SPEAKER_01: There's 50 million Americans on antidepressants.

SPEAKER_00: You know, there's been campaigns run in America and around the world, mostly by drug companies, that have convinced the public that they are suffering from chemical imbalances,

SPEAKER_02: that if they're feeling sad or they're feeling anxious, it's something that's gone wrong in their brains. Pharma had no interest in getting people off or doing any research to get people off, unless you had a business case for it. When people go to their doctor and they say,

SPEAKER_00: I've stopped my drug, I'm having panic attacks, I can't sleep, the doctor doesn't even let them finish the sentence and says, you must be relapsing. That is, your underlying depression or anxiety has returned.

SPEAKER_06: Do we know what it's done to people's brains if they've been on it for 10 or 20 years? We are running the biggest open-air experiment

SPEAKER_07: on the public ever conducted.

SPEAKER_09: Okay, everybody, it's Twist. It's May 22nd, 2026. Lon Harris is here. I'm your boy, Jake Howell.

SPEAKER_10: We have a lot on the docket today to get to. But before we do anything, Lon, I'm just going to grab my plod. Yeah, we should applaud the Lord.

SPEAKER_11: I always like to applaud, applaud. I take my little plod pen. And when I'm wearing a jacket like this, I use the magnetic. You use the magnet.

SPEAKER_13: I have the magnetic thing around here. It's right here.

SPEAKER_14: But you like to clip it like a dork.

SPEAKER_16: I like to clip it like a dork. Exactly. Yes.

SPEAKER_17: I like the more elegant magnet thing.

SPEAKER_09: And then you just press the little button, feel the haptic shakes. Now I'm recording. Now everything I say, I can do. Hey, make sure I go through a plan with producer Jacob and Lon about the community manager coming on board and just documenting where all the communities we have for this week in startups, this week in AI, all in, et cetera, founders, founding university. Make sure we just at least document where all the communities are so she can hit the ground running. Okay. Now I got that on my action item. I'm ready to go.

SPEAKER_10: When I drop the plod pin into my little charger here, that's just on my desk. Look at this, Lon. It's just a little tiny charger here.

SPEAKER_23: When I drop that in, it auto-syncs. Yeah. Gets onto the Wi-Fi network, the 2.4 Wi-Fi network, boop, right up.

SPEAKER_25: You can send it to an app on your phone or your device. You can send it into Google, wherever you want. You can port this information. I was explaining it to a friend and she was saying, you know, oh, I hate those kind of things that record you.

SPEAKER_28: And then you got to go through these transcripts. And I don't like listening to myself.

SPEAKER_29: And I was like, that's the great thing about plod is it really does all of that. It really does all of that for you. And then you just have this document that's what you said without the fluff or anything you don't need. It's really great.

SPEAKER_34: And I click on the most recent MP3 file or whatever file size it puts in there. I hit generate. And I go with auto-generate, but you can do custom. And it will make highlights. It'll make a summary. It'll put the transcript. It'll put action items.

SPEAKER_11: It'll make like five tabs about that file that you recorded, whether it's a meeting or you taking action items or you're writing a blog post, whatever it happens to be. They have a code. Give them a code. Oh, sure.

SPEAKER_25: So if your work relies on conversations, you got to get yourself a plod note pin S. Check it out at plod.ai slash twist.

SPEAKER_29: And if you use the code twist, they're going to give you 10% off your new plod pin.

SPEAKER_24: Okay. Let's get started with the docket today. What's on the docket? Sure thing. Well, first, we got a guest. We're very excited. Oh, great.

SPEAKER_25: His name is Lucas Zinger. He's the co-founder and CEO of Divergent Technologies, Jason. They are 3D printing lightweight vehicle and aerospace structures in an industrial scale using AI-generated designs. They've got their own three-part system. So they generate the designs in AI. They 3D print them themselves using their own proprietary alloys and systems. And then they've got a robotic assembly system that puts all of these components together in their factory. So they're sort of an all-in-one solution for 3D printing these large, intricate, formerly

SPEAKER_41: very component and little part-heavy structures. I'm aware of this because they had the Singer vehicle, right? They were building a hypercar or something. Correct. Yes.

SPEAKER_25: And we've got Lucas Zinger here, too, to tell us all about Zinger vehicles and moving from automotive into 3D printing for, you know, third-party customers and clients, including the U.S. military.

SPEAKER_46: Welcome to the program.

SPEAKER_48: Tell us a little bit about the hypercar. I remember that. And then how you went and pivoted hypercar into, you know, I guess, being a partner for the military.

SPEAKER_52: Divergent had really a industry agnostic and product agnostic brief from the very start. So we'll talk about the hypercar.

SPEAKER_54: We'll talk about the U.S. government work. But really, from the very beginning, the vision was, can we create a manufacturing business and an engineering business that ultimately would own and operate factories across the globe that hit low unit cost while being product agnostic? Right. What does that mean? So have a factory that can do a chassis for a city vehicle, a chassis for a performance car, and then a cruise missile airframe, all on the exact same hardware that's running in that factory and can switch adaptively between those product sets and then can scale to be very large and also hit a low unit cost. So, yes, we start in auto. And when we started in auto, we actually brought in-house a vehicle development, and that became a separate company, Zinger Vehicles.

SPEAKER_56: Yeah. And then we also started serving the large auto OEMs, including McLaren, Aston Martin, Bugatti, and others, making their structures and shipping those structures actually from our startup

SPEAKER_54: in Los Angeles over to the EU as a tier one supplier. And about three years ago, we got involved in the defense line of work using that same engineering practice, AI-driven engineering practice, to manufacture airframes, piece parts, and complex assemblies for the U.S. military.

SPEAKER_59: You're building Kinkos for 3D printing stuff.

SPEAKER_36: You know, we all laugh at it, but the idea is, hey, there's a global series of locations to print this stuff. So, my question to you is, there's 3D printing companies that make the hardware.

SPEAKER_34: You use the best of those. Like, I know we were investors in desktop metal. They were doing metal 3D printing. There's a bunch of different people using different types. So, you're the layer, the application layer, the service layer for those type of 3D printing

SPEAKER_64: technologies. Am I correct? We're actually both.

David Friedberg: So, we, you know, in terms of analogy, I would say, think about AWS or TSMC.

SPEAKER_54: Think about cadence as the front end. So, we are both the engineering application that's going to take functional requirements for something like an airframe, engineer that airframe very rapidly into a higher performance point. And then take that design and manufacture it, both the prototypes for competition. But then where we really specialize is in the high volume runs. We want to make tens of thousands, hundreds of thousands of parts. And we actually build our own 3D printers as well.

SPEAKER_56: So, we don't sell the 3D printer, but we are vertically integrated through the printer, through the assembly system, through our post-processing, all the way down to the material card where we generate our own material chemistries as well.

SPEAKER_22: What's the option and the number of units where people should use 3D printing versus going

SPEAKER_34: to a factory that either stamps it or uses other technology to make millions of items? Obviously, we're not 3D printing iPhones or we're not 3D printing our whoops. So, where is that number today? Because in the past, people used 3D printing to build the prototypes, but not to build the finished products. At some point, we moved over to the finished products being done on 3D printers, yeah?

David Friedberg: Yes. And a bit of education. When we started 10 years ago, we were printing at about 10 cc's per hour.

SPEAKER_54: We're printing now in the high hundreds of cc's per hour. So, there's been much greater, actually, than a 10x improvement in cost productivity. So, the market fit for 3D printing of metals has changed drastically in the last decade. And now you're seeing that in aerospace and defense work, we're actually more affordability typically per unit than any of the traditional manufacturing methodologies. And our highest volumes outside of class 1 drones, which I would say we don't specialize in, and usually those are plastic structures. But when you get into metal structures, maybe the highest volume runs you're seeing is around 10,000 per year. At 10,000 per year for defense-type geometries, additive manufacturing, but having designed for additive manufacturing, which is critical as well, you need a design for this manufacturing technology, and have the downstream processes in terms of post-processing and assembly, that methodology is actually going to win out at about 10,000 units plus. So, in A and D, you're covering the full spectrum. In auto, the geometries are simpler, which means you can actually get a cheaper cast part in auto than you can typically in defense. And the volumes range much higher, right, into the millions of units per year. So, in auto, I would say the crossover point today is in the high single-digit thousands. If you're above 30,000 per year, you're probably not going to look for additive manufacturing today, 3, 4, 5.

SPEAKER_61: Additive manufacturing means what in this? 3D printing of metals.

SPEAKER_77: 3D printing, yes. And 10 cc's, that's cubic centimeters, so I'm going to take a guess?

SPEAKER_78: Yes, yes. Okay.

SPEAKER_79: Starting a company is hard enough without having to take on a ton of time-sucking administrative work that doesn't move anything forward or improve your product.

SPEAKER_81: You need to get back to running your business and clear the back office work from your plate. Every can help you with that. Every is the all-in-one platform from incorporation to banking, payroll, benefits, taxes, basically your entire back office. And they've worked with over 1,000 startups. If you're just getting started, they're going to handle your Delaware C-Corp, your EIN, and getting yourself a registered agent, all with no legal fees or delays. And if you're further along on your founder journey and you need to scale up your team, Every will help you hire contractors in over 200 countries. Plus, you're going to get 3% cash back on every dollar you spend with your Every Corporate Card. So head over to Every.io and stop wasting time on the stuff that isn't growing your company. That's E-V-E-R-Y.io.

SPEAKER_34: So this has gotten much faster. The Fidelity has obviously always been pretty great, but it's perfect?

SPEAKER_40: Or is that an issue today? Or has that issue been put to bed long ago?

SPEAKER_54: That issue has been put to bed. And quality was another large issue. How repeatable and how certifiable is the end article, especially in defense where you're talking about a cruise missile on an exquisite aircraft, you're lug mounted, you need to make sure that unit is 100% safe for the pilot of that aircraft. And that's where we have put together years of data and performance history to be able to show the Air Force, the Army, the Navy, our end customers, right? As a tier one, we supply companies like Lockheed and Raytheon and the Neo Primes, that it's a high quality system that actually has better controls over it than traditional manufacturing alone.

SPEAKER_48: Lana, I'll let you ask a question next. But my final question on this technology is the speed at which you make, say, the drone.

SPEAKER_77: And if we were using the drones that Ukraine, Iran, and other folks are doing these commoditized drones, I don't know if you build those or just the ones that fly 10,000 feet in the air and are like the, you know, Desert Storm drones, the ones that were much more sophisticated than the standard ones that are being used in Ukraine today and Russia during that conflict.

SPEAKER_09: Are those getting to the point where they'll be made in the field, closer to the field? And what's the speed? How many come off the line, et cetera?

SPEAKER_52: For plastic printing, absolutely. I think you're going to see this forward deployed.

SPEAKER_54: You're going to see it in Connex boxes. You're going to be making your class one type drone in high volume in the high thousands out in the field. Now, that is one element that we need for air superiority and to win. But another element is largely munitions, right? And we talk about the munition crisis today from large, more exquisite systems like Tomahawk, but also the affordable mass programs like the family of affordable munitions. There you need a system that can carry 100 plus pounds of payload that can fly at hundreds of miles per hour and can go long distance, say a thousand miles. And that's where we really specialize in making those complex metal structures and getting those into the thousands. A single one of our 3D printers can do about 200 of your typical cruise missile airframes per year. So if you have a hundred of those running, you can do about 20,000 of your typical airframe out of one factory. And that is high volume scale. And that is a factory that you can quite literally copy and paste similar to a server farm. If you're thinking about infrastructure, you can set that up across the U.S. And the beauty is if you have a hundred factories across the U.S., you can share capacity across them and you can really have surge capacity. You can go from commercial work to military work. You can rotate in between the two to the demand signal of your end customer.

SPEAKER_94: Yeah. Wow.

SPEAKER_95: Incredible how far this has gone. Conex is just basically containers, right? The military came up with that concept long, long ago. Yes. C-O-N-E-X, which is just what is the standard issue today. I think they were different sizes when they started. But during the Korean War, they needed a standard box for shipping military stuff. And all the great shipping containers now, Lon, use that same standard.

SPEAKER_100: There's a book about this, too. I started reading.

SPEAKER_81: It's about the containers, really? About shipping containers? Literally just the shipping container and how the shipping container changed the world.

SPEAKER_09: Literally having that become a standard just made the entire world became like a unit of

SPEAKER_104: measurement. I'm trying to think of like a gallon of milk or whatever.

SPEAKER_09: It just became this crazy standardization that changed the world. Yeah. How the shipping container made the world smaller and the world economy bigger by Mark

SPEAKER_17: Levinson. This book went like, I read the first like three chapters and I'm sorry to Mark.

SPEAKER_106: I was like, I don't know if I have a thousand pages.

SPEAKER_25: Did you see that Wall Street Journal piece this week? Dad books are a dying breed. You're single-handedly keeping the dad book alive.

SPEAKER_29: That's the most dad book I've ever heard is like, here's the story of shipping containers and why shipping containers.

SPEAKER_84: History, military, these are dad books. Yeah, dad books. Yes. This is like the history channel of books. Yeah, exactly. Lon, do you have a question?

SPEAKER_25: I did. Well, one thing that is interesting to me. So earlier this year, Divergent announced the Venom prototype they've worked on with Mock Industries. I'll bring this up right here. And here's a picture of Lucas with our Secretary of War, Pete Hegseth. I'm curious. Oh, wow. How do you-

SPEAKER_41: Oh, look, they're doing- Are you guys doing shots in that? Is that- They're not doing shots. It's very professional.

SPEAKER_25: If you can't see the visuals, it looks very professional, folks. But I guess- I kid, I kid. My question would be like, there's got to be so many startups and companies that are trying to get that meeting, try to get into that room in front of that guy who's got the power to say, yes, I like this technology. So I'm curious, like, what was that? Is it really just pure meritocracy? Like, you have the tech that is so interesting that the military can't say no? Or was there some lobbying, marketing?

SPEAKER_29: Was there some, you know, like, how did you get into that room?

SPEAKER_52: It comes down to performance, really. I think they're looking across what companies are making an impact today that they're hearing about from the end services as well.

SPEAKER_54: And we're shipping thousands of units per year out of this Los Angeles-based factory. And they know where the hardware is ending up. So they're looking at, okay, Divergent is on over 20 of the programs that we care about with hardware on those. We need to go see what this team is up to, what this factory looks like. And then they think about the technology. But I think it's actually impact first, and then the technology as the second set. And we're known for our additive manufacturing work, but really, we engineer multi-material as well. We have multiple manufacturing methods. Sometimes we're combining off-the-shelf composites with additively manufactured structures and doing these hybrid systems for large airframes, say 30-foot large autonomous airframes. And I think that's really what the secretary was looking at is what companies are having the most impact, what companies should I be highlighting, and what companies do I need to scale so that we can hit our framework agreements and so that we can invite more competition and more volume into the defense space.

SPEAKER_48: And what magnitude cheaper are you at building a drone than Lockheed Martin or whoever else

SPEAKER_126: you're, you know, those major, what do they call it, the major primes?

SPEAKER_52: Yeah. Yeah. I think we say primes and that neoprimes is how I typically hear it. So you're a neoprime, you're a new prime.

SPEAKER_129: So we're, or a microprime, are you a mini prime? An antral's a maxing?

SPEAKER_131: We're actually not maxing.

SPEAKER_56: We're actually not a prime at all.

SPEAKER_54: So we are the infrastructure layer that the primes tap into. So we always are a sub supplier to the primes and the neoprimes and the big legacy guys. So just as much as Lockheed's a customer, Mach is a customer. And what we're doing is really trying to force multiply for those primes, right? Make their products faster to market, cheaper per unit, and then support them at scale to get to thousands per year. And that's also why I think why the US government looks at us in a unique way is they see we're not competing head on with the legacy guys or the new guys, but rather we're going to help that entire ecosystem get to a lower cost point and move faster. And when you ask, you know, what do we enable versus legacy designs, a couple of metrics that are typical for a cruise missile that we're supporting, an airframe might go from around 200 parts down to under 10 parts. Fuel volume is likely to increase 20 to 30%. Mass is likely to decrease 20 to 30%. So you're getting much higher performance, reduction and failure methods. And then you gave the Mach example, that was 71 days to flight. We're typically able to deliver mature CAD within about a month and a half, two months, get first hardware within a couple of weeks thereafter, and actually then just ramp continuously from there. And the same exact system that's delivering the prototypes is going to be used to deliver production hardware. So essentially your first unit is production grade. On the cost side, we've had programs where we've actually been 50% cheaper per unit. Sometimes we're helping engineer products like Coaspire's Rackham system that's, you know, in the low hundreds of thousands. What's that system, Rackham?

SPEAKER_134: What's a Rackham?

SPEAKER_54: Rackham is a adaptable, affordable cruise missile platform. And it is flying through an FMS contract right now, but also going into the U.S. Air Force

SPEAKER_56: here in the future. Lon, you had a last question.

SPEAKER_25: Yeah. One thing that interested me when I was doing the research, you start, you're now the CEO and again, the co-founder, sort of acting chief. You started the company with your dad. I'm curious, what's that like? And what do you think, what are some of the best messages or lessons that you got about running a company from the old man?

SPEAKER_141: I've always wanted to have a merch store here where we sell this week in startups gear.

SPEAKER_144: And once they get around to actually bringing my dream into a reality, there's no better partner that I'm going to use than Shopify. That's right. We're going to set up a Shopify store here for this week in startups merch. And when we do set it up, we're not going to have to wait weeks to get things up and running. Nope. I could just start selling our coffee mugs, crew jackets within hours and all the infrastructure, all the ecosystem from integrating third-party apps to Shopify seamless checkout process. They're all in place already. That's why Shopify is the e-commerce platform behind millions of businesses around the world. They are cooking with oil. They are getting it done at Shopify. In fact, they're powering 10% of all e-commerce in the United States from household names like Heinz and Mattel to companies that are just getting started. So it's time for you to turn your what-ifs into.

SPEAKER_146: Sign up for your $1 per month trial today at shopify.com slash twist. That's shopify.com slash twist.

SPEAKER_52: Yeah, it's been a tremendous blessing, I think, for both of us.

SPEAKER_54: Not something either of us had imagined, but something that has been just an incredible journey. Kevin, my father, really had the idea for Divergent, right? How do we hit low unit cost and stay product agnostic? I'd studied engineering, gone out of engineering, done some finance work, and then decided I'm going to roll the dice with my dad here and aim to help him build this startup. And my first five, six years with the company, I got to build our tech platform. So focused on the robotics arena, then that and manufacturing, then the software. Combined those three and then moved on to run operations first as president and COO, now as CEO. But it's been tremendous to work in the trenches with my father early on to celebrate some of our biggest highs, our first capital raises, our first assemblies, our first production shipments. It's absolutely incredible. And my father, more so than I say business, you know, truths that he's passed on, although he's passed some of those on as well, it's really been what is the character development required to run businesses and do hard things. And that's been a lifelong lesson from him to me and one that certainly has enabled Divergent

SPEAKER_149: today.

SPEAKER_95: All right, continued success and thanks for keeping the good guys armed and hopefully we

SPEAKER_48: can end some of these conflicts. I'm a bit concerned here about how long this Iran thing is going and how Ukraine never finished.

SPEAKER_64: I mean, in some ways good for the military industrial complex, but I think we need these things to wrap up. What's your take on that, Lucas? Is Iran going to wrap up or what is that?

SPEAKER_149: What's the word on the street? I think it comes down to what hardware does the U.S.

SPEAKER_54: have and how much can we deter conflict and put an end to conflict as well. So I'm a proponent of deterrence and making sure that we don't start any unnecessary conflicts or our adversary doesn't start them because they see the U.S. is so prepared. And that's what we're working on every day. That's why our team's also motivated. We can stop something here if we're prepared.

SPEAKER_36: Yeah, it's well said. Speak softly, carry a big stick.

SPEAKER_144: I think we speak loudly right now. Our current president's a little bit boisterous and loud. So it's speak incessantly and carry a lot of sticks and bring a couple of CEOs with you.

SPEAKER_25: Teddy Roosevelt, who said that quote, was also kind of a loud guy, but he figured out there's there's certain occasions for that. There's times for it.

SPEAKER_36: There is, you know what, yeah, move silently.

SPEAKER_48: Like sometimes you want to just say less and have that big stick. All right, Lucas, continue to success. Yeah, thanks, Lucas. With your starting. Great time here.

SPEAKER_158: Thank you. Yeah, great to talk.

SPEAKER_77: It's amazing how this military, you know, these neo-primes and the competition has been infected by entrepreneurship and also by the venture capital industrial complex. Those two things are absolutely fantastic.

SPEAKER_141: And it was an area where we believe that startup ecosystem would not play a role.

SPEAKER_34: And now startups are playing a role, which is pretty fantastic because if you look at that, it was a non-competitive space. Right.

SPEAKER_77: That now has massive competition from startups injected into it. And the way startups compete is they make things that are better, cheaper, faster, right?

SPEAKER_34: Yeah. Better, cheaper, faster has always been the mantra. Um, and that just happens to parallel what the game on the field is, which is commoditized

SPEAKER_95: small munitions, drones built to order, you know, in the field, not years before, not

SPEAKER_77: on decade long contracts, but literally being made in real time.

SPEAKER_171: So it's just a different, uh, it's a different, it's a different, it's a different front.

SPEAKER_29: It's a totally different way of thinking about preparedness. I mean, you read about like before the U S entered world war two, before the world, the,

SPEAKER_25: the Pearl Harbor attacks, the U S Navy had already been working on these huge aircraft care, these huge ships that they were going to need to send out. It, there was just foresight. If we hadn't done that before Pearl Harbor, the whole war in the Pacific would have been very different.

SPEAKER_29: And so today we're sort of, we don't have that same risk factor because we can more quickly catch up and ramp up. So even if we hadn't necessarily seen that coming, we would have been more prepared. Well, we were, we, I think before these Neo primes, we were having the problem of these

SPEAKER_36: things were getting built too slow. And the, the business model, this is really important for founders incentives matter. So the business model previously was called cost plus you, there's a cost for it and you get a little percentage above the cost. Cost.

SPEAKER_34: So then the incentive to drive down cost and to do it faster and cheaper doesn't exist. If you make something faster and cheaper, you make less money. Right. In other words, if I'm getting paid 10% over whatever it costs to make this missile, the missile costs a million, I get a hundred thousand. Yeah. Well, what if the missile could cost 250 K should, should I get instead of a hundred K, maybe

SPEAKER_77: I should get 20% of the, you know, the, the cost savings, which would be, you know, 20% of 750 K saved, maybe I should get 150 K. Maybe I should make 50% more by making something 50% more profit by making something more affordable.

SPEAKER_29: And it was these incredibly long timelines. It would be Boeing's working on this new kind of plane for 15 years and a startup can't compete with that. They don't have 15 years. They need to start selling things today. And so by shortening those timelines and changing that the whole landscape has now really shifted. I mean, we'll talk about it on the show all the time. We have companies that are innovating in ways to do missiles and drones and autonomous vehicles and all sorts of AI for, for the battlefield. And I mean, it's just, it feels like every week there's a new.

SPEAKER_64: And you know, I love to try to understand the world. And the first thing I do, we have a great partner in poly market.

SPEAKER_77: And, uh, I, I just go to poly market and I look at Russia versus Ukraine ceasefire agreement by, and, uh, looking at the end of the year, uh, people were thinking if you, if you go to that December 31st and you click on it lawn, you can see the order book over time. Yeah. And obviously the order book started high, you know, people thought it was an 80, 90% chance as we get closer, people are like, Whoa, this is dropping. And so the, uh, the number of people, you know, uh, involved in this million dollars.

SPEAKER_09: And now it's only a 43% chance that we have a ceasefire agreement by the end of the year. Majority chance is that we will still have this war going into it. It's I think fifth year. I can't believe it. Like with the first episodes of all in, we were debating this Ukraine Russia war and that Russia would be in Kyiv in four days. Yeah. They're like, they'll be in Kyiv in four days. I it's just unbelievable to me that people got it so wrong.

SPEAKER_29: The conventional wisdom was Ukraine is this small country. It's this relatively small military. They're going to be rolled over by the massive and, you know, incredibly potent Russian army. And yeah, I think very, very crazy to see this many years later, they're still holding their own and it's become such a pivot point for the entire world.

SPEAKER_142: Like all geopolitics at this point in some ways is navigating around this conflict. It's, uh, it's wild to see.

SPEAKER_48: Well, and it, uh, I think speaks to these paper tigers as I heard, um, you know, somebody, I don't know if it's Graham Allison or somebody referred to it as paper tigers.

SPEAKER_77: Like you think Russia or Iran are these incredible military forces, North Korea, obviously. And it turns out some of them may be there.

SPEAKER_144: They haven't been updating their weapons systems and technology.

SPEAKER_29: Yes, they're, they're incredibly potent militaries from a late nineties, early aughts perspective.

SPEAKER_25: But from a 2026 perspective, when yeah, a swarm of small drones can get through your defense

SPEAKER_29: shield that was designed for, you know, jets or whatever missile systems, uh, it's a whole new world and it really has changed the fundamentals of that, that fight.

SPEAKER_188: Yeah. So thanks to poly market for keeping us all informed.

SPEAKER_77: Yeah. It's just so great to know where the sharps are putting their money. There's always people who have analyzed these situations really deeply.

SPEAKER_47: I, I got another one.

SPEAKER_191: Do you want to, do you want to say? Oh, you do?

SPEAKER_29: Okay. Yeah. One more poly market that I thought was fun. They're obsessed with SpaceX. There's a hundred million poly markets about every aspect of the SpaceX coming IPO.

SPEAKER_25: Here's one I thought was funny. What will SpaceX's public ticker be so far? 98% are saying none of

SPEAKER_29: these options. The leading ones are X or S E X that would, which would feel very Elon to do, uh, yes, SpaceX S E X. But I, you know, I, I thought space or Mars makes sense. They're, they're three cents and two cents to buy space and Mars. What do you think the SpaceX, uh, public ticker symbol

SPEAKER_193: will be? I, you know, I thought this got resolved and they announced it. So I am.

SPEAKER_194: Well, maybe that's why maybe that, but this says, I think that's why.

SPEAKER_142: Yeah. So, uh, S P C X, I guess is what we're the conventional wisdom. Yes. I think what you're seeing is that it's resolved now.

Chamath Palihapitiya: Oh, okay. There you go. That would make sense. Uh, so the other ones now are no more, no more too bad.

SPEAKER_195: And that's a good one. Why wouldn't we do that?

SPEAKER_196: Yes. Big cloud providers may offer you cheap compute, but you'll end up paying the difference in engineering costs and hiring extra developers. You don't want to waste time configuring virtual networks. None of us do, or your access policies. You want your team building your product. So it's time to look at render render is the all in one cloud platform for developers that allows you to deploy scale and secure your apps and agents with zero ops. Most cloud platforms ask you to split your focus between product and infrastructure, or they force you into platform constraints. You'd know you're going to grow in six months, but just connect your GitHub repo to render and you are live web services, cron jobs, the whole stack in one platform. It's time to find out why 5 million developers are already using render 5 million. Go to render.com slash twist and apply for the render startup program. You'll get anywhere from $500 to a hundred thousand dollars in free credits, depending on your stage and who your backers are. That's rendered.com slash twist.

SPEAKER_05: Well, I actually, I think the funny one would be if everybody is, uh, in agreement that Tesla and SpaceX merge eventually, uh, which does make sense, uh, in many ways, then why not just make it

SPEAKER_64: dollar sign E-L-O-N? You're basically betting on Elon. So just make a dollar sign Elon.

SPEAKER_142: Yeah. It does feel like Musk industries, like Stark industries at some point is just going to happen.

SPEAKER_205: Like it's just, we're going to put them on or whatever. Okay. Yeah. It's also great for, um, sports. You know, I love watching the poly markets revolve, resolve themselves in real time. So I'm watching my next one again last night. I'm going to go Saturday to Cleveland.

SPEAKER_29: I was so mad at the, the, the final Knicks 76ers game. They were, it was almost even odds on poly market. And I did think like the Sixers have kind of not been doing so bad. It seemed to me like it was going to be the Knicks. I didn't bet. And then of course the next, the next one, I would have made the Knicks

SPEAKER_15: are rolling now nine games. They've won in a row and I'm going Saturday. You want to hear a crazy

SPEAKER_77: story. Talk about off duty. Uh, you know, I've sat courtside now. Sure. Twice in this nine game,

SPEAKER_05: win. I sat courtside for the massacre in Atlanta when the Knicks won. And then I sat behind the

SPEAKER_34: Knicks bench and then a little bit behind in MSG. And then I went to the Sixers and my friend Ben

SPEAKER_05: Stiller and I sat courtside there. They tried to stop me from buying those tickets at the Sixers. Um, Atlanta didn't seem to be, uh, care too much. Now there's a report, Dan Gilbert, the owner of

SPEAKER_129: Cleveland and I can kind of confirm this as a breaking news story here. I, I was acquiring

SPEAKER_212: a pair of courtsides for Saturday night. Person was going to sell them to me. He said, I got a

SPEAKER_77: problem. I said, what's the problem? He said, if I sell them to a Knicks fan, they're Dan Gilbert's making everybody sign an agreement on DocuSign before you buy the courtside seats that you won't give them to a Knicks fan or you won't take a Knicks fan with you.

SPEAKER_215: Some binding. Can you do that? I don't know. I don't, I think it's well, I, I think he can make

SPEAKER_09: rules unknown, I guess is the question. Once they're on the secondary market, he can't stop it, but this is the primary market. So I think he can make whatever agreement he wants with the ticket holders who are seasoned ticket holders or who have access to those tickets. And they also blocked with their partners, anybody from buying tickets who had a credit card address that's outside of

SPEAKER_195: the particular area. Yeah. If you don't live in Cleveland. Yeah. Which was pretty easy for my team

SPEAKER_09: to resolve by the way. I mean, I don't want to get into it, but like, you know, if you have a couple of banks who work for you, issuing you a credit card or popping up an office space for a day in another

SPEAKER_38: city, not really a big deal. We're going to bring on the co-founders of Outro Health. They're

SPEAKER_29: Brandon Good and Mark Horowitz. This is a company, they are a virtual health service, Jason, that is helping people taper off their prescription antidepressants. The SSRIs like Celexa, Zola, Prozac, Lexapro, you know, this category of drugs. So I know it well, and they were, I believe at launch

SPEAKER_129: festival and everybody thought, Hey, this is an incredible idea. And I was really interested in

SPEAKER_05: investing in it. Um, and meaning and spending more time with the founders. Why I think these

SPEAKER_34: things are massively oversubscribed. I have no issue with, you know, people with a doctor doing what they think is best. However, I do think the incentive system in America personally,

SPEAKER_79: professionally, and then economically is designed to push people towards quick fixes

SPEAKER_09: that are extremely profitable for the pharma industry. If you look at, uh, what happened with

SPEAKER_228: that Sackler family and what was the, Oh, you're talking about, uh, Oxycontin, Purdue,

SPEAKER_205: their, their companies, Purdue and right. Yes. Purdue making Oxycontin and I watched this, uh, dope sick film, uh, or it was a series. It was a Hulu series. Yeah. Incredible with Michael Keaton,

SPEAKER_09: I believe it was Michael and Peter's my friend, Peter Sarsgaard, um, who I was in center of the world with, look it up. This is, was an incredible story of just how the pharma companies,

SPEAKER_79: their sales channels, the incentives can get misaligned profits can ensue. And then anybody

SPEAKER_09: who subs their toes seemingly would get a bigger and bigger dose of Oxycontin, get addicted, then get removed from it and then wind up taking fentanyl or some street drugs and dying. I look at it now, young kids now are being massively overprescribed. A young kid has a problem. They have anxiety created by their smartphone, by school, by whatever it happens to be. And they just wind up pressuring you. And I've had many parents come to me. Hey, what do you think, Jake? They're pressuring me to put my kid on anti-anxiety meds on this and that. And I'm like, your kid's 10 years old and they're literally putting pressure. I'm, I'm not kidding. Nine, 10, 11 year olds, they're pushing them to scramble their brains. Now an adult,

SPEAKER_25: totally different, but I can say from personal experience, like even if you go on an SSRI and you go back to your doctor and you say, I don't like it. Like I I've been on Zoloft and I went back and said, I don't like this. I don't think it's working. They don't say, okay, let's try an alternate idea. They just try the next SSRI. Like, well, let's try Lexapro. Well, let's try

SPEAKER_29: select. So, well, let's try. Like the, the idea is one of these is definitely going to work. We just need to fine tune it. And I don't, I don't personally know if that.

SPEAKER_36: As opposed to what I told you, which was, Hey, go to dinner with your friends, work out, take three weeks in Italy. You know, like there's other modalities here.

SPEAKER_129: Anyway, let's hear from the founders how they're helping people taper off. Introduce our founders

SPEAKER_29: here. Yes. We've got Brandon Good and Dr. Mark Horowitz. They are our co-founders. Gentlemen, thanks for joining us here on the show. Thanks for having us guys. Yeah. Uh, I, I like Jason, they call themselves the inverse hymns. I'd love to hear you guys. The inverse hymns. Okay. Yeah. Unpack it. What, what's with the inverse hymns idea?

SPEAKER_244: Yeah, I can, I can speak to the commercialization side of it. So my background was actually, um, Canadian, moved to the U S learned what a copay was for the first time and thought that was strange. And ended up launching GLP ones at Nova Nordisk at their headquarters, uh, and realized pharma had no interest in getting people off or doing any research to get people off medications, unless you had a business case for it. So I came across Mark's work on hyperbolic tapering, which you can talk more about. And we've turned this hyperbolic tapering model into a virtual care platform that allows clinicians to help people safely taper off antidepressants. And we're expanding to other psychiatric medications that follow this principle. And as Mark will explain from this methodology, it goes beyond psychiatric medications, even to things like proton pump inhibitors, opioids, and even GLP ones all follow this hyperbolic curve of how you reduce the dose, monitor the patients to make sure that it's going at the right pace for them, supply them with compounded medications so that you can follow this pharmacologically accurate staircase and then help them with non-drug tools to replace the medication so that they can sustain their mental health long-term like going to dinner with your

SPEAKER_247: friends or going to Italy. Now there's good evidence around that. Thankfully that doctors can recommend.

SPEAKER_77: Well, I let's start with first principles, Mark. You're an MD, correct? And you've prescribed these drugs, uh, I'm assuming or worked. Okay. So you heard my little preamble there. I have a lot of friends

SPEAKER_09: who come to me and ask me, Hey, Jake, how you seem like you're pretty well adjusted and happy. Uh, how do I, how can I get there? And I say, okay, listen, here's my big five sleep, diet, exercise, meditation, socialization. I work with my kids on those concepts, my friends on those concepts. If I have an employee or a founder in need, and they say, Hey, I can't sleep at night, et cetera. I'm like, okay, here's the protocol. We're going to get you on a whoop or an aura. We're going to start looking at your sleep. We're going to get your eight sleep dialed in. We'll work on sleep. Then we're going to work on diet. Then we're going to work on exercise. Here's your meditation with calm or waking up app, whatever it happens to be. And then socialization, are you going out and breaking bread three nights a week with people who you love, et cetera. That's my protocol. But I want you to talk from first principles about how we got here. We'll get to tapering, of course, but how did we get here to where they're not aligned? And am I crazy sounding like Tom Cruise when I give my protocol, or am I prescient and I've refined this over time by helping my friends?

SPEAKER_00: So you're not crazy. Let me tell you how we got here, because it'll make sense of how we go to get out of here. There's been campaigns run in America and around the world, mostly by drug companies that have convinced the public that they are suffering from chemical imbalances, that if they're feeling sad or they're feeling anxious, it's something that's gone wrong in their brains. And that has been a very successful campaign that now most of the population believes they've got a chemical imbalance. And that story began as a hypothesis, you know, the 1960s by psychiatrists. And the short answer is it hasn't panned out. 60 years of research looking in people's blood, cerebral spinal fluid in their brains haven't found that depressed people have lower serotonin than healthy volunteers. You know, in fact, by the age of 45, about 70% of us will experience clinical depression or anxiety. In fact, you know, it's very, very common. It happens to more of us than it doesn't. And it's not possible that 70% of us have something wrong with our brains. You know, what the data shows is the number of stressful life events we experience determines whether we'll become depressed or not. You know, the sort of thing that your grandmother probably said, when life is tough, you feel bad. If you get divorced, lose your job, move across the country, and your mother dies in a year, chances are you're going to be miserable. If none of that happens,

SPEAKER_36: chances are you're feeling pretty, pretty okay. So life comes at you. And these pharma companies

SPEAKER_79: had a thesis, there's a chemical imbalance in your brain. These are SSRIs are inhibiting the reuptake of serotonin. And that means there's more serotonin in your brain, which means you'll be happier. And it turns out there is no blood test or scan or research that has proven this chemical imbalance

SPEAKER_34: to be true. Yet we put hundreds of millions of people globally on this, tens of millions of Americans

SPEAKER_212: on this without having actual proof of the chemical imbalance theory, correct?

SPEAKER_00: Exactly. I mean, you made a comparison to Purdue pharmaceuticals and opioids, and I would say, you know, opioids are more dangerous drugs in lots of ways, they make you stop breathing. But what is similar is the messaging. You know, what began the opioid crisis was messaging from the company that these drugs are not addictive. You know, if you're given it by a doctor, it won't be addictive. That was the sort of thing that allowed doctors to give out these drugs widely. And there are similar messages that have been put out there, mostly by drug companies that have allowed antidepressants to take over. And of course, they're used by way more people than are using opioids. And those messages are, depression is caused by a chemical imbalance. When you stop these drugs, you get mild and brief discontinuation symptoms. It's euphemism, you know, that it's easy to come off. And they're highly effective and life saving. And all of those messages have major scientific flaws to them. And that is what has allowed this to sort of take over society in lots of different ways.

SPEAKER_05: Well, and is that what you believed or the doctors told you when you tried Zoloft? Did they tell you you might have a chemical imbalance or was that your belief when you went in?

SPEAKER_29: No, it was very, it's very explicit. You're depressed because there this, this misfire in your brain. You don't have the right chemicals that you're depleted, you're whatever. And that, yeah, it's, it's nothing, it's nothing to do with me. It's nothing to do with the way I think. It's nothing to do with my, you didn't do anything wrong. It's nothing to do with the way I'm, you know, living my life or what's going on with me. There was most of the time there would not even be a talk therapy component. It wasn't like, we're going to put you on this drug and then you're going to go to a counselor once a week or once every month and tell them about your problems. It was just take these and then you'll come back in

SPEAKER_34: a month and we'll see how it's going. And it's not your fault. It's not your fault, Lon. Right. And that's the same thing, Mark, that they told people with their pain. Hey, listen, you have pain. We're going to give you these opioids. And this is sinister to me.

SPEAKER_28: It's less about fault. It's not like it's not your fault, but I, there, there are no other,

SPEAKER_25: there are no other recommendations. It's not, well, you should probably try to get out and be a little bit more physically active. Well, you should maybe try to like get a little more sleep or wake up earlier. There's none, there's none of that. Here's how behavioral cognitive therapy about how to approach triggers or stressful situations, all the things that would really help a person navigate negative feelings and emotions. You don't get any of that. It's purely take this pill twice a day. And then if you have side effects, they're like, well, maybe you need the other pill or the

SPEAKER_29: other pill or the other pill. There's never a chance of like, maybe it's something else.

SPEAKER_273: Unpack what you're hearing from Lon's lived experience.

SPEAKER_275: So I just wanted to say it's, Jason's comment is really prescient that it's, you know, it's not your fault. And I'm not here to say it is your fault. I'm not here saying, you know, you need to work harder, pull yourself up by your bootstraps. You know, I think the data shows we all have a

SPEAKER_00: breaking point when we have enough stresses, you know, we all sort of have a breaking point. And I think it's good to normalize that. There are, there's research that shows that people that are told they've got a chemical imbalance feel less in charge of their recovery. You know, they've been told there's something wrong with a major organ. They're more lucky to want medications because if you've got a chemical problem, of course, a chemical solution makes sense. And in fact, they're more pessimistic about recovery because now, you know, there's a major medical problem. If people are told accurately, this is a rough spot in your life, you know, you need some help. All the things that Jason mentioned, all the holistic things, people feel more optimistic about recovery. They're less likely to want to use a medication, you know, and they do better in the long run. So I think it's a very, some people say this, this reduces blame. And that seems like an ethical thing to do, but in the long term,

SPEAKER_276: it reduces autonomy, you know, and agency, that's the word. Exactly.

SPEAKER_34: Yeah. I mean, that's the, the amount of victimhood and lack of agency and self-reliance that I think is the trick. It's the, it's the magic trick that these pharmaceutical companies will then give to you like, Hey, here's the trick. You don't have to be responsible. I don't have to be

SPEAKER_77: responsible. There's another thing. And Hey, it's just, this is a, you know, a magic bullet. So

SPEAKER_79: tapering off of these though, I've had friends with the Adderall, you know, which is basically

SPEAKER_205: speed. I've had friends with the Zoloft and the Welbutrin and all this stuff really hard to get off of

SPEAKER_09: really hard to get off of. I think getting off of cigarettes, I've seen my friends get off of SIGs easier than getting off of Adderall and Prozac. Tell us about how hard it is to get off of these and then why this modality that you guys are working on in the systems that you're building

SPEAKER_40: works. And then how do you productize that into a business? I think we'll go back to you,

SPEAKER_283: Brandon, for that. What happens with any drug that you use, whether it's nicotine, alcohol, opioids, antidepressants, Valium, Alprazolam, Xanax,

SPEAKER_00: all those things, your body adapts to the drug. It's called homeostasis. You know, when it's hot outside, we sweat. When it's cold outside, we shiver. If you use a drug that increases serotonin in your brain, your brain becomes less sensitive to serotonin. And when you stop the drug, your body and brain is crying out for it. It's not the same as addiction. This is sort of an issue that people get a bit mixed up with. People aren't craving antidepressants. They're not out of the

SPEAKER_275: street trying to inject more antidepressants, but you don't need to be addicted to something to have withdrawal effects when you stop it. I'm going to guess that Jason, everyone listening is probably physically dependent on caffeine. If you stopped it, you would get a headache, you'd feel bad. I'm also going to guess you're not injecting it. You're not sort of on the streets trying to get more, you know, some occasional night.

SPEAKER_36: Depends on the day. No, I've had, I mean, caffeine is a perfect example. I have tapered off of it before. Massive headaches. And the way I did it, because I was drinking it well into the afternoon, I used to drink an espresso after dinner. The way I actually did it was I would have coffee up until noon and I'd stop or 1 p.m. Once in a while, I'd slip up and have a three o'clock. Then I just told my housekeeper, here's the bag of decaf beans. Here's the bag of regular beans. Mix them into a giant thing, shake it up. And then I started just making half cap. I may still be doing half cap. I don't even know, but I just lowered the caffeine per cup and I just tapered

SPEAKER_06: myself off. Is that the solution you have? That is getting into our territory. I mean,

SPEAKER_00: I guess the main thing to say is when people stop antidepressants, there's a whole range of withdrawal effects that occur. You know, people get emotional symptoms, panic attacks, low mood, anxiety, crying spells, even becoming suicidal, as well as physical symptoms, brain zaps, headache, dizziness. But it's the emotional symptoms that cause all the confusion. When people go to their doctor and they say, I've stopped my drug, I'm having panic attacks, I can't sleep. The doctor doesn't even let them finish the sentence and says, you must be relapsing. That is your underlying depression or anxiety has returned. And that has caused huge confusion because doctors have been taught, look out for relapse, people are going to get sick again. Whereas withdrawal is probably more common than relapse. And that has led to doctors not seeing this issue in patients, which is why there are so many patients that have sort of fled regular medical services to go looking for help online, you know, on Reddit threads, in Facebook, on social media, you know, it's sort of this absurdity. ChatGPT. ChatGPT, exactly. It's this absurdity because doctors have been taught from short term studies, withdrawal is not a big issue. But for long term users in America, there's 20 million people who've been on these drugs for five years or more, and the studies go for eight weeks. So there's a very big disconnect.

SPEAKER_05: Do we know what it's done to people's brains if they've been on it for 10 or 20 years? Because I remember when these were introduced, when I was in my 20s, in the 90s,

SPEAKER_36: we became like Prozac, generation Prozac. They were, I don't know, five or 10% of my Prozac nation. It was a book. Yeah. And literally like 10% of Gen Xers, then maybe 20% were starting to take Prozac, and people were sharing their Prozac with each other. It was just like this new way to feel a

SPEAKER_129: little bit happier, 10% happier. But yeah, speak to that, I guess. Yeah.

SPEAKER_283: Look, I mean, the real answer, and I don't alarm people, is we don't know what they structure in

SPEAKER_00: the long term. You know, the studies go for several weeks. There's a famous professor of psychiatry called Alan Francis that led the DSM-IV committee, who says, we are running the biggest open air experiment on the public ever, ever conducted. Because now in America,

SPEAKER_283: have 50 million people on antidepressants. What I see in my clinic-

SPEAKER_01: There's 50 million Americans on antidepressants? One in seven? One in six? One in six.

SPEAKER_29: I mean, yeah, I would say-

SPEAKER_295: If there's 50 million kids, 60 million kids- It's bigger than the population of Canada.

SPEAKER_29: Yeah. Think about your friend circle, and then think about how many of them have ever taken like Zoloft or- It's almost-

SPEAKER_142: Yeah, it's like a third. It's almost everybody. It's almost everybody I hang out with is on one of these.

SPEAKER_17: Wow. This is just, I mean, I hate to say it, but I feel like it's criminal. Okay. So let's

SPEAKER_40: talk about the business here and incentives, Brandon. Obviously, you've got-

SPEAKER_300: Should I just show the tapering?

SPEAKER_275: Please, yes. Show it, and then let's talk about business models. This is my obsessive, monomaniacal. You'll have to stop me talking about it at some point. You'll have to get a gun out. Can I share my screen?

SPEAKER_303: Because of course you will. You'll save a thousand. Please. Let me try this.

SPEAKER_304: We love a good demo here, and then just remember 80% of the audience is listening.

SPEAKER_00: Okay. I'll try to use my words to explain what you're looking at here. I am showing you, if you can see it, a picture of the way Celexa acts on the brain. X-axis Celexa, Y-axis effect on the brain, and it's a curve. It goes up very steeply, and it flattens out at the top, like the left-hand side of an archway. And the key issue here is not a straight line. The doses that are used most often in America are 20 milligrams and 40 milligrams, and it sounds like that's doubling the dose. But in terms of effect on the brain, because that's near the plateau area, it's actually a very tiny increase in dose. The smallest tablet in America is 10 milligrams, but even a two milligram dose has about half the effect of 60 milligrams. So these teeny tiny doses that some doctors might laugh at and call homeopathic actually have very significant effects. And the reason for this is the pharmacological law, the law of mass action. And what it basically says is, when there's not much drug in the system, all the receptors for a drug are open for business. A bit like the game of musical chairs, it's easy to find a seat. So every milligram of drug has a B effect at low doses. As more and more drug is in the system, receptors are saturated, the chairs are filled, you get less and less effect for your bang. Let's bang for your buck. And so you get this law of diminishing returns. Now, the way that most doctors recommend people to come off their medication is they go from 20 milligrams to 10, that's say half a tablet, then go to half a tablet every second day, that's five milligrams. It sounds intuitive and like it's spread out. But in terms of effect on the brain, it's anything but. Going from 20 to 10 causes a reduction in effect on the brain, a small reduction. Some people already find that quite unpleasant. Some people can handle it. The next reduction from 10 to five causes an even bigger change because you're now going down this steep curve. And the final reduction from five to zero is like jumping off a cliff. It has about 20 times the effects of going from 20 to 15. And so people are repeatedly pushed off this cliff by the way that doctors recommend coming off. And they end up in a screaming mess, headaches, dizziness, brain zaps, panic attacks. And the doctor says, they don't know that this cliff exists. And they say, you must have relapsed. You must need your drug lifelong. And that's what people end up in this revolving door on and back off,

SPEAKER_276: often back on the drug. Dr. Justin Marchegiani Crazy. Dr. Justin Marchegiani Sorry. Dr. Justin Marchegiani

SPEAKER_34: No, I'm saying it's crazy when you look at the, well, I mean, going from five milligram to zero is 58% change. Whereas going from 20 to 15, which is still five milligrams is a 3% change. 15 to 10 milligrams is 6% change. 10 milligrams to five milligrams,

SPEAKER_144: 13% change. That's all taking five milligrams out. But the last five milligrams, as you're saying, those are the important ones. And that should really be done much differently.

SPEAKER_29: Dr. Justin Marchegiani

SPEAKER_25: And it does feel like when you go through it, as I have, it does feel like depression. It feels like, oh God, I'm chronically depressed. I'll never be free of this unless I just go back on the drug. So it really is a, it really does fool your brain into thinking that you can't not be on Celexa or whatever.

SPEAKER_314: Dr. Justin Marchegiani That's a really good point. You know, I say that a lot to patients, our emotions feel very familiar to us. You know, when you, when you have say a hangover from alcohol, feeling depressed

SPEAKER_00: feels a little bit similar to how you feel if you break up with your partner. I mean, we're not so, it's sort of, it's very, it's very, it's very familiar, these emotions. And so it can be very confusing to people and they get caught in this idea. I must need the drug.

SPEAKER_316: Dr. Justin Marchegiani

SPEAKER_320: Going back to that fallacy, we talked about the fallacy of the chemical imbalance. Dr. Justin Marchegiani

SPEAKER_29: There is something wrong with my brain. This now proves it. And I've got to just do what Dr. Justin Marchegiani

SPEAKER_06: Reinforcement loop. Dr. Justin Marchegiani Exactly, exactly. The withdrawal, the idea that there's not withdrawal and the people have a chemical

SPEAKER_00: imbalance have a sort of synergy where they convince people that they need the drug. That's exactly right. So the clever trick that we published a few years ago is rather than reducing dose by even amounts on the x-axis, reduced by even amounts of effect on the brain. And so I've drawn four horizontal lines separated by 20 percentage points on the y-axis. And that corresponds to going down by smaller and smaller amounts in terms of dose down to really tiny final doses. So that, you know, if you imagine you're walking down this curve, it starts off like a country walk and it turns into, you know, coming down a cliff and you have to go inch by inch. And that means that you need to go down to these tiny doses that are much lower than commonly available tablets. So you don't end up jumping off that cliff. And because this, this curve is called a hyperbola, the jargon term that this is hyperbolic tapering. And we published this in the Lancet Psychiatry now a few years ago. And that is, that is the essence of what we do. Dr. Justin Marchegiani Makes total sense. Dr. Justin Marchegiani You know, people, people who come off in this way have a much higher chance of success than people who jump off the cliff. I mean, obviously. And, you know, we've, we've now, there's now studies more and more coming out of Holland, showing that this can help people who couldn't get off in the regular

SPEAKER_327: way, come off and avoid severe withdrawal effects. And that's the core of what we do at Outro.

SPEAKER_326: Dr. Justin Marchegiani I can tell you about this metaphor when it comes to skiing. When I ski,

SPEAKER_77: there is a 3.4 mile, you know, roundabout at the mountain I like to ski at. Very gentle, green and blue difficulty. And then there is a 1.4 mile straight down the mountain. That's a double diamond

SPEAKER_144: diamond. And you go very fast and you get down the same amount of vertical in a fraction of the time. But it is scary as all get out. If you're not an expert skier, do not bomb down those double diamonds.

SPEAKER_06: And I think that's, you know, what you're showing here. Dr. Justin Marchegiani That is exactly it. I mean,

SPEAKER_00: whatever that does to your knees, this is doing to people's nervous systems. And, you know, knees are precious, but nervous systems are a lot more delicate. And so, you know, people, I basically say to people, when you've tried to come off the drug before, you have jumped down, you've gone down

SPEAKER_283: the double diamond ski run. What I'm suggesting is, let's go out the back and go out, you know, on the kiddies snow patrol. So you're going down very slowly.

SPEAKER_331: Dr. Justin Marchegiani Yes, the wraparound is what we call it. It's the old wraparound. You wrap around the mountain, traverse it. Incredible. Okay. So this makes total sense. Brandon,

SPEAKER_09: how does it become a business? How do you productize this into a business? And then I'm curious about the incentives. We had the incentives of the drug companies, which is sell,

SPEAKER_34: sell, sell, sell, you know, increase the dosage, double the price, et cetera, et cetera.

SPEAKER_09: And then we got 50 million people on it. But I suppose that the employer is paying for health insurance. I pay for lawns as an example. We want our premiums to come down. We want to spend less. So now we are at odds with the doctors and the pharma industry, which get paid every time lawn comes into the office. Every time he changes that med, that rings the doctor's register, I believe, and it rings the pharma's register, I'm sure. So talk to me about these, if I'm correct, and I am a neophyte, these opposing forces and incentives matter. We always talk about that in startups, in government, everything in between personal responsibility. Incentives do matter here.

SPEAKER_79: How do you manage a business model with conflicting powerful forces?

SPEAKER_244: Yeah, I'll talk about the incentives and then get into the business model. So, you know, the lens that I apply it from was working on the market development market access side for the GLP-1 drugs, and therefore understanding the incentives from the patient, from the clinicians and the payers, whether public payers like the region of Catalonia in Spain, or Aetna in the US. And there's a really interesting why now when it comes to the pharma incentives, because most of these drugs are actually off patent. And there are some like newer drugs in market that aren't really as popular that are still on patent. Maybe their list price is $1,300, but Lexapro, Zoloft, Prozac, those are all generics by now that maybe cost $20 to $30. So that's really clinical inertia, why patients are still prescribed these things and why they've been on them for so long. The drugs coming down the pipeline in the mental health category, you know, I have my qualms, I think there may be a step forward, but they tend to be inspired by psychedelic molecules that follow rather than a pill per day paradigm, they follow a, you know, one, two, three intervention paradigm. Those drugs actually don't work very well when people are on antidepressants because antidepressants blunt the effect and dull the serotonin response. So essentially pharma, hopefully, you know, there's no black SUVs following me. I do get a little bit scared when there's turbulence on the plane nowadays, but you know, they don't have their sales reps out in full force when it comes to these drugs. When making the argument for insurance companies, we actually, what's interesting is our business model, we don't have to make an argument because what we are doing is medication management, which is already reimbursed. We're just managing the medication now. So we're already in process of getting a network with insurance companies for our care. But moving forward into the future, so we can expand our model, make it more comprehensive, enter what's called bundled payment models, which a lot of opioid use disorder companies do with insurance companies. We're collecting the data and partnering with universities to publish on this data to show that getting off antidepressants and other psychiatric medications with outro in the right way that's personalized and supported is much cheaper than getting off antidepressants the wrong way, which can lead to missing work, you know, lower productivity, going to the hospital, getting prescribed new drugs, entering a prescribing cascade. So that's part of the strategy there that we're employing.

SPEAKER_48: Yeah, really interesting, Lon. I think incentives matter so much here. I think we're going to enter an

SPEAKER_77: era where as healthcare costs are double, triple the rest of the world, you know, the United States has the best health care in the world. We also pay through the nose, like triple most other countries that have socialized medicine. So there's, you know, this disparity we're seeing. And one of the ways

SPEAKER_36: to close that gap is to not give care that's unnecessary and unravel that. And then of course,

SPEAKER_48: avoid downstream negative impacts. And so that's why I just love this company is, hey, it's clear that

SPEAKER_34: 50 million people don't need to be on these in my estimation, and many others. And if that's correct, and we can get that number in half, that savings goes right to the bottom line. And those people have, you know, a better life, then what are all the downstream things from their dependency on this?

SPEAKER_77: Oh, my Lord, you know, in terms of productivity, etc. And it's the same thing with GLPs, you know,

SPEAKER_41: which I'm a big fan of. It's also, yeah, motivation was always a big thing with me, you kind of it

SPEAKER_29: dulls you out in a lot of ways, which is great if you're feeling intense panic or anxiety or depression,

SPEAKER_25: but it also means your creativity, your passion about your hobbies or work, you know, doing your

SPEAKER_29: job, like all of those things also suffer. So I do think if we managed to get, you know, 20 million Americans off of these drugs, we would probably see a lot of new businesses,

SPEAKER_25: a lot of saving billions of dollars, saving the money, but also we would see people more active productivity in the world because they're not having their emotional state sort of

SPEAKER_48: blunted by these drugs. Mark, where can people find out more and maybe just some closing thoughts from

SPEAKER_00: you on the humanity at state care? So they can go to Outro Health, we've got a website online, you can look at my work, I've got a dinky website, markhorowitz.org. I'm on Twitter, Mark Horowitz. I just want to just jump onto that last point because that, you know, people, we talk about how to come off the drugs, we talk about why drugs might be overprescribed, but we haven't talked about the side effects of the drugs. So just briefly worth saying something about that and you put your finger on it well. You know, these drugs cause a whole series of side effects that people often don't pin on the drugs. Because they're on it for years or decades, they don't kind of attribute what they're experiencing to the drugs themselves. What these drugs cause, they cause daytime tiredness, we know that they disrupt nighttime sleep, they cause impaired concentration, impaired memory, that's been shown in healthy volunteers, that's not an effect of depression. They cause weight gain, they cause sexual dysfunction for most people who are taking the drugs. Less desire, less ability to get an erection, less ability to orgasm in both sexes. We know that for some people, even when they stop the drugs, that will persist. That's a big deal affecting a lot of young people

SPEAKER_326: at the moment. That's really upsetting. It's called PSF. Do you think this correlates with the

SPEAKER_34: population decline and the lack of people getting married and having families? Is that why this last

SPEAKER_283: couple, or is it contributing? I mean, it makes sense that it would- There's probably a lot of factors going on, but you know, the fact that 10 to 15% of different demographics are on these drugs that

SPEAKER_00: profoundly affect sexuality, and that has to be excluded. We're going to do a study looking at the relationship between asexuality and taking antidepressants. I don't want to jump to

SPEAKER_260: conclusions, but it's an obvious hypothesis that's worth looking at. Yeah, hypothesis worth checking out.

SPEAKER_77: There are these, this incel movement, people on Reddit, subreddits, you know, playing video games,

SPEAKER_34: not being interested in girls. You know, it seems to be like this male component of like,

SPEAKER_77: not pursuing the opposite sex. And then maybe that they don't couple or find a mate and make babies. It's really important for society to think about that correlation. I think the other thing that goes

SPEAKER_00: with sexual numbing that you've put your finger on is the emotional numbing. You know, people ask, people say to me, well, if these drugs don't fix serotonin problems, what do they do? There's all these theories about urogenesis, inflammation, stress hormones, there's a whole lot of kind of biology out there. But if you ask people on these drugs, how do you feel? Three quarters of them say emotionally numbed. They say that their normal range of emotions from very positive to very negative is squished into the middle. And, you know, as you said, if you're very, very panicked, turning the volume down from a 10 to a three can be a relief, but it also affects your interest in life, intimacy with your partner, relationship with children. It's a very big price to pay for a short term fix, which I think is why it makes sense if you use these drugs to think about short term use, not long term use. So there are a lot of effects, as you say, on productivity, on creativity, you know, on a whole range of things that I think has very big effects on the wider economy. And people should be thinking about what is this drug doing to me in the long term? You know, I say to people, open up the patient information leaflet that comes inside the drug packet. Most doctors say, look, don't read it. It's written by the lawyers. I say, read it. It's written by the lawyers. They

SPEAKER_352: wouldn't be writing if it wasn't real risks. So I think it's worth looking at those.

SPEAKER_353: Yes. Everybody check out outro.com. What a great domain name as well. And good luck, gentlemen. And we are rooting for you. Thanks, guys. Appreciate it. Thanks for being here. We got

SPEAKER_229: some crazy news. We should I we should jump in. Yes, let's get to it. I think we should jump into

SPEAKER_29: this first item. You were you were talking about this all week. We finally get a chance to do it as a segment meta. A lot going on this week with meta layoffs, meta hiring practices, meta firing

SPEAKER_25: practices. Let's just jump in. So everything sort of got started with a more perfect union at more perfect us on X this week shared leaked audio from an April 30th meta all hands meeting in the clip. CEO Mark Zuckerberg basically comes out and tells employees those of you who are keeping your jobs are not getting laid off. We're putting this monitoring software on your computers. It's going to watch your keystrokes. It's going to pay attention to everything you do. And then they're going to use it to train AI models based on the work that their employees are doing. It does sound a lot to people like they are being watched by an AI that is then going to take their take their job. The key quote, the here's from Zuckerberg on the call quote, the AI models learn from watching really smart people do things. The average intelligence of the people who are at this company is significantly higher than the average set of people that you could do tasks. So if we're trying to teach the models coding, then having people internally build tools or solve tasks that help teach the model how to code, we think is going to dramatically increase our models coding ability faster than what others in the industry have the capability to do. So kind of a dystopian message, right from right from the leader of meta. And then as a follow up, there was a Chinese language treat from AB Quaidong. Thankfully, now X is translating. So this spread all around the world in English and whatever other language you want to read it in. He claims meta laid off employees this week in a cruel, deceptive fashion. A number of employees have reported that meta told them to work from home. There was some sort of humanitarian concern. They didn't want them in the office that day. And then the next day, they all received layoff emails and their access cards no longer work. So they couldn't go back to

SPEAKER_29: their office and do anything. And then the other staffers who weren't laid off were told they now

SPEAKER_25: have their mouse movements, clicks and keyboard inputs are going to be used to train their potential future AI replacements. And then one more tweet, this one you brought up from former meta staffer, Jeremy Bernier. So he says meta was easily the most toxic company I've worked for. There's a reason the Chinese call it squid game. Other refers to it, others refer to it as hunger games or Lord of the flies. He's basically saying meta has too many middle managers, people who aren't doing the the work on the ground. These people are catching them up in endless meetings and making them do pointless AI tasks and basically wasting everyone's time. And that it has created this very toxic, competitive, every person for himself culture where people don't work together. And it's very, you know, sort of like unpleasant to work there. He also claims, I think this is really interesting, that there's so many Chinese immigrant or, or people from China now working at meta that it has created a sense of Chinese or Chinese American favoritism. And that non Chinese people at meta now find it harder to move up because there's so many there. They're sort of a class within, obviously we have no evidence for this. This is just one person's opinion. We have reached out to Germany about Jeremy about coming onto the show and sharing his personal experiences. So I guess my question is to you, what do you think? Are we getting a glimpse inside meta's actual culture? And what are your thoughts more generally about the ethics of using your employees work in real time to train AI that may

SPEAKER_24: one day take over? If the Chinese are appalled by worker conditions at your company, that speaks volumes.

SPEAKER_144: Like in China, people who work 996 factories, they're known for oppressive work conditions. If they're appalled by Facebook's work conditions, that might be worth some self-reflection from the management team at meta. At the same time, we have to answer a very basic question. Have we moved from big tech is bloated

SPEAKER_34: and they're using AI washing as cover for doing layoffs? Or can we take these leaders at their

SPEAKER_09: words and study their actions and understand that AI is creating job loss at big tech?

SPEAKER_34: And Sachs and I have debated this over and over again on social media, online, the administration, everybody. And there's all studies. And when you're in the early stages, what I've learned, in the early stages of a profound change, you will be getting mixed signals. You'll be getting signals. Oh, there's a ton of people crossing the border. Oh, those videos of people crossing the border were from five years ago. And it took a year or two to figure out what was going on in the Biden administration as to like the border controls, as just one example. And eventually we learned, hey, yeah, a lot of people streaming across the border. I think what we're learning now is that AI at big tech companies is resulting in a massive amount of job loss for certain jobs. Right. And that this is appealing to the leadership at these companies because doing more with less

SPEAKER_09: means earnings go up and means you can invest money in other places. Yeah. What Zuckerberg is the most

SPEAKER_77: cutthroat of any of the modern business people I've ever, you know, covered as a journalist or a commentator. He is cutthroat. He has obviously a long track record of screwing partners and people

SPEAKER_09: have complained and sued him over that. He people have complained that he steals everybody else's content. And this would be another cutthroat behavior, which is I'm going to study my employees. I'm going to see if we can train our large language models, our internal systems, our public systems, and make our employees more efficient and get rid of their chores and maybe get rid of the people who are unnecessary. Right. At the same time, all of this is coming out. It's just a really bad look. I'll say, let me take it one more time. It's a really bad look for Zuck to be doing these profound layoffs after forcing everybody to do like AI first hackathons, which they reported he was doing.

SPEAKER_364: Yeah. And then at the same time, say, we're going to monitor everybody's desktop,

SPEAKER_09: not for security reasons that you're not leaking information or data is not leaking. And we want to protect our users data like that. That's a valid reason to lock down computers. Every modern company has locked down computers with this software to make sure data doesn't leak. You have to, you have no

SPEAKER_79: choice. It's just part of business today. He's actually doing these things at the same time, having these conversations and putting them together, which is honest of him. It is. It's honest of him. He's like, listen, we're studying you to figure out how to make this all

SPEAKER_09: more effective because you're all so brilliant. It will result in more job levels.

SPEAKER_25: And that was kind of my other question is, do you think Met is the only one doing this? My suspicion

SPEAKER_29: would be Google, Amazon, OpenAI, Anthropic. Like, why wouldn't they also be doing this? They also have a keen interest in trading better and better coding models. So is it, is this maybe a case of like Zuckerberg saying it out loud? Dario is maybe behind the scenes, not making as big a big deal out of it. But I would be surprised if Met is the only one that's thought of this.

SPEAKER_34: I think you'll be able to answer this question in the next year or two, when you look at the number of employees working in big tech and what jobs people are looking for and which jobs they're no longer looking for. Product manager, designer, middle management. What Matthew Prince in his recent layoff manifesto, Matthew Prince of Cloudflare, I think a very considered person, I know him. He was saying, hey, listen, we got rid of the measures. What are the measures? It's middle management, the people who take reports from, you know, I don't know, an advertising system in marketing or, you know, accounting or whatever. Those people's jobs are easy to automate with the

SPEAKER_95: current tool set. And he said, you know, we're more profitable than ever. We're growing faster than

SPEAKER_36: ever. And we're doing these layoffs because it's the right thing to do in terms of efficiency.

SPEAKER_34: What does it mean for the average person in big tech is you're going to either sell the product or make the product. You're going to sell the product or make the product. But the people who are supporting those two groups of people are going to be less and less necessary. Agents will do that work better. And it's just the game on the field, folks. You have to embrace these tools. You have to get good at building products or selling products.

SPEAKER_77: I think a lot of the other jobs will be automated, just like the typing pool was automated. The mail room was automated by email. The typing room was automated by everybody becoming typists. It's

SPEAKER_36: over, folks. Like the writing's on the wall. I don't think we have to have this debate over, is this AI job loss? It is AI job loss. So then the next debate, the actual debate we should have is, are these companies going to tackle more problems and hire more of the AI builders? And I think the

SPEAKER_09: answer to that is, heck yes. I'm continuing to hire people because I see more opportunities to solve more problems, to build more products. And if people can build products better, cheaper, faster,

SPEAKER_79: well, that's an opportunity to build things that maybe weren't as economically viable.

SPEAKER_25: Yeah. Just to move things faster. I mean, we, we, you know, we, we've always known we want to launch a bunch of newsletters. We want to launch a bunch of new podcasts. So the idea would be, maybe we don't need fewer people. We just are now can launch 10 X as many projects as we previously could have with the same or meet more people each doing a lot more with these AI tools.

SPEAKER_77: Exactly. Yeah. So, um, this is, uh, I think going to be the game on the field. Zuckerberg's,

SPEAKER_375: you know, never been great at messaging this stuff in a, what it comes down to in a lot of ways. Yes.

SPEAKER_212: I mean, other, I think Matthew Prince's was more considered. Jack from square, uh, was more considered block. I'm not saying they're perfectly communicating it, but I think they're a point

SPEAKER_25: about the Matthew Prince one that I didn't like, which is labeling those people. Measurers feels a little pejorative. And like, it does feel like those people who go to get their next job, there's

SPEAKER_29: going to be this, like, Oh, you were one of those measurers at cloud flare that they don't need. You don't want a scarlet letter if I'm using that term correctly.

SPEAKER_379: I think Jack could have been handled a little better personally.

SPEAKER_34: I mean, I think sometimes people are candid and yeah, you could, you could, you could, um,

SPEAKER_380: unintentionally bump into people. Right. Or, or just, you know,

SPEAKER_25: it could sound like they weren't doing an important job. Not that, you know, we've innovated and we've shifted around and like, we need more people doing this and less people doing that. I think that's a fair thing to say, but just labeling them like, well,

SPEAKER_229: those measurers, they are kind of useless now. Like, nah, you're going to, going to impact people's.

Chamath Palihapitiya: I mean, middle managers, like everybody hated their middle manager. Like it's like the entire Dilbert cartoon. Yeah. Rest in peace. Scott Adams was predicated on middle management

SPEAKER_25: being dumb, unnecessary, not providing enough value. The millennial thing was emails and meeting jobs.

SPEAKER_29: Well, like that, a lot of people get cushy in those email meeting jobs where they're not really

SPEAKER_304: producing things. They're just commenting. I saw it at AOL. People were measured based on the, what percentage of their calendar was filled. And if you could go to your boss and be like,

SPEAKER_390: look, my calendar is swamped. I am super busy. And nobody ever said like of those meetings

SPEAKER_393: or any of the, how many of those were actually necessary? And it was like, none of them were

SPEAKER_142: necessary. Well, I worked at my space in their last year. There was a ton of, there was a ton of that. Just like tons of people who just, their whole day was just start a committee, do a meeting.

SPEAKER_25: All right. Let's keep moving to the news. All right. Well, we should talk about

SPEAKER_29: our friends at CrowdHealth. They let you ditch the bureaucracy with a peer-to-peer funding platform for your healthcare. Now, this is not insurance, Jason. That's important to note. It's a totally new

SPEAKER_25: approach for paying for your healthcare. You go to joincrowdhealth.com, you get set up with an advisor, and then they walk you through the whole process from finding a doctor, getting the best

SPEAKER_29: rates by using the sort of crowd to sort of negotiate the best rates on your care, and then even crowd funding through the site, your actual paying of your bills so that you don't get

SPEAKER_25: completely overwhelmed. You can get started for $99 per month for your first three months

SPEAKER_29: by using the code TWIST at joincrowdhealth.com slash TWIST. And if you're looking on your screen right now, you may notice a QR code. You could just scan that and it'll take you right to

SPEAKER_34: joincrowdhealth.com slash TWIST. You know, we need more business models, lower cost tools

SPEAKER_36: for healthcare so that people can get the care they want. And I love the idea of a peer-to-peer funding platform that helps people negotiate better bills and still find great doctors and this

SPEAKER_142: capitalism at its best. So there you go. Joincrowdhealth.com slash TWIST or scan that QR code

SPEAKER_25: and that's how to get there. All right, well done. Our next story, I want to hear your, this touches on,

SPEAKER_29: this is an all-in adjacent news story, so I wanted to get your thoughts on it. Politics, here we go.

SPEAKER_25: We'll do a little politics. On Thursday, President Trump announced he had planned to sign a new executive order. This would give the federal government the power to evaluate AI models before they are publicly released. Now the government was saying it's not just like they don't want, they don't want them to be woke. It was from a defense and national security perspective. They're worried about another Claude

SPEAKER_29: mythos. They want the government to be able to get a look at these models before they come out and

SPEAKER_25: upset the global order or cybersecurity or something like that. However, just hours before the EO is due to be signed at this big event, the White House canceled it. Trump told reporters that he quote, didn't like certain aspects of the executive order. Here's the full quote. I think it gets in the way of, you know, we're leading China, we're leading everybody. And I don't want to do anything that's going to get in the way of that lead. A few theories on what might have happened here. The New York Times said that Trump had invited a lot of big tech CEOs to join him, you know,

SPEAKER_398: stand behind him while he signed the thing. He loves being inclusive and loves business people

SPEAKER_77: coming with him to China, coming to the White House, coming to the Melania screening. It's all

SPEAKER_265: part of what he does. He wanted Jensen and Zuck and Satya and like a few of these big guys behind

SPEAKER_25: him. Most of them declined. A lot of them said, oh, it was last minute. They were sending surrogates or whatever. So one theory was he canceled it because he felt embarrassed that the tech CEOs

SPEAKER_29: didn't show up to be there with him. Axios though, is reporting that David Sachs, friend of the pod,

SPEAKER_25: and a few other tech industry insiders, Zuckerberg and Elon were also both on the list,

SPEAKER_29: met with Trump in these hours ahead of the signing and basically talked him out of it.

SPEAKER_25: Axios is saying David Sachs in particular hated this executive order and convinced Trump that excessive government regulations is one of his signature things that he's against and that

SPEAKER_29: this would have been going against his core anti-regulatory philosophy. So that is the leading sort of theory among the press that there are these sort of two sides. You know, obviously we know Steve Bannon and a lot of the MAGA leadership don't like AI and in a populist way, they're looking at the poll numbers, how much Americans don't like AI and the intelligence agencies as well are very worried about tech companies getting too much power with AI. So they're all pushing Trump to be more regulatory friendly. But then the Mark Andreessen's, the Elon Musk's, the David Sachs is the Peter Thiel's who also have Trump's ear are pushing him in exactly the opposite direction. So I'm curious if you, what, what you believe happened here and is this, is this idea dead in the water or will we see

SPEAKER_404: Trump maybe reconsider again? Yeah. Well, Sachs wasn't on all in, I don't know the details of it, so I don't have any inside information. If I did, I wouldn't share the inside information anyway,

SPEAKER_36: but you know, by the, while the time you're listening to this, there'll be an all in episode

SPEAKER_128: with Gavin Baker, not Sachs. So one might infer from that, that he was busy this week working on this.

SPEAKER_64: Uh, you can make whatever inference you want. I don't, I can't confirm, deny or whatever,

SPEAKER_79: because I don't know. What are we trying to achieve with regulating frontier models? Because

SPEAKER_212: from what I've seen about the original EO that got paused was this had something to do with frontier

SPEAKER_382: models, mythos specifically scaring people. The government and especially the intelligence

SPEAKER_29: community would get an early look at models before they're publicly released and that would make a decision. Can we approve this or do we need more time to make matches and prepare for it?

SPEAKER_24: The, the, the concept of an executive order doesn't actually accomplish all that much. It's kind of like,

SPEAKER_09: here's what the chief executive, the president of the United States, this is what they want to occur.

SPEAKER_393: And then it has to become either a law or be executed on in some way. He can direct

SPEAKER_29: the executive branch. So he could tell his cabinet, here's what I want. Here's your direction. And he

SPEAKER_382: could set up like a commission. Like I want the president's commission on AI models. He cannot set a law. Exactly. And the law is something. Yes. So just from first principles,

SPEAKER_34: what were we trying to do with this executive order? What are we, and what are we trying to do

SPEAKER_36: in terms of regulating AI? Are we just trying to get an early preview of these models? And if that

SPEAKER_79: is the case, there's nothing that says the industry can't self-regulate like the MPAA, the Motion Picture Association of America came up with a rating system in order, and as did the

SPEAKER_09: record industry, in order to make parents and adults feel comfortable with, hey, this movie

SPEAKER_95: is Caligula. It's Conan the Barbarian. It's Indiana Jones and the Tumbledore. I remember all these movies

SPEAKER_17: getting rated when I was a child in the seventies, eighties, and into the nineties. So that parents had

SPEAKER_95: some way of doing it. That's what needs to happen in our industry. A very simple solution is this self-regulating anybody doing frontier model says we, we have a group, we have an operating group.

SPEAKER_09: What this operating group does is we stress test each of these models through a dynamic set of tests. That's always evolving. We try to have it hack into people's personal accounts. We try to have it

SPEAKER_34: build a bio weapon. We try to have it, uh, do some kind of, um, social media misinformation campaign.

SPEAKER_22: And we see how resilient they are and what alarms, uh, doing, um, any kind of, uh, revenge

SPEAKER_393: or, you know, with, you know, all that kind of dark stuff. I'll just leave it there so we don't get

SPEAKER_79: tagged here on YouTube, but any kind of dark stuff there, just create your own rating system that says, Hey, this model is not intended for people under the age of 18. It has characters doing sexual stuff. This one's PG 13. This one, we tested it for giving information on how to recreate a COVID virus.

SPEAKER_09: That's 10 times worse. And that would basically, I think, solve the same problem. That's what the industry should do to get ahead of this. And we don't need to regulate them. I think at this point in time, I think this is political theater because right now AI is extremely unpopular and it dovetails with the story we talked about earlier, job loss, people losing their cars. These are all becoming one issue. Polarization of wealth. Okay. We now have a couple of trillion dollar companies with these AI tokens. We have jobs being retired at an alarming rate. And is every truck driver going to lose their job? Is every cab driver going to lose their job? Is every delivery driver going to lose their job? Every factory worker, everybody at Amazon going to lose their job? The answer is yes. All those jobs are going to go away where the majority of them will go away and those people will have to be redeployed in some intelligent fashion. You combine that with, Hey, the intelligence wants this and America first. This is really about the civil war inside of the Trump administration, the Republican party. The civil war in that party is people's jobs, America first,

SPEAKER_79: and these tech oligarchs, as they call them, are they running away with this and stealing everybody's

SPEAKER_393: money, quote unquote. And is it just going to create a permanent underclass that always loses

SPEAKER_09: their jobs? This is what Bannon, Fuentes, Tucker, they've all departed. They've, Meg and Kelly, they're all leaving the Trump camp or trying to put pressure on the Trump camp to stop kowtowing in their minds to big tech, to oligarchs. Trump loves money. He loves billionaires. He loves super PACs.

SPEAKER_77: Good luck with that. He's shown very clearly. He loves the military industrial concept, uh, complex. He's getting them a trillion and a half more dollars to build weapons. He started the war with Iran and he loves big tech and he loves super PACs. He wants to get hundreds of millions of dollars into the super PACs. He's pro crypto. The train's left the station. It's not going to be America first. It's going to be big tech and the military industrial concept first. It's just the nature of Trump's

SPEAKER_29: decision-making. Yeah, it definitely seems at least at this point that this is a huge victory for the sort of accelerationist side, and it's going to be pretty tough for the anti-AI folks to get through now that he's punting these executive orders. And there just doesn't, it doesn't seem to be a huge

SPEAKER_142: appetite among Donald Trump to regulate the AI industry. No. And I, you know, to be totally

Chamath Palihapitiya: honest, looking at it, like, do you see areas where it should be regulated? I'm, I use these tools all day. I'm like, what are we supposed to do here to regulate this? Obviously bio weapons and that kind

SPEAKER_41: of stuff makes sense and hacking. I'm not like a free market, you know, like absolutist. Like, I think there's probably regulations there to be had. I do think that the problem. Do you have one in mind? Do you have an example of one?

SPEAKER_29: Um, yeah, I mean, I, I do think that there should be more tight rules around these chat bots and what kinds of conversations you can have with a chat bot. And we should probably be forcing these companies for kids. You mean for anybody to shut down. If you're starting to have a conversation with a chat bot that is deeply like psychological, that is about. That is about violence. Like we heard that there was that, you know, shooter that was talking to chat GPT and chat GPT was suggesting,

SPEAKER_25: oh, well, if you'll get more infamy and that's going to lead to 30% more news coverage. And like, it shouldn't even be insane. It shouldn't even be having those kinds of conversations. And I think the whole idea of chat GPT or chat bots is like a therapist is probably not a good use for this

SPEAKER_29: technology. There are so many good uses for this. So that's a good point in relation to that.

SPEAKER_34: There's three ways that this could be regulated. One is in the review mirror after bad things happens, the companies get sued. Sure. Okay. That's slow. Right. But it is happening. Currently there is.

SPEAKER_128: That's true. Yeah. There are multiple lawsuits over this. The second is self-regulation. The third is

Chamath Palihapitiya: obviously. Yeah. Uh, regulation at the state and or federal level, all three of these things will

SPEAKER_29: be pursued. I think the one that makes. My other one would be copyright. I think you shouldn't train your model on a bunch of bootlegged books and movies. You should compensate the people,

SPEAKER_229: the artists who created the things you want to train on. And that also is being adjudicated in

SPEAKER_29: the course. Correct. Uh, yeah. I think the problem is that on both sides, we have people who don't fully understand what the tech is and what it does. Uh, and so I think you get a lot of like ban new data centers nationwide, which to me is just not a very practical response. As much as I, I sympathize with the idea of it. I feel like, well, that's not going to happen and we should be taking more

SPEAKER_77: common sense approaches to regulating. I have a really simple one, uh, that I think is happening and that we can find some common ground on, which is in the case that robots, whether it's self-driving

SPEAKER_36: or, you know, the figure robot sorting packages on Amazon, we say, Hey, uh, every time a robot goes

SPEAKER_34: into a full-time position or works eight hour shift, that equals one human. Every time a robo taxi works an eight hour shift, that equals one human. We're going to think about the pace of deployment here. And the pace of deployment is going to provide some soft landing for those employees. Yeah. Either the government handles it, the private market, because I think there's going to be,

SPEAKER_09: if we don't start this dialogue about what happens when every Uber driver, every truck driver, every door dasher loses their job. If we don't have that discussion now, it could lead to like riots in the street, civil war type stuff. Right. Which by the way, it's, I'm sounding like a lunatic. Just go to Europe and look at the riots in Greece, Spain, Egypt. They all happen when there's 20, 25% unemployment in certain demographics, typically young men. So we should just have that first principles discussion about that pace of deployment. By the way, this isn't my idea. This is an idea that Bernie Sanders, the Chinese government, the CCP, they're all floating this idea of maybe there'll be robo taxi or a truck driver, door dasher,

SPEAKER_366: medallions, licenses, and we'll only have a certain number per city, if that makes sense.

SPEAKER_138: Yeah. I think I, yeah, that's what I mean. I think we can, I think you could think about it piecemeal and there are plenty of ways we could probably come in and start,

SPEAKER_25: you know, regulating some aspects of these companies. We just always try it in these like big picture ways that don't even, even the term AI, like legally, I feel like is very limiting because it means that as so many different kinds of applications, there's some that I have no problem with. Like if you want to make a podcast on Google notebook LM, I think it's great. I don't need to regulate that, but you know, what podcast or licenses. Yeah. But like what Palantir and flock safety are doing, like, maybe we should be thinking about how to regulate that. So I think that

SPEAKER_29: to me is like, it's almost useless privacy. Sure. It's almost useless at this point to talk about like AI regulation, because like, well, what do you mean? There's a thousand different versions of it.

Chamath Palihapitiya: Anyway, it's going to happen. It's going to happen on a state by state level prop. And so

SPEAKER_77: as much as people want to like pretend it's not going to, I, I, I'm not advocating for one or the other. I'm just saying the game on the field is going to be, do we have cataclysmic or accelerated

SPEAKER_196: or violent job loss? Yes. And that's what everybody's fear is. The fear is just job loss. If you take job

SPEAKER_79: loss out, if people were make getting more jobs because of this, if they got higher paying jobs because of this, if inflation went down because of this, if food costs went down, if, you know, construction costs go down, if education costs go down and they should, then people will feel

SPEAKER_77: good about this technology. It's just going to be a crazy five to 10 years of transition where people lose a lot of jobs. Let's keep moving through the docket. All right. Anything else we

SPEAKER_466: want to cover? Well, I did. I want, we, we, we talked about the, uh,

SPEAKER_25: Lewill AI versus Kled AI. We had Avi from Kled on the show earlier this week. There is a new wrinkle. There's a new viral post, uh, crypto, uh, exec, Robert Chang was on the Rasmur report,

SPEAKER_29: uh, live stream podcast. And he, he suggests, I'll just, I'll bring up, I'll bring up the tweet so we can take a look at it together. He suggests, uh, they obviously forked Kled. I don't think

SPEAKER_25: Lewill AI has worked that much. He's saying they got funded because the founder's dad is an incredibly powerful person in the United States. That's why they were able to raise money. The company itself is basically just a shell. Uh, I did a little bit of digging into what he meant. Uh, and it turns

SPEAKER_29: out one of the co-founders of Lewill, this is the company that's accused of ripping off Kled and taking their technology. Uh, it is, uh, one of the co-founders is a young man named Inigo Lenderking.

SPEAKER_25: His father is private equity legend, Eben Lenderking, currently the CEO of the process and flow

SPEAKER_29: control company, FCD, previously served in roles at Harrington industrial plastics and various private equity firms. I think my question to you is, is this an overly cynical take? Is it theoretically

SPEAKER_25: possible that a private equity Nepo baby could sail through the investment process in his startup, even though he didn't have a very helpful product? What do you think?

SPEAKER_471: Um, I, is there favoritism in the world? Is that what I'm asking to adjudicate here?

SPEAKER_29: I mean, obviously if your dad is a wealthy and well-connected person in private equity, you get a leg up in, in pitching your startup. Nobody's arguing that, but do you think that that would be enough to get funded by like real VC companies?

SPEAKER_77: No, real VCs, um, have to answer to their LPs. Any conflict of interest could open them up to lawsuits and things like that. So there's pretty deep diligence that's done on these. Um, you do have a portfolio and you do expect a certain number to go to zero. So it's not like you're going to be super obsessive. However, I do think, uh, yeah, you're, you're, you're not

SPEAKER_477: placing bets with tens of millions of dollars based on Nepo babies, right? The majority of

SPEAKER_79: the time, because you need to get returns and venture so hard that you don't want to waste a

SPEAKER_77: bullet. Like I, I, if I had family members come to me or friends of family members, which happens

SPEAKER_34: sometimes I say the same thing to them. Like we invest in one out of 150 companies that apply 200, you know, the chances of you beating out all those other companies is extremely low statistically.

SPEAKER_77: But if you want to go through the process, that's fine. And then if I were to invest,

SPEAKER_481: you know, and there's some conflict or whatever, I have to get like a waiver from like my LPAC, my,

SPEAKER_77: my, um, committee that, you know, uh, represents the LPs. It's just not worth it. Yeah. So I, the only thing I might do, let's say I had a cousin with a great idea. I would probably forward them to friends of mine and say, Hey, you might want to take a meeting. So a good idea.

SPEAKER_38: I think, I think the Nepo Bay, like it's a, it's a real phenomenon, but I do feel like sometimes we,

SPEAKER_29: we invest too much importance in it. And it's like, you know, no, like Jack Quaid, isn't going to get all these roles every year, just because his mom is Meg Ryan and his dad is, uh, Dennis Quaid. He also, it's a leg up, it's an advantage, but you also have to have something else going on. You're not just going to sail through the whole system. All right. That was fine.

SPEAKER_485: That was what I thought too. Uh, let's go off duty. You ready to go off duty?

SPEAKER_25: Oh yeah. Tell me what you got. All right. So, uh, the first one I wanted to share, uh, this is a very viral clip this week. Uh, Christopher Nolan, uh, one of my favorite

SPEAKER_29: directors, I think one of yours as well. He's got that Odyssey movie coming out. So he's doing a lot of press. He explains here that, uh, he does not have a, he doesn't use email and he does not own a smartphone. He talked about this on 60 minutes. So I'm just going to take a

SPEAKER_487: second and bring this up and then we can, that's interesting. So he wants to be super present and

SPEAKER_25: not distracted. Right. He explains, he explains his thinking and I think his thinking is, is pretty

SPEAKER_491: interesting. So here's, here's the clip. Here we go. I do not. I never have. And why not? I don't know. No interest. Do you get letters? I mean, postal letters with stamps on them?

SPEAKER_493: I get a lot of, I get a lot of printed emails where people are like, you got to take a look at this.

SPEAKER_496: All right. Uh, but no, I, I, I just never been particularly interested in that as a form of

SPEAKER_501: communication. No smartphone. No, I've never had a smartphone. Uh, it's getting harder and harder. The return of the QR code has been quite, quite tricky. Ah, QR code had sort of gone away, COVID brought it back and now it's kind of everywhere. And if you don't have a smartphone, you can't, you can't do much on the QR code. So, I carry a, a flip phone when I travel.

SPEAKER_510: Is this a way, and I don't mean this in a pejorative sense of, of any kind, actually in a very complimentary sense, is this a way of keeping the world at a appropriate distance and not letting it cave in on you like the rest of us do in the digital world?

SPEAKER_496: I mean, I think it's more a thing of, I, I haven't let the rest of the world catering on me. I'm just living the same way that, that we all used to. So, you know, to me, it's just life as life as normal. And I envy you deeply.

SPEAKER_129: Yeah. I mean, I don't know how he does it. If you're going to dinner and you got to take a picture of a QR code to order from the menu. I mean, what are you supposed to do here? It's like, I, I don't, he must have his spouse or, you know,

SPEAKER_25: his security team have his phone. A lot of people are pointing out, if you're a very wealthy person,

SPEAKER_29: you could afford to have an assistant following you around who does have a cell phone so that it, you know, you would, you would, it would be easier, but I don't know that. I think that's

SPEAKER_25: kind of a cop out, like almost every prominent person I know that has an assistant also has their own phone. It definitely adds a layer of complexity to the modern world. I just think it's so, I just think it's so interesting that he would be such a hardliner about this in 2026, like still no email,

SPEAKER_09: nothing. I mean, he, I'm sure there's the people who work at his companies have emails and they're, you know, feverishly using them. Um, it's probably a situation like Donald Trump. Donald Trump gets printouts of everything. There's a person who travels with him. You can go look it up and she has a portable printer with her and she just sits there next to him wherever he is. And according to the public reports of this, you'll find it on social media. She's just printing out his emails, printing out press clips. Then he'll take the press clip and say, congratulations, Brett Ratner and

SPEAKER_34: sign it. Yeah. And then hands it to his assistant to mail to the person. So it's like, almost like he's turned this into a way to curate his network or whatever. And then whenever he gets deposed or there's like a discovery, they're like, Donald Trump has never sent an email. So it's almost like a mob

SPEAKER_41: boss situation where there is no paper trail of anything he's ever done. Paul, he's sending

SPEAKER_29: his assistant out to go use the pay phone and then come back to the house and tell him what the, what the message was. Uh, yeah, I will say, I agree with Chris about the QR code menu things that was in Italy and Greece. Every menu is now on a QR code. They're, they're done with paper menus.

SPEAKER_77: And it really was driving me crazy. I hate having to. It's I am, I'm torn about it. I love the fact

SPEAKER_95: that we're not just wasting these printouts constantly. Yeah. Um, I love the toast system,

SPEAKER_77: which allows you to just order without waiting for a person. I love the fact that I don't have to flag a waiter down when I'm with my kids and somebody wants to another set of dumplings, you

SPEAKER_393: know, like you, you got three daughters and like the dumplings come and there's like five of them. And all of a sudden it's a civil war. You're going to need to get their second. You're going to need double dumplings for sure. I just love to double down on the dumplings and just, yeah, man, it saves time. You get in and out of things. So I, I, I am for convenience and efficiency. Um, I wanted to

SPEAKER_405: share some sites aren't, aren't, aren't well designed. I, I, I, I, it never works the way I

SPEAKER_77: will as smoothly as I want it. I have to do something about my addiction to my smartphone and social media. I'm thinking for my six weeks when I'm traveling this summer of deleting all the social

SPEAKER_95: media off my phone and then maybe going to an outbound only where I make videos, send them to

SPEAKER_126: you and the team and you guys make content out of them and post to my accounts. And that's it.

SPEAKER_77: Yeah. I'm just off of social media. Cause I am, I am doom scrolling every night for an hour. I'm addicted to TikTok, Instagram X. Obviously you see me on these platforms. I'm addicted to it all because my job is to talk about the news and it's like, I have to be involved in it, but there's something lost and I want to read a couple more books and just slow my brain down a bit. So I need to come

SPEAKER_142: up with a, a system and I need to be vigilant. Yeah. I was not on the phone very much during my, my trip and it, it really does open up like, oh, there's a whole world still out here. I'm not,

SPEAKER_224: I don't need to constantly be looking at Reddit and like, what's new, what's new. Uh, so yeah,

SPEAKER_482: I, and there's the taking of the pictures. So I'm thinking of maybe just bringing a camera with me and just getting into taking pictures with like a digital SLR and enjoying it, you know,

SPEAKER_77: kind of situation. Um, for me, um, I wanted to share, uh, one of the companies we invested in is called Terra cafe. And I've had a Terra cafe for a long time. I love a one touch espresso machine

SPEAKER_09: had done the Jura before that. Uh, but they came out with this great one called the Demi D E M I incredible for making an espresso and they took the, um, stick for frothing milk out of it because there are frothers now that are so much better. So I have a smeg S M E G frother for milk makes perfect froth milk, cold, hot, and you can pick how much froth you want to have on. If you're making a flat white, you want less. If you're making a cappuccino, you want more, et cetera. And so they're coming out, they have the Demi and then they have this arrow milk frother coming out. That combination

SPEAKER_95: is absolutely fantastic because you separate the two and, um, it just makes the, this Demi espresso one touch is now half the size and half the price. So you can get into this thing for like 600 bucks or

SPEAKER_09: something, 700 bucks. Well, and you think about espresso based beverages, you can make a long,

SPEAKER_95: um, Americano, longo, espresso. Then you do the froth. I was trying to explain this to somebody,

SPEAKER_48: I think it was one of the, uh, staff here I have at the ranch. And I was trying to explain to her how much the pods were costing her. I was like, how much of these pounds costs, like the pods,

SPEAKER_29: you probably use a pod system at home. No, no, I hate the pods. I have a, I have a coffee, I have a black and Decker, like drip coffee machine. And I also have a grinder, 30 buck drip coffee, finishing a grinder, but it makes great drip. I do have a grinder. And so I buy the ground, the fresh beans, and then I grind my own. I think that makes the biggest difference.

SPEAKER_01: It's like, these pods are like 75 cents a dollar, you know, you're buying them, you're destroying

SPEAKER_29: the earth. And it's not even like it's a better flavor. It doesn't taste good.

SPEAKER_34: A terrible flavor. You'd have two cups a day. You're at two bucks. That's, you know, $700 a year. You have a couple of people over. Now you're well over a thousand dollars a year. Over three years, you're at $3,000. You buy one of these, you buy whole beans on a subscription. You're going to be

SPEAKER_552: way in the black with much better taste, much better, fresher, more delicious coffee. Exactly.

SPEAKER_34: The problem is you have to pay for the machines upfront. Right. I think they're doing like a subscription coffee thing, or you can, this is the perfect example of if you can't afford the $600, $700 espresso super automatic Demi ahead of time, this is where using a firm on one of those services

SPEAKER_77: to pay for it over two years, if they don't charge, I don't know if they do have that offer.

SPEAKER_553: Right, like a buy now, pay later thing. Yeah, yeah.

SPEAKER_95: Well, that's how they get you with these pod systems. The pod system is cheap upfront, and then they get you on the razor blade. It's just much better doing it this way and having better tasting coffee. Thanks for coming to my TED Talk. Let's get to the weekend. Bye,

SPEAKER_555: bye, everybody.

SPEAKER_558: Oh, sorry. One day. We'll see you next day deal flow, apply for Jason's Angel Syndicate at thesyndicate.com. We find two to three deals a month. And check out This Week in AI, Jason's experts-only roundtable with top AI founders and operators every week. Find it thisweekinai.ai. Check out The Twist Ticker, our daily newsletter, at thisweekinstartups.com slash ticker. Thanks again to our sponsors for making today's show possible. Follow the show on Instagram, follow the show on x.com. This Week in Startups publishes three days a week, Monday, Wednesday, and Friday at 5 p.m. Central Time. You can submit an audio or video file question by emailing it to thisweekin.com.