Open transcript
The AI explosion is massive. AI has gotten so good at development that there are tons of billion-dollar companies built all around it, including Builder... Oh, what? Builder AI fake business with Indian firm Versed to inflate sales? Apparently, they did a lot more than that. They were technically using AI a bit, but it seems like a lot of the code, if not the majority of the code, coming out of this AI app-building company was actually being built by a bunch of Indian developers. Kind of insane. There's a lot to dive into here. I know it's been covered by a few others, but I wanted to do a deeper dive on how these economic disasters happen, why SoftBank seems to be involved in as many of them as they are, and what this means for the industry. What the hell happened to Builder AI? Very interesting question. Excited to dive into it with all of you guys, but none of these companies are paying me, and I certainly didn't raise a billion dollars. So let's have a quick word from today's sponsor, and then we'll dive right in. If you don't care about your users, you can skip this ad, but if you actually care about what your users are doing and how they use the things you build, this product's gonna help you out a ton. PostHog has made my life so much easier as a person who cares about their users a lot. They're an all-in-one suite of product tools, everything from feature flags to surveys to, most importantly, analytics, and I use them a ton for everything we build, especially T3Chat. I could tell you all about how we use them, but I'd rather just show you, because again, we use them for everything. I recently covered how we rebuilt Captchas for T3Chat in order to make things more reliable, and I only knew it was more reliable because of how easy PostHog made it to check. This is a dashboard I made to keep track of all the things going on with Captchas. We had ReCaptcha on the main site and HCaptcha in the beta site. I was routing all of it to the same project, so it was really easy to keep track of. And we can see here the amount of successes and fails with different things. So we have the ReCaptcha success and fail rate, and you can see the red there. It was failing quite a bit. And you can see HCaptcha was a lot less aggressive. And you can clearly see where the cutoff happened where we moved to HCaptcha in production. It was super easy to create all of these charts, track the failure rates, track how many users are being challenged by our setup, and so much more. Creating those insights is surprisingly easy to do as well. I had the different success events for when the Captchas passed, as well as a bunch of different failure events. Rather than having to log and compare all those different things or try to force everything to be one event type, I just added a formula. A plus E is success. B plus C plus D plus F plus G are all the different fail cases. If we scroll down, now it's a chart. What if I want it to be, I don't know, a bar chart? Oh, cool. It's that easy. What if I want a table for how things behave over time? There we go. It's super easy to set things like this up. I would know. I spend way too much time in the post-hog dashboards. I honestly spend as much time in here as I do in my editor and the rare times I do have problems, I hit them up and they're fixed within a day or two. I've been super happy with post-hog. I have a feeling you will be as well. Check them out today at soidiv.link slash post-hog. The big two things that went wrong here that people are talking about are the straight up financial fraud that they committed, as well as the chaos that was not actually having AI build the things that their customers were requesting. We make building an app so easy, anyone can do it. Whenever someone makes a promise like this, you know they're full of shit. Even the vibiest of vibe coding companies like Lovable and Bolt don't pretend that you aren't a developer. They show you the code right in their UIs and they generate it alongside you. These guys pretend that it's that easy. And it's particularly funny because it wasn't even that easy for them. Also want to make a statement here because I've made this mistake multiple times already. Builder.io is a different company from Builder.io. They were originally a web tooling company trying to make it easier to reuse components across different frameworks. They're also the place that Mishko is applied. And Mishko was originally the creator of Angular. Now he's the creator of Quick, QWIK, the framework. And now apparently they're a vibe coding platform too. So yeah, that's where we're at as an industry. Builder.io, Builder.io, nothing to do with each other even though they now look like the same product and they basically have the same, I hate everything. Is AI Vibecode Builder the like crab of application development now? Is it the thing everyone keeps accidentally making? Oh God. They're all building the same thing. They're all building the same thing. I'm gonna go insane. Anyways, Builder.io scam. The two major points are one, legit financial fraud. And two, legit AI, aka actual Indians. The fact they actually had Indian devs as the people doing the stuff, like it's the meme everyone's made forever, but they actually did it. It's absurd. Yeah. So how did we get here? The first financial fraud bit is actually kind of interesting. It's a thing that every founder has thought of doing, but most of them know better than. Let's say you have two companies. You have Filder BA and you have Per Se. Now let's pretend these two companies are friends. Filder BA and Per Se are homies. The founders like each other. They work with each other. Good vibes. Now let's say that these two companies decide that they want their MRR to go up. Something that's important to know about all of these venture-backed companies is that the valuation that they have is not based on how much money they are making. It's based on how much more money they are making than they were some amount of time ago. The thing that determines how valuable these companies are isn't how much money they make a year. It's how much growth are they showing month over month and year over year. So a company that's making a million dollars might seem much more valuable than a company that's making $100,000. But if that million dollar a year company was actually making 1.1 million last year, or maybe they were making less. They're making 800K last year. But the company making 100K was making nothing last year. That's much better growth. And you are investing based on the potential for this company's growth to continue or improve further. If you were to do a chart like this for my business, ping.gg, that is now also the business that T3Chat is based in, the chart would have looked funny. So it would have looked like this. It would have went up to a pretty high amount at the time, especially, because we were only two or three months in. It pretty much flatlined. Then it dropped a little bit. Then it bumped a little bit. Then it stayed with a very, very, very slow trajectory. Very, very, very slowly going up over like three years. And then T3Chat dropped. And this is our chart. And as such, the handful of people that invested here, they felt like they were being scammed for three years. In fact, some of them asked to pull out their money. And a couple of them were annoying enough that I let them pull out their money. I let them pull out their money. Right? Fucking here. And I hope they are kicking themselves because they were assholes. And I don't feel bad about kicking them out of my table. Imagine you held your money from here to here in a company that was not really growing. You somehow think it works like the stock market. This is the biggest thing. The reason you need to be an accredited investor in order to invest in these earlier stage companies is because you are expecting that your money is gone forever. When you invest in an early stage company, there's a 99% chance it goes to zero. And you have to expect that. As the company's stage gets further and further along when you get to series A, B, C, D, then the likelihood goes down. It also means the multiplier goes down. Because if you invest really early like here, and then the company keeps going crazy like this, maybe it gets a little less crazy at the angle, but the type of distance keeps going up. And you want to invest at these other places on the line, like you want to invest here and here, you're investing a lot more money, which means you're getting a lot lower return if the company goes public or gets bought out or whatever. But you're making a bet always. And that bet is largely driven by the trajectory of the company's growth. And when you're investing this early, you're taking a huge risk because there's a good chance the company just stays flat forever like we were, or goes down, or just fails outright. And the reason you have to be an accredited investor is you have to be liquid enough that you can eat this if it does go belly up. Accredited investor in the US means you meet one of two qualifications. Either you've made 250K a year for the last two years and expect to going forward. Big deal. If you made 250K because you got some crazy grant or bonus two years in a row, and you don't believe it will continue, you technically are not an accredited investor because it's basically just like a, it's a vibe check type thing. The other option is you have over 1 million of assets that are not your primary residence. This is a big deal. It's a lot of people that are worth this type of money, buy a really fancy fucking house, and then they go gambling on these markets. They lose all their money, or worse, they're sitting on this all, waiting and hoping that one of those companies will pay out. None of them do for five years because you're lucky if one of these pays out in five years. Usually it's like 10 to 15 at best. Now you're not making enough money. You spent it all on these startups. You can't get it back. And you don't want to move out of your house because who wants to move because they're running low on money, especially if they just invested millions into these startups. So your primary residence does not count towards your assets. And I think that's a very good call. I think the house you live in is not a fucking investment. It's your house. It's a utility to you. If you're not willing to sell it, it's not a goddamn investment. Stop pretending it is. So for those reasons, being an accredited investor is important because again, you cannot touch the money once you put it in. When you give money to a startup through the traditional investment mechanisms, it is their money on an IOU where eventually they will give you some stock. When you do these investments, it's not your money anymore. When I look at what I am worth right now, all the money I have in these early stage startups, I do not count towards my net worth because I have to treat it like it will go to zero. And maybe someday it doesn't, but it will effectively be as though I won the lottery if those companies hit. So you need to know that going in. And as the company gets further and further along, the risk goes down, but the amount of money you are putting in goes up. There are different investment firms that have different strategies here. Some of these investment firms, like Y Combinator, do what I call the shotgun approach. You know how shotguns work? It's kind of like this. They send a scatter to a bunch of different places at once in hopes that enough of them hit the target versus something like SoftBank, which is the, what I'll call money bags approach. SoftBank comes in way later stage, finds a company that seems like it will grow a little bit further than it currently is, maybe two to 10X. And then they bankroll them with an ungodly amount of money, hoping that there's a 50-50 shot that they succeed. So instead of betting on two to 600 companies at a time, like Y Combinator does, they used to do it twice a year. Now they do it four times a year. They're over 100 companies a batch. So that's minimum 400 a year, probably closer to six to 800. Versus a company like SoftBank that will invest in two to five companies a year at best. They invest the same amount of money year over year, I'm guessing roughly. SoftBank might even invest more actually. The gap between these two would be that they're roughly the same or SoftBank invest significantly more. I'd have to look at the numbers to be sure. But Y Combinator ends up with a similarly absurd return because they're investing so early that if just a couple of these companies hit, they print. Because they're getting 7% of a company when it's worth nothing. And 7% of $10 billion is a lot of fucking money. And Y Combinator is getting that 7% for 500K. SoftBank is spending 10 to $100 billion a year to get smaller percentages of these companies in hopes that that will turn out to be much bigger someday. They're very different strategies. This is a spectrum and they're the opposite sides of it. And other firms fit different places along the spectrum. And knowing that Microsoft has their own ventures program where they'll invest into businesses, I hope you all have the intuition to know where this would fit. It's on this side. Microsoft goes in on the later stage companies, not the earlier stage ones. And that's why Microsoft and SoftBank are investing in these safer bets. And here's where the controversial takes start. If the Builder AI story was an early stage Y Combinator company, it'd be like a nice funny joke and it wouldn't surprise me that much. The fact that they're doing the types of scams, allegedly, that we're about to talk about, makes it way fucking funnier that it was some company on this side of the spectrum, not on this side. But I've been dilly-dallying too much. The reason I went down this whole rabbit hole is to show that growth and consistency of growth and the proof that you are improving your revenue month over month and year over year is essential to being able to raise money. That's how you show these companies that you are worth investing in. Not that you make money now, but you will make more money tomorrow or next month or next year. That's why these investments happen. So what happened between our alleged theoretical companies, Filder and Per Se? Well, both of these companies made some amount of money. Let's pretend that they both make $5 million a year. Now let's also say these two companies are struggling to grow. They want to make their numbers look better. They want to show that this year they actually grew and now they're at $6 million a year. But in reality, maybe the revenue plateaued. Maybe it dipped. Maybe something else happened. They both want to show growth. So they come to an agreement. What if we over at Filder pay you, allegedly, some amount of money? Let's say we do $1 million a year for some on-the-books thing that we need. And now, per se, allegedly, pays you back $1 million a year for other services. This on the books now looks like these two companies are making more money because of one other important detail that I probably should have brought up earlier. this chart is not profit. This chart is revenue. The thing that investors look for is not increases in profit, it's increases in revenue, which is disappointing because our profit at T3Chat has doubled month over month for quite a bit. But investors don't care because they want to look at the revenue number. Thankfully, our revenue has been going up quite a bit as well. But our profit margins have been getting better as we've made changes to the service. I'm proud of that because I want to make a business that will survive. But this isn't about businesses that will survive. This is about businesses that will crush everything in their wake and lie in order to do it. And that is what these companies decided to do. Imagine that you made $5 million a year and now you're making $6 million a year and your actual take-home, your profit, hasn't gone up. In fact, it might have gone down a bit just because of the transaction fees and other bullshit you have to do for this agreement. Terrifying. And that's a very real problem. As such, there are people who care about this problem. Investors, if they knew about it, would. But you can't do this for a long time and not get caught. And that's what ended up happening. They got caught by one of their creditors because even if you have investment money, you have to put it somewhere like a bank. And you'll often also take things out like loans from the bank or use a credit card that you rack up a huge charge on with the bank. So even if you have venture debt, which generally means the money's yours until you go out of business or you become profitable, you still also have some amount of credit debt. Almost every business does. They had a lot of credit debt. And one of their creditors noticed that this bullshit was going on, decided that they didn't want this company that is committing fraud, allegedly, to be part of their bank and a liability for them because if this whole pool of bullshit currently actually owes you $37 million and you gave it to them because you looked at their numbers and thought they were a real business and then you figured out later, allegedly, that they are not, you want your fucking money back. And thankfully, if you're a bank that controls the money, you can just take it. And that is what they did. So allegedly, the two companies routinely billed one another roughly the same amounts between 2021 and 2024. Documents reviewed by Bloomberg Show as part of an alleged practice known as round tripping. People said Builder.ai used to inflate revenue figures that it presented to investors. In many cases, products and services weren't actually provided to either company for these payments, said the people who asked not to be identified discussing confidential information. Versailles is saying it's baseless and false and Builder.ai refused to comment. They were previously valued at $1.5 billion. The fact that multi-billion dollar companies are fudging their books when I am being so disgustingly transparent, it just hurts me. I'm thankful more of these companies are getting fucked over because you don't really get a benefit long term for being dishonest like this. I personally like to believe that honesty wins long term. It's why I do what I do and why I am who I am and why so many people fucking hate me because I can't help but talk shit when I see shit and I'll even do it with myself. People think deep rechat is less stable than our competition because I actually explain what goes wrong when we have outages and everyone else just fucking lies and hides it. It's annoying but I know long term that honesty will win as we see here with a $1.5 billion company not being able to cover their fucking asses enough to not go bankrupt. It's insane. Bloomberg reported earlier that Builder.ai overstated its projected 2024 sales to creditors by 300% which contributed to the lender's decision to seize the company's funds as they fucking should. Builder.ai collected close to $60 million in revenue from Versailles in the four year period for services such as application development according to the people with knowledge of the situation. In turn the AI startup sent funds to Versailles and its subsidiary Quark Media Tech for services such as marketing and I guess that's all it is. The documents were showing this pretty clearly. The companies appear to have interspersed the timing and amount of the invoices to avoid suspicion though each firm ultimately spent approximately the same amount according to people and documents. What's fun here is I actually have a problem with this myself. I am two entities to an extent. I have T3 content which you can guess what this one does. It's not T3 chat. This is my YouTube channel, my advisory, all of the other stuff I do as an individual. This is the one that pays Ben, my channel manager, to help run the channel. This is the company that pays FaZe, my editor, to edit my videos. This is the company that gets paid when sponsors come in to sponsor the channel. This is the company that gets paid from YouTube and Twitter and all the other places I get ad revenue from. T3 content is the media arm and it is a fully, entirely, unrelated, separate business from Peng, aka T3 Tools Incorporated. This is the business that I formed in 2021 when I got into Y Combinator and made a real company and raised real money and did real shit and three years later we finally actually started making money. T3 Tools is the company that owns Upload Thing, it owns T3 Chat, it owns Ping.gg, it owns Pick Thing, all the things that I build and charge money for and have my team maintaining. That is my business, my startup, that I am at 50-50 split on with my CTO, Mark. So Mark and I both own about 46% after dilution and whatnot of the business. And that's an important detail because if we didn't own 50-50, I'd be the majority owner and if I was the majority owner, these now become the same taxable entity, which was a very annoying thing when T3 content was making money and Ping was not. And realistically speaking, I am not the majority owner of Ping. There are no decisions I can make that Mark can't override and vice versa. These are two entirely separate legal entities. One is me, T3 content is an LLC that is mine. It is my business. Ping is a venture-backed company that I and others own significant portions of. These are entirely separate businesses. So here is the problem. Because I cover things like the T3 chat stuff we build on the content side, there are companies that I will use the product of, like, and do a video on. And since I like them, I'll reach out about a sponsor deal. And their favorite thing to say is what if instead of paying you, we give you the service for free? Or even better, I shit you not, one of these vendors that we spend three to four hundred bucks a month on offered to give us their $60,000 tier for free to get a bunch of ads, to get more than 60k worth of ads. And I had to explain to them, I'm sorry, motherfuckers. I gave you free coverage on my business, on my YouTube channel, because I thought the product was good, because I used you and I pay for you on this side. These are separate entities. They do not cross streams. These do not, they don't pay each other, they don't work with each other, they're different fucking things. I'm also paying you way less money than that tier, and I don't need any of the features in that goddamn tier. I don't care. If you want to convince me to subscribe to that tier for my company, try it later, because I'm not interested right now. You're not getting out of paying me if you want the sponsor coverage. If you want me to cover your thing in my videos going forward, the thing I actually like and use, and basically did a video promoting from, just because I used it and enjoyed it, they're like, okay, but what if we give you it for two years? And I just didn't have the fucking heart to respond to the email because they read an email where I said, you can't get out of paying this by paying this, because then I am crossing streams. I am committing a similar type of fraud to what we are talking about here. These are different fucking businesses. And they were so willing to just do it anyways. They were pressuring me to do it anyways. And now I'm not going to work with the brand. And now I'm hesitant to even leave the video up where I promoted their product because this was skeevy enough that I didn't like it. I am as honest and real as I can be with this, both to all of you people that I talk to and share my thoughts about these things with, but also to my investors, to my creditors, to my bank, to everyone else involved in the chaos that is T3 chat, my business, and T3 content and my YouTube stuff. It's just so frustrating that people default to committing fraud. The number of stories I've heard of creators being reached out to, being offered to, to be paid money to review a product and not disclose that it was a paid review, it just makes me sick. If I'm ever paid to do something, you will know it as soon as it happens. And I try to limit what brands influence me to do by telling them what the fuck I'm going to do. Do you have any idea how many brands have hit me up and offered to pay me five times more money if I did a dedicated video that no one would watch instead of a two-minute ad that's actually useful? I'll never do it. I learned my lesson. I might, if the product's really good and I have an angle that's useful, briefly consider it. But I like the new model. And do you guys like the new model? And our goal with all of this is to be honest. The goal of this fucking company is to look like they are successful more than to be successful. It's also the same thing I talked about in the ARC video. Cosplaying CEO. At this point, it's cosplaying a successful business. You know what successful businesses do? They hire a lot of people. They spend a ton of money on marketing. They spend a shitload of money doing all sorts of things that don't really matter. They have offices all over the world, really big ones that look super cool. They also have revenue that goes up. So Builder AI managed to make all of those things happen. But they forgot to actually make money. And it is so goddamn common. And it hurts to see. Honestly, the fact that they were faking the AI part by having a crappy scaffold really early before Claude got good and just hiring a bunch of devs in India to go build the code for them, that's not even the interesting part to me at this point. It's like funny. But that's not the thing that is as pathetic here. Obviously, you guys almost certainly think that is more pathetic. But doing the obvious scam, the one everyone sees and thinks about and knows better than to do. Do you know how easy it would be for me to pay myself from T3 Tools every time I mention T3 Chat and then just spend it on the unlimited tier on T3 Chat and buy a bunch of credits? I could so quickly 10x R-A-R-R by just paying myself to make ads about T3 Chat and then paying the company for a subscription to T3 Chat. I could make this business look absurd and it just feels fucking stupid and like it's obvious you don't do that. It's so stupid fucking obvious. T3 Tools pays me $50,000 a year to be the boss. That is my salary. That is what I make off of T3 Chat. It doesn't matter that the company is profitable. I don't see a goddamn cent of that money until maybe someday I sell some of my shares pre-market, post-market or we exit we go IPO or I sell it to a business. That's how I make my money on T3 Chat. Okay, I did just bump my salary because apparently I was paying myself so little that I'm not technically honoring my role as an employee and that can mess up the relationship between the business which is fucking stupid but I'm paying myself I think I bumped my salary to $100,000. It's annoying I have to, I don't want to but that is my salary as an employee of the business and it has nothing to do with my coverage and my content. My content site is entirely separate. It is obviously separate. It should stay separate. It will stay separate. These things have nothing to do to each other. So if you're one of those people that I talked about sponsorship stuff, you're trying to get me to take a discount for the T3 Chat usage of your product to trade a sponsor, fuck off. You don't know how any of this shit works. Please stop annoying me about it. And I'm going to send this clip to the company that I am talking about here because I need them to know how fucked they are and I need them to realize that they're on notice now because what the fuck? Anyways, this has been quite a rant. I hope you guys understand where I'm coming from and why I think this part is much more interesting than this part is. But yes, it is also very funny that they couldn't get the AI working so they just paid a bunch of devs to fix it anyways. That's how these things go. That's how this all went. Let me know what you think and until next time, stop committing fraud. Allegedly. Allegedly. I'm going to fucking die.
