally serious.
[00:43:43] Speaker B: That's right. You have to burn fuel to slow yourself down.
[00:43:45] Speaker A: You need to save half your fuel, you need to save half your atomic bombs. And now they have to go off in front of you.
They're not that crazy.
[00:43:57] Speaker B: Literally what they do is they do the mathematical calculations ahead of time to use whatever the. The celestial body at the destination to slow you down. So, for example, like with this mission, right. Half the journey back is free. Basically what that means is they calculated the trajectory of the slingshot around the moon, uses the moon and Earth gravitational field to basically slingshot you back so that you don't, you know, go off into the deep space. So like you can do all of this calculation ahead time, and it's literally the best thing ever.
[00:44:32] Speaker A: All right, we got one more and then we'll end the show. Are you ready for this?
[00:44:36] Speaker B: Yes. The meme of the Week.
[00:44:38] Speaker A: Retardmaxing. What is your perspective on this, Paul?
[00:44:41] Speaker B: All of the terms. Okay, so for anyone that doesn't know what retard maxing is, it's basically this guy or this guy on YouTube, he does like 30 minute videos just talking about life, work, or not working, anything. Basically his entire philosophy is you could live life however you want. You can get a job, don't get a job, work.
[00:45:04] Speaker A: It's just basic stuff.
[00:45:06] Speaker B: It's basic stuff. Why it's kind of funny or important. Is Marc Andreessen.
[00:45:11] Speaker A: Yeah, he's a fan.
[00:45:12] Speaker B: He's a fan. Basically. He's a fan.
[00:45:14] Speaker A: He's a fan.
[00:45:15] Speaker B: It's so against the tech bro philosophy of you gotta work 16 hour days, pump virus.
[00:45:22] Speaker A: Yeah, yeah, yeah. So I, I mean, like, I don't know, I don't know why Andreessen decided this was interesting. But like, I do think at a higher, like cultural level, single is a big shift. So we went from like hustle culture, airy V. Like I, I unapologetically subscribe to this. I've always been this kind of person. Just like, yes, worked super hard. Like, worked 24, 36, 48 hour shifts, ride my bike 300 miles. I've just, I remain that person. Just there's this intensity inside of me to just come out and it. And it just comes out. Right? And it's just so, it's just very natural actually for who, who I am. And I think that that's been like the archetype of the San Francisco tech bro. Right. Just like I'm locked in. I'm just going to stay in this room and drink a lot of cola and like make my startup and stuff. So this retard maxing is, is a, is a completely different approach from that.
[00:46:14] Speaker B: So. Okay, I have a better form thought after our discussion yesterday.
All right. Okay, so let's hear it. Margaret Driessen is one of the proponents that says software is no longer a blocker for productivity. He basically says software is now a commodity.
[00:46:34] Speaker A: Okay.
[00:46:34] Speaker B: Actually most like productive or what we used to consider productive work. So design engineering software, because of AI, is now completely commoditized. Right. That's no longer a blocker.
[00:46:46] Speaker A: It's been retard maxed.
[00:46:47] Speaker B: It's basically.
So, basically what he's saying now, and I think we talked about this before jumping on the show, is you need to be retard maxing to kind of feel out or figure out what you should be working on at a constant moment. I think a lot of people using AI has already realized this. It's no longer about doing the thing. It's about figuring out what you should actually be doing.
[00:47:10] Speaker A: So, like, I'm still. I'm still not convinced, like, why that is maxing.
[00:47:15] Speaker B: But it's basically okay. You know how when you need to
[00:47:18] Speaker A: be creative, like, I agree with the premise, but just. I thought the premise of Retard Maximum was like, hey, I'm just going to, like, go with the flow. That's something. Like, there's all go with the flow, or is it quiet quitting or. I think it's more go with the flow, personally, is what I took away from it, but correct.
[00:47:32] Speaker B: I don't think it's the quiet quitting. I think it's the going with the flow now. So, for example, I'm terrible, terrible at this, but I don't think so, though, because, like, say you're being. You need to do something creative, right?
[00:47:43] Speaker A: Okay.
[00:47:43] Speaker B: The worst thing that you can do is just type lots of words or draw lots of things or design many iterations. Right? Like, a lot of the time, creative work comes from just you not tackling. You're not actually physically doing the thing. It just comes to you as an inspiration. So, like, Tej, basically what he's saying is like, you know, you need to be retard. Maxing, smoking a cigar on your porch, talking about life, doing random things, and then suddenly you'll have an inspiration that will actually move you forward. I think that's basically what he's trying to get at. And I. If that's the case, I completely agree. We all need to just do less.
[00:48:21] Speaker A: Keep going. I. I think. I think that is a fantastic way to end the show.
[00:48:27] Speaker B: Chillax. Go for a walk on a Saturday morning.
[00:48:30] Speaker A: Exactly. I'm gonna go get some coffee. I'm gonna go for a walk. Sun's outside. Well, this has been awesome, dude. I'm very excited about our new live stream setup, and we've done 40 episodes, and I can't wait to do another 40 more.
[00:48:41] Speaker B: Let's do it. All right.
Okay, we're good.
