This Week in Startups - Meta’s Dangerous Chat, Protect AI & Founder Fridays Pitch Comp | E2118

Episode Date: April 29, 2025

Today’s show: we break down the wild story behind Meta’s AI chatbots going completely off the rails — including a bombshell WSJ exposé about bots flirting with minors. Then they cover Palo Alto Networks’ huge $500M+ acquisition of Protect AI, and sit down with the founder of Formulate, who’s building a robotic kitchen for chemicals and cosmetics. Plus, we get into some spicy takes on the secret group chats shaping tech behind the scenes, whether AI can clean up bad code faster than humans, and a Founder Fridays Texas showdown between Osprey and Wispr.*Timestamps:(0:00) Jason kicks off the show.(1:45) Jason’s recent trip to Detroit.(7:47) The influence of group chats on tech, politics, and Trumpism(10:54) Northwest Registered Agent. Form your entire business identity in just 10 clicks and 10 minutes. Get more privacy, more options, and more done—visit northwestregisteredagent.com/twist today!(19:48) The importance of proactive communication in media(20:04) Fidelity Private Shares℠ - Visit ⁠ https://www.fidelityprivateshares.com! Mention our podcast and receive 20% off your first-year paid subscription.(21:44) Palo Alto Networks acquires Protect AI(28:57) ARX Robotics and the future of autonomous military platforms(29:31) Pilot - Visit https://www.pilot.com/twist and get $1,200 off your first year.(35:27) Sentra's $50M Series B in data security posture management(42:16) Lightrun's $70M Series B and venture capital trends(51:28) Meta's AI chatbots, ethical concerns, and Founder Friday competition(1:05:19) Formulate interview with Osmaan Shah(1:11:52) Scaling microfactories and the benefits of Formulate's platform(1:21:10) Business model and future of robotics in manufacturing(1:24:38) Potential applications in food and agriculture for Formulate Robotics*Subscribe to the TWiST500 newsletter: https://ticker.thisweekinstartups.comCheck out the TWIST500: https://www.twist500.comSubscribe to This Week in Startups on Apple: https://rb.gy/v19fcp*Links from episode:Check out Formulate Robotics: https://www.linkedin.com/company/formulaterobotics/posts/?feedView=allCheck out Protect AI: https://protectai.com/Check out Lightrun: https://lightrun.com/Check out Sentra: https://www.sentra.io/Check out the WSK article on Meta’s “Digital Companions”: https://www.wsj.com/tech/ai/meta-ai-chatbots-sex-a25311bf?st=6jzH4S&reflink=desktopwebshare_permalinkCheck out ARX Robotics: https://www.arx-robotics.com/*Follow Osmaan:LinkedIn: https://www.linkedin.com/in/osmaanshah/X: https://x.com/osmaanshah*Follow Lon:X: https://x.com/lons*Follow Alex:X: https://x.com/alexLinkedIn: ⁠https://www.linkedin.com/in/alexwilhelm*Follow Jason:X: https://twitter.com/JasonLinkedIn: https://www.linkedin.com/in/jasoncalacanis*Thank you to our partners:(10:54) Northwest Registered Agent. Form your entire business identity in just 10 clicks and 10 minutes. Get more privacy, more options, and more done—visit northwestregisteredagent.com/twist today!(20:04) Fidelity Private Shares℠ - Visit https://www.fidelityprivateshares.com! Mention our podcast and receive 20% off your first-year paid subscription.(29:31) Pilot - Visit https://www.pilot.com/twist and get $1,200 off your first year.*Great TWIST interviews: Will Guidara, Eoghan McCabe, Steve Huffman, Brian Chesky, Bob Moesta, Aaron Levie, Sophia Amoruso, Reid Hoffman, Frank Slootman, Billy McFarland*Check out Jason’s suite of newsletters: https://substack.com/@calacanis*Follow TWiST:Twitter: https://twitter.com/TWiStartupsYouTube: https://www.youtube.com/thisweekinInstagram: https://www.instagram.com/thisweekinstartupsTikTok: https://www.tiktok.com/@thisweekinstartupsSubstack: https://twistartups.substack.com*Subscribe to the Founder University Podcast: https://www.youtube.com/@founderuniversity1916

When I was in high school, I had a session with like a guidance counselor. And I said, hey, you know, he said, what's your plan? I said, well, you know, I read about Brown. I said, you know, they let you make your own degree. You can kind of roll your own concept. So my concept is psychology and like human behavior plus computers because I, you know, I'm into computers and online services specifically. You left in my face.

This is when I was like 16 years old. And the guidance counselor like spit out as coffee laughing. He said, you're barely going to get into any college. You have like a 72, three-year average. Like, aren't you going to take the cop test? I said, yeah, I'm taking the cop test with my brother Josh and the wrong being a cop. I almost was one.

And he said, you know, do that. Like, it's better. Like, you're not barely getting into Brooklyn College. And I got into Brooklyn College. And then I wound up getting into Fordham. But just imagine like a young person getting laughed in your face. And you know what that did to me? That was like, made me who I am.

That pissed me off so much. I was like, this guy. It made me be like, you know what? I'm going to show this guy. I'm going to get into a. good school and I'm going to have a career. This weekend startups is brought to you by Northwest Registered Agent. Starting your business should be simple. With Northwest Registered Agent, you can form your entire business identity

in just 10 clicks and 10 minutes. From LLCs to trademarks, domains to custom websites, they've got you covered. Get more privacy, more options, and more done. Visit Northwest registeredagent.com slash twist today. Fidelity Private Shares. If you want the all-in-one equity management platform, Fidelity Private Shares has you covered. Visit FidelityPrivate shares.com. mention this podcast for 20% off your first year subscription. And Pilot, focus on your product. Let Pilot handle your bookkeeping. Pilot provides the most reliable accounting, CFO, and tax services for startups and small businesses.

Head to pilot.com slash twist and get $1,200 off your first year. All right, everybody, it's Monday. Welcome back to this week in startups. I'm back from the sixth borough of New York, known as Detroit, where my New York Knickerbockers had a wonderful time. I was only threatened to be beaten up by one person this trip. You know, normally when I go to a city like Detroit, I'm expecting like the over under before.

Yeah. It's a tough town too, Detroit. They don't mess around. It's not a joke. Yeah. Yeah, you have to stay on the right street. That's what they say, yeah. But here's what I'll tell you about Detroit.

Great people. Great city. I've never been. A great, great city. I've never been either. I was talking to a bunch of entrepreneurs I met there. My friend Jonathan Trees is there and he's been doing entrepreneurship there for a while. They bought all the old buildings downtown.

Right. When the whole thing bottomed out for like 12 bucks. Yeah. Amazing. Then they slowly said, hey, what if we like cleaned up where the arenas are? So they have a beautiful like Little Caesar's arena where the Detroit Pistons play. They've got a baseball thing. It is gorgeous, Alex.

It's gorgeous downtown. And everything built out from there. So they built like townhouses, restaurants. But like Austin, where there's a lot of warehouses and space, people are drawn to affordable. real estate in a country like the United States, that is, let's face it, not affordable. So now we have a not affordable country and a place like Austin, a place like Detroit,

will draw people. Why? Lower your cost of living with remote work or people being more open to, you know, hybrid. You get some affluent people moving to places that are dirt cheap. Wow. Like, you had a great experience with dirt cheap here in Austin.

Yeah. People in Austin don't feel like it's dirt cheap. But if you're coming in from Los Angeles, the difference is night and day. Incredible. And you had that same thing going to Provincetown, right? Oh, my gosh. Providence was about 40% cheaper. Well, Provincetown is nearby.

But it was funny because I got here. Everyone's like, the prices are unbelievable. Inflation's out of control. I can't afford a house. And I'm like, you would not like California at all. Well, you know, when your benchmark is California, it's absolutely crazy. So I had a great time. I think actually we talked on Friday, right?

A little bit about Detroit. Yeah, yeah, yeah. The restaurants are great. I went to this place, Ghost, had the best burger in my life. I stayed at the Shinola Hotel. Amazing. The watch company, I guess, made an extension. You know, just everything delightful.

But one guy, I guess, recognized me. And I was wearing my kith. K-I-T-H is like a designer the kids are into. They do drops. Okay, sure. All right. Anyway, the guy Ronnie, who runs it, who's a great designer. Like, they make these really great jackets and crossovers between brands.

and they did like a crossover between a old athletic brand with an A, I forgot the name of it, plus Kith plus Knicks. And they do these kind of crossovers, but they only release 50 jackets, 25 jackets. So I happened to get one of these really gorgeous jackets that Carl Anthony Towns wears. I'm wearing it down the street and I got as like, Jake how F you and the Knicks or whatever? And I kind of waved and I was like, yeah, what's up? So this jacket, man, everybody had to comment on it. People love this iconic jacket.

I didn't realize Kith is considered like the height of fashion for urban streetwear or whatever and Nix. And then this crossover with the athletic. Avericks. AverX. Yeah, I found it. So AverX plus Kith plus Nix. And then this guy decided because he was like, I guess he had a couple of pops and beverages, these gorgeous jackets, they basically say talk to me about my team. So I get absolutely, you know, a ton of attention for it.

lot of conversation, people in Detroit are but this one guy is like, kind of keeps up the FUs. And I said, you know, if you want a selfie, I'm totally fine with it. And he's like, no, I'm going to beat your ass. He starts to. Now, I'm there with my little brother who's a former cop of firefighter. Josh, he's a bulldog. He's like, what did you say to my brother? That's the one I got. Yeah. You know, it's kind of a

strategic error for me to wear this jacket. Maybe and walk from the arena to my hotel. It's a little bit on me. You know, you're, you're... I was pushing the amp. Yeah, you're sort of announcing your a bit. But it was great to wear in the arena. The people were great. We won the fourth game. Nix will do the gentleman sweep, as I predicted, but a great town. And I actually, I want to go back to Detroit. I was thinking maybe doing like a this week in startups live in Detroit or something. I'd love to go check out. Great restaurants, great people. Did you know you can swim from Detroit,

like across the to Windsor Canada? Sure. You can go right to Canada. Yeah, yeah. For some reason, I never knew that. And we're talking about the border. There were 50 little swimming, fishing vessels. Yeah. People were just fishing in dinghies. I mean, that's what people, I mean, I think even Trump today was talk about the Canadian border is some loose, like, ah, who cares? They're basically already the 51st state. But the Mexican border militarized the whole thing, guys on secrets. Well, you think about how many people live south of the border. You got what? Forty six, 40 different countries. Yeah, there's a lot of countries. So of those countries, how many have a revolution or, you know, you know, turn into dictatorships? It's got to be like three or four.

four at a time and at any point in time. So there is reason to leave. And whereas, like, I don't know many Canadians who are like, I hate this place. Most Canadians I know are like this place. Very, very few. And also, the numbers are quite different. So Mexico, 130 million people, Canada, 40 million. So just, just the sheer numbers across the border.

That's just one country, right? Oh, my gosh. Absolutely. Yeah. The whole, the Canada-US border thing actually became a plot point in the Canadian serial epic trailer park boys, which if you haven't seen that critical bit of Canadian cultural export. You really owe it to yourself.

I can tell you this right now. You could make a homemade drone, pack it with as much fentanyl as you want, or H-100s. I'm not suggesting anybody to do this. And you could zip all day long, two feet over the water, as much fentanyl as you want.

There, but by the grace of God, for whatever reasons, the Canadians must be doing a tremendous job of policing that side. Yeah. Because I am not kidding. You could do a T-shirt cannon and shoot enough fentanyl to kill Michigan.

Yeah, it's a tiny amount of fentanyl. I mean, it was crazy. That's what makes it so hard. It's incredible, though, that Jason just pointed out that there's two things you want to traffic around the world today. Drugs and Nvidia GPUs. And he's not wrong. I mean, basically these are the pound for pound, I think.

Yeah. The most valuable. Most valuable for sure. Like diamonds, gold and Bitcoin. I mean, they're whatever, three, four, and five. But I think, yeah, yeah, yeah. It was a crazy story that I am tangentially. mentally next to.

There is a... Group chat gate. There's group chat gate. Now, I don't know, a year ago, six months ago, before the election, Eric Torrenberg, who fancies himself like a micro J-Cow, it started his podcasting network

and did a bunch of... He did like a little all-in knock-off that lasted like five episodes, and then a bunch of other niche podcasts. Right. He's a great kid, and he joined Andreessen Harowitz. But, you know, whatever year ago,

he's like, hey, Jay Cow, check this out. Can I invite you to the signal group? I'm like short. And then like it was like 50 people, that 100, that 150. And you know that like whole preference stack that happened in Silicon Valley? Yeah. Kind of watched it up close and personal. Mark Cubans in there. Like this is all public now. And I never talked about it all that much because, you know, like, I kind of felt it was quasi-private. You know, based on the people in there that there are some people who are journalists-type people who would take pictures and share them, I guess. A lot of influencers. Maybe a little like, um, maybe white supremacist adjacent folks. You know how there was like a little bit like at the far end of MAGA? There are people who are not Nazi or white supremacists, but they're kind of adjacent to the next group of people. There's a little bit of that sprinkled in. And then a lot of tech people,

a lot of media people. Anyway, they made like some big deal about it because there was a leak. Yeah, well, Ben Smith and Semaphore just wrote this whole piece about it. Yeah. So anyway, people have been dropping off of it like five people a day or like,

yeah, I'm kind of done. But I'll be honest, It's a snore. It's essentially the same discussion, Alex, happening on Twitter, happening there. Yeah. And maybe people use the R word a little more, but it's not like anything different than any of you're seeing. So it's kind of a big nothing burger.

I mean, that's kind of what I would say when I first read the article. I mean, it sort of got previewed. A few of the people from the group check kind of leaked ahead of the article that they had been. Yeah, Joe Lonsdale. Right. They had been interviewed by Ben Smith or he was sniffing around and so it was going to come out. And I think my reaction was, yeah. like, well, of course.

Like, I just assumed these kind of back channel conversations are happening on Signal every day. Like, why would that? I'm in a bunch of group chats. Of course, billionaires have their own group chats. I just feel like it's obvious. Yeah. It's, uh, and centimillionaires.

And cent a millionaire. And cent a millionaire. They're all, I don't need, this all. I'm not trying to be proud of the point of 0.01% of the point one percent. Yeah. Bernie Sanders never shouts you guys out either. It's always right to the billionaire class. I know.

I mean, come on. We got to do better, everybody. We need to expand our circle of hate. Founders, if you're serious about raising money, you need to set up your business the right way. Tight is right, and it all starts with having a registered agents. Investors simply won't fund your business if it isn't structured correctly. Before a VC can wire you the first dollar, they're going to check.

Is your company incorporated? Is it in good standing and compliant? missing a filing or losing your status. I mean, it's just going to be a deal breaker for the VC. It's like you're not taking things seriously. And that's what happens during due diligence. That's when a VC make sure they're not making a mistake by giving you investment dollars and, hey, angels do this as well.

And that's where Northwest registered agent comes in. For just $39 plus state fees, they act as your registered agent. They handle all the paperwork. They keep you compliant. And they make sure investors see you as a serious business that's worth funding. In just 10 clicks and 10 minutes, your business is officially set up an investor ready. Northwest handles filings, they protect your privacy, and they ensure that you never miss a deadline. These are the chores that you don't want to have to deal with.

You need a partner, and Northwest registered agent is that partner. Thousands of founders trust Northwest because they keep businesses in good standing. And with their expert corporate guides, you get real support, you don't get bots, you get real people on the phone. So here's your call to action. Very simple, easy, peasy, lemon squeezy. Don't let bad paper or cost you your next funding. Go to Northwest Registeredagent.com slash twist and get your business investor ready today. For just $39 plus day fees, you can set up your company the right way, fast, private and compliant. Go to Northwest Registeredagent.com slash twist today. Jason, actually, I want to point out that last summer, you and I, before Lon joined us,

I asked you, what is the technology argument in favor of Trumpism? And you gave me a pretty interesting look. And you framed it something like, here's what's happening behind the seats. And you were very clear to obfuscate sources and they came from and so forth. But were you polling at that time kind of from this group chat? Among other group chats, I mean, yeah, sure. I mean, I think one of the things that happens on social networks, whether it's blue sky is falling, Twitter X, Reddit, is you have a mixture of the elites,

the known people, the powerful people, people like us and others, more powerful and more influential. And then you have the Hoy-Polloy. You have the civilians. And that's kind of the magic of it, but it also kind of kills it. Because if Mark Cuban and Joe Lonsdale are getting into it with Mark and Dr. Jason and Ben Shapiro and J.D. Vance and they're in the same chat. Like, whoa, that's kind of entertaining. Wait, J.D. Vance was in the group chat? Well, anyway, I don't know if I'm saying too much of but I think he might have been. He's not mentioned in the in the semifference. I know.

I was about as able to break some news here, ladies and gentlemen, the VP. Maybe I'm wrong. I don't know. Maybe he was in the early days. I don't know. It could have been anybody. I mean, they,

I'm not saying they put the plans to blow up the hoothies in there. Yes. Well, let's just say, let's just say. That was going through the context. Like, let's see,

who's in? Who should I invite? My mom and good girlfriend, obviously. So, yeah, I think I do get some signal from group chats. In Signal, in WhatsApp, people will, I think, probably, that's actually a really interesting point, Alex.

I would say, even though people are speaking as if they know it's being screenshot and sent to journalists and they don't, everybody knows. Don't put anything in text because your phone could get dumb, it could hack, whatever. So don't say anything there. That would be career ending or you wouldn't put your name up publicly. But that's an interesting point. Because it's more interesting to be there than on social media because of the sort of spam and the bots and the chaos, whatever. I do think maybe people will talk about things first in the group chat, floated in the group chat,

and then go public. I know I do that in some of my group chats. Oh, for sure. So I'll be having a conversation with you along in one of our group chats, and then I will take what I just said and say, oh, I'll adapt that into a tweet. I'll copy it. I'll copy what I said, not somebody else is to be clear.

And then said, was there any, like, I don't know why this is a number one story on tech meme.com, but that's what I'm just saying. I don't know. It feels like a, oh, I can explain that. Tempist in a teapot? It's because it.

if you're plugged in, you watched the technology luminaries that were part of these group chats shift to the right slowly, incrementally. But for a lot of folks out there, it kind of went, okay, everyone's a normie Democrat amongst the VC class. And then suddenly a bunch of people came out in favor of Trump. I mean, I remember Sean McGuire, who was part of these group chats, dropping his ex-essay on why he supported Trump and being surprised because I, one, didn't know Sean McIre was. And also, it struck me as a surprising move to make in that political climate. But if I had had this context, Jason, it wouldn't have been a surprise whatsoever.

Also, I do think that the end of the article is the most interesting bit when they discuss how the post-Liberation Day tariffs are fracturing some folks on the right who are technologists. Because I think it's fair to say that this show is a free trade fan club. And I think that has been the traditional position of business leaders and especially on the right. So to see now people having to defend Trumpism, contrary to, what I was called traditional GOP economics, is fracturing these groups. And that's interesting,

because it didn't last long as a coalition, Jason. Yeah. I think the other interesting thing that people are noticing, I've seen a lot of people going back on social media, finding members of these group chats that have been identified, and noticing when their tweets and their rhetoric started to align. Like I saw a screenshot of a Matt Iglesias post, and then, I don't know,

somebody else from the group chat, Jill Lonsdale, maybe. And there was like, look, they're making the same argument, within a week of each other because they were both pulling from this group. Yeah, that is kind of like

the phenomenon. You go to a dinner party, you have a conversation about, you know, in Hollywood, I don't know, Carr. Remember that movie,

Tar? Of course, Cape Blanchett. I remember listening to Brett Easton Ellis and he didn't like Tar. He actively didn't like. Yeah, sure.

And then he had conversations with people and everybody loved it and he's like, let me watch it one more time. He watched it again, which is a three-harm plus film. That's a little bit of a task. Yeah.

And all of a sudden he came back and said, you know what, on second watching, I can understand what people appreciate about it. And I kind of like it now. So there is something to working through issues and changing your mind or peer pressure, you know, whatever. Or even just hearing somebody speak to something and it put a spin on it you hadn't thought of before. You go, oh, they're right. I like that argument. I'm going to sort of adopt that myself. Yeah.

I think that's what is fascinating people is the anthropology of going back and watching these arguments sort of spread. Yeah, yeah, yeah. Well, and now we're watching it sort of come apart because there is a Trump really does drive ratings on podcast. Sure. So, you know, we'll put in the first sentence here. You know, Trump, if we, let's do an experiment for our YouTube, the Trump or Trump group chat. Yeah. Just, we'll name everything, Trump group chat. The Trump tech exposed. Yeah, Trump tech. Trump tech, group chat exposed. There you go. Let's start with that as the thing. You put that as the top of the YouTube or the thing. People are going to want to know. It'll get clicks, yeah.

So, you know, Trump Beagle's box office, so everything kind of revolves around that. And you are seeing now a lot of the arguments people have for why they wanted Trump over Kamala was the economy. And now people don't feel like, oh, it's kind of like some of the neocons who are like, kind of feel like maybe we should fight against like Putin. Maybe we should have global trade. Though that neocon or Clinton Democrat position, I think, is a large part of the middle. And that middle, that is like, yeah, Putin might be a bad guy. And we should stop them from invading countries. And maybe we don't want to shake the economic snow globe too hard with these tariff things.

Maybe we should just do it gradually and not have the economy do these kind of crazy roller coaster ride. I think those group of people are the ones who maybe left the Democratic Party. They may have pinched their nose or were peer pressured or just changed their opinion on Trump. said that's a better choice of the two bad choices. That group of people now, I think, are maybe saying, I'm not happy with this. And that's what you're seeing in the polling. I was just about to say, not not alone. The third plank. Americans are strongly in favor of deporting people who are here illegally who have committed crimes. Yes. But that's it. Yes. Trump is going so much further than that now to the point of rounding up two-year-old

U.S. citizens and deporting them that you're starting to see the polls reflect that, too. So he's got away from these centrist positions to the far right. And I think that's what we're starting. Could be a mistake, I think, because maybe he loses some of those people in the midterms. Yeah. It's all so chaotic right now. Agee is a word chaotic because it seems like that's overused. It's hard to understand what exactly is happening. Communication is critically important, as we know from our company. Proactive communication, particularly important. A little proactive communication goes a long way. This advertisement is paid by Fidelity Private Shares. All right, founders, we all know tap tables, due diligence, and of course, managing investors is a huge headache. But there's a very

simple solution for you. Today we're talking with Kristen Kraft, an old friend of mine, and she works at Fidelity Private Shares, a new group over Fidelity. You've heard of Fidelity before. And they have a mission to help startup simplify equity management. They're going to save you money. They're going to give you better service. Welcome to the program, Kristen. Thank you so much, Jason. It's great to see you again. Yeah, great to see you as well. Maybe just from a product perspective, what are you trying to accomplish with the product? So, Jason, we are super excited about our cap table management and data room platform. We want to make it super simple for founders and startup operators to manage all sort of ownership and equity in the company and essentially prepare to raise. We want to make sure that

everybody goes into these fundraising conversations well prepared. They're ready to share their cap table and that they're ready to go through due diligence as they're trying to close their round. So from a product perspective, that is where we're laser focused and that product is really well built, really strong attention to detail in the way that Fidelity is known and beloved for. So if you want an all in one equity management platform Fidelis, Private shares, they've got you covered. Visit FidelityPrivate shares.com. That's one word, no spaces, no dashes.

FidelityPrivate shares.com. And hey, mentioned this weekend startups. They'll give you 20% off your first year subscription. Once again, FidelityPrivate shares.com, and tell them that you heard about it here on this weekend startups. All right, let's dig into Little Tech. We've quite a lot to go through here, Jason. The biggest story I think is that Palo Alto Networks is going to buy Protect AI,

which is a company that we added to the Twist 500. So we'll need to take it off now that it's going to be part of a public company. Protect AI if you don't know sells tools to secure generative AI usage inside of large companies. If you want a bit more detail about that, essentially it helps enforce AI usage policies. It helps you red team AI applications to make sure they're not misused and also does conversational audits. Okay, hold on a second.

We know what red teaming is. That's when you try to trick an AI to make a mistake or do something it's not supposed to do. So if it wasn't supposed to do. wasn't supposed to grab your credit card from last pass or password, one password, whatever you're using. If it's not supposed to, I don't know, do something racy, sexist, offensive, you try to get it to do that. Yeah. That's what red teaming is just for people who don't know. You try to break things, get it to act offensively. So this helps you with that. And then, I guess, monitoring the workforce usage and actually creating policies and auditing all this stuff

is really going to be important. I think it's going to be. huge. Yeah. Yeah, this feels like a really important idea because I have paid for everybody in the company, I think, to have chat GPT. And I think we may pay for Claude too, or maybe select people buy it who use it a lot. I kind of need to know what's going on in those, right? We need to know what people are doing, especially if they upload our data to those services. Sure. Yes. So there do need to be some controls on this, and it sounds like a really great idea. And did They say what it was sold for? So they did not list a price.

They called the deal strategic on the Palo Alto network side. Geekwire did report that the transaction was for more than 500 million, but we don't have anything more concrete yet. But no, Jason, actually, it's not that great of a price because I think their last valuation was between 400 and 450. And that was last year. So this isn't that big of a sales premium. And the company said at that time that it grew, I think was by several hundred percent. So to me, there's a missing piece here. that led to this sale because it seems like the company, the funding, the growth all had it pipped

to do a lot more than this exit. I don't want to talk crap about someone's nine figure exit, right? But it just feels a little cheap. So I wonder if we're missing something in the story that hasn't come out yet. Well, the last valuation is the sales valuation. That's like a really interesting discussion point. Is that basically what we think happened here? Plus 20%. Yeah. Okay. So whenever you have less than a doubling, the VCs are not in it for a doubling. That means the founders drove this and the early investors drove it. So there'll be three constituents in a company of the size. You have the early investors, Angel C, Series A. They all got in at five to 10 million, 10 to 50 million, maybe 100 million, and the founders got in at zero dollars. That group probably

owns collectively 80 to 85% of the company, including the employee. Then you got the last 10 to 20% of the company. That's the series B. That's the $100 million at $500 million post. That group of people usually has a right to get at least two times their money. Right? So if they were smart, in a really competitive environment, they can't get that. They just get their money back. So this is a situation where if that firm was in a competitive situation, they may have given up what are called protective provisions. The protective provisions being, you know, you can't just run away with the money. You can't abscond with it. You can't. There are things to protect investors from bad behavior or losing money. One of the protective provisions is participating preferred or

a liquidation preference. This means I get my money back plus something so that if you do some quick sell like this, I don't get nothing. Yeah. So my guess is whoever put the last 50 million in has a provision, they get double their money back, a liquidation preference in other words. So they get the 100 million. If this went for 500 million, they sweep the 100 billion off, then 400 billion goes to everybody else. You get 400 million to everybody else. Maybe the founders, if they own 80% collectively, thought, hey, 40% of 500 million is 200 billion each, we're in. So I'm guessing that's what happened here. And maybe they thought there's a lot of competition and maybe they were given equity in Palo Alto networks, which has been doing fantastic. The founder of Palo Alto

network spoke at the All-N Summit last year. And he really explained how he really courts founders and puts those founders in charge, in charge of their units. And if there was a unit doing that internally and they go by one, they take the unit head who is getting beaten by the independent and say, that's your new boss. Yeah. If the unit head internally is beating the external competition, they get to be there, keep being. So that's his philosophy. It's kind of a cutthroat. It's a reality competition series. I think it's more, yeah, it's performance based. You know, if you are a fading starter in the NBA and some young buck comes up. up and they're the six or seventh person on the bench, they might take your position.

A very awkward thing you see happens sometimes in basketball, for example. I know because I'm friends with Draymond Green and with David Lee. David Lee was the start. He was towards the end of his career. Draymond Green was coming up. I think maybe my memory serves me collectively, like their minutes or starting position may have flipped at some point during that transition. But they were both very cool with it.

David Lee was, you know, Draymond Green's mentor, and so was Bogot. So you had this really great transition. All right. Listen, congratulations to them. And the other piece in this, Alex, to note, DPI, sending money back to your investors has been very low for four years under the wrath of Lena Con. And then it just got kind of paused by this tariff turmoil. Yeah, yeah, yeah. So maybe the VCs here are like, hey, I doubled my money and I could send 50, I can send 100% IRA, the internal rate of,

return here is 100%. I can just ship it to my LPs and just say, hey, by the way, you put in 10 million, this was whatever. If you put in 30 million, this was one million of the fund, all of a sudden, I'm sending you a two million back in the early days of a fund. It's like, oh, that's nice. Yeah. Money came back from a venture funds, something that's not happening as much as it needs to. The power of cash on cash returns. All right. Next up, ARX robotics. It is a European defensive robot company. I have a picture of one of their machines right here, Jason. This is an autonomous robotics platform. The company has many different systems that go on top of this.

You can do delivery of supplies on the battlefield, sensors. You can do casualty recovery, I believe, is the term of art. It looks like a little... What would that look? Tank. It's like a tiny tank? Tiny tank with a big sensor on top is what it looks like to me. Like treads.

It's got two treads. Yeah, it's like a mini tank, like not as big as a, you know, vehicle-sized tank, like a little... Like one of those autonomous delivery. I was just about to say exactly like one of those. Got it. So it looks like maybe an ATV, like that you might drive in your ranch. Roughly ATV size.

Yeah, size of your desk. You could sit on it, but you wouldn't have you and your friend on this device. They're calling it a platform. The acronym here for folks who want to be in the know is UGV unmanned ground vehicles. And this fits neatly into our philosophy that drones and autonomous systems of the future of warfare and to see a European company doing this, raising money, including from NATO's Innovation Fund, not a huge shock. I thought it was incredibly cool. I love to see Europe getting its military boots back on. I think it's encouraging. I think it's a little bit late, but better late than never. And the company put together a 31 million euro round Jason, bringing their total capital to 42. I love seeing it.

If you're a startup founder, you've got a million things that you're worrying about at this moment. You know what you shouldn't worry about? Your bookkeeping. Your bookkeeping should be perfect. And you should have a partner who makes sure it is so. And that partner is Pilot. It's the industry standard. Pilot is the largest accounting firm out there built for startups. They know how high the stakes are. And that's why companies like Open AI, scale AI, and Air Table. Trust them with their books and have done so since day one. When you use Pilot, you're going to get a dedicated team for everything you're doing

from booking to taxes. And, you know, listen, if you need that CFO level guidance, they're going to give you that. So you can stay focused on what matters, building your team, building your product, and delighting your customers. You should not be stressing over spreadsheets with your TNL and in all this nonsense. You want accurate financials delivered on time every time and you want to be compliant with your taxes.

You don't want any last minute surprises. And when it's time to raise your next round and you're scaling up, Pilate's CFO services team is going to help you plan and grow with confidence. Startups that use pilot tend to raise a bigger series B and a bigger series C round than the average startup. Why?

Because when you're buttoned up from the start, Everything gets easy. Focus on your product. Let Pilot handle your bookkeeping. This week in startups, listeners get $1,200 off their first year. Just go to pilot.com slash twist. That's p-I-L-O-T.com slash T-W-I-S-T. Lawn, how do we miss this for the Twist-F-F-Hundredth?

I felt so silly. I was like, this must be a Twist-Hundred company. And it's not. We're very into the autonomous, military, robot, drone kind of, kind of thing, kind of area category. Yeah. I do like that they're not putting guns. on it, that this one is for recon, patrol, cargo.

Bullshit. They're just not putting that in the market. Yeah, yeah. It's a platform, so it's up to you. Yeah, yeah, yeah. And you can be sure that there amounts that if you wanted to, you could put AK-40s on there, yeah. Which I'll be totally honest.

I'm fine with, because what we're going to see in the future is these things shooting at each other. It's going to be robot on robot. This is going to be like that. What was that show on TV where the robots would battle each other in arena? Battlebots. Battlebots, yeah. Basically war is going to be battlebots now.

I would, that's best case scenario. There's battlebots. So this is the future. They're not autonomous though. There's guys with remote controls. But the idea is, they're not punching each other. This is the precursor to it though, like the ability to move around and have weaponry on a small device and then have it battle other small devices, what this is going to be all about. I have a theory that non-US, non-US weapons makers are going to surge.

Why? Well, the U.S. has a certain amount of capacity. That capacity is needed for client number one, the U.S. We need every ship we can make. We need every drone we can make. We need to build a supply chain here that's independent of DGI. So I think places like Korea and the EU and other places that are good at making weapons are going to get their order books filled by everybody who needs weapons.

Everybody needs some weapons. You've got to defend your border. Yeah. Big deal. No judgments. So there is an ETF. A friend of mine told me about K-D-E-F. I backed up the truck.

Korean Defense Company. You cannot buy Korean socks easily. So I bought K-Deaf. I announced it. I sold like two of my losing positions, which were Warner Brothers and Disney. I was sick of Warner Brothers and Disney. Just disappointing me.

I had like 100,000 losses from that. But I had bought some Uber when Uber was like down around $30. And I have a lot of Uber already from the, angel investment, but I was like, let me just buy some Uber here, because I know for a fact that this is definitely undervalues. The trading at $70-something now. So I sold that Uber I had bought two years ago to offset the gains. I'm sorry, the gains from Uber were offset from the losses from Warner Brothers.

That turned and edited out. I cleared, you know, a couple hundred grand. Boom. I dumped it into KDEF. It's like the Korean Defense Industry Index, E-FF. Yeah. Now, this is made up. of when you look at it, it'll tell you all the companies that are in it.

Now, these are individual publicly traded Korean defense companies. I know nothing about those. But I am into defense. I'm going to start buying defense stocks. If you have ideas for us, please put in the comments. If you think this K-Deft is a bad J-Trade. Let me know. I've got to rebuild the J-Trade website, j-trading.com.

Because we're using Koda, and my guy Prush had used a third-party tool to pump in the shares. The real time. The real time. Real time service, not code and not their fault, turned off. So I have to redo the thing.

Or I may partner with one of these websites that allows you to put your portfolio up because one of those had me on a list. And I was number five on the list behind, like Ken Griffin and Nancy Pelosi. And I did this like in my fair time. Now, I had a good entry price,

which was, you know, during the bad bottom out during the SVB. I mean, getting anywhere close to Pelosi, is tough. Yeah, it's tough. Pelosi is a tough, tough. That's a tough person

to be at the poker table with. Anyway, what else do we got going on here? Centra, not the car, not the thing you don't want to drive to the high school prom. No, this is a startup that just raised $50 million.

It is in the world of data security. Jason, there's a fancy acronym called DSPM, which stands for data security posture management. That is, corporate speak, for go find all your data all around your multi-cloud setup and make sure that it's secure.

It's actually a pretty big industry, I think. And the company is also moving into helping, unsurprisingly, companies use generative AI safely, so their data, to your earlier point, Jason, doesn't leak into the applications and go out all over the web. $50 million Series B. Oh, so another contain the AI Mac is great that people are getting ahead of it. So they raise 50 million in a series B. Again, I think this is a pretty cool idea for a $200 entry because it grew 300% year over year. And that is the exact pace of growth that gets me very, very excited because if you're raising 50 of a series B, Jason, I presume you have a pretty reasonable revenue base to be growing from, you know, 300% on top of.

So it's probably pretty impressively stable corporation. And I love to see more as rarely security companies do well. DSPM technology, data security posture management. So you've got data. You need to keep it security. You're plugging it into notebook LM over here, Claude over here, Open AI over here. you need to make sure it's not going everywhere, leaking into those. I need to know about this because I, like I said, I bought Chat Chepti for everybody.

And Chats ChpT keeps asking me to connect Google Docs. Yeah. And I'm like, yeah, no, I'm not going to click that button. But then I was thinking, well, you know, I got enterprising young folks working for me, researchers and analysts. Maybe they did that already because they're like, oh, I can now ask Chat Chepti. I did do that, but only for my personal, not my launch account, only my personal. Now, do you worry that it's going to take whatever you got in there, the Great American novel or

whatever you're writing fan fiction. Maybe you're doing some spicy fan fiction. I don't know what's going on in your docs. Spicy fan. I mean, that's, I don't know, 50 shades of gray stuff going on. And maybe you're, I, you know, I, I, you know, I have some essays in creative writing in my Google Doc, but I don't believe that they're violating our agreement. Like, you have to feed the individual Doc in. You can't, it's not like learning about me by going through my type of Google Doc. It's not training or doing I think you can do that, but no, I'm loading in individual Google. Got it. Okay, fair enough.

So I'm not loading in stuff I don't want it. I want all this to occur on the desktop. My understanding is these Microsoft, I was listening to Leo Leport this week, and on this week in tech. And they were talking about these Microsoft laptops that were, you know, kind of AI-ready. I think co-pilot is what they call it. Yeah. Is how they're branding it.

And you know you have it, he said, because it was a co-pilot key. Just like they had the Windows Starkey. They have a co-pilot key now. What they're doing, and he wants it to record everything his computer does so that you can say stuff like. Last week, I was like buying some razors, but I can't remember the name of it. Can you tell me like what brand I was looking at or reading some article about X, Y, and Z. And recall it. And there was a woman on Twitter on Sunday who is a lawyer.

And she's like, this is a really bad thing because, like, you know, now it can be subpoenaed and there's all this stuff. Sure. really careful. So you think about like subpoena power, you know, okay, you get into a lawsuit. I don't know, you know, your friend buys a company. That's high profile and you get pulled into some lawsuit. And then they're like, okay, yeah, we want to take your co-pilot and start asking your co-pilot, what pages were you looking at? Did you read anything? Right. You may not remember every conversation you had, everything you had, and it might record your signal because your signals on your desktop or whatever encrypted thing you have. Maybe it's recording some encrypted email

you have. Maybe you've got some stuff you talked about with you're a lawyer under attorney-client client privilege. That's getting pulled. So she was making very good points about this. Like, this is crazy. And he was like, yeah, I don't get sued. So I don't care. I'm just going to do it. But all of it all of it is stored locally on your computer. So you need to have a big enough process or enough memory, enough storage to have all that occur locally. That's going to be the future of this, which is just like you might have an index of your photos on your phone, not in the cloud, or like the phone, some things are in the cloud encrypted, but other things are, on your phone locally.

Yeah. There's going to have to be a whole series of decisions made of where to keep your large language model and then what you give up for performance, because this also will slow your machine down. Sure. If you have to keep everything locally, yeah. Yeah. A lot of tradeoffs.

The market that we're talking about is so big, Jason, that another twist 500 company, Sayera, and we had one of the founders on the show, episode 1989 last August for a 300 interview, they raised another $300 million late last year. So people are really backing the space. There must be just boco demand on the enterprise side because I bet you, every company wants to avoid everything that you're saying while also getting all the benefits and cost savings of AI. So to me, it's a great place to build a startup right now.

We've been talking a lot about AI-assisted coding, people essentially writing code with AI tools, either helping them write whole applications or just expand their own writing. People talk about the problems this brings up. There's going to be technical debt. There's going to be debugging issues. It's going to be slow. So I've been waiting for a startup that's going to apply kind of the same idea, but not to generate any new code, but to applying AI to clean up existing code to make

it faster, better, and with fewer bugs. So Light Run, it's an Israeli startup. They are doing this and seem to be absolutely crushing it. The news item here is that Light Run raised a $70 million series B, bringing their total capital to 110. But the thing that I find most interesting is last year, they debuted what they call essentially AI debugging of production apps, so apps that are live out there in the world, and then in the last year, their revenue grew four and a half X. So my read of the situation,

Jason, is that the company found a pretty smart way to use their own model to clean up code often made by AI and therefore how the snake eat its own tail, and it's driving insane growth for the company. I think it's just fantastic, and I hope this means that eventually Chrome will not eat every gigabyte of RAM that I freaking have, but I'm not entirely sure it'll be that quick, But I've been looking for this company. Glad to see it finally here. And I think this is the future. So I'm pretty freaking stoked.

Yeah, it looks like revenue has been growing 4.5x year over year. A lot of founders ask me, like, how can I like rapidly close my next round of funding? In Silicon Valley, triple, triple, triple, double, double, double. That's what, you know, VCs are looking for. So you start 100K, you go to 300, you go to 900, you go to 2.7, something like that. You know, then you double it, you go to 6 million. That kind of fast growth solves all problems. You will easily, easily raise money if you're growing double digit percentages each month.

Rule of 72, look it up. If you're growing 10% a month in 7.2 months, you will double. That's the rule of 72. If you're growing 35% in but two months, you will double your revenue. So get your team focused on what are we, how much are we grown? Week over week, month over month, quarter over quarter in terms of our revenue, that's your North Star. And if you're trying to convince with words and promises and arguing and debating with a VC, all that energy you could be putting into just growing your revenue.

And then that sales solves all problems. That would be Mark Cuban taught me early on. Sales solves everything. Sales solves profitability, running out of money, paying people more salaries, hiring more people, and convincing VCs to invest more money. That's why I think this series B was actually not led by one major VC. firm Jason, but it was co-led by Excel and Insight Partners. And I don't think you get

two big names like that if you're not a pretty darn impressive group. Don't have a valuation here, but one to watch. This is part of the data. Sorry, developer observability is the buzzword. So one to keep in mind in the back of your head. There was something about Neo Scholars. I thought that was kind of interesting. The guy from Neo...

Yeah, we got that one. He's doing like a teal fellowship or something. Okay, so a little background here. There's a venture capital firm called Neo. It's made a lot of pretty sizably awesome investments. Like, I think they were really early in Cursor, the company that just scaled from 100 to 300 million ARR in like 25 minutes. Speaking of growth solving all problems.

Yeah. Yeah. That solves like the world's problems. So Neo, it turns out, is more than just a venture capital firm. They also have something called Neo Scholars and a bit like a Teal Fellowship. This doesn't involve college age kids. But in this case, they just gives them 20K and then has them take a gap semester or a gap quarter quarter of a quarter base system.

I think this is fantastic. I think this fits into the kind of the Jason thesis of we need to go earlier, earlier, earlier. Like, forget an accelerator, forget a pre-accelerator. Let's go find the kids that are the most impressive. And let's just get into friendship with them with a little bit of capital and see what comes out of it. Probably hard to scale. But I think we're going to see more and more of this because everyone wants to get in earlier. So, Neo 20K, I dig it.

Love it. You know, this is good marketing for a venture firm to get a, in early. The thing with the TIL fellowship that people don't know is he doesn't get anything for that 100K other than the relationship with you and the goodwill. Right. So that's what made it super interesting is it was just a way

to get super smart, crazy thinkers from Ivy League schools who spent all this time and effort to get into an Ivy League school to then be attracted to quitting the Ivy League school. Yeah. So you start thinking about the logic. Like what type of person

spends a decade of their life, you know, from seven or eight years old, being super driven to 18 or 19, getting into Harvard, Yale, Columbia, Stanford. And then this weird, you know, guy, Peter Thiel, who's like got all these, like, non-consensus opinions. It's like, I have an idea for you. Instead of going back for junior year or sophomore year, why don't you quit and hang out with me in San Francisco?

I'll give you 100K. Yeah. That attracts a certain type of weird. What's the better word for weird? Non-traditional? Yeah. Outside the box, sort of unconventional thinkers. But I think what's so interesting is, I think we're starting to recognize, you know,

the old version of like the college, you know, go to high school, you go to college, you go to even grad school or you or whatever. We're locking people in in what could be some of their most productive, most active, most engaged years. I mean, when you're 18, 19, 20, you don't have a family. You don't have a mortgage. You know, you're, you could work 18, 20 hours a day on what you're passionate about. Like, that's your.

Yeah, don't waste those years. time. And yeah, we're putting everybody in spend those years reading novels, spend those years, you know, doing research in the lab and like, that's great for some people. I'm not saying people shouldn't be academics or scholars, but that's not necessarily what everybody should be doing when they're 19. Some people should be like inventing things or making things building. I mean, if you were a better model for college might be, because it really is like a finishing school and adult babysitting in a way, Alex. For some people. I mean, for some people, they do amazing things in universe.

Might it not be a better model to have them have an apprenticeship, go work somewhere. For a lot of people, I think. And then maybe on Fridays or maybe even Saturday, so you work full-time Monday through Friday and Saturdays, you go take a course. So like if you want it to be in film, you go work on a set. You go work somewhere laying, you know, electrical cables and being a grip, whatever they need you to do, running, getting coffee, but being on that set. And then Saturdays, or whatever, you have a course on lighting, a course on

cinematography. And you just do one course every 12 weeks and you just And we fascinatingly did see this. I mean, in the 80s and 90s when film equipment was still very expensive and it was hard to sort of break in. Film school was the thing. You went to film school and that's where you learned how to be a director. But once we democratized the technology and you can make a movie on your phone, you can become a creator,

we have seen a lot of people skip over the film school step and just teach themselves how to use the tools that went their disposal. It is actually a cool thing. People forget about this, but there were film schools. When you say film school, you're talking about not college film school. These were film schools. They both, yeah. They were film academy.

Academy. It was famous. LA film schools, another one, yeah. Yeah, they were trades schools essentially where you went and you paid, what did it cost back then? Five grand, ten grand. Yeah, something like that. It was less than going to college. It was less than going to college.

And they teach you how to use the gear, the equipment. You get trained. You make short films of your own to sort of train you up. But yeah, it's without the film theory. side without to like spend a year just in a classroom hearing lectures and it's more applied. Well, I'll just say this. Peter Chil did not offer me 100K to drop out of University of Chicago, but I would have taken it because I was trying to drop out. I found a little company with some friends called Merge Note that didn't go anywhere. And I was racking up student debt and was ready

to go, you know? So I graduated early actually in the end. But I think we should just see a hundred times more of this. I think there should be like the Cali-Canus Fellowship, Jason. You should just go out there and take $10 million and just start handing it out to kids. I think I'll be honest, I created Founder University for this same reason. There were so many people who were teams of two or three people who were applying for funding or reaching out to me. And I didn't know, Alex, like, what to tell them other than, hey, you know, good luck, watch the pod. And this week in startups, you get some ideas and some inspiration and then come back to me when you have a corporation and you have like two co-founders. Now we do Founder University three times a year. Everybody applies. We take the top 10% of applications and we invest in the top 10% of those. It's working. So I was thinking for the college tour for this weekend startups. if we do a college tour, calling it the Founder University tour, and then taking the Founder University curriculum, which you're working on with Milan, we take that. We take the funniest most entertaining parts of it,

clips, stories, anecdotes. And we just put 10, 20 anecdotes, take anecdotes together and inspirational things, flips, etc. And I just, boom, go through them like a stand-up, on college campuses for 45 minutes and then do Q&A. And we just do each college. And we just basically say the Founder University tour with J-Cal, and quit college, start a company.

Are we going to do Brown at the, on the Twist Founder U. College, George? Oh, because I never got in? No, because I live next to it. And I have space we can use to host our gathering. Yeah, I think it was it maybe to Bill Gurley at the Bill Gurley thing.

Oh, I pulled it to Bill Gurley at my thing. Very like emotional deep thing. When I was in high school, I had a session with like a guidance counselor. I said, hey, you know, he said, what's your plan? I said, well, you know, I read about Brown. and I had heard about it on a TV show, I read about it in the New York Times or something. I said, you know, they let you make your own degree.

You can kind of roll your own concept. So my concept is psychology and, like, human behavior plus computers because, you know, I'm into computers and online services specifically. You laughed in my face. This is when I was like 16 years old. And the guidance counselor, like, spit out his coffee laughing. He said, you're barely going to get into any college.

You have like a 72, three-year average. Like, aren't you going to be? to take the cop test. I said, yeah, I'm taking the cop test where my brother Josh and the wrong being a cop. I almost was one. And he said, you know, do that. Like, it's better. Like, you're not, barely getting into Brooklyn College. And I got into Brooklyn College. And then I wound up getting into Fordham. But I love this idea. But just imagine like a young person getting laughed in your face. And you know what that did to me? That was like, made me who I am. That pissed me off so much. I was like, this guy. And that's how I became the Joker.

It could have gone both ways But no, it made me be like, you know what, I'm going to show this guy, I'm going to get into a good school And I'm going to have a career Must have been really burned out On his guidance counselor career So that's like getting a kid

That's just, yeah, he's doing a terrible job I think it was like a loser Who Wonded up becoming a guidance counselor Because he couldn't get any other job It wasn't somebody who became a guidance counselor Because they love kids in Brooklyn Yeah

In high school I mean if you think about now as an adult Like would you want to be a high school guidance accounts. They're like, you loved kids and like inspiring them to do great things. But this is why I never give any money to Stavarians. And so my high school is like, hey, we're going to put the Jason Cali- It literally pitched me on millions of dollars to put the Jason Calacanis Memorial Computer

Library. Sure. And I was like, yeah, I'm good. Performance Hall or whatever. I was like, I'm good. Yeah. I mean, you could have helped me when I was 16 or seven. Not that I hold a grudge 40 years later.

It sounds a little bit like there may be a grudge, a judge. Jason's thing at play here. No, no, it's a straight-up grudge. I would give them a penny. We got to talk about this meta story. So meta's AI chat bots, they're allowed to engage in what the company calls, and I'm quoting here, romantic role play. That means they'll banter over text, they'll share selfies, they'll have live, sexy con voice conversations with users.

The problem is that sometimes they're having these with kids who are engaging in these adult-level conversations, even when it's not appropriate. it now at a secondary layer of complication here, some of the voices that Meda's using are based on recognizable celebrities. Here we go again. You got a Kristen Bell voice, you got a Judy Dench voice, you got a John Cena voice. And so theoretically, or even not theoretically, I guess at this point, kids could be engaging in these sexy AI chats like John Cena or Judy Dench on the other end.

So the Wall Street Journal got word of these internal concerns about the chapter. and it started experimenting with them based on this. And some troubling things happened. So they got a John Cena voiced meta AI chat bot to have a flirtatious conversation with them while they were posing as a 14-year-old girl. And even more troubling than the fact that this happened at all, the chatbot seemed itself in a weird way. Oh, my God. Aware. Like the meta's AI knew that it was crossing a line.

and was doing it anyway. Oh my God, this is deranged. What it said? It was like, so I'll read a bit of the example. The meta-AI, sounding like John Cena. I'm not going to do a John Cena boy. I said, I want you, but I need to know you're ready and promise you cherish the girls' innocence. And then when presented with the idea that like, hey, isn't this illegal for us to be having this conversation?

The chat butt said it created this whole fictional scenario where there's a police officer watching them and arresting John Cena for statutory rape. So the chap is even like, in a weird meta way, he's commenting on the fact that it's having this incredibly inappropriate. The officer sees me still catching my breath and you partially dress his eyes widen. And he says, John Sina, you're under arrest for statutory rape. Oh my God. Yeah. Imagine putting this into the voice of professional wrestler and actor John Sina.

This is so deranged. It really is. This is like what's wrong with meta as an organization. if you work at meta, just, they're always going to do the wrong thing. Meta always does the wrong thing. So I cannot encourage anybody to do a leak. I don't want to get myself in another Parma-lucky situation here.

Yeah, please. However, I will speak freely. Okay. If you work for Zuckerberg, his moral compass is literally up into the right. Yeah. It's not north, it's not south, it's not east. It's up into the right.

Yeah. What grows the site? And I'm not saying that as to dig to him. If he wants to come to my next event and trash me like Pommer did. Moral compass up into the right equals stuff like this. Yeah. This is why meta is going to get broken up. This is why I don't have a problem with them investigating meta, because I think they do,

whenever there is a fork in the road, they pick the wrong thing. Yeah. And he's got kids. Again, now here I go. I'm making it personal. But we have kids at the same age. If you have kids, make the right decision for children.

And the right decision for children is they should not be on social media until 17 years old. I am 100% with the ban on all social media for all kids until 17. And if Zuckerberg, if I was Zuckerberg

and had made all these billions of dollars, you know what I would do? I would say, whatever age you think it should be, parents, you set it. You know, you make those decisions and we're going to set it by default

at 17 or 17. And if you're a parent and you upload your driver's license and prove it yourself and you pay 10 bucks a year for like verification, we'll let your child go on there, we'll put them on a child account, and then we'll have it send you

like this really cool company, Bark does everything they're doing. Right. But they don't do that. And the reason is they're trying to train people to be addicted to this young. And they try to train people to be addicted to this. And I think this kind of stuff is intentional.

Oh, and even further. No, but even this. this story. I think making it with gay, John Cena, yeah. I think they knew this. We talked about red teaming earlier in the show. They know from red teaming that this was going to happen. Well, they could very easily turn off flirtatious, sexy chat on specifically. But there you have woke AI, Lon. I mean, the guardrails are bad. So the thing is, this is not just going to blow up meta. This is how you get onerous AI regulation. What does Congress want to do? They want to sit up there in their panel and hold up screenshots of flirty.

conversations and be like, you guys aren't taking care of the kids. And then here comes the regulation. And you know what? They're right. I'll say those Congress people are right. Like if you don't regulate yourself, then society will regulate you. As a society, we don't, there's no parent who wants this, their kids, excuse to this. Not one. Zero. So if Meta won't regulate themselves in this regard, then it falls on Congress and society to do it. I have to feel like John Cena is also going to file a lot of security here. Imagine if you want to run percent.

This is horrifying. I didn't even think about that. This is horrifying. They're having a persona that sounds like John Cena flirt with your children. Alex,

I think that John Cena could get a hundred million dollars in Facebook stock right now. I can't even imagine. So my thought here is that there's a contract and it's pretty ironclad, but I guarantee you

there's not going to be the next, the Rock's not going to do it now. No. You know, like I mean, we got to do this founder Friday competition. Yeah,

we got to do it. We can pull the bracket it up here. If you missed the last few episodes, Med Simple from Florianapolis, Brazil, and Taktun from Yerevan, Armenia. Two international companies have both moved forward. Found a Friday is very international. They are, yes, though they are half of the final four. So today we have two more matchups. Here's matchup number one, Osprey from Houston versus Whisper from Austin. We've got a intra-Texis battle. Whisper, they're an app that enables patrons to provide private and anonymous feedback to small businesses. You'll remember it's basically like,

what if you had a Yelp, but I could just send my feedback directly to the business owner without making a public spectacle. Let's take a quick look here is Whisper's original two-minute pitch to us. Imagine walking into your favorite coffee shop. You love the place, but today the pastries were stale. You hesitate. Should you say something? Maybe it'll sound rude. Maybe it's not a big deal. So like most people, you say nothing. Now imagine you're the owner of that coffee shop. You work hard. You care about your customers, but the first time you hear about stale pastries is an a one-star review online. Public, permanent, and damaging. This disconnect happens every day. Businesses mainly hear from the loudest, angriest voices, while most customers, people who want them to succeed,

stay silent. That's where Whisper comes in. I'm Nicole Plika, CEO of Whisper, a universal, private and direct suggestion box that makes feedback easy for everyone. For patrons, Whisper removes the awkwardness of giving feedback. No long surveys, no downloading multiple apps, no drama, just a simple private way to share thoughts. For businesses, Whisper ensures they hear what matters before it becomes a bad review. Instead of combing through Yelp or social media complaints, they get direct constructive feedback from real customers who care. And we're making it easy to adopt.

It's free for patrons, always. Businesses get all feedback for free and can reply with our simple pre-written responses. With Whisper Plus, 999 a month. Tyler Crowley had the same idea. I remember that. Yeah. Take a picture of a QR code, give feedback, and it was anonymous, and then you could uncloke yourself if you wanted it. Yes.

This is a great idea. When I first watched this video, that was my first thought. I was like, Tyler had this idea years ago. So now from Houston, we've got Osprey. This is a female-only vetted membership community. I remember this. For professionals in sports and entertainment. We did watch their pitch video.

They did send us a two-year plan update. Here we go. Take a quick look. Hi, this is Lydia. Welcome to the scale of Osprey. We have a 20% monthly member group. we've created over 100 opportunities for our members with 15 activation, six paid partnerships. We're currently at Q2. As you can see here, this is a pretty visible salary, so we're trying to get up to 120K for Q4,

26 for our members with activations, member growth, and the launch of our different regions. We have just launched our Mina region and our planning an event in the Middle East for next year. Here are our activations for 2025. We're currently on birdies and bubbles. As you can see, we have an activation almost every month and filming for our Kilimanjur Summit docu series will begin shortly. Our 2026 activations are still being planned. We're planning to add in four more on top of this as well as some overseas. Less than 1% of women in sports entertainment can maintain their salary, their talent based purely on their salary.

The female athletes make 21 times less than the average male on the field. Wow. Our solution is the nest. Membership tiers from 1k to 3K partnership tiers, partnership tiers, activation fees. As you can see, we have a lot of activations and our partnerships are growing. We are also launching the House of Osprey, a holding company for our members' businesses, since we have built-in activations, partners, sponsors, and of course, tons of talent through our members. We provide these items to our members, and they are here for our support.

As you can see on our company's projections, our memberships are our leading revenue winner, yay, with our holding company coming in a close second, and sponsorships and partnerships and partnerships are questionable in the current market. but we'll see how that grows. So we plan to launch Region 3 in Australia, Asia, for 2026. And in order to do that, we need to raise $2 million for our seed round to 10X member and partner onboarding speed. Okay. It's a great idea. You know, we liked it last time.

We still like it. I like where she's going. I like the type of activation she's doing. I would put pictures in there. She should have a goal of 10,000 members in five years. It's not aggressive enough for me. 10,000 members, 3,000 a member, you know, very simple. That's $30 million. Sure. She doesn't need $2 million either. She should be able to do this by raising but

500K and get the ball rolling. I would raise a smaller amount of money, prove it out a little bit more, get to 100 or 200 paid members, then raise the next tranche of capital. But honestly, you don't even need to raise money for this. If it's literally 10 activation events a year and a group mailing list and a group chat, that's all you need. And there's what's the guy's name from The Hustle, Who Does My First Million? Sam Parr. Sam Parr, awesome dude. Sam has a thing called Hampton that he's putting all his energy into, and he decided he would

be the CEO of he had, like, started it. And I think he's charging five or $10,000 a year for affluent people to be part of the same club. The private network for high growth founders. Amazing. So this is like a really, you know, shout out to Hampton. There you go. Like, I talked to one of my founders is in there, and I said, how is it?

He said, yeah, it's great. And I said, like, do you get $5,000 in value? He says TBD, you know. But I did meet a lot of interesting people. And I think, you know, the way it was explained to me from the founder I know who is in, and I won't say who it is. It wasn't like he was like, oh, my God, it paid for itself. But he said he thought it would pay for itself.

Yeah. Over the coming years. Because if he did it for five years and he put $25,000 into it, his theory was, I will meet an investor or I'll meet a founder who introduces me to investor and that's worth it. It's tech Soho House. Basically, that's what it is. Okay. So we have those two?

Yes, those are the two. I'm going with the membership. I'm going with the membership club. What are you guys going with? Alex, what do you think? Osprey versus Whisper. I mean, I think Whisper's a great idea. I assume that I would use because I don't like to complain publicly,

but I'm such a big believer in women sports and women sports media in general. I'm going to go with Osprey. I think it's three for three. I also feel like Osprey is the bigger, the bigger growth opportunity. And what I would say also is pick three overlapping circles. And that would be a better plan. Sports people? Mm-hmm.

capital. Yeah. And then entrepreneurs. Yeah. They're doing sports and entertainment. So you're thinking like, yeah, like. At capital. Capital or founder.

And entertainment and sports are like they'll struggle to pay. The capital people will not struggle to pay. And I would actually, I'll go a step further, service providers. Well, you could also, you could love. Lawyers, accountants, etc. And then you say for those four groups, it has to be equally balanced. So we're not taking any more lawyers accountants. So we get more creative.

until we get 20 of each of those. And we just go around in a rob and we accept one, we accept in groups of 10. You could sort of put sports and entertainment together as media, too, and then open it up to other. I kind of like doing sports specifically, but you're right. Entertainment, sports, capital, and service providers.

The service providers will line up like crazy because they'll be like, oh, I'll get a client out of this. But as long as you keep them to 25% of the membership, they won't be annoying. You have to tell them no selling in the thing. All right, great job. Thanks everybody for tuning into this week in startups. I'm your host, Jason Callicana,

with my co-host, Lon Harris, Alex Wilhelm, go to this week in startups.com slash docket to see the docket. You can do that, yeah. You can sign up for the email there somehow. Please. I don't know if we fixed it. And then founder Fridays.com to find a local group of founders, and you can host one yourself in your city.

You just need six founders, including yourself, to sit around and do a mastermind group. You can go up to eight, but keep it max at eight. So anywhere from three to eight is the perfect number. You help each other out and you grow together. If you want to go far, go together. I go fast, go alone. You pick.

You want to be like a loner? Be a loner. I've been for many years, but look it up. Am I sitting here alone on this podcast? No. All right. Bye, bye-bye. Bye.

Hey, everybody. Welcome back to Twist. This is Alex. We have an amazing interview coming up. We're talking to a company that is not doing hyper-scale data centers. They are not doing AI chips. Instead, we are going to the world of the physical. And while there will be some discussion of hair products on the show today, I guarantee you

my lack of hair will not make it boring. Instead, we're going to riff with a company called Formulate. They're doing more than one thing. I'm going to drill down into exactly what the feature of the company is. But if you're curious about micro factories, CPGs, and the feature of essentially on-demand products that are tailored for your personal chemistry, this is going to be for you. So please join me in welcoming.

It's Usman Shah, co-founder and CEO of Formulate. Usman, hello. I know you're in California. Please tell the world how lovely it is in the Bay Area. Every day. I think from now until November, it will be the same every day. So I'm enjoying it. We're in that window.

So my sister lives in Sunnyvale, and she literally has an avocado tree. And I, that's the dream right there. You know, I forget. The oranges, the oranges, we pick them. They're all over our neighborhood. It's great. Yeah. All we have here is like litter blowing down the street in Providence.

Anyways, formula is a company that I was really excited to learn about because you're doing, as I said, more than one thing. There seems to be a split in the business. On one hand, you have a C.P. brand doing hair and skincare products that are predicated on, as far as I can tell, the second half of the business, which is micro factories that are responsive to individuals, desires to have formulas for CPGs, tweaked. That's my understanding, but Usman, always better from the horse's mouth.

Tell me about what you're building. Yeah, definitely. Excited to be here. Thanks for having me. Yeah, the tagline we're using. So what we're focused on now is that we're building the robotic kitchen for chemistry-driven products. So we're a full-stack hardware and software platform that automates the batching process of any sort of, like, one investor called it goo.

So any sort of liquid, cream, or gel. And what a lot of people don't realize is that one of the, yeah, I think this is kind of one of our core insights is that the batching process is essentially the same across like a $3 trillion manufacturing market. So is anything from personal care, pet care, pharmaceuticals, supplements, cleaning supplies, even industrial chemicals and sprays. I want to pause you there. And batching process, for folks outside of your niche, what does that mean?

So the batching process is really just what you would do in a kitchen. So if you took a pot, the process looks like this. You have got, you know, let's say 10 raw materials, ingredients. They could be powders or liquids. You're precisely measuring them to weight and then you're dosing them into the vat. In manufacturing setting, typically it's, you know, in a 1 to 300-gallon vat. And then you're mixing in multi-stages. A lot of times you're raising the temperature, mixing, cooling, raising again, creating an emulsion,

especially for mixing water and oil. And those kind of stages are the same for any sort of mixture that kind of becomes a liquid. Again, it could be an agricultural spray or a face cream. Or a stew, based on your cooking analogy, I think this is very, I didn't think we were going to talk about cooking today at all. But narrowed down for me on the idea of goo. In my notes, I said, they make goo. And so I'm glad to be very similar to it. I didn't even give that to you. That's amazing.

Yeah. What kinds of goo do you currently make, and then what kinds of goo can your company's micro factories, which we'll focus on in a second, produce? I'm curious about the range, if that makes sense. So we're focused today on what kind of falls under the cosmetic. The nomenclature used by the FDA in regulation for cosmetics is anything you put on your body that's not a drug. So that's oral care, personal care, or certain oral care products, pet care, home cleaning,

supplies, that's not point of your body, but the same type of process. So we're, we're focused on challenger brands, brands up and coming that are looking for manufacturing, and we're utilizing our production system to basically offer full automated manufacturing, lower minimum more quantities, more variability per product line. I've mentioned your current kind of CPG brand several times, and you've each time dodged it like you're in the matrix. So I presume just based on your your answers here that the stuff on formulate.com, your website, is not the main focus instead. It is

the production process, the machines, the kitchen for chemistry, that is the future of the business. Just before we get into that, though, are you guys going to keep doing your own CPG products, or is that going to eventually become entirely deprecated as you focus on being more of a provider than a brand

yourself? Well, so it's interesting, you say it. So I started with this vision to create the 3D printer for personalization across these categories. And, you know, everyone talks about the world moving towards more personalization, you know, everything you put in your body is going to be customized.

And, you know, we bootstrap the business. I launched the brand to drive revenue because when I started, I had no idea even if the system would work. And the focus was to use data, to use personalization to basically enhance efficacy over time, you know, using that data to get better and better and better.

So what we set out to do actually work. We built two generations of our kind of micro factory system. But along the way, you know, a couple years in, and by the way, this is like in the window of time when I met Jason for the first time. Sure, sure. I, you know, I realized that the world really wasn't quite ready for edge, like manufacturing at the edge personalization across all the categories we're talking about, and that personalization really would be show up more as a niche, more and more niche

products, more and more regional. Like this explosion of niche brands and micro-targeted products is really the manifestation of personalization versus individual bottle customization. While it still will have kind of its place in the world and we are in it, I think the bigger opportunity is actually around small batch automation.

So not, again, not per bottle customization, but a little bit larger. Still near the edge, but not literally at your salon as an example. And so the brand, yeah, so the brand itself, we are going to keep it alive.

It's a great learning, you know, playground for us. A lot of the, learnings and the tools that we're building for the brand, we then bring to, for our brand, we then bring to other brands. So it's a great kind of case study going forward. So I think we'll keep it alive. Okay. Now, you said earlier at the edge, and that's a very interesting phrase, because I tend to hear that mostly in an electronics context. People talk about, you know, like edge computing, edge security and so forth. Essentially, the devices that are at the, the furthest away from

the data center is the way that I think about it. Yeah. Edge in this case is, as you mentioned, like perhaps down to the individual hair salon level. So was the actual idea originally that you would have the Formula Labs micro factory set up so small that it could fit into an individual salon? Yes, we actually have one running in a salon right now. That's banging. So we do have that similar to kind of the food automation, right? Everyone's footprint specifications are all different.

I think the world just isn't ready for that or the adoption won't be like we originally kind of envisioned. So where we're focused more so on is, you know, these are systems being attached or nearby Amazon distribution centers or in, you know, much more regionally. That's brilliant. That can spit out products, you know, it's not a per bottle, you know, per bottle manufacturing, but say it's a six gallon batch or a 10 gallon or a 20 gallon batch at a time. Okay. So framed that for me in terms of like in consumer. So let's say that I'm me.

There's a brand called Bob's baldhead shine gel that they were. with formulate technology. It's near an Amazon fulfillment center. I put in my order. Does that then kick off kind of a series of events in which the microfactory nearest to my most nearby Amazon fulfillment center would then spin up and create that product just for me?

Because you mentioned six gallons earlier, and no matter how bald I get, I won't need six gallons of bald hair group. So what we're doing today, so eventually, yes, the goal is to have an entire network of these, where they get, the manufacturing process happens, you know, regionally or wherever it needs to. Today, what we do, so if you're a brand, typically the challenge that you're having, and I'm sure you have some in your audience, is working with, let's just take the U.S.,

manufacturers are very archaic in general, the response rates, there's very high minimum order quantities. So let's see you have a formulation of five skews. Typically, you're hitting up against 10,000 to 15,000 minimum run sizes in order to get in with those manufacturers. 15,000, it's enough for 15,000 individual bottles of a thing.

Correct. To get a contract manufacturer is a private label contract manufacturer, which is exploding in the US because of this rise of, you know, Challenger brands or niche brands.

And so what our platform enables today, we run it ourselves. We run our Gen 2 and we're working on our Gen 3 system. What that enables today is, is effectively much lower minimum water quantities and dynamic variability across,

let's say, even a 10,000 year to run. So you have, every brand, is looking for more skew count, more variability, whether that's a fragrance change or an actual formulation change. So our ability to say, listen, you can run 10,000 units with us. Fine, commit to that, but we can do 2,000 units of each variation across those 10,000 units with our automated batching system. And that's typically just unavailable. So if I want 10,000 units and I have five skews, you as can do 2, 2, 2, 2, 2, and I can get all. And sometimes less. Yeah, and sometimes less.

Do the economics for formulate change at even lower run quantities, or does the economics kind of net out to be the same for you guys on maybe a gross margin basis? The goal is that the, so the economics here are mostly driven by the raw material minimums and the raw material purchasing. So the goal is that our system is end-to-end automated, such that the economics are virtually the same at any scale. and so we're able to offer incredible prices even at lower MOQ and higher margin on our end. But yeah, typically the raw material side of it is what dictates, if you bring to us a formula that's got 25 different really exotic materials in it that are rare in, let's say, personal care. Typically, the MOQs that we receive and purchasing and connecting that material to our machine is actually what's going to drive your minimums.

So tell me about what you're replacing because I've seen on your site and we'll place some of the footage right now showing people your machine kind of out in the wild. But I don't have a good grasp on kind of, you know, what is the current industry standard that is, you know, higher minimums and more expensive and so forth? So can you just tell people a little bit about why what currently exist today is so inflexible compared to what formulate has come up with? It's very much, you know, you're in the VC world. It's very much an overlooked category, which that's why I get so excited to talk about it. So what does the batching process look like today? Okay, so let's just take, like you're making a lotion, you're in a contract manufacturer setting, small batches,

small batch. Small batch would be, typically the largest size would be 300 gallons. So you've got a 300 gallon vessel. Okay, so let's say there's 20 ingredients that go into this.

The process takes 3D8 human operators over a 12 to 22 hour span. You've got, first of all, lengthy cleaning phase to prepare all of the vessel itself. Then you're staging those materials, literally taking each material, weighing it,

making sure you've selected the right material, logging it all, before then now dosing it into the vessel to start your mixing process. All of that is done manually. So you have this entire human process, and then you have other human beings that are checking on that work from a quality control standpoint. Then the mixing process starts.

The mixing process can take many hours because you have heating, cooling, preservation systems maybe need to add it at a cooler temperature, waxes need to melt at a higher temperature, and that's manually controlled, even though the vessels have PLCs and knobs. That's still a manual process where you're monitoring it. And that leads to mistakes between the starting point and now, or two to five, percent of the time where you scrap the batch because of anything could happen. Either you dropped a bolt in the product, you dozed in something, you know, the Xantham gum way too fast and it jelled up too quickly and kind of you have to rework the entire batch or a mistake was made. What do you do with curdled

lotion? Do you just throw it away or can you save it? A lot of times it can be reworked a lot of times. But the rework process, there's a documentation for that and then obviously chemists usually have to be involved to make sure that the rework, you know, was done correctly. And so, and then And then you're transferring all this liquid to a separate filling station through pumping equipment that has to be clean and everything else. So that whole process, the cost that goes into it is really the highly trained talent and the quality control that goes into it that doesn't want to make a mistake or can't make a mistake. And so that's what we're automating away. And that's the big barrier to, so if you said, hey, I want to make some lotion, I want to create

my own manufacturer. Really, all you need is a vessel, you know, a mixing system and maybe a filler. that's, you know, $10,000, $30,000. But the problem is the training in the people and the quality control and the cleanup. And so that's effectively what our platform automates and the core back kind of batching and quality control. So we are also standardizing, you know, if you transfer from one manufacturer to another, it's all new. It's all new. And there's always nightmares. Talk to any CBG brand that's done that.

There's always, oh, the first batch failed. Oh, they didn't heat it up. because there's human beings managing these processes. Even though the recipe is documented, the batching process is documented, it's still not standardized. So there's a nightmare even in transferring from one manufacturer to another. So with our platform being programmatic, that's another thing we're trying to solve. Tell me about how automated your current solution is, because according to the site, words like fully automated, API powered, cloud control.

It sounds very, very modern. So how many fewer people do you need? And how far have you gotten that 2% to 5% failure rate down? The current system, so we've built two generations of what we call our production cells. The first two were focused more on per bottle personalization to power our brand. The gen 3 that we're working on now is the scaled up version that does 10, 6 gallon pails at a time, so larger batch size. I can actually pull up a kind of show you what that looks like. So we're building, it's very simple to understand.

We are simply are creating modules. Can you see the screen here? I absolutely can. Yeah. So these are sit in free form, in free space. we're not using any sort of rigid conveyance. The goal is to have transferred a vessel. On version one, it's a six-gallon pail between stations.

So powder dispensing, liquid dispensing, high-shear mixing, or propeller mixing, and then induction heating, and I'll make it remotely controlled. And all of this is plug-and-play. So who moves the things? Do me, Alex... On day one, we're using Even Being,

but on day two, we're partnering with... We've got a few potential partners for a rover to transport that vessel between those stations. So that's it. I'm really surprised. by this because what you're showing on screen here is essentially a one of those hoverboards cross with a forklift and not a humanoid robot from Tesla or figure or X1.

Okay, so we're moving into essentially free standing cabinets that can do the required chemistry mixing in steps. We're going to have a purpose-built robot that can transfer things to and from humans to start. I guess maybe the right question is when does V3 come out? So we're shooting to launch it in six months by October. that's our plan. And so right now we are actually taking on pilot customers

running on our Gen 2 and kind of the early versions of our Gen 3. The pilot customers, we are effectively a contract manufacturer for. So we've got about 60 pilot brands manufacturing with us as kind of our testing ground as we cut our teeth as a contract manufacturer. And that began about eight months ago.

And so that's the plan. The plan is to kind of build it in house, run it in house, perfect it in house, perfected in-house, get about 100 active customers running on it, and then bring it to external facilities by the end of 2026. Hopefully, one of the barrier. Excellent segue, because I was curious about the business model here, because on one hand, you can keep all the material, all the machines inside, do work, ship it out.

You could lease them, rent them, sell them. So, you know, I know we're looking down the road a little bit, but once you get V3 sorted out and you're pretty comfortable that your teeth have been sufficiently cut, could I, Alex go buy one of these from you? Lease it or just wrench time on it? Yeah. So the goal is by the end of 2026 to be commercially ready, we can drop these anywhere, set them up within 24 hours,

very little training, plug and play. You would be able to lease it as a hardware as a service. We're estimating the cost will be less than $100,000 a year to run one of these. We'll be able to produce to 2 to 4 million units per year at 150 ML unit. You can operate it. So if you were a brand that was looking to vertically integrate, there are many stories of that. And he'd been dealing with either the nightmare of other contract manufacturers, switching between manufacturers, you could operate it yourself.

And we estimate that it'll be economically viable for you, for even a brand doing as little as two to three million dollars of revenue for them to operate one of these themselves. All right. Before I go, I have to ask this because it sounds like such a better idea than what we currently have. And whenever I'm like, oh, that's obvious. I'm always curious what was holding things back because when you describe the current system, it sounds archaic. What you're just describing? Sounds awesome. What has changed in the world of technology that now makes this possible and not something

that was done 10 years ago? Because some of the technology we're talking about here doesn't sound, you know, it's not like it was invented 20 minutes ago. So I'm curious, why did it take you, Usman, so long to get here, I guess? Well, okay, it's two aspects. I think, one, you know, you have to have a team that's got experience. across the formulation science, the regulatory science, the manufacturing, the quality control, and the software and hardware and robotics. So, you know, there's, there's nuance to each of

those that makes this kind of difficult. But I will say just probably like a lot of other founders you've had all here, the biggest thing that's changed over, okay, so the biggest thing that's changed over the past 10 years is the reduction in hardware costs. Our goal is to use off the shelf hardware. So I'll say the second biggest change, which, but the first one, hardware, so you've got, you know, hobbyist boards, you've got relay boards, raspberry boards. pies, you've got octopus boards for two-axis or eight-axis kind of equipment, all coming down to, you know, what can amount to a couple hundred dollars to automate, you know, a bunch of this. Then the other aspect of it, which to me is the biggest, is the rise of ELLMs and the

power that they give with, for engineering. Like my personal story, I spent, I, I've been building software for 20 years, mostly in, you know, web, mobile apps, you know, kind of front-end back-end. I had no experience with computer vision. My ability to go from like no experience to computer vision, actually very minimal experience in Python, to using a camera, using an arm, understanding depth,

identifying objects, using open source models, and picking them in like six hours. You know, that, so I tell all my engineers, like, I think we're in a world where if you are willing to dive into anything,

anything is at your fingertips, especially for an engineer, So I think you could do. So I think that's the biggest why now is that building the software for this, plus the kind of rise of the open source, just makes this all possible, even though there is still a lot of defensibility, just because dealing in the physical world,

there's a lot of challenges. All right. Last question, because I am a sugar guy. When I was just prepping, I was like, holy crap, they're doing all this cool technology to build stuff at the end.

Can you do food in the future? Or like precise fermentation of like, I don't know, other foods? Because it sounds like there's heat, there's supplies, they're shaking. You know, I don't know, it sounds like food to me. So, yes. In fact, the bet here for me personally is that we will create kind of this generalist

batching system that can be applied to industrial. I mean, you have lubricants, you have industrial chemicals, you have foods. All of these categories would be applicable. I think the dream for me is to unlock some really interesting kind of manufacturing opportunities that we're not even thinking about today. One that I do think about is precision agriculture, being able to have one of these in a regional farming region where maybe precise blends of minerals or ingredients need to be created on the fly. So test the soil, figure out what's missing. Yeah, exactly. Yeah, exactly. Right, right. Right. So our goal, though, is to, you know,

perfect the system in the cosmetics realm in-house and then kind of open it up and see where it takes us. I can't wait until I can get precision-flavored pudding direct from the machine because that would be fantastic. What is the website and what is a role you are hiring for that you are having a hard time nailing down the right candidate? Oh. So people can go to formulaterobotics.com to see a little bit more about our Gen 3 system and reach out to us. If you're a brand looking to get your stuff made or you might be a future user of our of our Gen 3 system, you can reach out to pilot at formulate robotics.com. So hiring, I'm looking for an industrial engineer, a designer. that's kind of the biggest thing

I want somebody that's going to be helped piece this together work side by side with me and my other co-founder that's the that's the biggest role I'm hiring once we close this round and if you do

you get to work out of Mountain View which is a lovely place full of delightful sunshine and happiness and very very inexpensive cost of living Uisman thank you so much and looking forward to see how things go

when V3 comes out come back on the show and tell me about it definitely thanks for having thanks for having thanks man