This Week in Startups - Drones, Deepfakes, and Disruptors with Lon Harris, Remento and Cuby | E2058

Starting point is 00:00:00 What they've done here is they're turning memories into a keepsake book, no writing required, preserve a loved one's story the easy way. Remento turns their spoken words into a personalized keepsake book of their stories. Now what they've done is with language models, you can select prompts that you want your parent or your grandparent or whoever to answer or your family. They record responses in their app. Then they create a book based on it. And then you can scan the book to listen to whatever QR code. Is there something a little creepy about it? I don't know.

Starting point is 00:00:32 It's definitely better than using AI to like recreate your dead loved ones. Like I know there's that thing too where it's like, feed their voice into our AI and you could chat with your grandma. And like that really is troublesome to me. Like this at least your- Why is that troublesome to you? It's just creepy or you have a concern about it. It is creepy.

Starting point is 00:00:50 I don't have like a concern like it's going to ruin the world or anything. But I think it's if there is something a little ghoulish about it, because I don't believe that AI is really able to, like I think it's a lot. lie. People are supposed to die and then you remember them and that's how it works. And this idea that we need to like hold on to them forever. You're raw dogging. Grief. Yes, I'm grotting. Grief. Yes, exactly. I want to feel it. I don't need to augment

Starting point is 00:01:16 it. I respect it. This weekend startups is brought to you by Squarespace. Turn your idea into a new website. Go to Squarespace.com slash twist for a free trial. When you're ready to launch, use offer code twist to save 10% off your first purchase of a website or domain. Lemon.io. Hire pre-vetted remote developers. Get 15% off your first four weeks of developer time at lemon.com slash twist. And Oracle. Oracle Cloud Infrastructure or OCI is a single platform for your infrastructure, database, application development, and AI needs. Save up to 50% off on your cloud bill at Oracle.com slash twist. All right, everybody, welcome back to this week in Startups. It's a Tuesday.

Starting point is 00:02:01 It's a Tuesday. And we've got a great interview with the founder on the back half of the show. But at the front, we're going to do some news and the original news reader from This Week in Startups is back. My guy, Juan Harris. How are you doing, pal? Good, good. But many years, 15, 16 years later, I'm back to do the news.

Starting point is 00:02:18 I think 14 years, 2,000 episodes. Boy, my arms tired. Jeez. What a, what a run. What an impressive run. It's one of those very interesting things about podcast. You know, if you love doing it, it becomes like exercise. I consider, you know, doing this week in startups a way for me to just really connect with people and to stay in the flow of the news and the zeitgeist.

Starting point is 00:02:40 And then just build a community of people who are really into startups and tech. Of course, all in has made that go supernova as well. Well, those people, they're into startups and tech and other topics. Let's say there's a few other topics. Yeah, there's a few international relations. It gets a decent amount of traction. And so what was your take on seeing Sacks named as the AI? Wow.

Starting point is 00:03:05 What, yeah, I've never had any kind of this kind of proximity to power. Like working on a show with a guy who's in the White House. It's very, I'm going to try not to rub him the wrong way anymore. Well, you know, it's great. David is an awesome executive. So I'm very excited. I don't have anything to add to it other than, I'm so proud of him, and I just love the fact that, you know, if Trump is the president,

Starting point is 00:03:31 you know, people know, wasn't exactly my favorite candidate. I'm just glad that my friends are around him and hopefully steering the country towards smaller government, less regulation, and just more efficiency, which, hey, who can get behind that? Listen, I don't agree with David Sachs on a whole lot, but I feel better having him speaking into President Trump's year than some of the other people he's going to be surrounded with, for sure. You know, I think this is such a great point you're making. We don't all have to agree on everything in politics, but you can know a person and trust them and think highly of them and know they're effective and, you know, have good intent. That's really like one of the great things I think about what we're going to see in 2025 is hopefully a great run there. God, there's so much news going on. It feels like the world is spinning very fast right now. Normally, this is when everything kind of slows down until the next year.

Starting point is 00:04:25 Not 2024, though. We're barreling through. I think this is going to be like another frenetic, you know, a couple of years, which was a hallmark of the first Trump presidency. I think it's going to be frenetic as well. Yeah. I mean, aside from anything practical that he actually does once he gets into office, like there's all that stuff.

Starting point is 00:04:46 But there's also just, he is such a divisive figure. He divides people and they create, and I'm not saying it's all his fault personified, but he creates this environment where it is really like combative. And I feel like we did have a few years sort of a little bit off from that and we're jumping right back. It's a good observation. I did say, I don't know if it caught Trump on the Sunday shows, but I listened to the first half of his interview, it was his first beat the press interview. And I was pleased that he said a lot of the things that I was concerned about the deportation of 15 million people. He said, like, listen, we're going to start with the criminals and get that done. You know, it's like, okay, well, everybody agrees on that.

Starting point is 00:05:25 And then he said, yeah, I'm not doing retribution. That's not my interest. My retribution is going to be the performance of the country. So I mean, you know, Donald Trump just says things. So who, who knows? We'll have to wait and see. It's a, you can't, you can't trust like, well, he said he's not going to do retribution. So I take him at his word. Like you, right, Donald Trump is word. Or you can't, uh, or the other way around when they say they're going to deport 15 million. and then he says, well, maybe we're just do the 500,000 criminal. Absolutely. Right.

Starting point is 00:05:51 And I think this is an important thing. Exactly. Like, I think what gives Trump so much of his power is that whether you support him or don't, at this point, everyone knows he just says things. And so you can pick and choose what things he says that you like. And that's what Donald Trump stands for in your mind. But it could be different for everybody. Like, I was just reading this, like, you know, they interviewed voters who voted for Trump. like why did you vote for Trump? And it's all over that. He's going to fight insurance companies.

Starting point is 00:06:21 He's going to end homelessness. He's going to end all these foreign wars. Crypto. Yeah, he's going to support. He's literally said he's in favor of and against everything at this point. Like TikTok is another great example where he ran now on I'm going to save TikTok, but the TikTok ban started in the Trump administration. That's not a Biden thing. Here we go. I mean, so what's going to, I mean, it's really interesting on a business. basis. I have friends who are shareholders in TikTok, you know, bite dance, the parent company. Sure. How do you even, you know, know, know what to do with your shares? He's against it. He wants to ban it or divest it or keep it. I mean, it's just hard to keep track of, you know, a little bit of chaos right now in New Jersey. I hate to get all conspiracy there here. Well, I mean, this isn't even a

Starting point is 00:07:08 theory. Just look up. Yeah, there have been a bunch of drones, large ones, in fact, in New Jersey. Tell us what's going on with this story line. Yeah, so starting in... It is a technology-related story, yeah. Yeah, so starting in mid-November, people in central and northern New Jersey, specifically Morris County is where it started. They started noticing a large number of drones in the skies at night over their heads. You can see from the video, these are larger than the usual commercial.

Starting point is 00:07:38 We've all seen drones in our neighborhood that hobbyists are flying around, but these seem much larger in scale than those kind of hobbyists. drones. So that's raising concerns. But also, they seem to be flying around sensitive areas, military installations, President Trump's Bedminster Golf Course, other places that, it seems a little suspicious, why would there be a bunch of drones swarming around these places at night? So earlier this month, the mayors of 21 towns across New Jersey sent a letter to Governor Phil Murphy calling for an investigation into the drones, specifically their origin, their purpose, and like, are these complying with all of our local regulations? And if there aren't regulations

Starting point is 00:08:21 against this, why not? It's kind of like, shouldn't this be illegal? Like, it's a very murky sort of area right now. It is very murky. Yeah. And I think that's what a lot of the mayors were concerned about. So on Monday, Murphy gave a press conference, Governor Murphy, and he said, they still don't really have answers. They don't know whose drones these are, why they're flying around. Murphy basically said he understands why he's like, I hear your frustration. I get it. So we was doing that kind of political thing. He did say the drones seem to be very sophisticated and that they have the ability to when we start looking at them, they can go dark. Turn the lights off, the safety lights. Okay. You know, so it's harder for us to get a clear look because

Starting point is 00:09:07 once they realize they've been detected, they can sort of shut themselves down or go dark. But so far, Governor Murphy said, Homeland Security, the FBI, the Secret Service, the New Jersey State Police, they're all looking into it. And none of them say there's any immediate reason for public concern. Like, these don't seem to be a threat to public safety. They're not going to start spewing out Joker toxin or whatever any second. But it's hard to know how certain they really are if we don't know who sent them and what they're doing there. The FBIAA has also imposed flight restrictions on the area pending further investigation. Some witnesses are saying they've also seen a fixed-wing aircraft flying near the drones,

Starting point is 00:09:48 but we do not have confirmation. That is just witness. Okay. It sounds like a military group or some commercial venture is doing this. Like they're playing with their drones. The reason I say that is if they're larger, those are very expensive. You're talking a $5, $10,000 quadcopters, those big large commercial ones. Those ones are also customizable or self-built or assembled in many cases, as opposed to a DGI one that I ordered off of Amazon I played with.

Starting point is 00:10:16 And I was shocked that you could just fly this thing in Manhattan. I was in Manhattan flying it in between buildings, looking at rooftops. And then I was like, I didn't think that was allowed. How was this even possible? My friend had one and he was having a good time with it. And so I just bought one and started playing with it with my daughters in the summer. It was fun to do. We were flying it all over the piers on the west side, making videos.

Starting point is 00:10:35 it's tremendous fun. They feel relatively safe for something that's flying above you because you're so stable now. They will not slam into an object as one example. But you do not have complete control over your drone. It's not allowed to hit things. It's not allowed to turn its lights off. So that leads me to believe that this ability to turn the lights on and off would be something that a more military or handmade drone would be able to do. This week in startups is brought to you by our friends at Squarespace, the all-in-one platform for you to build a beautiful website and for you to grow your business. Whether you're starting out or you're scaling up, Squarespace is going to make it easy to create an incredible website. That engages your

Starting point is 00:11:20 audience, allows you to sell whatever you need to sell products, content, or even your time, all in one place. And here's what makes Squarespace a game changer. First, their new design intelligence feature. This combines two decades of design expertise with cutting edge AI to help you craft a stunning personalized website in just minutes. There'll be no other website like it. Basically, it's Squarespace putting a high-end designer right at your fingertips. And second, Squarespace payments. This makes managing your payments seamless. With fast onboarding and support for popular methods like Apple Pay or Quarna, you'll be up and running in no time at all and ready to take payments however your customers prefer. And that's what you want to do. Meet your customers

Starting point is 00:12:03 where they are. Finally, built-in analytics. Hey, if you can measure it, you can manage it. And the built-in analytics at Squarespace are second to none. It's everything you need to scale. From understanding your sources of traffic to understanding sales trends, Squarespace is going to deliver insights that help you drive growth, all fully integrated, all for the same price. Here's your call to action. Squarespace.com slash twist for a free trial. And when you're ready to launch, go to Squarespace.com slash twist to get 10% off your first website or domain purchase. That's squarespace.com slash twist. Make sure you go to that URL so they know you listen to the pod. The ones that are made for retail use, let's say, as opposed to business insider use or military use, those commercial

Starting point is 00:12:45 drones as opposed to the retail ones, consumer ones, I think are very different. In fact, you're not allowed to fly my drone in certain areas. So if I wanted to fly it, I think, at my ski house, some of that land is like protected federal land and in federal land they will not let you fly it in certain places if you try to fly it over certain places. In LA too, like there's some parks and areas where you're not allowed to have them, but like at the park across from my street

Starting point is 00:13:10 from my apartment. The software line doesn't allow you to go there. So they have in the map and the GPS, you cannot fly your drones over certain parks or certain military things. So if these can, then that means it's one of those handbelt ones, which is what's, you know, you see in Ukraine,

Starting point is 00:13:24 when people are using them in the military sentence. They make those themselves. You just buy the parts, you assemble them, you have your software kits, boom, and you just drive them out of there. This is how the next terrorist, I mean, I hate to say this, but terrorists already know this. This is how the next terrorist attack's going to happen is people are going to start using these things. Yeah, well, you know what this reminding me of is that Chinese weather balloon story? You remember that when all of a sudden we started seeing this Chinese weather balloon over the country and why don't we shoot it down and we don't know what it is? And I think we imagine that we live in a world where everything in our skies is just, everything in our waters is just like we could see it.

Starting point is 00:14:00 We know it's there. We're constantly on guard. We've got perfect vision of everything happening around our country. And it's really kind of the illusion of safety. Like a really dedicated person who had the technology and knew what they're doing could definitely fly a drone over an American city with some sort of payload in it. Like it seems pretty obvious. You could use this for something to fit. areas. You know, the thing about the balloon story that's super interesting is we actually shot

Starting point is 00:14:26 some of those down. Well, we did after it was already over, it had passed over, I believe, North Carolina and was headed. They didn't want to shoot it down when it was over land in case it crashed on somebody. Yeah. It went over Montana where we keep our nuclear missile silos. Right. So they waited until, I think it went over like North Carolina like over the water and then they shot it down. That's when they got it. Yeah. Right. I mean, this does speak to we probably need to have more insights into what's going on in the sky. Local police need to have that. And a lot of local police now have their own drones. So there's been a bunch of startups that have been making drones. We've had a couple of them on this program. And it's really interesting when the police have a

Starting point is 00:15:03 drone, they can send the drone to the location. It gets there before the squad car arrives, typically. So when the squad car is arriving, now they've got an overhead view of, I don't know, if there was a burglary, like in a home or a store, you have the overview. It's almost like the military when they are going to a compound to whack Osama bin Laden. Now you got like the ability for the local police to see over the whole space. And on the other side, I mean, listen, I want to be safe from drone warfare. But the flip side of that is it does feel like a privacy issue as well. Like, the cops aren't supposed to be able to see into your house unless you let them in or they have a warrant that kind of baked into the whole system. And now if they could just fly a drone, look into

Starting point is 00:15:45 your windows, look in your backyard, like, eh, that kind of disappears. There are a lot of instances of kids taking these drones and being peeping toms in cities and whatever. So they're flying them all over the place. I'm really interested when if and when we'll even see these drones start delivering food for DoorDash. I know the DoorDash team was testing it in a location. There's a company testing it in Ireland. And it's really interesting that you can get things there quickly. But with autonomy at the same time. We're now going to have a race for who can get you your burrito or your mocha and your chocolate croissant. You know, I do the double mocha sometime. I'm only a little skeptical about this

Starting point is 00:16:25 because I live in, I live in Silver Lake, which is a very high traffic delivery area. Like, I live among, you know, like, I'm not one of them, but I live, a lot of my neighbors are well off and they're people who order a lot of their meals, Amazon constantly, all sorts of delivery. So we're constantly, like, anytime you go out onto my street, there's, two or three delivery drivers arriving to drop off food or a package or something. When that's drones, isn't it going to be like living on an air strip? Like, am I going to constantly have drones whizzing by me, bringing people their burritos? That doesn't sound ideal.

Starting point is 00:16:59 The good news is they are quieter than cars, typically, like an ice combustion car, they will be quieter on average. So I think if you were to choose, if that door dash could be delivered in somebody's backyard by drone or by a car on the front of your house, you probably go with the drone. You probably go with the drone backer. Yeah, I mean, I guess we'll find out. But that is a thought that I've had is like, are the skies above us just going to be constantly buzzing with drones because everybody wants their, that is the vision.

Starting point is 00:17:24 Rush from Amazon. Yeah, it really works. It's not necessary in Manhattan. It's not necessary in Silver Lake. But if you were to live. Well, LA is very spread out. So it would be helpful here. Something less than Silver Lake, what would be a more spread out area?

Starting point is 00:17:38 Even Santa Monica, too dense, you don't need it. It needs to be. Right. But like in the valley. And, you know, once you get up towards. upland and Seamy Valley, then it's, you know, yeah, it would be more practical there. Yeah, I think what we're seeing here is a zip line. Is this the zip line, folks? Yeah, this is zip line. Yeah. These things are going to work really well in areas where the houses are a mile

Starting point is 00:18:01 from each other. Right. When you're out in the country or the suburbs where, you know, people have a couple of acres, you know, let's say two, three acre lots and you're a half mile, quarter mile from your neighbor, these things are going to be able to zip right to your backyard and save you a, you know, like the 45-minute drive. That's what this is for. So if you were in Lake Tahoe, let's say, or Sacramento, and you lived in that, the burbs, I think the area that would, I think even the valley is too dense. I don't think something like east of L.A. would be a better example or out in New Jersey somewhere, anyway, where these drones are flying. It's going to be interesting to see which proves to be the better solution. But the truth is, you're going to be getting

Starting point is 00:18:40 things cheaper and faster than ever before. Without a day, I mean, already Amazon can do same day delivery here. So I can order something right now and it'll be here in two, three hours. I tell you the thing I noticed when I was in Austin is Austin has better delivery than the Bay Area. And I was like, how is that possible? In Austin, I don't know what they're doing here, but you'll order and they're like, can we deliver it from 4 a.m. to 8 a.m. They kind of ask you if you want your stuff during that window.

Starting point is 00:19:09 I wonder if it's the Whole Foods thing. They already had a huge infrastructure. That's where Whole Foods is from. Yes, they have the infrastructure. I think what it is really is they have now found that there are drivers willing to work that shift, say, 4 to 8 a.m. shift. And there's no traffic. If you have another job, that's perfect.

Starting point is 00:19:26 You can free before work, you pick up four hours. Yeah. Precisely. Yeah. Great story. Let's keep going. All right. So our next story, SORA, Open AI finally launched.

Starting point is 00:19:37 They announced it first in February. let some beta testers and other people sample it. But their AI video generation tool, SORA, has formally launched this week. It's available right now in the U.S. across Asia. We're waiting to hear details about launching in Europe and the UK. And current chat GPT Plus and pro subscribers don't need to pay anything extra. They get access to SORA included in their subscriptions. All right, founders, are you tired of doing all your own?

Starting point is 00:20:09 software and development. Do you need help? But you can't afford all this time it takes to find great talent? Are you dreading the endless interviews and email chains just to find somebody great? It takes six months. It takes a year. Well, what you need is lemon.io. Lemon.i.o has thousands. That's right, thousands of on-demand developers who can help you, and they've done the work already to vet these developers, making sure that their results-oriented, and that's They're super experienced. Of course, they got to have competitive rates, so they're going to take care of that as well. And great developers are so hard to find it integrated into your team unless you're using lemon. com, because they handle all that for you. They only offer handpicked developers with three years of experience at a minimum.

Starting point is 00:20:57 And they have to be in the top 1% of applicants, right? Something goes wrong. Don't sweat it. Lemon.io will find you a replacement developer ASAP. So many of. of our launch founders have worked with lemon.io, and they've had great experiences. So here's your call to action. Go to lemon.com.com slash twist and find you're a perfect developer or the perfect tech team in 48 hours or less. That's right. And Twist listeners get 15% off the first four weeks. Stop burning money. Higher developers smarter and faster at lemon.com slash twist. Interestingly, Open AI, they helped to kickstart the entire generative AI app fanfare a few years ago with Dolly. That was their image generation tool. So this is kind of

Starting point is 00:21:45 the next logical step. Now, instead of just an image, you can input a prompt or even use existing videos or images and feed those in as prompt. And SOAR will output a high definition clip based on your specifications. In a blog post, Open AI says they want to give society, and I quote, time to explore the possibilities and code develop norms and safeguards for AI generated video. So their pitch is like, this is not really a complete product. This is a experiment. They want everybody to play around with it in test so that we all get a better idea of like what AI video is all about and what it can do and what kind of rules and regulations

Starting point is 00:22:24 we should one day think about applying to it. Several other AI companies and startups already have their own text to video generation tools. Google announced their product, Lumiere, Amazon has create with Alexa, stability has released its stable video diffusion product. Open AI, the big difference there is what I was talking about that, what they call multimodality. So the idea that this can combine text, image, and video generation sort of all in one product. So you can feed in anything as a prompt. You can get back any of those kinds as a prompt. It's sort of working from all of those different sort of formats at once. It does seem to me that they're doing a watermark on it. So that to me seems like a really smart idea is to add the watermark to the bottom right so everybody knows. And then if somebody clips the watermark out, you kind of might know, hey, they're probably up to something. Yeah, a bad actor. But I do think that touches on something interesting that a lot of the reviewers are talking about, which is we keep talking about this in terms of like photorealism. Like people are going to mistake Sora videos for the.

Starting point is 00:23:31 real life and the dangers of it are all tied up and like, you could make it look like Donald Trump is saying anything. But to me, when I look at all of these examples, they certainly look interesting, but they don't look, they all have that. It's like halfway between animation and reality.

Starting point is 00:23:47 Like there's an uncanniness. It doesn't cross the uncanny valley for you. I don't think any of these. I mean, you could take a look and see what you think. What about the shots of like, well, I think the establishing shots of like flying over a town or if I saw it in a

Starting point is 00:24:00 movie or a TV show or going down a park. I would actually think, yeah, that's really, I'm looking at the website for Sora right now. And there's, to me, that that's, it speaks so much about human development and our brains developing over hundreds of thousands of years. We are evolved to recognize a human face and like read the expressions on a human face. Like, that was so vital for our survival in the wild is like to be able to distinguish between a person and a bush and an animal and a shadow. Yes, we're good at it. And be able to like, see a person and be like, they're angry, they're happy, they're sad, they're confused. And we, our brains are so fine-tuned to that, that yeah, it is much easier to fool us with like a field than a person's face. But isn't that

Starting point is 00:24:45 the, but isn't that the goal? Like, over time, we have to figure out that angle of it or the tech is not, can't do its job. Precisely. And if you were to look at the SORA homepage right now, there's one here of those two towers. I don't know if that's Qualil and Poor. The two towers that have the bridge between them with the snowy, right. Let's zoom in on that one there. It might not even be, that might not even be a real place. I mean, that might just be. It may not be real. But that one looks real enough to me that I wouldn't do it. The Commodore dragon obviously looks fake. Some of these are designed to look fake. But, you know, when you see a horse going across the field or the rabbit, you know, maybe I could get fooled. But I think

Starting point is 00:25:20 they're close enough that they're going to get there where they were crossing the Uncanny Valley. Right now, it feels like Luke Skywalker in the Mandalorian when he shows up. Spoiler alert. Right. Like, you remember, like, the de-aging didn't work perfectly? And then there were some kid who did his own AI software. This is, like, what, four years ago, five years ago when that happened? But it's still, even the best version of that, like, to this day, you could go see movies where they do that. You can always tell.

Starting point is 00:25:47 Like, Robert De-Machus just did that one home where he de-aged Tom Hanks. You could still see. You can still see. Okay. I mean, I think it's, to me, I feel like we've undersold that as a challenge. Like, our brains are so. It's the same thing. with virtual reality.

Starting point is 00:26:01 When they're first pitching virtual reality, it's like you'll feel like you're in a cave or a castle or outer space. And it's like your brain is so good at that. It's so good at existing in 3D space that it knows you can put the headset on and maybe your eyes are fooled, but your body knows you're still in your room.

Starting point is 00:26:17 The trick is to downgrade the video quality. So I saw people were taking the CEO shooter's image and him going to get his degree. I guess he was like at the graduation. We know his name now. Luigi Mangione. Luigi Mangione. And he, they have him walking up to give the, like, the VAL of Victorian speech, but they kind of put it through a VHS filter.

Starting point is 00:26:42 So it looks like a little old. And I got tricked. And I was like, oh, my God, this guy was the Vala Victorian. Okay, that checks out. This guy's like super high IQ. It does seem like it's pretty smart. And then he gets on there. And of course, he says something inappropriate or whatever.

Starting point is 00:26:53 And it was like a short clip. And I was like, oh, okay. So I was fooled. But now when I see anything on some. social media. My guard is up. I'm just like, right. Is it fake or not? To, yeah. And that's, I'm not the only one who has pointed out some of the limitations. In one of the first big reviews on YouTube, Marquez Brownlee notes that SOR has a lot of

Starting point is 00:27:15 problems with physics, like objects in motion. It doesn't necessarily know how things move through space. And so some of the physics of 3D objects looks off. And, and Open AI has sort of conceded to all of this that they're saying their goal is not to create fluid, perfect, like movie quality images. They describe it as a co-creative dynamic. Users can use this video to like brainstorm and explore ideas. It feels like a storyboarding tool, and it's awesome for that. Or if you were making a comic book, you know, when we had animation first come out from Disney, then you had Pixar and then you had industrial light and magic. you had this whole progression of people saying, you know, is it real or not miniatures in Star Wars

Starting point is 00:28:01 and Blade Runner giving way to the pre-equals and Blade Runner 49 where they kind of did a little bit of both? Long story short, we have this discussion every time. And then these things kind of fall into place as its own genre. So I wonder if this is going to be its own genre in the way Pixar, computer animation, versus traditional animation Snow White, versus. you know, claymation as a genre, you know, Coraline, whatever, they kind of fall into like, oh, you can appreciate a marionette, you know, South Park, you know, rough animation, anime, Pixar.

Starting point is 00:28:38 Do you think it's going to fall into like, okay, this is a CGI thing? I will go see, you know, a AI-C-GI. Let's say it's AI-C-GI. Would you go see, do you think it's going to fall into something where it's like Pixar, where everybody was like, ah, this is just nonsensical, it's never going to work. And then you're like, you know what? I got lost in the story. I get lost in a Pixar story where I forget I'm watching animation.

Starting point is 00:28:58 I don't, I don't, I don't know. I'm, I, I, part of me, the, philosophically, I'm tempted to say no. Because I think that part of animation, any kind of creativity is a brain. It's sensory experience. It's, it's, it's conscious thought. It's living in the world. world. Like, it's not just drawing, but every time you draw something, you're putting your, the way you see it and your emotionality and your perspective on it into the drawing. And

Starting point is 00:29:34 even on a computer, like a toy story computer animator is still doing that. They're using digital technology, but they're still designing and creating its creative expression. A computer, it's always just running a calculation or an algorithm. It's a predictive model. It's a predictive model. It's looking at what other art looks like and guessing what the next thing it should do is based on that. So I feel like I'm just a, I know getting like a little existential here, but like just on that level, I don't feel like a computer's ever going to be able to do what a person is because it doesn't have emotions and experiences. It doesn't have a perspective on the thing it's drawing. It's just drawing it based on specifications that were given to it.

Starting point is 00:30:17 Having said that, living in this world and seeing how quickly this tech progresses, I feel like it would be foolish to say 100% definite like no computers ever going to animate anything that looks cool that I'm going to enjoy like to me it feels like I don't I can't say that with certainty like in this world

Starting point is 00:30:37 where you know like there were all when you first showed me Uber we'll get to more Uber on the phone you first showed me Uber on your phone before it launched I was like well this is stupid who's going to get into a stranger's car what a ridiculous idea I felt the same way about Airbnb I'm going to sleep on a serial killer's couch like are you crazy I've learned humility in my time in the world of tech, that my first instincts about something are not always, sometimes things work out in a way that I'm not prepared for.

Starting point is 00:31:04 All right. Let's talk about AI, everybody. It's everywhere self-driving cars, breakthrough medical research, and of course, business tools that make everybody better and faster at their jobs. If AI hasn't hit your industry yet, trust me, it is on the way and it's happening fast. But AI requires. a ton of computing power. You know that. And those costs will skyrocket if you're not careful. I have seen these bills come in. It can get crazy quick. And that's why it's time for you to upgrade to the next generation of cloud. And that means Oracle Cloud Infrastructure or OCI. OCI is a blazing, fast and secure platform that covers all the bases. Infrastructure databases, app development, and of course all your startups, AI and machine learning needs. Now here's the kicker. OCI costs 50% less for compute and 80% less for network compared to other cloud options. That's a lot of savings, especially when every minute counts in the cloud. Thousands of businesses have already upgraded to

Starting point is 00:32:04 OCI, including MGM Resorts, specialized bikes, and fireworks AI. So here's your call to action. Right now, if you switch to OCI, Oracle is offering to cut your current cloud bill in half for new U.S. customers with minimum financial commitment. But you'll need to act fast. The offer ends December 31st, see if your company qualifies for this special offer at oracle.com slash twist oracle.com slash twist. My gut is telling me we're going to see a new medium emerge here from these tools. I think some artists are going to get their hands on it. When I say artists, I think maybe like authors, right?

Starting point is 00:32:42 And now they wouldn't have the ability to or the, they might not want to go do a TV show or make a short film. But you might get some artists who's at home, you know, like let's say Quentin Tarantino was you know, working at his video archive store. He's written true romance. He's written Reservoir Dogs. He's writing like little scenes. I could see him and Roger Avery using a tool like this, if it existed back then, in the way they used final draft to write stuff or whatever they chose to me, they wrote it by pen and paper. But I could see them making like little short animations telling a story and then sharing them on Twitter or Instagram and saying like,

Starting point is 00:33:16 here's an interesting scene. And that's like I'm starting to see people make little creative vignettes, five seconds, 15 seconds, 30 seconds that are interesting, conceptually interesting, like a storyboard might be. Right. And so, yeah, I like that as a possibility. I definitely like this idea of integrating these tools into the creative process. That just fundamentally makes a lot more sense than to me, than them being the final, like final step has always got to be human enabled.

Starting point is 00:33:49 But, you know, if along the way, these two, are being utilized in an interesting way. I think that, you know, who knows? Your creative process is your own. Who am I to say you're doing creativity right or wrong? If you're a person, it's up to you. Yeah. I think it's going to like, you know, in the same way, the digital camera allowed a lot more people to get creative and make stuff. But most people didn't because most people aren't creative or don't have the energy or the desire to do it. I think it's going to be kind of like that. People play with it, throw it away. But that's going to be like 1% of the audience who maybe figures out what to do with this. We've all got an iPhone, but Sean Baker was like, I'm going to make a movie on this thing. And now he made Enora, one of the years most respected movies.

Starting point is 00:34:30 Oh, and it was shot all on iPhone. I don't know if Enora was, but Sean Baker, that was his big claim. He did Tangerine, which was shot all in an iPhone, and then Florida Project, which was mostly shot on phones. So that's like his, I don't know. I think Enora probably was mostly shot on iPhones. I just saw the trailer for 28 years later, obviously building on 28 days later and 28 weeks later. weeks later. But that was also 28 days later was one of the very first like DV features.

Starting point is 00:34:57 Yeah, yeah. Where you shot all on digital video. Yeah, they did a bunch of these digital video, which when I did a digital video movie back in the day, center of the world with Wayne Wang, that film, which I'm in, and I was a consultant to, that film with Peter Sarsgaard, they shot on these digital films. But when you blew them up, they were a little pixelated on the big screen. Yeah. But they were kind of like that is. They had that kind of fake digital, you know, look. when you projected, but that was part of the aesthetic.

Starting point is 00:35:25 Yeah. Yeah, it's interesting. Like Figgis did that one time code back then. It was an interesting time. I think now with these iPhones, you're going to get like some shots that might be as good as you would get on a red camera to the audience. They might not actually know the difference.

Starting point is 00:35:37 The other interesting thing about all that shot on iPhone stuff is that now iPhones have so many add-ons. Like you can take your professional grade lenses and attach them to an iPhone. So you really can. Like, at this point, the difference between like the core camera you're using, in the iPhone, so many add-ons and other accessories

Starting point is 00:35:56 that it doesn't make as big a difference anymore. Got it. Oh, so the lens can be excellent. If you've got really good lights and lenses and everything else in your camera department is top grade, people won't even be able to tell it was shot on an iPhone.

Starting point is 00:36:08 An iPhone's as powerful as a lot of standard digital cameras, especially if you go back a few years. You know, I wanted to show your startup here. This is, you know, something we used to do on the show all the time would be the startup of the day, product of the day. And this one's called Ramento.com. And I saw this on product hunt. And this is a service that I have been pitched over and over again. So this idea to make a book or, you know,

Starting point is 00:36:36 a memorial of my, you know, deceased parent, my aunt, whatever, has been around for a long time. But what they've done here is they're turning memories into a keepsake book, no writing required, preserve a loved one's story the easy way. Remento turns their spoken words into a personalize keepsake book of their stories. Now, when I've been pitched this, it's been a marketplace where documentarians would interview your family and then write a book based on it, and they would charge you two to $10,000 to do this. It was like a bespoke service. Now what they've done is with language models, you can select prompts that you want your parent or your grandparent or whoever to answer or your family. They record responses in their app. Then they create a book based on

Starting point is 00:37:21 it and then you can like scan the book to listen to whatever QR code. But this isn't like a super expensive thing to do now because they handle the writing. So here on this page you can see, our speech to story technology turns voice into narrative, choose the writing style and customize the final result. So if you use grammarly where you've seen the Apple intelligence writing tools, any number of these, it will take, you know, whatever your parent says from the first person into the third person. You have the transcript. So kind of like the script, I think you're able to go in there and polish it up. And I just thought, this is a very interesting idea because it's better than nothing. If you can't afford the two to $10,000 person, I think that

Starting point is 00:37:59 this is going to get you kind of halfway there or maybe two thirds of the way there. And I think, you know, because it's more accessible, maybe more people will try it. What are your thoughts on this one? Yeah, I mean, it is, it is interesting. I don't know. I feel like there's something, Is there something a little creepy about it? Like, I, I don't, I don't know. It's definitely better than the AI, using AI to, like, recreate your dead loved ones. Like, I know there's that thing, too, where it's like, feed their voice into our AI, and you could chat with your grandma. And, like, that really is troublesome to me.

Starting point is 00:38:36 Like, this at least your- Why is that troublesome to you? It's just creepy or you have a concern about it? It is creepy. I don't have, like, a concern, like, it's going to ruin the world or anything. But I think it's, if there is something a little ghoulish about it. it because I don't believe that AI is really able to, like, I think it's a lie. Like when the guy Nightmare Alley and pretends, the mentalist pretends he can call up the spirit of your dead wife from beyond. I feel like that we're lying to be.

Starting point is 00:39:03 We're telling them the AI is actually able to recreate the mind and the consciousness of their dead love one. It's not an impersonation. It's a trick. And like, I feel like it's, I feel like it's kind of a cruel trick. So that's not this. This is thankfully using the real people. Yes. Like they record the stories themselves.

Starting point is 00:39:24 So like I, you know, it doesn't, it doesn't have that element to it. But I don't know. I don't know if you need this kind of an interactive digital archives so you can always go back and like, I don't know. I don't know. I'm just thinking about it from my grandparents every day. But like, that's part of being alive as you miss people. I loved my grandfather.

Starting point is 00:39:47 stories and we never recorded them because, you know, he, he, he died before we all had phones in our pockets. So you would have had to set up a VHS. So just that like simple structure of like, here are prompts that, you know, documentarians use. Yeah. You know, where you say what's happening in this photo or do you remember this day or, uh, you know, just when did you meet grandma? You know, like all those kind of things. I love it. I get it. I get it. And I totally get like if I could go back could do this with my grandparents, you know, it would be nice to have now, especially because now I have, you know, my, my little nephew, Dougie never got to meet them. Right.

Starting point is 00:40:27 That's the other piece. Yes. So he would have gotten to see this were your great grandparents that you hear us talk about all the time. But I don't know. There also is, part of me feels like, but people are supposed to die and then you remember them. And that's how it works.

Starting point is 00:40:40 And this idea that we need to like hold on to them forever. You're raw dogging. life. Grief. Yes, I'm raw-dogging grief. Yes, exactly. I want to feel it. I don't need to augment it.

Starting point is 00:40:53 Like not watching. I respect it. I respect it. I tried to do that on the flight. I went to Florida. Did you read that all those Twitter threads about X-threads about raw-dogging flights? Yeah, that's kind of what I was making light of. Yeah, where you just go sit there the whole time.

Starting point is 00:41:05 You don't watch a video. You know, you just watch the map. You just watch the flight map that it tracks you across the country. I made it for about an hour and then I gave up. I couldn't do it. It is like a very very. very interesting mental exercise because we now have done the opposite. You and I are film fans as part of our friendship is based on that.

Starting point is 00:41:23 You know, if you think about the ultraviolence, if you think about clockwork orange, the torture they give him is to pry his eyes open. Right. Make him lock him in a chair and make him watch dopamine hitting clips. Sound familiar? We're literally optimizing now for flipping through dopamine. dopamine hits in scrolling vertical video, TikTok, YouTube, Instagram. They all have it now.

Starting point is 00:41:51 And I think that's like very interesting. Yeah, we're all getting the Ludovico treatment. Oh, is that what they called it? Ludovico? Yeah. In the movie, that's the Ludovico treatment where they, they, they, but they also, they're key, they're playing his favorite, uh, they're playing Beethoven, his favorite music. So they're associating the sickness that he, the revulsion that he feels from the violence

Starting point is 00:42:13 with listening to Beethoven, which ruins his, because that's his favorite music. Wow. It's like, would be like making me watch like that stuff while listening to Dyer Strights would absolutely kill me. Yeah, really.

Starting point is 00:42:23 Exactly. Associating Mark Knopfler with feeling nauseated all the time. God, please know. Super interesting. That was also a device in the fifth element. When the fifth element, played by Milo Jovovich,

Starting point is 00:42:36 I believe, when she is... That's what she's learning about our world, right? She learns about our world and she just like downloads it all at once and you just see her cry. because she sees Hiroshima, Nagasaki, Holocaust, and then everything in between civil rights movement.

Starting point is 00:42:51 That's that classic sci-fi trope of the outsider watches, like instantly catches up on our world through the TV or through the internet, yeah. Yeah, pretty good trope. I like the, I like the truck. All right, let's wrap up here before we get to our amazing startup interview. I think there was one more story on the doc that we wanted to talk on. Yeah, well, we just had this tweet from our Uber CEO. Oh, yes.

Starting point is 00:43:13 Dara Khashahi. So new first for Uber last Friday, one million concurrent trips. That's bizarre. More than the entire population of San Francisco all at once using Uber, getting rides at Uber, more than 3,000 processes, notes that AI also helps to ensure matching, routing, safety, payments all working at scale, you know, like tremendous amount of rides happening every minute that they've got a service. Very proud of our tech team who effortlessly translates huge digital complexity.

Starting point is 00:43:44 into real world convenience for hundreds of millions of people around the world. So I'm assuming that means just people riding in Uber's, not also food delivery. That would be a separate division. Well, you know, when they say trips, I think they include food in it. Okay. So this is both people getting food delivered and people riding in Uber. That's my guess.

Starting point is 00:44:02 I could be wrong. But I think the routing of food is the same as the routing of people as far as they're concerned. Because the infrastructure is the same. Yeah. I mean, this is a massive achievement. to be able to do that at once. I think he is also, maybe he's always been let the results, you know, speak for themselves and the results at Uber, since they got to profitability, raise the prices and, you know,

Starting point is 00:44:28 basically scaled to such enormous, and obviously I'm long, I'm still long with stock, although I'm diversified. So it's not like an acute thing for me or a religious thing. You know, he's just letting the fact that they became a money printing machine and their plan on buying back their stock on a business basis. speak volumes, but now it's self-driving in Waymo and Tesla Cybercab and B-Y-D and self-driving little robots that deliver your food in cities. I think he's maybe wants to share more of what's special about Uber, just in terms of marketing or, you know, corporate communications maybe, and I haven't spoken to him about or anything, but I do think he's maybe just trying to explain to people how big Uber is, so they appreciate this, you know, because I guess, he's quiet about it. He's just like, look at the numbers, the results of the results. Everybody

Starting point is 00:45:17 relies on this. There's hundreds of millions of customers and a million at the same time. You know, if there was a 150 million customers, it's almost 1% of them using it at the same time, which is kind of interesting just in and of itself. I wonder what the denominator there is for a million concurrent rides. But it also proves is if there's a million concurrent rides going around in the world right now, that means you need a million cars at any given point in time to just do those rides. A million cars, a lot of cars, folks. To be on the road at one at any one moment, yeah, because we're not talking a million cars in the fleet. It's a million cars working that shift right then.

Starting point is 00:45:54 Yeah. It's fascinating to me. Like, Los Angeles, I feel like Uber as a transportation option has really fallen off. There's fewer drivers. Rides are a little more expensive. It's Uber Eats is the breakout. Everybody's constantly using Uber Eats here. You got such great food there.

Starting point is 00:46:13 I mean, when I was staying in Manhattan Beach over the summer, I mean, there was a Gwyneth Paltrow's Goop Kitchen was really good. You know, if you want to eat healthy, there, I mean, it's just, it's also all spread out. The famous sugarfish sushi was on it. Right. You might be in Santa Monica and want something from Hollywood or downtown. You might be in the valley and want something from the South Bay. I mean, so it is the perfect sort of Uber Eats city.

Starting point is 00:46:39 I had a Hi-Hoburger. That was great. The Ihoe burger is great. Really good and sent them on. But like, I'm curious, like, I'm sure every, this is another complexity to the Uber business. I'm sure every market is different. Like the, how many people are taking rides versus how many people are ordering sandwiches, it's got to be, you know, incredibly diverse.

Starting point is 00:46:59 The riding is on the wall, though. You know, I think we will be sitting here in another decade. And it's going to be, instead of 1% of rides in the United States are ride sharing or, you know, whatever, 2% of them, it's going to wind up being like 10 or 20%. I think car ownership. We need to get these cars off the road. However we do it, I'm fine with ride sharing as an option, whatever we do. Ride sharing mass transit. For ride sharing is going to be incredible because you'll be able to recapture that time. And so, and then you take out parking and storing your car. Those are major costs in cities.

Starting point is 00:47:34 And so all that's going to get. The worst thing we ever did in cities was fill them, fill them with cars. Now we're stuck with it, but it's a nightmare. It would be so great to get all these cars off the road. In Manhattan, they, it was, always amazed now, I guess Bloomberg did this towards the tail end of his tenure there. But in Manhattan, they, you know, you get, you're coming down Broadway and you get to like Harold Square or whatever, you're usually, you leave, you leave Times Square 42nd, Harold Square is 34th. It's just, Broadway's closed. It's all people. And they've shut down the roads and they have tables and outdoor cafes and pop-up boots and coffee. And you're like, oh,

Starting point is 00:48:13 this is beautiful. It's like the San Maraca Promenade. Well, then you think about all these like great world cities. How did they become great world cities? It's like public squares and plazas and destinations where people can go walk around and not necessarily like every street is bordered by this expressway that you have to like run across. Yeah. It's going to be fantastic. I'm really excited to see more. I am concerned now, though, I have to say about millions of people losing their gig jobs. We have all this controversy lawn about, oh my God, you know, people are being abused by, you know, the ride-sharing companies, the delivery companies, DoorDash, whatever, Amazon, they're not paying people enough while people flock to that as an

Starting point is 00:48:54 option for extra employment or, you know, part-time employment, whatever. Very few were actually full-time statistically. The prices keep going up is, you know, they have to compete against working at Starbucks, so the salary has to beat or be as good as a Starbucks or an Apple store job, in my mind. Now those jobs are going to go away. We're sitting here 10 years. The idea that you're delivering food is going to go away. Ten years, probably going to go away having a human driver and then a truck driver.

Starting point is 00:49:24 I mean, this is you're talking about millions of jobs in the U.S., like millions and millions. And people pay their rent. People pay their credit card bills. People pay for their kids' school by doing an extra door-dash shift or on the margins or being an Amazon, picking up an Amazon shift like we talked about earlier in the show. I'm a little worried about what happens to when we eliminate those jobs. I don't think there's like a quick plug-in. There's so many things we sort of could do, but Americans just have this huge resistance to anything that could even be remotely mistaken for like welfare or, you know, like socialism or. Or like, all you have to do is say, well, that's socialism and everybody gets upset about it.

Starting point is 00:50:06 But like, the answer to all of this stuff is probably going to be some kind of universal basic income. Eventually, we're just, we are innovating everyone out of their jobs at a rate that's far beyond how we can train them for new jobs or even needing them in new jobs. You know, you know, once you combine AI, robotics, self-driving cars, all these things, you're eliminating millions of jobs at a go. You know, and, you know, people that you can't just make them into, like, you can't just teach them all how to code. There aren't jobs available. We already eliminated the retail sector, so they can't go work at a store. Like, that used to be, when I was a kid, that was the go-to. It's like if you lost your job, well, you go.

Starting point is 00:50:48 I go work at the gap. You go work at the gap. You go get a job at Macy's. You go get a Barnes & Noble gig for a year or two to tie you over. Right. And we were now innovating ourselves out of all, like Amazon ate all of those businesses. DoorDash and Uber Eats are eating all the restaurant businesses, turning them into ghost kitchens.

Starting point is 00:51:07 Yeah. You know, we don't need delivery drivers. We're not going to need warehouse and factory workers. Like, what are all these people going to go do? All right, everybody, another great episode for you today. And now we're going to have our incredible interview with QB technologies. What do they do? Well, they build homes.

Starting point is 00:51:24 And they build homes with mobile micro factories. It's super interesting. Their business is to build a factory. And that factory's business is to. to be on-site, running 24 hours a day, building homes that are delivered within 100, 150-mile radius. And this allows you to build homes faster, quieter, cheaper, with higher precision. I know because I was an investor in a company called Blockable that tried to do something similar. You're going to love this interview with QB technologies. Enjoy. And thanks for Lon for coming on the

Starting point is 00:51:54 program. He is X.com slash L-O-N-S. See you next time. Bye-bye. All right. So next up from the Twist 500, we have QB Technologies and we have yet another Alex, in this case, Alex Gumpel. He's the co-founder and COOJs. And this company founded in 2021, based in New York City with manufacturing operations over in Eastern Europe, and a city called Minsk that I have actually yet to visit, but is on my list, only raised $6.4 million in known capital.

Starting point is 00:52:23 They're a building mobile micro factories. I think this company is awesome. Ah, there he is. Alex himself, how you doing? Welcome to the show, Alex. What I love about what you're doing is I had an investment in this space. I'm sure you know about them blockable. And it didn't work out very well for that.

Starting point is 00:52:38 Maybe they were too early, but it's a really hard space to build homes. And you have a similar model, which is you like to build homes in factories. If you build a home in a factory, well, boy, you are not constrained by the weather. And you can work 24 hours a day because you have sound debatemen going on in there. And you can be warm. and you have all these incredible tools that you can bring that you don't want to have on a work site with a bunch of mud. Correct me, if I'm wrong, the key difference here,

Starting point is 00:53:10 and I'd like to see some videos, is you bring the factory to the work site as opposed to delivering the homes that had been built in a work site or a factory 500 miles away. Am I correct in the basic overview here? Yeah, first of all, thank you for having me, big fan of both of yours.

Starting point is 00:53:28 I'd say we're the anti-fews. thesis to anyone trying to industrialize construction, which is where you're placing the category. Our antithesis is really tied to three core pillars. One, yes, our product is a factory and we bring it to site. So we built a mobile micro factory that gets deployed near where there's construction demand, and it makes a kit of parts that then we use our own labor to assemble that kit of parts into homes. We have a second antithesis, and it's around changing what a home is. As soon as you start creating volumetric boxes, 3D printing, etc., meaning the more you deviate a way to from what a traditional home is, the more you run into regulation risk and incumbent adoption risk.

Starting point is 00:54:04 So we said, you know what, we're going to build homes on site traditionally. We're not going to shove electric plumbing and whatnot behind panels. We're literally going to build a home the same way that Lanar builds a home, except we're going to do it with better process information. We're going to do it with a better supply chain procurement management system. And the third pillar is really around cost. No existing group in the space and probably by your investment you've seen has done this in a cost-effective way that combats where existing costs are. So the sales cycle becomes, hey, I'm more expensive, but I'm better and there's better quality assurance. We took a different approach. You know, you guys spoke about Tesla a lot. We don't like that

Starting point is 00:54:41 analogy. This whole thing of Roadster to Model 3 isn't going to work in this space. This is a highly conservative space. If you're not cheaper than existing home builders today, you're almost irrelevant. You'll never go that curve from Roadster to Model 3 where eventually you hit critical scale. So we told ourselves we can't go to market unless we can consistently figure out how to build homes at around $100 a foot. National average is around $152. But in short, we're a full-stack technology company. We build hardware and software. It's all for the purpose of improving building processes in new home construction. In a TLDR, we design, develop, and deploy mobile micro factories, exactly what you see behind me. It's a containerized solution to a factory.

Starting point is 00:55:20 So that factory is like a pop-up building that's got one of those inflatable roofs, I think, or a boatrous roof. Like, you might see a tennis court, yeah? The opposite of a tennis court, because if you've played inside a tennis court, it's super loud, you know, it's pressurized. So as soon as the doors open, it deflates. We actually use a repurposed air hangar technology, the pressures in these little tubes. So you can keep the doors wide open. But yes, it's a containerized solution. So we build about 50% of the hardware in-house, and we deploy that hardware into containers.

Starting point is 00:55:54 So then we can send, call it 60 to 100 containers later, anywhere in the world, a factory that gets erected in eight weeks. Eight weeks. Okay, that was my first question was how long does it take once you pick a site to unload all the shipping containers and then get it set up? How many houses do you think one of these mobile microfactories needs to construct to be viable economically? Because there's costs with bringing the equipment and getting it set up, training, and so forth. Yeah, we no longer think. This is past theoretical. We've deployed one of these already. So it's past kind of this technical risk stage. So we're at about 400,000 engineering hours, TRL of 8, which means we're far enough where we're ready to commercialize this and we are commercializing this. But each one is about $10 million in terms of resources and uses, which if you guys know manufacturing or spend time in manufacturing, that's an insane achievement to be able to launch factories at that cost basis. That allows us to bring down the, output of one of these to be profitable at about 200 homes a year. So about 430,000 square feet of output. And each one of these service is 150 mile radius. That's a perfect throw distance for us to deploy our own labor onto each construction site to put together the home, which is why we

Starting point is 00:57:04 replace a general contractor and all the 22 different subcontractors involved, which most folks don't do in our ecosystem. How many people would work to be the crew for one of your MMFs, both at the factory itself and also on the actual job sites in that radius? Yeah, so all in all, each MMF is its own standalone business. Think of it as its own standalone business that's capable of producing end-to-end 200 homes a year. Okay. It means that we're employing about 260 folks. The way it works is there's 35 folks on the factory floor in two shifts,

Starting point is 00:57:36 and now there are a dozen folks that are GNA, and then imagine you can have 20 different homes being built within 150-mile radius at any given time of an MMF. Per each site, you have about four unskilled workers in two shifts. So that's a total of around 260 folks all in all at skill. Four unskilled workers on site? That's it? In two shifts.

Starting point is 00:57:56 Yeah, building a home over the course of 60 days, essentially. That feels incredibly minimal on the human capital side. So I presume Alex then, therefore, there's going to be quite a lot of repeated work that there's probably some panelization of these products. you would kind of take out of the factories that are easier to put together? So the factory really consumes about 600 different bill of materials. Those bill of materials, they do one of three things. One, we're producing items from scratch.

Starting point is 00:58:24 Think windows, think foundation systems, think framing, certain non-structural wall elements, everything else for just prepping. Sheet rock is technically a finished good or a pex pipe, but we're prepping so there's no dirty work done on site. And then there's things that are off the shelf, like kitchen cabinets or toilets or certain fixtures. But all in all, we're breaking up a home into 35 stages, sending just in time last mile those stages to be put together by unskilled folks. We have one KPI and one thesis as a business. There's not enough young people that want to go swing hammers. It's a massively retiring workforce,

Starting point is 00:58:58 meaning that about 50% of the workforce in construction today is going to retire in the next 10 years. We're already at a massive housing crisis. Doesn't matter where you sit politically. Bipartisan issue, we're screwed pretty much. And most of it is because we don't have enough folks to swing hammers. So our thesis is around deskilling the labor required to go build homes. So how do you build more homes with fewer hands? That's what we're trying to do. It's really interesting, Alex Wilhelm. And with these MMFs, I think that means micro. Mobile micro factory. Mobile micro factories, this term is the term you've come up with. But Alex Wilhelm, this is something we had with Blockable, which was a lot of people who are in

Starting point is 00:59:37 construction have to then go on the road. They leave their family. They do. a construction job in another state. They stay at a dingy motel with four people per room. And it's kind of not a fun life. And you're outside in the elements. If it rains, you stop working. You get the idea. When you're in the field and you're making cuts,

Starting point is 00:59:56 you cannot use the equipment that you see behind Alex, you know, this high precision equipment. You're literally using exact own knives and saws and whatnot. So when you make these in a factory, you've now eliminated weather, you now have shift work and people can work

Starting point is 01:00:15 in a factory near their homes and they could pick up shifts, work two shifts, a double shift, work at night, etc, work seven days a week and you don't have noise as well bothering the neighbors,

Starting point is 01:00:25 which is always a major issue, but you do eliminate somebody having to draw lines and look at documents to figure out if they've made the panels correctly because I would assume every panel is made

Starting point is 01:00:38 with one of these high fidelity machines, correct, Alex? Kind of. We actually have a bit of a slightly different thesis as well. Everyone talks about roboticizing the space. The only reason why Tesla was so quick to roboticize is because they had 50 years of Toyota's production system to play off, right? Lean manufacturing. No one's even done lean manufacturing construction.

Starting point is 01:00:57 So we're first about taking away skilled labor. How do you make that skilled labor hyper-efficient? Then you can roboticize in an ROI positive way. What you're describing, Jason, is kind of interesting because we extend the elite manufacturing process beyond just the factory setting, but also on site. No one's done that before, and I'll talk through an example of how we're doing that. Usually in building a home, that skilled labor typically brings their own tools, which is crazy. Imagine a Tesla factory worker showing up with their own screwdriver. It's weird, right? It doesn't exist in other industries.

Starting point is 01:01:29 So what we've done is we said, okay, we're going to deploy on-site a container where the home is being built, bathroom, shower, locker, storage for all the equipment and pools. for those four unskilled folks to then follow a process instruction saying go put together this stage, this kit of parts at that specific stage, here's the tool you use, and here's the Unreal Engine augmented instructions that tune every time you build a different home. So they're pulling tools based on what they need at that time. They don't need to bring their own tools anymore. So you want to extend the lean manufacturing process on site as well.

Starting point is 01:02:01 So it's kind of like putting together IKEA furniture. It's not easy necessarily. It takes some thoughtfulness, but if you're given really good instructions, clear instructions, and it seems like you're building 3D models and videos for people to do, and you're numbering the tools, and you have consistency with the tools, it gets a lot simpler, huh? Yeah, it's a hardware software-enabled process improvement to traditional construction versus going a full 180 of let's reinvent materials, let's reinvent what a home is. Again, thesis number two, the more you deviate, the harder,

Starting point is 01:02:35 is to sell into the system. Okay, so that's where I wanted to pick up because we saw a clip earlier that we had on the screen of a beautiful home that you guys had built. And I actually watched a video of one of your founders walking around showing off this house. But my question is about customizability because that house looks great. It matches my own personal design aesthetics. I love wood.

Starting point is 01:02:53 I love right angles. But what if I was a more fussy home buyer, Alex, how far can you customize these homes inside of one MMF, inside of one of these radius as you're describing? Because we're not volumetric modular, which Jason, you said you invested in blockable in the past. They were partially volumetric modular, from my understanding. So a lot of folks are approaching the space where they're designing entire home offsite, which is finished product. We're not doing that.

Starting point is 01:03:19 We've standardized inputs that go into a home of a non-proprietary supply chain. That means we can have different iterations of those inputs, which means when they come together, that's hundreds of permutations of what can be built on site. Think of it as like Lego blocks, but not reinventing what traditional. construction is. We can go to Home Depot, get those same materials. We use very standard inputs. There's really two differences in our homes. By definition, if you're going to be standardizing, you limit yourself from hiring a home builder that like Alex or you Jason can hire to go build you a custom home. But that's what's wrong with construction today. Every project is a one-off

Starting point is 01:03:53 relative to every other industry that's had this type of standardization, lean manufacturing capability. So yes, hundreds of permutations, but still more limited than traditional construction. But for the most part, we can back into, you know, pretty much anything the production home builders can build. That's awesome. You're doing this in Europe to start while Europe as opposed to... No, no. So, so let me explain. And Jason, you saw this from Blockable. This space is inherently hard. Whichever cohort, whichever company gets this right could be the biggest company in the world. Naturally, you know, you look at the U.S., there's 3.8 million homes missing. You multiply 500K, you know, a home that's $2 trillion of market demand today that needs to be and can be absorbed.

Starting point is 01:04:33 but we physically can't build enough. So we're talking about a company that can have a massive outcome. Oleg and I, my partner, we didn't want to play the venture game early on where we would have needed to raise, let's talk about 400,000 engineering hours we've done. We would have had to raise $50 million at seed. It seems kind of a bad bet to make with the value of death that's notorious in deep tech. So we said, okay, what's another angle? We can go to get through our technical risk.

Starting point is 01:04:56 Eastern Europe is a 7 to 10x cost saving in terms of great engineers. So we set up a big engineering R&D hub in Eastern Europe. We're obviously a U.S.-based company. Our first commercial deployment is in Nevada. And we're looking to put up 275 mobile micro factories. I'd say most of our target markets are first world countries where the coefficient of skilled labor is much higher than the cost of materials relative to the way you build.

Starting point is 01:05:20 Like cost of labor, higher than the cost of materials relative to the dollar that it costs to build. So naturally, it's Europe, U.S., Australia, Canada, etc. And these are single-family standalone homes. You're not doing multifamily, you know, three, four, story, you're going to make your bones in the single family American classic home, row houses, essentially. At our cost basis, we hope to bring back a lot of master plan developments, which I think are much needed with an aging millennial demographic that, you know, is being stuffed into studios and, you know, one-bedroom apartments because that's the only

Starting point is 01:05:52 feasible thing anyone can build in an urban core. So we're hoping that folks through our process and the cost basis are incentivized to build more, build to rent, for sale, single-family homes. Okay, so then I want to know how many homes are going to be constructed using KB Technologies technology in the next, I don't know, three years. Alex, I'm just trying to figure out how far away you are from kind of early mass commercialization of this technology. Yeah, so we didn't historically focus on like building homes out of our first of a kind because we're in the factory business. We're not in the home business. It's super easy to build a home. So for us, we kind of have grief around companies that raise a ton of capital, build one home, show it off. to investors, put in an architectural digest, the hard thing is the process behind building hundreds and thousands of homes. It's not necessarily the home itself. So we focused on, let's call it, you know, as corny as it is, the machine that builds the machine, right? That's what we focused on. Our first commercial deployment is in Nevada. That factory will output for 10 years, about 200 single

Starting point is 01:06:51 family homes per year. So we're looking to do that 275 times over the next decade or so. So, 2,000 homes, 200. 300, 300 factories, you start to get to a lot of homes. Hundreds of thousands of homes you hope to build in the United States at a minimum, yeah. Yeah, I'll tell you something interesting about the market. At scale, if we get to 275 mobile micro factories, we're only building about 55,000 homes per year. What's ironic is the U.S. builds about a million homes a year today.

Starting point is 01:07:22 That's what they build, and we're undersupplied. So we're like less than, you know, one and a half percent of the entire market share in the U.S. that's at today's at today's under supplied volume that's built. That's why it's such an interesting problem to work on is just you can't claw back enough. You can't get big enough to even be, you know, remotely. So your customers will be the home builders or landowners or are you going to pop up like an LLC where you have a home builder and a landowner and yourselves sharing economics? How do you think about your business? Are you just providing the lowest per square foot price and beating out other construction companies? Are you going

Starting point is 01:08:04 to sell direct to consumers? What's the eventual business model here? Who winds up being your customer? Jason, we should get you as an investor. You seem to get it. It's all the bump. So we're very much focused on correct cost of capital. Venture at some point is not a creative cost of capital to us. Our business is built perfectly for project finance and eventually debt. The way we grow is tied to that specific cost of capital. Today, because we don't want to use our balance sheet, we partner on our factories. So, for example, in Nevada, our partners control 6,000 acres of land. To them, the economics are compelling enough to say, guys, no problem. We'll back, you know, a big portion of that 10 million required to launch your factory. You run it for us,

Starting point is 01:08:42 and we're the embedded customer off taker looking to build 2,000 to 3,000 homes. Eventually, what we think is interesting is to launch our own factories and then have a bunch of Bob Smith, the home builders in a local market that build 20, 30 homes, dying to compete at the cost basis that Lenark can build at, because of, their, call it band-aid to their inefficiency, which is scale. And everyone's not, everyone's overlooking kind of the backbone of American housing market, which is small to medium-sized home builders. They build 50% of the U.S. market. They're the ones building 20, 30, 100 homes a year in that market. That's who we want to unlock cost competitiveness for. Eventually, though, our master plan entitles

Starting point is 01:09:19 us to launch a home builder ourselves, you know, call it five to 10 years from now. Yeah, I mean, this is one of the challenges of this business, having been on the board of block is if you sell to, you know, people who are building these projects, they just want the lowest cost per square foot and you're in a race to the bottom. And then there are landowners who, you know, they have land and they're like, great, we'll provide the land. You build the home, but you then need to finance it. That's not how venture capital works.

Starting point is 01:09:49 Venture capital is not looking for a coupon, you know, or, you know, some tap return. But there are investors who are interested in that. So, you know, I think popping up, dare I say, an SPV for each project or for different projects and having the core IP and the main company is potentially a viable solution here where you could do a JV, use your technology, you own all the technology, all the innovation, all the patents, everything to do with your factory. Hey, and that landowner is customer zero and customer 100. They just take 100% of the output of the factory. Maybe they give you a little taste of some of the upside. in the homes when they sell. And it could be super powerful.

Starting point is 01:10:30 I love the idea also of having watched some videos of your product. There's a lot of innovation that can happen here in terms of customization where panels that go on the inside or the outside of the home can be standardized or used industry standards, can use magnets, can use lug nuts, you know, that are very easy to hide. And aesthetically, they can look radically different. You know, many people know this because they put in flooring, now, that's high-old-based flooring. So it looks like you put in this incredible wood flooring, you just put in panels. You put in a driveway, right? Some of these driveways look

Starting point is 01:11:06 incredible like you're going to the palace of Versailles and it's just panels, right? Yeah, exactly. It's, it's interesting because we want to be, I don't want to say it, it's very corny, but we want to be like the AWS of home building. Yeah, no, I don't think it's corny at all. I think that's exactly the right. You go do whatever you need to do. Yeah, and the way we think about it is 275 MMFs later. If we just traded a 15x on EBITA, we can build a $50 billion business, which I'm conservative. I come from private equity. It always erks me to say things like that, but it's possible. And I think whoever does it, maybe it's not on us, maybe it's the next cohort. These are massive businesses in the making and obviously much needed.

Starting point is 01:11:43 Hard to call the U.S. a superpower when we can't, you know, when 50% of the U.S. folks can't afford, you know, $250,000 homes. Yeah. One last question before we let you go, Alex. You keep saying 275 MMFs. Why is that the number you keep coming back to? Is that based on upcoming financing? Is that just based on near-term goals? It feels a very random number to me. It started out as a joke because I ran out of space in our corporate financial model.

Starting point is 01:12:08 So that's like the max I could draw it down over 10 years. But then it fell perfectly to what will be constrained against the Papa factory that makes the factories. We'll get to like we're looking at a time horizon, but eventually at scale we want to be doing 20 of these per year in terms of production. Okay. So you've got to build a factory to build factories. And, you know, the great news here is Alex Wilhelm.

Starting point is 01:12:31 When you look at a business like this, we always talk about, you know, standing on the shoulders of giants. There are people doing things in fabrication, in water cutting, you know, in material science. And AI, of course, that are going to just be wind at the backs of this company. just like batteries from iPhones led to Tesla having an easier time with batteries because so many people were innovating on battery technology. There's so much innovation going on in material science. The problem is, you know, if you've got a labor force that knows how to use sheet rock, knows how to use studs, and you introduce stuff that has to be cut in a factory, that cannot be cut in the field, but it's this new material and it lasts 100 years. and it costs twice as much. You have economic supply issues

Starting point is 01:13:18 where the people who are building it only care about the per square footage price or the per square foot price, but they don't care about the maintenance and the maintaining of the home. One of the great innovations that blockable kind of reached at different points in time were how much more affordable it was

Starting point is 01:13:37 for each block, and they would stack these blocks into little studio apartments, one bedrooms. You know, they could make an innovative in the HVAC system that could save, you know, $100 a month across 100 apartments, across 100 apartment complexes,

Starting point is 01:13:54 across 100 years. They could make innovations in, you know, just building a trow underneath each apartment that kept it from flooding. So when there was a flood, the water would flow underneath the blockable unit and go outside of it as opposed to warping the wood. And they could use wood

Starting point is 01:14:11 and new material science that could be cut in a factory, that didn't warp and that was water resistance because the two, what I learned was the two biggest problems in apartments, fires and floods. And they kind of go hand in hand. You have a fire and the spring goes up, you get a flood. These things cause massive damage. If you could eliminate those two things, you would eliminate like 90% of the damage cost of homes. Painting a home is easy. Warped wood floors are getting ripped out. It's kind of interesting because like logically, the world would think like that, but developers have one constraint. It's yield on cost, right? They're looking for a return metric into the underwrite hard costs, meaning day one

Starting point is 01:14:48 cost. They don't think about carry costs, that future efficiency, because it's not about finding the most innovative developer. Even if you find the most innovative developer, people don't realize they're constrained by their lender. They're constrained by their LPs. So it's a really multivariate equation with lots of layers to the onion, essentially. We actually, the way you think is interesting, because we want to be the conduit. Eventually, like, we have great investors. And I think it's important in Deep Tech to have great long-term investors, but some of our investors are very much focused on climate and just generally deep tech, like type one, out one, et cetera. The way they look at us is if we hit our mission and we get the scale, we become the biggest conduit of some of these alternative

Starting point is 01:15:28 materials because it's hard for them to sell directly, but eventually if we're the engine, we can just feed all those alternative materials to us. Yeah, I mean, this is, it's almost like a tragedy of the commons. The incentives matter so much. and nobody's thinking about the incentive of, hey, what is the third owner of this home in year 40 going to experience with the home? Are they going to experience the windows falling out and cracking and warping? They're not thinking about that. But if you were thinking logically and you could spend, I don't know, 10% more money to get an extra 10 years life out of those windows, it'd be well worth it. But it's the opposite. Spend 10% less on it. And you make a

Starting point is 01:16:12 a little bit more or you win the job. And then maybe the windows start warping in year 20 or 30, but it's not your problem because you as the construction worker will be retired by then. And the housing company will have sold it already or maybe the housing might have sold it twice. I forgot who said this, but this is wild. I think homes in the U.S. are one of the only products where today you can buy new pay more, but it will be a worse product than before or something you can buy used from, say, the 60s, the 70s. It's kind of wild because no other product you pay more and get worse quality.

Starting point is 01:16:46 So wait, wait, what's an example of that? Like, uh, you're saying if you bought an older home, it would be a higher quality than a newer one that's made with more disposable. It's like sheen or sheen, what is the, shen where everybody buys their, you know, Coachella clothes. You buy those clothes. You wash them twice, Alex, as you know, have, you know, going to so many music festivals, they just disintegrate. Whereas, you know, some, my mom's, you know, my mom's, you know, leather jacket, her fringe jacket from the 60s or 70, she just gave to my daughter. It looks great. It's going to last. And then my daughter will give it to her daughter. Yeah, no, cost on the home building front will keep going up, which means home builders will keep cutting corners. We'll

Starting point is 01:17:23 keep finding ways to use crappier and crappier materials. Yeah, it's a disaster. A real tragedy of the common short-term thinking, hey, Alex, we wish you great success on this. It's awesome, and I can't wait to meet you and go see the factory in Nevada. And yeah, I'm, you know what, I would You know what? I think my, the scar tissue of blockable might have healed by now. I might want to get back in on this. I might want to write a check and meet you, Alex, and come on this adventure again, because I'm, I'm realizing how enthusiastic I was about blockable in being in this space, because it's so obvious, somebody will figure it out. And this is, like, as a capital allocator, Alex Willhelm, one of the things you have to deal with is, you know, you might have a really

Starting point is 01:18:04 traumatic experience, lose millions of dollars, and then the next person up the hill actually figures it out. So you have to keep that resiliency and enthusiasm for big markets and solving big problems, which Alex, you're doing, cannot wait, Alex Gample. Am I pronouncing your last them, correct? Gample? That's correct. Yep. All right. Well, let me know when you're Austin. I'll take you for some brisket and a beep over at Terry Blacks or some brisket at LaBarre, and let's break bread and talk soon. Perfect, guys. Thank you so much. much. Really appreciate you reaching out and doing this. Talk soon. Congratulations on being in the Twist 500,

Starting point is 01:18:39 created by Alex Wilhelm. So I don't, people don't know my secret plan, Alex, but with this Twist 500, I want to have an event where I invite all the Twist 500 founders to come hang out for a weekend. It would be awesome. Yeah. And, uh, it's going to be a, uh, index fund eventually. So my, my idea is to try to buy, you know, raise $100 million and buy $500,000 worth their secondary and

Starting point is 01:19:03 200 of these companies, eventually all 500 of them, and then have a proper index of all 500. It'd be pretty hilarious if you could actually own a little bit of each one. It would be fun, too. Have like a private market index. All right. Well, we'll see you soon, Alex. You guys. Talk soon.

This Week in Startups - Drones, Deepfakes, and Disruptors with Lon Harris, Remento and Cuby | E2058

There aren't comments yet for this episode. Click on any sentence in the transcript to leave a comment.