Riley (2:37)

Yeah. It's as though they made a product called, like, Therados 2 colon. Better than the gods, maybe. Was astonishing. Okay, so. But Look, I, I actually, I actually want to talk before we get into delve, before we get into their awful, awful campaign. I'm just going to link the blog post in our chat window here so we can all follow along with it later. There are a couple of bits of news I wanted to share. First is I have this sense that we are in an accelerating golden age of shitty memes breaking containment from the Internet and becoming bricks and mortar locations. In real life, Elon Musk's Diner did this. But now Polymarket, you may remember them as the. The platform that allows Trump insiders to profit off of drone strikes has created a Monitoring the Situation bar in dc. Cool.

Hussein (3:37)

Okay, so what do you do? What do you do there? Well, you monitor the situation, but I. But I suppose the question is, what situation are you monitoring in this place?

Riley (3:45)

Any, any situation you wish to monitor, you can monitor there. And it's a bit like, I don't know, this is a bit of a stretch. Has either of you seen the movie the Sting?

Hussein (3:55)

I haven't actually, no. So you're going to have to tell me.

Riley (3:57)

So. Good. So if for all these of you out there who have seen the Sting, a lot of the action takes place in a old timey 1930s bar, like a gambling den. That is a place where you go to bet on horses and they have a wire that they like, get the horse race odds and results. Now polymarket is basically that. It's like a wire. It's like a wire room for horse racing but for like drone strikes. And they've put it in Washington D.C. and it's like fun and you get free mojitos. And polymarket does this all the time. And honestly, it's like going for a drink inside Elon Musk's Twitter replies. They say, imagine a sports bar, but for situation monitoring. Live feeds of X flight radars, Bloomberg terminals, Polymarket screens. This is from a piece in Wired. Meanwhile, mocked up newspapers with headlines about monitoring the situation are passed around. An enormous illuminated LED globe served as a rotating centerpiece for people to pretend that they're in the war room from Dr. Strangelove. But as gamblers, I have to say

Wendy Liu (4:56)

working there as a bartender sounds just like a total nightmare. And I don't. It's hard to imagine how much money you would have to be paid to be able to work there and deal with the kinds of insufferable people you have. I'll probably ask you to change the screen to show some other thing that they're betting on. And then as soon as they lose money, they're going to be belligerent and start fights. So, yeah, that sounds like a really fun place to work.

Riley (5:21)

Like, hey, can you please, this, this video that's just in your eyeline. Can you please change this to a live feed of something that looks like the Collateral Murder video? I stand to win upwards of $10,000 if this, if this bomb goes off in the next hour. This. They enter. The person who was doing this interviewed some of the people who were there. Tobin Stone, 24. Tobin Stone. He should go work for Greg Stubby, 24, works at a think tank in D.C. he admits to having placed a bet on, quote, whether or not something would happen in Congress, having been certain of the outcome beforehand. Others are quick to reject the notion that speculating on war is in any way wrong. If you're living in a war zone, having accurate information is really important, said Ben Shindle, a 28 year old blogger. He argues polymarket can provide those caught up in conflict with information about their livelihoods, whether they should relocate and how long a war will go on. These are really good things to be able to know and understand. Shindle says.

Wendy Liu (6:09)

I wonder if there's another way to get information about what's going on in the world that I, you know, I wonder if there's, if anyone's ever thought of that before.

Riley (6:18)

Wendy. See, we did try that, but apparently everyone involved was an inveterate lefty who was, you know, bought and paid for by Soros or whatever. And you, you could never get accurate information because everyone was too ideological. It was the failing New York Times. You need the neutral hand of Polymarket. Incidentally, I don't think anyone's noticed this yet, but if you wanted to make a lot of money on Polymarket because, like, these guys would argue, like, look, actually insiders betting on like government activity is good because it creates a stronger signal than a reporter might get from talking to someone, right? Because this is just simply money, right? This is, there's no spin, there's no political agenda. It's just someone wants to make money. They're incentivized as a rational actor to, like, signal what's going to happen by betting on it. The problem is, I don't think anyone thought about this, is that if you wanted to make a lot of money in polymarket, all you would need to do is, is place a bet that looked like an insider trade, right? All you would need to do is create a new wallet and make a high conviction bet on something that, like, only, say, a Trump Administration insider would know. And then try to cover your tracks, but not too well. And then a bunch of people just pile into your bet and then you sell your positions to them. So it obscures much more than it reveals. It is utterly socially useless, even in the way that they say it's useful, basically.

Wendy Liu (7:40)

I think in that sense it feels a lot like the crypto bubble and honestly, like a lot of the bubbles that we've seen in the last few years where it is just a matter of this new technology that's not really that new, but new enough to evade the usual regulations. And as time goes on, it just becomes more and more scammy. People abuse it. People make a lot of money, obscene amounts of money from it, not providing anything. And then the behavior that evolves, it ends up really just like resembling scammy behavior that has been outlawed.

Hussein (8:10)

Right.

Wendy Liu (8:10)

It's like we have laws against insider trading those are supposed to exist. They still do, but they haven't yet caught up to this new area. And it's just so weird just to see everyone act like this is normal and this is fine. And of course there is an administration that is probably not going to do anything about it because maybe it's good for them. It's sort of in line with the ethos of the current administration. So we are just still living in this weird twilight zone and it is lasting too long and I think we're all losing our minds. We're all losing our sense of reality and how things should be.

Riley (8:43)

Oh, I mean, it says though you had a bar themed around. I don't. I don't know. Yeah, it's. You have an insider trading theme bar on Wall street essentially is the same thing as having the Poly Market bar and in D.C. and you know, this again, this is quite serious because when you're talking about like matters of life and death, then you know, the Poly Market guys have shown that they are completely happy to threaten to kill someone will report something that makes them lose a polymarket bet.

Hussein (9:13)

Look, I also just want to say, right, here's my controversial take, but these guys have made gambling really lame and really like lose because, like, this isn't a novel thing. You know, you could sort of like go to a regional party. I don't know what it's like in like San Francisco, but like in the uk, in London, like there was a time where you could sort of make these types of very strange bets with a guy in the pub who's like, sort of sits in the corner and he like takes out his little notebook and will kind of like, sort of like let you kind of, you know, he'll, he'll run his own little gambling ring. Right. And you can sort of like make the bet that you want usually on like, you know, the dog races, but you can also do it for like general elections and stuff as well. And there was at least like a sense of, you know, forming like a type of, you know, camaraderie.

Riley (9:55)

Right.

Hussein (9:55)

There was a very social element to that. Right. And this kind of like the thing about like polymarket and Kalshi and all these places is that like, yeah, like the gambling, it's not good to sort of encourage this level of like degenerate gambling. It certainly isn't good to sort of like say, hey, whatever's left of kind of like reality. Why don't we monetize, why don't we monetize time and space itself and try to manipulate reality through the lens of kind of micro betting? And I'm sure there'll be no long term consequences of that. But above everything else, it also is deeply antisocial. And the thing that I sort of felt reading about the Poly Market Bar, this kind of gimmick that this company was doing to I guess appear to be sort of normal, is that I can't imagine any type of socializing happening there. I want to know what types of conversations were happening with people, whether or whether, and this is probably more likely to be the case whether it was a group of people who went there to buy like a drink or something and just spent the entire night looking at their phone.

Riley (10:56)

Yeah, it's the looking at your phone bar. And you know, again, like the other thing is threatening to hurt people who made your bet not work out. That used to be the province of Italian American social clubs, right. And now it's been distributed onto the Internet. This, this is the Polymarket monitoring the Situation Bar, the Situation Room. Probably the worst, the worst hanging imaginable. But there's another thing that I have been obsessed with news wise that I wanted to talk about before we carry on. And this is like some, I want to call it second order effects of what is happening in Iran. I mean, the Trump administration has said that the situation is largely resolved and they're winding down their military presence. Again, a bunch of people want a huge amount of money on polymarket betting that this was a lie. He said. Trump said on Truth Social the US was quote, getting very close to meeting our objectives as we consider winding down our great military efforts. But like, this is the Third time he's announced that the military operation will be winding down. And it's on the same day that he's asking Congress for more money and is also sending like 2500 marines. That is how, exactly how I wind down before, like a long. After a long day before bed, just ask Congress for money, deploy some Marines, and then, you know, in bed with a little glass of milk and a peak nightcap.

Hussein (12:09)

Yeah, great, great, great stuff. I don't, I don't really know what to even say to that other than I just. Yeah, let's, let's keep the winning happening, I guess.

Riley (12:17)

No, there's these second order effects that are becoming very obvious. So Mike Leskin in The United Airlines CFO said our plans currently assume oil goes to 1,75 a barrel and does not get back down to $100 a barrel until the end of 2027. And for context, United spent $11 billion on fuel last year. Under these prices, it would spend a further 11 billion. And the company has never, ever, ever profited more than 5 billion, ever. Which means under a situation in which the Strait of Hormuz remains closed or even because, like, all those gas production facilities in the Gulf require so long to get taken back online, even if the Strait of Hormuz only remains closed for a little while longer, it doesn't look like it's economical for United Airlines to exist.

Wendy Liu (13:02)

It's interesting because I feel like they will, of course, raise prices. And in some ways I'm like, this is a terrible thing to say. Is that good? Is that maybe good? Should we maybe be trying to deincentivize flights? You know, and it's, of course it's not good for prices to rise in the abstract, but it's, it's almost like, like, what is the opportunity in this to. I don't know. Yeah. Because it does feel like flight prices are way too low every time I look at it. Like, it's. It shouldn't cost, like, especially in Europe, it's like, it shouldn't cost like £19 to go somewhere on a plane. Like that is. That's probably not a good thing.

Hussein (13:37)

Yeah, it should cost £19 to go on a train.

Wendy Liu (13:40)

Exactly.

Hussein (13:41)

This is the insane thing. And I think, like, it's not like a new point, but it's just like it costs more to go from London to Manchester than it does to fly London to, like, Barcelona.

Riley (13:48)

Well, no more, though.

Hussein (13:50)

Now, now it'll. Yeah, now it will cost the same price to go.

Riley (13:54)

Finally.

Wendy Liu (13:56)

It's especially weird here in the US because obviously we don't have great train infrastructure and you know it's so cheap to get on a flight to go from SF to la, but it kind of shouldn't be and it should probably be really cheap to just take high speed rail and like this is the world we want to manifest. And I'm not saying that this is the, this is a way it will happen. Most likely we won't get trains and flights will just be expensive. But and so I help but hope at that, you know this, the oil crisis does at least encourage a certain kind of investment in alternative energy sources and you know, building out trains. I maybe I'm too optimistic but this is just all I want.

Riley (14:31)

What you've forgotten of course is like petro machismo which is anything other than an oil belching gas guzzler is feminine and therefore something that cannot be acceptable to the sort of average chud. But I go back to United's comments, right they purely whether or not it's desirable that the era of cheap FL is ending and I think it's actually arguable to say maybe it is. It is still a secondary effect of like of the insane war being prosecuted on Iran that the era of cheap flights is globally over it seems, unless something changes. But also I don't know how you square United's comments with the administrations unless United has like some knowledge about Iranian military capability that the White House lacks. What we know is that they have a duty to be honest with their shareholders about material factors that could impact the share price and an incentive to make those announcements before others do. Like you know, Deutsche said that the jet fuel price hike poses an existential threat to the whole European airline industry. SAS Finnair, like all the Nordics, have slashed thousands of flights. It is a whole pillar of post 1967, like post 1960s life when air travel caused places like Benidorm to replace Blackpool for British holidaymakers. That appears to be under threat like a given of a prior a thing that you just assumed to be true and will be true forever seemed it could be gone. And you know the United has to tell the truth in ways governments don't. And you know, and the effects of the, of the war in Iran that you're feeling right now like interest rates rising or for homeowners mortgage rates rising, those are the effects of predicted higher prices, not even the higher prices themselves. So right now the liquefied natural gas from the Gulf, the last tankers that have the last ordered shipments in them are going to get to their destination in 10 days. Like, do you understand what that means?

Hussein (16:23)

I think, I think what it means is everything's going to be fine.

Riley (16:28)

Of course. Or more specifically, it's that whatever normalcy was sustainable at a time when higher prices were predicted comes to an end when you can no longer disguise the difficulty of getting the thing with the prices and abstraction of it. Right. Natural gas, energy, energy costs, they go up and up and up. But at some point there's no amount of money you can pay that will cause natural gas to appear in the amount that it's needed. And so there won't be any. Right, because you can't order anymore because the last natural gas that there is is already in the ships and the ships are already going to the places that they're going. And once they drop it off, that's it, there is no more.

Wendy Liu (17:07)

It's done. What if we just created a prediction market that incentivize people to bring us natural gas and had like a bar to monitor that situation? I just, I feel like there's a solution.

Riley (17:18)

I'm going to place a bet that nobody delivers me a tanker load of liquefied natural gas in the next two weeks.

Hussein (17:27)

There you go.

Riley (17:28)

Yeah, perfect.

Hussein (17:29)

I do like the idea of just like human produced natural gas and you sort of make something like the matrix, you know like the matrix with those matrix pods but like. Yeah, you know, they're just sort of like extracting natural gas from the human body to like, I don't know, fucking power. Yeah, you have to use 50 people to power a very small Toyota.

Riley (17:49)

Perfect. I'll tell the Bones crew. But it's not just natural gas that passes through the strait. This hasn't all been covered together yet generally. But all of this is affected helium. Most of it comes from Qatar that's used for chip making. So what are you going to do? You have to keep making the new Blackwell chips.

Hussein (18:05)

The gamers are going to rise up because I feel like the game, I feel like genuinely, you think about 2016 and it's the real gamer moment in the sense of Gamergate achieves a sort of political win in the form of Donald Trump, all the sort of culture war stuff. And now 10 years later, the gamers have got nothing. Right? Their chips are fucked. Even before all this, their chips were fucked. RAM just doesn't really exist anymore. The sort of professional gaming industry has moved away from America and America is not really doing very well in games in general. And now this. I feel the gamer uprising may be the sort of consequence of all this. Right.

Riley (18:43)

Guess what else? If you want to sort of carry on this particular line of thinking, plastic is a crit is a hydrocarbon byproduct. Like the, the feedstock that goes into making plastic is a hydrocarbon byproduct. A lot of it comes from the Gulf. And that's like plastic food containers, for example, the plastic elements of a Doritos bag, whatever. Sing use medical devices, I guess the plastic of a game controller. Like 80% of this stuff comes from the Gulf. And like, imagine if you don't. It's not just that you can't power anything, it's that you also can't make anything because you don't have any plastic. And you can't grow anything because 30% of global fertilizer imports come through the Gulf. And like the UAE is a huge smelting hub for aluminium, which is what all this shit is made out of. So if you want to play games and eat Doritos, all of the things that make that possible are currently held up in this Strait of Hormuz. The modern world is held up in the Strait of Hormuz.

Hussein (19:37)

The gamer uprising is. It's ripe for this moment, right? Like, look, Reza Pahlavi, right, had a dream of becoming like the leader of Iran. That's not going to happen. But what I'm saying to Reza Pavlak Pahlavi is that you have a constituency of people who will kind of like stand behind you. You just have to like, it's for a different cause. Reza Pahlavi, you should sort of represent the gamers in their uprising against the US government.

Riley (20:01)

But look, we're here to talk about Delve. We're here to talk about Delve and their billboards and their strange business. Wendy, can you tell us about some of what you have seen of Delve in the last year?

Wendy Liu (20:13)

Yes. Okay, so Delve has, I will say, a very effective marketing campaign in the sense that they are everywhere. Billboards are eye catching, at least to me. They're interesting, they're strange, they're confusing. They have billboards, these little ads all over all the buses that say things like compliance done before your AI girlfriend breaks up with you. Or compliance done before your. Oh God, shut the fuck up.

Riley (20:37)

That is horrifying.

Wendy Liu (20:38)

Compliance done before your co founder drops. What B2B Sass taught me. LinkedIn post, something like that. And they had one that was really obnoxious. That was like a checklist. And it was like, drop out of Stanford, check. Get into YC check. Raise 50 million from a 16Z check. And then the unchecked one is get compliance done with Delft, something like that. It's like they're playing into the mythology of what it looks like to start a startup in the modern age. And I think they're trying to position themselves as a key part of that ecosystem. They're trying to ride that wave. And it's entirely just selling shovels in a gold rush. Even more abstract than that, selling food to the people who are selling shovels during the gold rush. It's so disconnected from anything actually real or valuable. And one of the things that they were doing with their ads is they would have the logos of other companies that were using them. And I looked into all these companies, and if you look at any of these companies, right, it'll be like AI code review, AI lead generation, you know, like AI marketing automation. It's all just like AI to do other things that are so, again, still disconnected. And it's like, how many levels down do you have to go before you get to something actually valuable and real? And it just feels like it's this whole ecosystem of companies in the cloud. So what Delve actually does, which they kind of say in their billboards, the billboards often have a lot of confusing terminology and acronyms that are deliberately obscure. But then they also mention SOC2, they mention HIPAA. They're positioning themselves as this company that can do compliance fast. And so they make it seem like everyone else is doing compliance in a really slow, inefficient way. But they. Because they're young and scrappy and they have AI, they come along and they help you do it in days, not in months. And of course, this. To anyone who knows anything about compliance, that raises some red flags. How do they do it so fast? What corners are they cutting? And then, of course, as we've seen in the last few days, they're cutting a lot of corners, which isn't really surprising, but also still kind of shocking.

Riley (22:32)

It was so much fun. I had so much fun looking at it. My jaw was glued. I think I mentioned this at the top of the episode. My jaw was glued to the floor. It was astonishing what they claimed. But let's. Let's go through it.

Wendy Liu (22:47)

So.

Riley (22:47)

And by the way, I'm sorry this is a little arcane, but we can't not talk about it. The fact that this company was allowed any airtime at all among serious investors, serious like, you know, purchasers of services or even people selling fucking billboard space, it is just a horrific indictment of the venture capital system as A way for allocating resources.

Wendy Liu (23:07)

Yes.

Riley (23:08)

So it was founded in 2023 by two MIT AI researchers, Karun Kaushik, who was. I need you to note this if you have a bingo card at home. Forbes 30 under 30. He was a Forbes 30 under 30. Like joining fucking Sam Bankman, Fried joining Elizabeth Holmes and how many other people have we talked about who are Forbes 30 under 30? Who have like. And the fucking. The lady who sold the student financing app to JP Morgan that turned out to just have. No, Troy. Yes, she was Forbes 30 under 30. It has to be looked into. But anyway, by these two MIT researchers, Karan and Celine Kosalar. They say Delve uses AI agents to help companies achieve and maintain compliance without the busy work. So Wendy referred to a few like, compliance frameworks. Like, it's all a little bit arcane, but these are basically things where you have to show that you're doing it and then an auditor will say, yes, this person's compliant with this framework. So like, if you're a government, you can contract with them. If you're a healthcare. If you do use healthcare data, you know that you can like work with them because they are as secure as you are. It's basically just ways of credentializing what it is you're doing so people can work with you easily without having to investigate if you've like turned on your multi factor authentication for your passwords. I say the platform now serves 500 plus companies across industries, eliminating hundreds of hours of manual processes while helping them get certified faster. Backed by Insight Partners, Y Combinator, Funders Club, General Catalyst and leading compliance executives. So they raised 32 million in 2025 against a 300 million valuation. That above text, by the way, was from Insight Partners, which hilariously still has the copy up and accessible. Based on what we talk about in the rest of this episode, I think it's soon to be accessed only through the Internet Archive.

Wendy Liu (24:47)

Yeah, actually Riley, I checked last night and it looks like they've scrubbed. They've scrubbed it from their website.

Riley (24:53)

Whoopsie.

Wendy Liu (24:55)

You predicted it. Your prediction was too good.

Riley (24:58)

Venture capitalists only do this when they're like, well, we've achieved our mission with this company, I suppose we don't need to waste. We don't need to waste valuable RAM on it anymore now that no chips are being made. So Insight Managing Director Praveen Akiraju said since compliance touches every part of how a business runs, from scaling operations to closing deals, to building customer trust, modernizing this function will modernize the entire generation Em dash. That's what makes Delve's approach so important. I can't not see it. Now we can go through what it's supposed to do and what they actually did. And again, like, Wendy, I know you know about this as well, but again, like, when we talk about compliance, it sounds very dull. And that's partly because, like, it is. And they're trying to sex it up. And when someone's trying to sex up a dull process, you should always assume that they're scamming you. Always.

Wendy Liu (25:46)

Yeah. I think a good analogy for this is, you know, like a health inspector in restaurants where it's like, it actually is really important to have a person who could walk in at any given moment, look around the restaurant, check if there are mice, or make sure people doing things properly. Like, you kind of need an actual person there. Imagine if someone were to be like, we're going to do health inspectors, but with AI. And so what they do is like, they have. Have you send, I don't know, like a photo or something of what's going on, and you upload, like, a photo of the food, a photo of the menu, a photo of the people, and then some AI is just like, check. You know, it's like that defeats the whole purpose. The point is to have the possibility of something going wrong and being spotted. And it's like these guys, these are just like children, right, who dropped out of mit. They see a shortcut they can take. They don't understand the purpose of this system.

Riley (26:32)

Well, I think it's more than that, though, with what they claim to be able to do. I mean, we'll go into it, it. But, like, most of what this would involve, right? It's, let's say the health inspector analogy, as you say, is if you're saying, okay, well, we, we, we're a. We're some, like, technology company, provide, like, and we need, we want to work with a hospital. We need to get HIPAA compliant before that hospital will share medical information with us. Fine. So we're going to need to, like, we have all these checklists of things we need to do, and then we're going to do them. And then we hire. And then there's an auditor that comes in and says, yes, you've done that. Yes, all the evidence has been given to me. So I can see this screenshot that says of a little tick box in AWS that just says, have you turned on MFA? Yes. Okay, tick, you've turned on MFA.

Hussein (27:13)

Great.

Riley (27:13)

And in reality, most, it'll be like okay, MFA is on for like 95% of people. We now have to go chase those other 5% of people to turn it on. And then we're HIPAA compliant and then we can, you know, get this done. And they're saying, okay, well what to go back to your health inspector metaphor, what if instead of taking a bunch of pictures of the kitchen and then saying, okay, well, there are no rats here. We've. Our AI has seen no rats. They will. Will have an AI agent walk in, start taking pictures, and then you don't even have to do that. So, yeah, it'd be like, oh, you need to comply with gdpr. Great. We're going to hit a little button and then an AI agent, which is barely a real thing, is going to get to work automatically scanning everything about everything you're doing. It's going to create all of these wonderful, you know, like, artifacts that's like, here's a list of all your data sources and so on and so on. You need to comply with SOC 2. Great. We'll auto test your security controls, take screenshots. We'll auto fix, like, patching issues and stuff this. Fuck. That's fucking crazy that they said they do that. But like, like you said earlier, Wendy, right? These are people getting into a complex industry, deciding if it can be done simply because they have contempt for it and don't understand it. They're MIT dropouts in the Forbes 30 under 30 list, starting a company that's being discussed in this podcast. What could possibly go wrong?

Wendy Liu (28:26)

It feels a little bit like TEMU for compliance. That's what they're doing. You know, it's like you just, you, you look at the screenshots of what they provide and it's just so shitty. Like, it's so obviously shitty and cobbled together. The report we were looking at that leaked, all the things are going wrong. We see all these Google spreadsheets, Google Docs, typos everywhere. It's just so obviously terrible quality. And it is kind of, I think what's impressive here and what's really sad and like you said, Riley, an indictment of the VC ecosystem at the moment is that this company did manage to raise so much money. And in a way it's like it was only 32 million compared to all these other AI companies, not even that much money. That is also an indictment of the moment we're in now where this sort of TEMU for compliance fly by night company can raise 32 million. And that doesn't even feel like a lot of Money, because it is. And we should think of it that way. If we calibrate it to earlier moments, Even just like 10 years ago, I think 32 million would have seemed like a huge amount of money for a company that, that was founded a few years ago and barely has anything to

Riley (29:28)

show that doesn't even have any underlying technology, which we're also going to get into. So basically what Delve says, right, is say, okay, look, you have to get SOC2 compliant. Cool. We've got an agent that's going to do this. You don't need to re engineer your backend to support this being done with AI, because the whole point of AI agents is that it's a way to just wish away the problems with ever having to do any engineering, because you can just be like, okay, well, it's a robot that works in a physical setup that's meant to accommodate humans. So again, by the way, this is. This claim is insane. The idea that, like, I'm going to handle all my health information and the thing that's going to see if it's being protected properly is like the QWOP robot, essentially, that. That is able to, like, choose the right thing in a highly controlled, structured experimental environment 14% of the time. When, when Delve made their funding pitch, that was the widely known benchmark, and Insight Partners was still like, fuck it, 32 million. Why not not take the money? And this is an insane claim. So they'd even claim like, okay, well, we have a framework of security controls. We're going to send an AI agent through your environment, not your test environment, your production environment. It would find servers, maybe servers that hadn't been patched or had passwords rotated. Then it would just patch them, which is also an insane claim. And anybody listening to this with any kind of technical background probably just barked like a dog in public. And I'm sorry, this is like, what I'm driving at here is like, if you were a venture capitalist who was even minimally technically mine, did, you would have seen that that was lies. You would have seen it was lies. But. But again, like, you're not. Because last time Marc Andreessen touched a keyboard was when he was making Netflix, or when he's like, you know, posting his epic essays or what have you. These are no longer people who understand the nuts and bolts of what they're investing in, I think.

Wendy Liu (31:10)

I mean, there's definitely that, but there's also the sense in which, I think even if the investors know that it's a lie, I think the market right now, reward a certain level of audacity and fake it till you make it. And I mean, it's always been the case, but it feels like right now with anything that involves the concept of AI, there's so much more of an appetite for a company that doesn't really know what it's doing, that doesn't really have the right technology. But as long as the founders are really ambitious people, they're going to raise money. As long as they can pitch something, they're going to raise money. And the idea is that, well, at some point someone will figure out, the tech will just get better and then some sort of use case will emerge. And that is a really, really weird behavior. Like, I think that is a thing I want to emphasize. Like, that's not how. It's kind of not how it should be. And it's definitely not how it should be if there's this much money at stake and this much of a gap between the promises and the reality. And I think, I don't know, it's just. It is such a weird moment that we're in where a company like Delve can have billboards all over the city, raise all this money. All of its companies are basically other AI companies that don't really do anything themselves or have, you know, grandiose claims. And there's a whole ecosystem of which Delve is just like one tiny part that rests on this. This gap between what AI is capable of doing in the moment and what it's actually doing. Because, I mean, I'm sure, as. I'm sure we'll go into it, but if we look at what Delve actually does and how little AI does, I

Riley (32:34)

was gonna say, let's do this next.

Wendy Liu (32:35)

Yes. And that's fucking crazy. It's okay. Yeah, please, Riley, please go. Go through it.

Riley (32:40)

Oh, my God. Okay, so. So the reason I've sort of gone through, like, what ordinarily it would look like to try to get compliant with something is like, okay, you first have to, like, go manually, do things, talk to people, look at stuff. Right. It's boring, it's repetitive, but that's how you find things. Like, oh, hey, we've accidentally left all of the pictures of people's like, you know, broken legs or whatever on a publicly facing, like, Internet of open server. Okay, well, we're going to have to fix that. Obviously what they say is, okay, we're going to just do this automatically with an agent. What they actually did, and this is so, so, so fun, is they would, they would Just auto generate all of the ev. They would auto generate fake evidence. They'd be like, oh yeah, here's your board meeting minutes that you didn't have. Because I don't know, let's say it's tf like this trash feature, right? It would generate our board meeting when we decided our security policy. We don't have a board, we've never written that kind of policy. Background checks for employees that we didn't perform. It would show us having completed completed full vulnerability scans 20 minutes after we signed up. And there's just a big box that says you are now certified. And then 90 days later you get an a quote independent US based auditor report through where there's leaked spreadsheet of 493 companies, 492 of them. The auditor reports were just find replace identical documents that even had like the same spelling mistakes in the same places.

Wendy Liu (34:03)

I really liked the phrase I use which is this is an incredibly high quality report. And it's like every person gets that email saying that it's an incredibly high quality report. That kills me.

Riley (34:13)

Yeah. Oh this is so good. And so like there was never an AI or the AI that there might have been would have been to like automatically generate fake reports that an auditor based out of a post office in India would then send back to you saying congratulations, you're compliant. And what I find really ironic is you know, you note this a few times that like their client base is other fake companies largely pretending to do something stuff. You know, one of their clients was Cluly.

Hussein (34:39)

Oh my God.

Riley (34:41)

Cluly is not all the, all that cheating that you're doing through Cluly, none of it is protected by anything.

Wendy Liu (34:48)

I mean is it basically open? Are we so like, isn't that what we expect from them? Isn't this the Cluly way? You know, just to. I don't know. I mean the other companies got whisper what a just bland, lovable. They're, they're all first of all terrible names, terrible products. Just everything about them is terrible. I think the other thing that we should, we, we should at least mention is if you look at the fake addresses because all these companies, it's like they have, they've registered addresses in the US sometimes in the UK they had. There's one in Canada, it looks like there's one in Canada that was like a vape store which was amazing. And then there's a Wyoming address which was the same registered mail address that a lot of scammers use. And, and so the people in that town are like, oh my God, it's so frustrating that, that all these scammers are signed up. It makes our town look bad, but there's nothing real about it. It's all virtual. It's all just this sort of shell game, these shell corporations. And I think that's what's kind of amazing because that on the one hand feels very yc. It feels very much like what a lot of YC companies try to do. Just fake it till you make it. And it completely violates the spirit of what this company's supposed to do. Because compliance is supposed to be about more than just faking until you make. It's supposed to be about int. About ensuring that things are done properly. I think that's what's so funny. I think that's what's so interesting about this specific case because they're operating in an industry where the whole idea of it is about ensuring that things are done well. And they're deliberately trying to do things as not well as possible for the purpose of making a lot of money, building some sort of empire. I mean, they talk about expanding into other areas like cybersecurity. And it's like, guys, please don't expand like you have nothing to offer beyond this sort of like temu ass model of how to do anything. But that is their whole business model. And I think that is probably how they pitched it. Like we, you know, it's like they, they probably pitch it and I think investors probably understood that these guys don't really have anything real, but that eventually they will. And it just feels like a giant wrecking ball going through society and you know, just knocking down all these things that have been built and refashioning that or like building out of the rubble this sort of awful, you know, inferior version with the purpose of just distributing profit among a smaller group of people who went to mit.

Riley (37:03)

Yeah, I, I mean you, you want to so combine a couple of, a couple of things together, right? We're in the era of the ends of things, right? The era of cheap commercial air travel. It's over. The era of I just sort of assume. And again, when I say trust, I don't mean trust to be good people, trust to be non exploitative, but just trust to be regulated by things that will force them, whether or not they want to, to like protect my medical information. Right? All of this is being undermined by the same stuff because, because those things were boring, those things weren't exciting. And the whole idea that these quite like again I say respectable, advisedly I don't mean respectable to mean great. I mean respectable to mean they have authority, these authorities, whether that's Trump and the administration saying, oh, yeah, well, we're gonna be gone in three weeks. Everything's gonna be fine again, you're all gonna be able to take short haul flights. It's gonna be cool. Don't worry. Don't even worry about the TSA people or whether that's like, like these, you know, tech lord investors saying, oh, yeah, compliance is, it's, it's securing. Yeah. Your medical information, whatever. That's like we're doing this a new way now. And it's just. It's like a body that is starving and consuming its own muscle. Right. It is the stuff that made the thing kind of reproduce itself. You know, again, not saying the reproduction was good, but the stuff that made it actually function is getting either beaten by the same kinds of motives and it's happening in the same way, which is like people in power don't know or don't care what it is that they're doing.

Wendy Liu (38:35)

Yeah. I think what we're seeing, and I think this really does connect everything we've been talking about with polymarket Delve and United Airlines. I mean, it really does feel like there are all these things that are kind of. They're sort of like replicas of things that already exist, but shittier and less regulated and less good for the world. I mean, I think we're seeing this in every sector. I mean, the sheen ification of everything. Right. It's like there's so much money to be made in producing products that are much worse than what already exists, but, you know, more convenient. I think you can kind of see AI in that vein as well. And it is just really disheartening to look at the world that we've built that all these other people around us have built and see what is happening to it. And it's sort of cannibalizing itself because we have just too much capital swirling around the system, looking to make a quick buck. And it's. I mean, it's. It's like. I don't want to be too pessimistic, because, I mean. Well, this is kind of a pessimistic podcast, but at the same time, it's like, it's. It's like we did all this for what, you know, it's like people. People pour their lives into building things and, you know, it's like we have a whole economy that's structured around work and people building things at work. And it's like what? For what? Everything is just getting worse, you know. And the only people who are benefiting from it are the. The sort of. The people who get to party on Epstein's island, right? It's like. It's. It's like we're doing all this for what, what. What is the point? This. And it is really enough to kind

Riley (39:58)

of just make you despair, you know, when this is. Again, sometimes I say stuff that's laser targeted to Hussein. But when I. When I look at the economy behaving this way, what you're. What you're talking about, right? The. This just replace this shinification teamification of everything. And I, I. You look at this especially at the. The way that the AI is being driven into things. Whether that's movie making, whether that's writing, whether or even whether that's like non creative stuff, non obvious stuff like Sock 2 compliance, all of it. I'm put in mind of Tetsuo at the end of Akira.

Hussein (40:34)

Yeah, that's right.

Riley (40:35)

You got me.

Hussein (40:35)

You got me. Very good. Very good.

Riley (40:39)

At the end of Akira. If you've seen Akira, this is the Tetsuo is growing in every direction and he is crushing. He's literally crushing the people inside of his like rapidly expanding body. And it's quite grote supposed it's quite grotesque monstrous mutation, but where he becomes this gigantic all consuming baby that's made of just whatever's around him that just is completely unstable under its own weight and melts in a way that destroys everything around it. Like we're living what delve and clulee and all of this polymark, all this fake shit represents is the Tetsuofication of everything that these things touch.

Hussein (41:19)

Yeah, except big. Big Tetsuo is kind of cool and interest and these guys are not.

Riley (41:25)

At least Big Tetsuo was cool to look at.

Hussein (41:28)

And also it's very much the case of just like. Well, at least Big Tetsuo is doing something in service to the story and stuff. But these guys are sort of like. They're kind of growing, but they're not. I think I was going to make this point I can kind of completely forgot, but I think like Wendy's point just about like. I mean generally it's just like how they emerge out of an existing fiction and their job is to sort of like continue and continue to spin this fiction. Like there is an extent to which everyone sort of, you know, admittedly will sort of know what's going on, but no one can really say. No one's really allowed to say that, hey, this stuff doesn't really work or it's fake, or it doesn't really sort of serve a purpose. But I think the biggest thing is it's objective. It does not sort of solve. They'd sort of stated objective. But I think one of the sort of big issues at this point is like, in a lot of ways these companies are allowed to exist. And I guess this is the interesting. This is where it sort of diverts away from your therapy Theranos and stuff, because it's like, well, Theranos kind of purported to do something that it didn't do. And the sort of scandal was how it obfuscated that. But it was always sort of held. Ultimately this obfuscation would always kind of come back to bite them in the ass because it was like, well, you promised this thing and it didn't work and you sort of did a lot of kind of crime in order to sort of pretend that that wasn't the case. But I think with these guys, what's sort of worse about it is because they already exist within a system in which everyone is producing fake stuff anyway. No one can really sort of say to them, hey, what you're doing isn't a real thing. You're like creating. You're sort of creating a fiction because they're not, they're contributing to a larger fiction that everyone sort of has to sort of believe in because the moment that they don't or the moment that the tide sort of turns, everything collapses.

Riley (43:06)

I'll take it and turn it a little bit, which is that they are falling apart because they couldn't do the fiction well enough, right? Okay, fine. Yeah. But to, to your point, right, if you're at Insight partners and these GU walk in, right, they say, we're going to do AI agents for SOC 2 compliance. We're, we have all these witty billboards and we're going to put like, I don't know, Deloitte or whatever out of a job. Then they're like, okay, yeah, great, we'll, we'll fund that. If you're the person who says, hold on, let me look at how you do that, then what you're doing is you're going to send them to the next guy who is going to invest and they're going to allocate the capital. So you're incentivized to believe the fiction. It just. The problem is the fiction has to be good enough that maybe someone could believe it. And it's like these guys, it's A there. And there are a lot of other examples like it. Like, I always think about wework as you guys were 5% too lazy and 5% too weird to get away with your gigantic scam. These guys, it's the same thing. You are actually too lazy to do a good enough plausible fake company to get away with this cam for a little while longer. You only got like a year out of it. This amateur hour.

Wendy Liu (44:18)

They could have spent maybe like 500k less on billboards and 500k more on just like a slightly more reputable registered address in Wyoming, you know, and it would probably would have been better, like just they could have allocated their money just like a little bit better. And I think, I mean, what's kind of funny here is that once you see the scam that this company is, it really makes you wonder, all the other companies that rely on them, that have used them to a degree, are they scams too? And is this just the tip of the iceberg? And I do wonder if any of that's going to come out. It's also funny because RSA is this week. It's like a big security conference in San Francisco and Delve was supposed to be there. It looks like maybe they're not going to be there. It's sort of unclear, but I feel like this is probably going to be a talk at the conference. And I wonder what else is sort of under. Under that. What comprises the body of this giant iceberg that is maybe about to melt or something.

Riley (45:10)

Like we talked about before, that the other thing AI represents is the dream of executives, which is to do, to produce things. Even though something like compliance is so abstract for its role in production. Right. It still has something, which is that it enables people to, like, work together easily. It's super, super abstract and like, very, very far away from, like, any kind of value. But, like, it is there. It does do something. So I want to go back to Insight writing about Delve, which, by the way, you can't find anymore because they deleted all of it. This was actually posted on the LinkedIn as, as you know, I'm very committed to being a hater and I will go on someone's LinkedIn and find embarrassing things they've done. In this case, Insight Partners Managing director Devin Paris Rec took Karen Kaushik from Delve for a conversation with Satya Nadella on the importance of building and deploying AI that businesses can trust and people can rely on. And there's a big picture of Satya Nadella, CEO of Microsoft, shaking hands with this kid being like, yes, you are the future of trust in organization. And he knew. God, it's like. And fucking like, Parak knew that. What? That he. That he didn't know what was going on under the hood. He doesn't want to know because that means, like, the opportunity to make money is going to walk out the door. Door. And like, fucking Karen Kaushik, he knew that. He was promising things that, based on current technology, cannot happen. Right? There is no way an AI agent. Again, this sounds a bit arcane. There's no way that an AI agent can wander through a bunch of, like, servers patching stuff without like, saying that's what's happening. That would be insane if that happened. It would be. That would be genuinely crazy. That's not. And the idea that it's like, oh, yeah, it doesn't matter what you use. Our AI agents are so perfect. No, we talked about AI agents before. You turn them on and they start ordering eggs for you to a house that you don't live in. This is literally what they do. But seeing like, okay, well, but I know that maybe sometime in the future someone's going to figure it out and then all have been here first and I'm going to get rich. And everyone has an incentive to believe one another, right? Everyone has an incentive to believe one another. And no one has an incentive to ask any questions. Everybody's just like, if we just smile, look at the camera and shake hands, hopefully we will be rescued by something else. There it is.

Wendy Liu (47:24)

Yeah. Yeah. There's this whole ecosystem where all the incentives are to just keep building, keep pretending. There's really no way to verify. And even if there was, no one really wants to because they all want to keep this thing going. And I think it's. In some ways, it's like we can look at it from a distance and kind of make fun of it and laugh at it, but I think we can't forget that this is just like. Like, this is like our world, you know, like, these are the people who are taking the real money that has been generated by the people who have been working, you know, on this planet, creating things. And this money is being allocated towards these people who are not doing anything that is good for the world. And I, like, can. It's such an elementary point, but I cannot stress this enough. And I think that is a thing we should be mad about. It's kind of crazy making to just. Especially for me, living in San Francisco, seeing all these stupid billboards all the time and just remembering the that all of the money here is being funneled towards these stupid, pointless, awful things that are just making the world worse. If they're doing that, if they're doing anything, they're making the world worse. They're all, you know, they're all about how can we, how can we do customer service without increasing headcount? Which just means like shitty AI generated customer service and firing people in the process. And it's like they're not even pretending that they're trying to make the world better anymore. I think that is the weird, the other weird thing we should note. There is kind of a vibe shift. Shift since the kind of 2000 and tens era of techno optimism where it feels like people back then were at least saying they wanted to make the world better. And now I don't know if anyone is even trying to say that. They're not even talking about anything as grand and abstract as that. It's all just like, here's how we use AI to streamline your business processes without increasing headcount. It's always about increasing headcount. It's always about profit margins and revenue and costs. And there's no pretense of doing anything bigger. And I wonder how far this goes, how far can this keep going before people stop accepting the tech sector as the kind of the keystone of the economy? Because it felt like maybe the tech industry felt like it had to legitimate itself for a while and now it's like it's dropped any attempt to legitimate itself. And so how long can this go? How long can the hype and the bubble and the scam go on before people get fed up and like, why do we need all of our money to go into this stupid, stupid scan filled industry?

Riley (49:40)

I mean, you know, you say, oh, it's, we're all about, you know, doing things that increasing headcount, it's expensive to pay people. It's even, it's just expensive to do anything. And you realize, well, we don't have to do anything. All the only suckers do stuff we don't want to do. We just want to own.

Wendy Liu (49:56)

But it's like it's, it's so strange because entrepreneurship is supposed to be the thing that legitimates capitalism, right? It's supposed, we have this idea that, that you put money into. Especially, you know, with the whole YC myth. You put money into these smart MIT dropouts and they'll create useful things that will make the world a better place. That's the whole myth. And it's like we're at the stage where it is sort of like the sort of vulture stage or the almost like auto cannibalist stage where they're not even. They're undermining themselves, they're undermining the whole mythology behind everything they do. And it's, you know, it's just so strange. It feels so short sighted and I just don't know how long this can go on. But I don't know, maybe it can go on for a long time. I don't know. That's. That's kind of a scary thing to consider, I suppose.

Riley (50:40)

It's like systems like this can go on until they become decrepit, right? How long could the likes banking sector, you know, of the 1980s and 90s go on until it became decrepit? Well, it went on until about 2008 and then it fundamentally changed. You know, it's. But how long can this go on until it becomes decrepit? Well, certainly threw its lot in with people who can protect it politically, but God fucking knows how long it can go on before it just is pushed over and replaced by something else. But you know, in every case. But I think the cycles are just getting faster where the new productive paradigm comes along and then it is at some point in that paradigm it just gets more and more efficient at creating growth with like the fully fictitious companies now that are being created with AI or the rise of gambling or whatever as part of all of society. Society. It's getting so efficient it can produce growth with no inputs. Right? It's, it's like a, it's like a black hole of incredulity where not no piece, no work can enter or leave. It's just, it's just numbers and stories and the suckers who don't live in Elysium can go on and do the stuff, whether that's taking a screenshot of an MFA password or flipping a hamburger or teaching a class or doing a surgery. It's like, no, that's. All these things that make life happen are for rubes and suckers and I mean Sun Ti. You can almost see it as like courtly rituals, right? Where instead of creating justifications for their own power, what guys like Andreessen and Musk and the companies that they choose to fund do is they are merely engaged in displays to one another of power. There's no more legitimation. It's just, it's just rituals that reinforce their power among one another of this existing sort of, you know, rulership. Legitimation is gone. The idea that you have to do something is, I don't know, it's for the birds. Because you just tell a story now. You just tell a fun story that says you too. You executive at some company that does something that makes movies say, because we talk a lot about AI in film. You can have the efficiency that we do. You can have the efficiency of growth without production because you can just work, wish it and it will happen. You startup, founder or company that sort of wants to handle medical information. You can just wish that you were compliant with all these very abstract frameworks and it'll happen, don't worry. It's just people in the club inventing numbers that they have. That's it. I want to say one thing, read one thing about Insight writing about Delve, which is so fucking arrogant. This has now also been scrubbed from their website as far as I'm aware. Maybe it's a backup, I don't know. Customers were as obsessed with Delve's pitch as our team at Insight was. Startups posted their SOC2 certification announcements on LinkedIn, publicly thanking Delve for making the process painless. Oh no, no, that's gotta hurt because you just. They've just lost a lot of business. Founders went out of their way to recommend Delve within founder networks, incubator communities and so on. These companies saw Delve as abstracting away the remediation and monitoring work that frustrated them from for so long, thus turning compliance into a low cost, low effort process. Again, it's frustrating because it's hard. That doesn't mean it's nothing, it means it's hard and maybe it's worth something. But if you are these people, you are not able to comprehend that if something is hard that might be good, if something is difficult that might mean. Doesn't mean you have to outsmart it. You might just have to have some bit, some domain knowledge that you don't have.

Wendy Liu (54:14)

I think this is just the sort of extreme end of the kind of YC model where the idea is like anytime you let at an industry and you see something that seems inefficient, this is your chance to disrupt it and to create a platform that takes all the money in that industry and you have this kind of like modern YC brain approach and like that's. It kind of works in some ways. Although of course we know that there are all these terrible side effects that are produced often in terms of labor. You know, it just. Conditions get worse for workers, quality gets worse for consumers. But this is kind of the logical endpoint of everything that People like Paul Graham, advocate. And it's weird because, because I'm sure it probably seems weird for people like that to look at what these founders are doing and to see it that way, but I think with these, it's like they're young, they dropped out of college, they've never really been in the real world. They're interpreting the words of investors and people in the startup ecosystem and they're filtering it out. And this is what they're getting, right? They're getting that, okay, we have to go into this industry and find a way to modernize it, which means just cutting out the stuff that we don't think matters. And we're going to raise money from people and we're going to create a platform underneath everything. I mean, this is the YC dream, right? This is what everyone wants to do. This is what's incentivized. And I think that's kind of the thing we should remember here that this is what the investor ecosystem is incentivizing at the moment and that Delve is the logical endpoint of that. Right? And all these other companies have some aspects of what we see going on with Delve. All the companies that rely on Delve, if they're using Delve in the first place, you know, they have to be cutting corners. They have to see the value in cutting corners. And it's like, of course we all know this, but at the end of the day, there's nothing we can do about it. I think that is the frustrating thing to see this whole ecosystem of fraudulent or sketchy fly by night companies taking all this money all around us and just to feel like there's nothing we can really do about it except just

Riley (56:07)

to watch this happen, watch it happen, or have it happen to you. You can't pick your hospitals like cloud storage provider, you can't pick their auditor. So it just, it just, it just happens to you. So the insight goes on with Delve. We knew we were betting on the team that could reinvent compliance. They sort of did, I suppose. Covered and Celine have a special combination of entrepreneurial scrappiness, customer dedication and technical acumen. They built integrations that customers requested overnight. They hand delivered 10,000 donuts to early adopters as a marketing move. One customer told us, this team will document for you. So we are thrilled to lead delves $32 million Series A. If you're an executive looking to get and stay at compliance seamlessly, be sure to check out Delve, to which I of course added the final editorialized sentence. We sure fucking didn't. Anyway, look, last word, Wendy. And then I think we're gonna. We're gonna wrap up, I guess.

Wendy Liu (56:59)

Yeah. My last thought is, you know, like Riley. Like Riley, like you were saying that, you know, you and I don't have a choice. I guess consumers, we don't have a choice. I think. I think here we see this weird paradox of the free market where the idea. I mean, the whole idea behind government stepping back, letting markets regulate it themselves is like, well, companies can kind of, you know, the free market somehow magically works. Customers will just choose different companies. You know, it'll just. There's some sort of magical mechanism by which everything will just sort of become the best or rise to the top, whatever. And I think what we're seeing now is there is a limitation to that, especially as companies are getting bigger and more complicated. Complicated. More obfuscated. And you have all these complicated supply chains that rely on technology that no one really understands. No one is really in a place to observe and to audit even. And so it really does feel like we're reaching the end of days of this kind of the free market will regulate itself mythology. It doesn't seem like it really holds up. It doesn't really feel true in an era of giant monopolies of these really complicated products and just these supply chains that aren't really visible to anyone and where there's no locus of changing anything for anyone on the outside. And I wonder what the question I have is, how much worse does this get? Because I don't think we've reached the bottom yet. And I think we have only a more trash future to expect, unfortunately.

Riley (58:20)

Would you say it's time for people to crack each other's heads open and start feasting on the goo inside? Look.

Wendy Liu (58:26)

Yes, exactly.

Riley (58:28)

And hey, you can apply the same logic to private credit, by the way. Awesome. Awesome. Especially private credit now that interest rates have gone up and suddenly some companies are going to be unable to service all those debts that they've taken on to keep going while not having any business. Awesome. Anyway. Anyway, that's a whole other different part of the economy that is equally fake. Wendy, thank you so much for coming on. It's always a delight to talk to you. As a reminder, where can people find you?

Wendy Liu (58:52)

Yes, I write a column for Bay Area Current about billboards, about the strange, bizarre, awful type of billboards I have to see every day. And I also have a personal delsystem me. And you can also find me on Instagram, Twitter, bluesky, whatever. Just all Elsystem.

Riley (59:09)

Wonderful. Anyway, we will see you on the bonus episode in a couple short days. Bye, everyone.