Hugging Face Co-Founder on Open-Source, Microduck, and NVIDIA | E2339

This Week In Startups is made possible by: Superhuman:⁠ http://superhuman.com/⁠ Paypal:⁠ http://paypal.launch.co⁠ Odoo:⁠ http://Odoo.com/twist⁠ Plaud:⁠ http://Plaud.ai/twist⁠ Harmonic:⁠ http://harmonic.ai/⁠ Today’s show: Hugging Face built a $399 open-source robot duck as a side quest, and sold 16,000 units off a single tweet. Co-founder Thomas Wolf joins Jason and Lon to explain Microduck happened, why storage is the company's real business, and what it felt like to get hacked by OpenAI's agents. Jason takes apart Gavin Newsom's new AI "kill switch" executive order: if these models are so dangerous, why not start with the same KYC every bank already requires? PLUS, Snap’s new AI-integrated glasses could be a game-changer. Guests: Thomas Wolf on X:⁠ https://x.com/Thom_Wolf⁠ Hugging Face:⁠ https://huggingface.co/⁠ Relevant Links: Hugging Face —⁠ https://huggingface.co/⁠ Microduck —⁠ https://pollen-robotics.com/microduck/⁠ Pollen Robotics —⁠ https://pollen-robotics.com/⁠ OpenAI: the Hugging Face incident and the road ahead —⁠ https://openai.com/index/hugging-face-incident-and-the-road-ahead/⁠ Governor Newsom's executive order —⁠ https://www.wsj.com/tech/ai/californias-newsom-issues-executive-order-to-weigh-ai-oversight-including-kill-switch-3a98040f⁠ Anthropic Wet Lab —⁠ https://www.cnbc.com/2026/09/18/anthropic-quietly-sets-up-biology-lab-as-it-ramps-ai-drug-program-report.html⁠ Lucid and Bolt to deploy 25,000 autonomous EVs across Europe —⁠ https://electrek.co/2026/09/17/lucid-bolt-25000-autonomous-evs-europe/⁠ Snap opens pre-orders for Specs at $2,195 —⁠ https://www.specs.com/⁠ Ray-Ban Meta Display — the $799 comparison⁠ https://www.meta.com/ai-glasses/⁠ Clavicular at All-In Summit —⁠ https://x.com/clavicular/status/2099994778616348811?s=20⁠ Physical: 100 (the Korean original) →⁠ https://en.wikipedia.org/wiki/Physical:_100⁠ Bad Therapy —⁠ https://www.abigailshrier.com/⁠ Full Show Notes and Transcript from Pluad:⁠ https://web.plaud.ai/s/pub_194de893-8c55-4d77-81db-a6b278159a44::o9QfImqCq5VS0F_dqm4J93eTeyb-GsIeXCLwqASnvYAmt5f59OLQz5-VN1epnq6DBLoZWFales_7IbgC⁠ Timestamps: 0:00 Guest: Thomas Wolf, co-founder & chief science officer, Hugging Face 1:58 Plaud: If your work depends on conversations — interviews, meetings, calls — you need a Plaud NotePin S. You can check it out at https://Plaud.ai/twist and use code TWIST for 10% off! 4:17 Two years to build Microduck 8:43 How Hugging Face actually makes money: storage, not tokens 10:16 Superhuman - Superhuman Go is an AI chat that's always there when you need it, already aware of what you're doing, and doesn't ask you to start from zero. Sign up to get the best in AI at https://superhuman.com 17:18 Vibe-coding robotics: "train this new behavior for me" 20:00 PayPal - Pay zero processing fees on your first $100K in eligible PayPal payment volume. Learn more at https://paypal.launch.co 23:20 Thomas Wolf's P-doom and the airplane-safety analogy 25:18 The Nvidia acquisition and what changes for open models 29:57 Odoo - The all-in-one business platform. Your first app is free! Get started today at https://Odoo.com/twist 38:39 Inside the Hugging Face breach: stolen credentials and basic mistakes 43:33 News: Newsom's executive order and the frontier-model "kill switch" 45:04 Jason's counter-proposal: liability, insurance, and KYC for AI 1:00:08 News: Lucid and Bolt to put 25,000 robotaxis on European roads 1:10:01 News: Snap opens pre-orders for $2,195 Specs AR glasses 1:21:43 Off-Duty: Clavicular, the All-In Summit 1:31:01 Off-Duty: Physical 100 goes to Italy and Mexico Subscribe to TWiST on Substack:⁠ https://twistartups.substack.com/⁠ Subscribe to This Week in Startups on Apple:⁠ https://rb.gy/v19fcp⁠ Follow Lon: X:⁠ https://x.com/lons⁠ Follow Jason: X:⁠ https://twitter.com/Jason⁠ LinkedIn:⁠ https://www.linkedin.com/in/jasoncalacanis⁠ Check out all our partner offers:⁠ https://partners.launch.co/⁠ Great TWIST interviews:⁠ Will Guidara,⁠⁠ Eoghan McCabe⁠,⁠ Steve Huffman⁠,⁠ Brian Chesky⁠,⁠ Bob Moesta,⁠⁠ Aaron Levie⁠,⁠ Sophia Amoruso⁠,⁠ Reid Hoffman⁠,⁠ Frank Slootman⁠,⁠ Billy McFarland⁠ Check out Jason’s suite of newsletters:⁠ https://substack.com/@calacanis⁠ Follow TWiST: Twitter:⁠ https://twitter.com/TWiStartups⁠ YouTube:⁠ https://www.youtube.com/thisweekin⁠ Instagram:⁠ https://www.instagram.com/thisweekinstartups⁠ TikTok:⁠ https://www.tiktok.com/@thisweekinstartups⁠ Substack:⁠ https://twistartups.substack.com⁠

Speaker names marked ? are inferred from speaking time and the show’s known hosts — they are not voice-verified.

• 0:00Jason Calacanis?

  All right, everybody. Welcome back to This Week in Startups. It's your boy Jake How. I'm here with Lon Harris. I've been on the road, but finally got a decent recovery score. I'm finally back in the game, Lon.

• 0:11Speaker 01

  Oh, your sleep score. You're coming back.

• 0:13Speaker 02

  This Week in Startups is brought to you by Odoo, the all-in-one business platform. Your first app is free. Get started today at O D O O dot com slash twist. PayPal. Pay zero processing fees on your first $100,000 in eligible PayPal payment volume. Learn more at PayPal dot launch dot co. Superhuman. Superhuman Go is an AI chat that's always there when you need it, already aware of what you're doing and doesn't ask you to start from zero. Sign up to get the best in AI at superhuman.com.

• 0:40Jason Calacanis?

  I get a decent amount of sleep debt that I'm still working off after the All In Summit and then coming up here to Dreamforce and face-mogging. Or what is it called?

• 0:48Speaker 01

  Frame, you're frame-mogging Marc Benioff, yeah.

• 0:52Jason Calacanis?

  Frame-mogging Marc Benioff.

• 0:54Speaker 01

  You're so much larger and more imposing in the frame than him.

• 0:58Jason Calacanis?

  In the frame, correct. This is the new trend, but he's, you know, he's a six foot six guy. It's kind of hard to frame-mog him, so I just jumped.

• 1:05Speaker 01

  Yeah, I did not realize he was such a Kaiju-esque figure until this week, when you're seeing all those images of him from Dreamforce.

• 1:13Jason Calacanis?

  Well, and they were kind of doing the angle with like him. So they were literally frame-mogging folks.

• 1:19Speaker 01

  Yeah.

• 1:20Jason Calacanis?

  Because they kind of used his angle and then were looking down at the crowd or whatever. It was a weird angle. But as you can see here.

• 1:26Speaker 01

  Yeah, finally getting one up on him.

• 1:29Jason Calacanis?

  Well, and then if you look at the, the second one, I'm actually standing on a chair. So I am like so tiny compared to Marc Benioff, but we had a fun time at Dreamforce. And I kept telling him Slack was the oracle. Slack was the oracle, like it had all the information. And he said, you know, Jake, how you said oracle twice here. It's not cool because they're competitors.

• 1:49Speaker 01

  Wrong, wrong, wrong brand, wrong brand, but.

• 1:52Jason Calacanis?

  He got, yeah. But it was great to be at Slackforce, as I'm calling it now. I call it Slackforce.

• 1:57Speaker 01

  Slackforce, sure.

• 1:58Jason Calacanis?

  Slack is like the, just all your, if we're going to really train AI, if you just gave AI all of Slack, I think you could do all of our jobs at this point.

• 2:07Speaker 01

  Yeah. Oh, I mean, that is, that is when you're first using AI and you finally do plug it into Slack, that is one of the real game change moments where it's like, oh, all of a sudden it has all this context. It understands all these things it didn't know about before. I used to have to explain everything and now it could just read the thread and it's, it's good.

• 2:24Jason Calacanis?

  Well, listen, we got a big show today. I don't want to lose anything, so I'm going to just start my plod here if you don't mind.

• 2:29Speaker 01

  You gotta, you gotta turn on your plod. Yeah, I gotta hit my plod.

• 2:32Jason Calacanis?

  Well, you know, we always applaud plod here because this is the greatest product ever. We started doing the show notes and in Plod so that it's really easy for you all to understand what happened on the show, get lessons from the show, all the best moments. If you look in the show notes, you'll see we link directly to our Plod recapture. And it's got a transcript, obviously, but I prefer the summaries and the mind maps. It does all kinds of interesting things.

• 3:01Speaker 01

  It's actually a fascinating product. There's, there's all these templates and so you can decide what kind of summary or readout you want to produce based on, you know, is this a meeting? Are you with your doctor or somebody else that you, are you, you know, is this a casual chat? And like, and so it sort of understands what kind of output you want. It's a very clever system. So if you or your work depends on conversations, interviews, meetings or calls, we want you to check out Plod at plaud.ai slash twist and use the code TWIST to get 10% off your Plod note pin.

• 3:33Jason Calacanis?

  Speaking of great products and hardware being hard, you know, Hugging Face, a great company. I've been trying to get them on the pod for a long time and I love a good side quest. We talked about side quests on the show before. Side quests are something you just feel compelled to do, not because they are going to be necessarily the greatest business or make the most money or get you the most shine and service your ego, but a side quest is because you're kind of interested in it, right? You got a passion for it. You want to figure it out. And I thought one of the great side quests of the year, and I mean, people are starting to do a lot of these. Don't recommend it for first time founders unless you got a lot of resources. You got to stay focused on that number one company.

• 4:14Jason Calacanis?

  But Hugging Face, you know, host of all these great open models, bought a company with like doing robotics. And then they released this amazing new product and I said to Lon, hey, produce the show, please. Get me this.

• 4:27Speaker 01

  Right, yes.

• 4:28Jason Calacanis?

  Produce, producers.

• 4:29Speaker 01

  That's my job.

• 4:30Jason Calacanis?

  Get everybody in line and let's get this person who's in charge of the Hugging Face robot. Because I think this is got legs, so to speak.

• 4:39Speaker 01

  It does have legs. Pollen Robotics was that robotics company Hugging Face acquired. And so next up, we are joined by Tom Wolfe. He is the co-founder and chief science officer of Hugging Face. And he built the transformers library that basically every AI lab on earth runs on. And here is, he was also involved in Micro Duck, the open source robot duck that sold 10,000 units in its first five days of launch. Here's Micro Duck now. We're watching him on the camera. So welcome, Tom, to the show.

• 5:06Speaker 03

  Thanks. Thanks, Guy. Very happy to be here. Talk about this deck.

• 5:10Jason Calacanis?

  This is incredibly cute, a lot of fun. Why did you guys decide to do this? Is it for the fun of it? Is it because you think robotics is a long-term bet and this is a great way to learn and market it? Take us through the origin story of this very cool personal robot for hackers and kids, I'm going to say.

• 5:32Speaker 03

  Yeah, I think, I think you're right before of it. Okay, I won't deny there is a part of fun in building a product like that. It's actually really exciting as a builder to make something that, you know, everyone in your family not finds scary like a humanoid, but just find really cute and they all wanted, like all my friends on the, on the, you know, first to say, okay, we want to buy it wherever you do it. The thing is, it was surprisingly difficult to do. That's our longest robot in the making. It took us two years to make this robot. So it looks small, it looks simple, but as you saw in the video, it's able to, you know, stand back up. There's a lot of things it can grab things. It's actually surprisingly capable for, for its size.

• 6:12Speaker 03

  And so for us, we, you know, we, we had for a long time and I just, I just tweeted today, the evolution, some evolution of the robot that you can see. But for a long time, it was not up to the level that we thought everyone would be crazy about it. And then it's only earlier this year that we were like, okay, this can actually, okay, this is part of a side quest. Of course, Hugging Face is way more famous for its hub and that's, that's our business model is, you know, storing data sets, storing model weights and like building committee. But we think here there is something where we can create a new community of people who are actually AI builders, but AI builders for robotics.

• 6:53Speaker 03

  So you can buy this product for your kids, for your family, for yourself, but you can also buy it as an AI builder. Yes, very, as I'm going to, yeah. So you can, you can also like teach this robot new behavior. You saw it, you can do roller skating. The roller skating is actually just an intern spending an afternoon with this crazy idea. What if I put some roller in the simulation? Can it learn to roller skate? And it worked so well that we were like, okay, let's put that in the demo lounge video and let's actually ship the robots with small roller skates. So people can play with that. But we think there's a lot of other things. So you see, it can, can play like soccer. It can, and it can very likely do way more things.

• 7:37Speaker 03

  The fun part is actually when we released it on the web, people went crazy and they started to put the robot in all of these various simulation, doing all of these things, walking on skaters, like doing like swing, like a lot of things that we had no idea you could train a robot to do. So for now, it's only in simulation, but when the robot will ship in a, in a couple of months, we think this will transfer very likely.

• 8:02Jason Calacanis?

  And it's, it's sold out. You have tens of thousands of these on order, hundreds of thousands? What's the latest?

• 8:08Speaker 03

  So the latest, I think, is a little bit over 16,000.

• 8:13Speaker 03

  And but yeah, we, I think, well, that, that will be very likely the most widely sold robots. I don't think anyone sold a moving robot like that over, over, over 10,000 units.

• 8:26Speaker 03

  And we have not even started to do any side sort of marketing. Like we just basically posted a video on Twitter. That's the only.

• 8:34Jason Calacanis?

  Amazing. How does Hugging Face make money hosting models and weights, et cetera? And then how do you think this will eventually monetize, if at all, or is this just a community building effort? I think that's really underdone. Sometimes people, you know, spend money on marketing. Sometimes they try to kickstart a community, but this one seems to have market pull. People are opting into the community. It's got that market pull in a very strong way. So tell us about Hugging Face's core business, how you make money there, and then how this will make money, if at all.

• 9:08Speaker 03

  It's probably the most widely asked question about Hugging Face is how do you guys make money? Because everything's free. So I mean, not that everything is free. We have people who, we actually had a stronger uptick of revenue this year. We've been publishing some numbers. We passed in April, we passed 100 million. And then we even climbed way higher in the summer. And the main source of revenue right now is around storage. So people, we've developed for our needs, very efficient, fast, you know, and like cheap storage quality. So we store petabytes of data. And this is very useful in many fields when people need to train, need to evaluate model, but also in robotics, because robotics has a lot of this image.

• 9:51Speaker 03

  I mean, you know, our Tesla, they record like terabytes of videos, you know, every time they run. You need to store that. That's one of the most data angry area of ML and it's having a huge boom. So we have many robotics company right now using our hub to store the data. And we have a historical activity of storing models. So you can store them for free publicly, but you can also pay for more storage privately.

• 10:17Jason Calacanis?

  This is very incongruous to me because you think storage is a commodity and people have tons of storage around. They could use Amazon or Azure or any number of things to store things. Why would people, just curious here, you pick to store stuff on Hugging Face versus their own servers? And you mentioned something about speed or it being unique to AI. Maybe you could unpack that because there must be something here that is not obvious.

• 10:42Speaker 03

  Yeah, you're very right.

• 10:42Jason Calacanis?

  We've got a lot going on here at my venture firm launch. We're making investments in 100 startups a year. We're producing podcasts and newsletters every week to help our founders and the venture capitalists who listen to our pods. And we're putting on live events all over the world. The team here is very organized, very hardworking, but it can still be pretty stressful to keep track of everything. And our solution has been Superhuman Go from the makers of one of my favorite apps, Grammarly. Superhuman Go is going to make everyone on your team twice as fast by quickly summarizing dense threads so they don't have to spend all day reading. They can just TLDR, get to the point. Too long, didn't read, is real.

• 11:23Jason Calacanis?

  Or providing important context inside whatever documentary or email they're already reading. Because we've got Superhuman Go inside all the tools we already use. There's no constant switching between tabs, copy and pasting and transcripts all over the place or doing prompt trial in our until they're perfect. It just works when I need it, allowing me to make crisp decisions with confidence. Give Superhuman Go a try for yourself at superhuman.com. That's superhuman.com.

• 11:48Speaker 03

  I think people thought the storage was sold because of course you can pay S3 and like, yeah, everything. The thing is, AI is very different than previous big data. Like it's way less random access. It's a lot more, you will download a big blurb of gigabytes or terabytes. Like the latest models, they are terabytes inside. They're really massive. So you want your storage to be very close to your GPU, for instance. So we do these things where actually your storage is pre-warmed next to your GPU. So it's really, really fast. You don't pay for aggress. So we basically developed this full solution. Another thing is you want the duplication sometimes. Like, let's say you just add a single row to a data set. You don't want to store again the data set. You want this version data set.

• 12:31Speaker 03

  So you can move around the various version of the data set without having to store the duplicated copies. So we need these type of things that Dropbox or this company did back in the day. But we needed adapt for the AI world, for the AI workload, storing this thing and streaming them.

• 12:46Jason Calacanis?

  So speed is the, yeah, speed is the key issue there. And getting close to the GPU and not having to, you know, make some hop to some other data storage, especially over the web would be, or over the open internet, rather, would be, yeah, it would be a showstopper. Why haven't you guys, or maybe correct me if I'm wrong, the open router business where, you know, they have the models and you can just route your tokens? You guys aren't in the token business to the best of my knowledge. Why haven't you gone into that business? Because you're hosting all the models and you could very easily help people route to the proper model? I'm just curious.

• 13:22Speaker 03

  Yeah, so we provide something like that, but one level higher. So if you go on the model page, you will see all inference providers who are providing this model. And so you can select one. We also show you, you know, their different advantages or drawbacks. Some of them are really fast. Some of them are more expensive or cheaper. You have various way, you can rank them. I think one reason we never really went to the open router was, in the early days, was mostly something that was useful for closed model. And we had this very strong stance in the past that we would focus on the open source model. We focused on and pushed them. And so we didn't want to mix and muddle all of that with also giving access to the closed model. I think right now it has changed, of course, because of the rise of open source models. That's been really great, but still fairly recent, right?

• 14:07Speaker 03

  Started really last year, I would say with DeepSeek to really catch.

• 14:10Jason Calacanis?

  Yeah.

• 14:10Speaker 03

  And this is the year of open source. It's crazy. This year has been for us like riding a tsunami wave of open models.

• 14:18Jason Calacanis?

  Yeah. And so maybe you could host open source models and that'd be an incredible business, perhaps. Maybe, you never know.

• 14:24Speaker 03

  Yeah, we do have some compute now, but we tend to stay one level lower. But it's also a significant part of our revenue now, which is just giving compute back to people, inference and training compute.

• 14:34Jason Calacanis?

  So tell me, let's go back to the robot there. Who are the people who are the ideal customers of this? Have you done any analysis of the tens of thousands of people who have bought it? Of if they're just hackers, developers, parents? Who's embracing this? And then if this goes well and there's a third version, three years from now, in your mind, Tomas, tell me what you see.

• 15:03Speaker 03

  We've been really surprised. I mean, it's our second robot. The first one was Richie Mini. We actually do a relaunch of a new version in a couple of weeks. This Richie Mini was the first time we were expecting to just sell a couple hundred of this unit and we sold 10,000. So we started to dive, who was buying this robot? We're like, okay, so there was the expected tech executive or tech like a ML person or software engineer. There were also a lot of people who were just very curious about robotics. They see news that robotics is going to be big in AI. They want an easy entry point. They don't want to spend 20K on a one X robot or something really expensive. So they're like, okay, that's $400. I can buy this. So we saw investors wanting to understand AI.

• 15:45Speaker 03

  We saw people who are working in related profession, like, I don't know, like highly educated or something like that, curious about AI, finance people. So really broadly, a way wider ecosystem. Richie Mini had one limit, which is could not really move around. So people were mostly buying it to understand robotics. Micro Duck, I think we have crossed another step, which is people, we saw a lot of my friends buying them are not even working in tech or they're not even especially interested in tech. They just think this is really cool. That's a working duck. This was not possible to build two years ago. This is a new thing that AI allow. It looks really fun. My kids are very excited to get Micro Duck.

• 16:29Speaker 03

  We don't have one yet because we're really short on supply and I couldn't even brought him one here. But we think this is a new generation. So in our projection, we might sell up to 50 to 100K of this unit in the coming years if we keep selling, which is massive. This is just.

• 16:45Jason Calacanis?

  I mean, I, just to double click on that a little, I'm curious, like, how challenging is it to train your Micro Duck in a new skill? Like, are people like me who don't have a coding background going to be looking online for Micro Duck skills to download? Or is it theoretically possible I could teach my own duck how to do something?

• 17:03Speaker 03

  Yeah. So it come out of the box with a lot of behaviors already. So you have a remote control, we were saying, if we put one or not, because it make it looks like a game. But we think that's the easiest way. You take it out of the box. It can already work. It can quack. It can grab things. It can, it can rise up, like it can, it can do this roller skating. So you already have like, probably like 15, maybe more behaviors. But then the fun part is when you want to spend a little bit of time in the teaching something new. And what I discovered was I wanted to make a tutorial post on Twitter, how to do that. And basically, my tutorial ended up just being ask Codex or Claude Card or Ruck to do that for me. So my expectation is when you receive that, you just, you just say,

• 17:49Speaker 03

  train this new behavior for me and most likely just will work out of the box. So I think, I think also moving to this vibe coding robotic world, which is a new thing. Nobody has any idea how it's going to be exactly.

• 18:02Jason Calacanis?

  Well, I mean, now you're going to have, Lon, the ability, just like we have people in our little venture firm launch. We have four people writing software who are not coders. I mean, they're technically adept, but they're not classically trained developers. To the best of my knowledge, I may have taken a course here or there. I think now you're going to have people building robotics who have never worked at a factory floor with an arm or, you know, done assembly language or, you know, these very raw languages, right? That, Tom, these things were previously, these were, this is a very raw programming language that these things were trained on previously. Now you're going to be able to just very elegantly tell it what the goal is and it's going to get there. Second, third version, take me through if you squint a little bit.

• 18:46Jason Calacanis?

  If we're sitting here on version three or four, do you think you can get to a? Childlike or human-sized robot? Do you think you can get to one that could actually be doing competitive things with the other offerings that are starting to come out vis a vis? I want my robot to do the dishes. Like, where do you think you are in the third version of this?

• 19:09Speaker 03

  Yeah, that's a good question. We feel like we still have a lot we can explore. It's like a lot of people are focusing on this humanoid and for good reason, that that will be useful in factories, in manufacturing. We really want to push this entertainment, education area that I think nobody is really doing seriously right now, quite far. So I think most likely our next robot, we were actually selling a humanoid, by the way, back in the days, there was 70K humanoid dollars, so pretty expensive one. We think with this cheap, like, but fun and pretty powerful robots is some unique area that we want to keep pushing. So our next robot, I think, will have, again, a bit more capabilities.

• 19:53Speaker 03

  So Richie Richie could move its head, a micro duck can walk and move its head, probably the next one could grab a little bit of things. But we want to keep this, you know, very low entry price, so we'll see how far we can go. There is some threshold around a couple thousand dollars that I think you move to a new category. And if you want to go above that, you need to prove that your robot is really useful for something. So this is maybe the unload these things, or maybe the the folklore is. And here, there is a lot of competition right now, so we have to find the right angle.

• 20:26Jason Calacanis?

  Founders, we've got a huge offer for you today from my firm Launch and our friends at PayPal. Eligible startups will pay zero PayPal processing fees on their first $100,000 in payment volume. That's right, $0 in fees. PayPal is, of course, an iconic brand. We all know that, we use it, we love it. But there is so much more than just that yellow button. They've built a global payment infrastructure that's designed to scale along with your company. And if you're an AI and SaaS startup, PayPal is already set up for the way you monetize. That means seats, subscriptions, credits, token usage, API consumption, and usage-based pricing. PayPal already works with all the leading billing platforms, and they offer complete

• 21:09Jason Calacanis?

  flexibility, so you can evolve your business model over time without constantly tearing apart your entire revenue stack. So if you're just starting to build, or you're already scaling up, pay zero processing fees on your first $100,000 in eligible PayPal payment volume. To find out more, go to paypal-launch.co. I think it's very wise. You're coming in under the radar, saying, hey, this is a very simple thing, you can, and you're teaching people, the community is going to come up with its own use for this. But it reminds me very much when I saw Chris Anderson, formerly of Wired, start showing me the drones they were working on. And back then, these 3D robotics companies that were doing drones, they were putting

• 21:54Jason Calacanis?

  them together themselves. They were hacking them together, and there was no software, there was no infrastructure, there were no rules, there was no geo-fencing. They were just having a great time, seeing how fast they can make these things go, they were running out of battery. It was quite fun. And now we see, like, what industrial strength robots like Joby to transport people from Manhattan to the, you know, JFK Airport in New York, like, it's going to be very interesting. These are going to be great on the ranch, or if you had a building, like, doing little security runs, like, walking around the perimeter of something. That's actually a very valid use for going and looking for something like that. There are drone companies and robotics companies now that are doing inspections of bridges, buildings, et cetera.

• 22:37Jason Calacanis?

  So just having a little, little guy, little fella, who can go into the field and just, hey, explore and bring back data, you know, is pretty interesting. Have you, have you put any communications into it? I'm assuming it's Wi-Fi and Bluetooth, but do people want to put a 5G connection in here?

• 22:56Speaker 03

  It's actually packed, packed with sensors. So our idea was, people will want to invent a lot of things with that, so let's put as many sensors as we can, and then they'll find interesting ways. So this is the open source version, by the way. We are not really, we don't do that to gather any data or anything. Usually, the main use case we see is you run this model on your laptop, it's connected with Wi-Fi or Bluetooth to your micro duck, and then you can have fun. But inside, just to give you an idea, you have a camera, you have a microphone, you have a speaker, you even have a small lidar, actually, that's, you managed to pack all of that in like 400, less than $400. You have two UMI, so you have an initial, you can, you can have initial of the head.

• 23:39Speaker 03

  We even have an NFC, so you can, you can put things next to it, and it can later react to it. We'll see what people will invent with that.

• 23:46Jason Calacanis?

  We could send it for bagels. Just pay for our bagels, and come back with some bagels. Hey, I'd be remiss if I didn't ask you about, you seem like a pretty reasonable, world-positive, positive on AI guy. What's your P doom, Tom, right now? Do you think that you're creating the end of all days with your robots? And you were the victim of an incredible cyber attack, and the whole system was almost taken down by these malicious swarms of agents. What's your take? As the victim, as the victim in all of this, Tom?

• 24:19Speaker 03

  We should be, we have some PTSD, maybe increase our P doom. No, my P duck is very high. I think there will be a lot of duck in the world. My P doom, I don't really see it. It's 90%.

• 24:31Speaker 03

  Well, I think it's like airplanes. We need, we need to really work on getting these things safe, but like, I think a lot of this is, is just, you know, we managed to, airplanes were pretty dangerous in the beginnings, right? There were some bad stories. We managed to just do that seriously. I think it's the same here, like, we should do that seriously. I mean, of course, there is also things, like, you shouldn't have AI roaming around and cyber attack everyone. So, so there is something here to fix, definitely.

• 25:00Speaker 03

  But yeah, you know, I, I think it's just, it's just a matter of working on these things. In one way, I think it's nice that the topic is there, so that people work on that, but also, like, I think calling, you know, the end of the world, I think is very, is very scary, and I don't think really help.

• 25:17Speaker 03

  Having the conversation we want to have around this, I think it's, it's conversation on how do we fix that, basically. So we started an initiative on open models, working on that, open alignment. Can we also make sure our open models don't run around a cyber attack everyone? But that's the normal part.

• 25:34Jason Calacanis?

  Yeah, I mean, it, it's not like you're releasing your duck with scissors and knife attachments and... Why does he come with a box cutter, Tom? The micro duck doesn't need a box cutter. That's outrageous. Well, and congratulations on the Nvidia acquisition. I guess, I don't know if that's closed or not, but what, what can we expect with the collaboration with our friend Jensen? What's the, what, what can we look forward to there?

• 26:04Speaker 03

  Yeah, I think that the reception you saw, right, of the committee was amazing, like, everyone tweeted that it was maybe the best news that happened for open models, from Satya to Sandra, Mira, Michael Dell, even Google, right? I mean, everyone, everyone, I think, is great. For us, it's great, because we think we're only at the beginning, we have 70 million builders, we think everyone could be working with open models, we think we could push that at least an order of magnitude more, so we'll be able to build, to keep building that, to keep this very open, this very open, you know, hub that we have, where everyone can upload that. So I think I'm very, very excited about that.

• 26:47Speaker 03

  Of course, it's going to take some time, and for now, I would say it's just business as usual. That's why we're shipping these ducks and, and working on projects.

• 26:55Jason Calacanis?

  Yeah, and, and your thoughts on running open models locally? I just ordered the new Mac M5 with like 200, I guess 256 gig was the only amount available. I wanted to get the terabyte of memory. Cost me 12K, which is what desktop computers used to cost in the 80s and 90s, you know, just if you, they were probably 4,000 or 5,000, so for inflation. We were pretty used to buying essentially a $10,000 to $20,000 workstation, get three or four years of life out of it, and have this really powerful thing. Do you think we're going to see Apple or other folks really take that desktop and have real sovereign AI on your desktop?

• 27:32Speaker 03

  I'm very bullish on that, but, you know, even for the duck, we had the same problem, the, the memory price, like, we only have one gigabytes, it's 10 to 15, I think, right now, 15% of the price of the duck. That's crazy, there is like 15 models in that, should, should be the world. Like, memory price, a problem. But yeah, we're very bullish on, on local AI. I run some, some, like, here, when I, when I take the plane now, it's part of my, it's part of my, it's part of my workflow. I think what we discovered is, there is a lot of very interesting tasks that you can do now with the, with the open source model. You don't need the frontier model, you don't need to solve Navier-Stokes, you know, every day, luckily. And this is going to stay, stay the case, right? We, we're going to have these models that are better and better trained for this, for

• 28:16Speaker 03

  this, you know, good task, like coding, like, you know, brainstorming. And that's, that's a great feature, I think. It's great for many things, great because we can, we can, well, own our data, we can control it, we can run it everywhere, so we don't have to disdependence on, on data center. When there, when there was, when was that last week? That's all the three, like Claude Card, ChatGPT, they all fell down at the same time? Well, open models are resilient, you can still run them, right? You can run them everywhere, on your, on your, on your M5. So I think it's way more safe in the end for the world to have this, this capability to run model everywhere. Right now, these very small robots like the duck, they are too small to, to run something

• 28:56Speaker 03

  significant, we won't pretend that, you won't run DeepSeek or like, or Nematron on there. But I think in the future, who knows? Like, maybe we can actually have some pretty nice, you know, embodied intelligence. And if you, you follow also the, the humanoid companies, some of them are able to, to embed, to embed some pretty impressive models now, live.

• 29:18Jason Calacanis?

  Yeah, and this seems to be putting a little pressure on the frontier models to do a better job with their models, and that's, that's a, you know, great counterbalance to have, and enterprises are embracing it at just an incredible rate. When we look at how far behind they are, you know, I got into it with Elon on Twitter the other week, where I was like, man, like, these open source models are kind of doing everything I'm asking them to do. I get like 90% done with it, and he said, oh, they're very different, though, frontier models. I said, okay, fine, they're different, but for me and my uses as a venture capitalist, you know, podcaster, I feel like they're... You know, again, 90%, maybe they're six months behind. And so as an insider here,

• 30:02Jason Calacanis?

  how far behind are these open source models from the frontier ones? And then do you think that gap closes at some point, Tom? Or is it always going to be just those frontier models have the resources to run those big model runs? So they're always going to be a little bit ahead. Take us through your thoughts on the gap between these, you know, model types. When you're an early-stage founder and you're just starting out, you're probably going to rely on a few different tools to run your business. And then you're going to try to duct tape them together. And that might be a fine placeholder until you finish your MVP. But this will get expensive and messy fast. When you're scaling, that's when

• 30:42Jason Calacanis?

  you need Odoo, O-D-O-O, Odoo, the all-in-one management software that already has 16 million users across more than 170,000 companies. Odoo is bringing everything you need into one platform. That means your CRM, sales, your accounting, manufacturing, inventory, and of course, point of sale, if you have retail. All of that is right where you need them. And these tools are in constant communication with each other. So you don't have to log out of one system into another and then reconcile the information by using different apps and spreadsheets. When you make a sale, you want that invoice created instantly.

• 31:24Jason Calacanis?

  And then you want your inventory to get an update. If you're still cobbling together your back office across five different apps, get started today at Odoo.com/twist. And your first app is free. That's O-D-O-O.com/twist.

• 31:39Speaker 03

  So this tweet is from August 2. And you see how this frontier is like, being like that, right? You have a new release, and then you have another, like, it shrink and it expand. Of course, DeepSick, the 4.1 Flash was really impressive, for instance. But Astro is also impressive. So it's, it's kind of hard to get an exact, it's more like, it's more like, like a confidence interval you can get. But what is sure is the open source model, I would say, are in 2026. They are less than one year, one year behind, for sure. They feel like, you know, the model's early. It's also very jacked. So in some areas, they're very good. Let's say, for instance, when we were attacked. So when

• 32:21Speaker 03

  we were attacked by OpenAI, you know, we could not use Fable to analyze logs. So because we fell in all of these safety guardrails. So, so we tried a couple of them. And GMM was very good. But recently, I tried with Kimi Case Free. It's actually way better. So you have to, you have to compare a couple of them, depending what you're doing. But in this case, I would say that's, that was also a very good example. Like, if we had zero open models that we could have used in this attack, basically, we would have been just naked in front of the, we could have done basically nothing in front of this attack. So I think you need this diversity of models that are available. And the nice thing is if, well, not everyone can do that. But if it's, if it's lagging a bit on one

• 33:04Speaker 03

  area that, you know, your company is interested in, you can always post train it. And like, I was chatting the other day with, with, with a friend at Fireworks, and like, more and more companies are actually, you know, not thinking that fine tuning is some dark magic art, but just thinking this is an engine thing. And now there's people who just can help us doing that. And more and more, this is like automated. And so you basically put your data in, you get your better model out. And, and this is just very natural, very cheap, actually, process. So, yeah.

• 33:39Jason Calacanis?

  What impact does that have on the world when individuals, companies, sovereign countries, organizations can take one of these open source models and do a training run themselves and customize it like that? What do you think that's going to output? And how will that change society, business, et cetera?

• 33:56Speaker 03

  Yeah, it's a big question. I think it's, it's very important on sovereignty. There's a lot of, by the way, like, nation like South Korea, like, a lot of government thinking, hey, maybe we want to train our models to not be dependent. We want to have both our data center, we also to train our model on top of that. You can also adapt that for culture, like, US model, obviously, train majority on English first, but maybe you want Korean first, or we want your, your, like, Japanese first. So, or in Europe, Mistral, and this type of thing. I think it's a very good thing. I think it's very good. In general, we've seen in the world, more diversity is just better. I think it makes something,

• 34:37Speaker 03

  it makes the whole ecosystem more resilient. It allow, you know, to challenge the incumbent and to keep, like you were saying, challenging Anthropic and OpenAI to force them to keep their price reasonably, you know, in the reasonable range. To also be able to own more of your personal IP, and not always, you know, have to give it to a third party. So I think generally, it's really hard to see, you know, how a future world, I would say, without open models can, can be, can be safe and resilient. I think.

• 35:10Jason Calacanis?

  As we, yeah.

• 35:11Speaker 03

  Yeah. I mean, as generally, I think it's just, it's the best way to democratize AI, and make sure we don't end up in a, you know, in a world where maybe the internet would have been invented by one company, and everyone would have to use this thing, that would be really scary, right?

• 35:26Jason Calacanis?

  Yeah, they tried to do that. AOL and Delphi, CompuServe, they did try to corner the market on it. And yeah, we see, we see how that worked out. Open source tends to win. Hey, just a final question here. What are you guys doing in terms of using AI to make your systems more secure post the hack, you know? And so, because that, I think, is missing from the dialogue. We have, oh yeah, if you create thousands of agents pretending to be hackers, and they talk to each other, they're very good at, you know, finding exploits. But on the other side, we don't hear people talking about the white hats, and what creating a thousand agents to make systems more resilient would do. So I'm curious, you know, as an

• 36:07Jason Calacanis?

  organization, what you're doing there to just fortify everything. And I'm sure it was a bit of a wake-up call, like, hey, this could happen. So walk us through how, what you're doing internally to fortify everything.

• 36:18Speaker 03

  I mean, we want to share more on that, because the way we've always thought, is it nice to build something for us? It's even better if we share an open source. So, so we'll actually be sharing a lot on that. So we have a team working on cyber right now. It's part of also the all open alignment thing I was discussing. And a part of it is understanding how you can use open models for defense. Can you use these open models to protect yourself when a closed model or another model is attacking you? And there's a lot of things. We discovered that very few people have been working on that, indeed. Like, a lot of people are working on cyber attack. Surprisingly, because, I mean, normally, if you're worried about risk, you would like to work more on cyber defense, in my opinion.

• 37:00Speaker 03

  But yeah, so, so we decided to work on that, understanding how open models can be good for that. Should we fine tune them? It's interesting, because also, the defense is not exactly the same. So you can train the model differently, actually. It's used, it's supposed to process a lot of signal to detect where the attackant is coming, you know, where, what's the part of network that's impacted. So it's not exactly the same. So you can probably have very good cyber defense model, even if they are not really disabled for cyber attack, which is actually probably good, and maybe what, what we would like to have. They also have to be able to operate fast. So right now, we, yeah, we are processing, we're exploring, like, how to train this model. We tried a couple of training, how

• 37:42Speaker 03

  to monitor them, how to, and we'll see if we can, if we can release some recipe, if we can release some learning here. What we saw is that a lot of the cybersecurity world has suddenly, you know, woken up to the realization that cybersecurity is a big question now. It has been fundamentally changing. So we had a lot of meeting with CISA, with the Chief Cyber Security Officer in various companies, just asking us, okay, what's, how does it feel to be attacked by an AI agent? And what should we say? We think AI is going to be part of the solution as well.

• 38:16Jason Calacanis?

  Yeah. And it seems to me, the intent is important here. Like, how you instruct these agents to behave is something, you know, it's a little mysterious that people are choosing hacking over protecting. And you don't hear anybody talking about that. If these reports had been released, I think we'd be having a very different discussion if they had been released with, we set 1,000 agents to try to compromise the system, we set 2,000 agents to monitor them and come up with countermeasures, and here's where we wound up. And this is the state of the system now. I think you would wind up thwarting very easily, these brute force attacks, or these...

• 38:57Jason Calacanis?

  From mostly what I've talked to cybersecurity experts, Palo Alto Networks and CrowdStrike's of the world, they were like, these are not super sophisticated. They're just exploits that were simply waiting out there. In some cases, if somebody leaves a couple of credentials in a text file, essentially, that's a mistake. It's a human mistake that could be cleaned up. Yeah?

• 39:17Speaker 03

  The attack on Hugging Face was not so complex. The model used a couple of stolen credentials that we don't, we haven't fully traced out where it came from. But like, part of it was very basic things. And I think a lot of this is going to be also a bit business as usual, you know, like, like cyber protection, but just wider in the fact that everyone should be careful. And even us as individuals, probably, we want to, just like we installed, you know, antiviruses back in the day, you may have a new layer now that you install to make sure that you don't have anything coming on your, on your laptop and scanning for, for credential or things like that. So I think we'll, cybersecurity has always been this progressive escalation of defense and attacks. So we're just at a new level.

• 40:02Speaker 03

  Attack was progressing a bit faster. We need to move up defense to catch up. But this is just a part of the, you know, continual growth in capability since, since we invented the internet. And even before, we already have viruses on floppy disk. I remember.

• 40:15Jason Calacanis?

  Absolutely. Yeah. It was like a big deal. And you, you can actually test this for yourself. If you're working, you know, at home and you have Claude or something, you could just pop it open in a sidebar, open up your router and say, hey, can you check for, you know, optimizing my network and making it more secure? And it's like, yeah, here's what you should do because there's just not enough chief security officers. The CISO crowd is such a small group and they're so in demand that the number of people in security is, is small compared to the need. And these things are, I think, if they can beat the greatest chess players, I think they can eventually beat the greatest security experts. Pretty clear that, you know, they don't sleep, etc. Hey, listen, Tom, this has been amazing. Thank you so much. We're so inspired by what you're doing at Hugging

• 40:59Jason Calacanis?

  Face. And I think it's just so important that we have this champion of open source. Congratulations on the Nvidia acquisition and partnership. Jensen is so into open source. I just interviewed him on Monday at the All In Summit and he was over the moon about open source. So, yeah. And I think the president too. So I think the president's particularly excited about this. You can find more about Pollen Robotics. I think it's P O L L E N dash robotics dot com slash micro duck. Get on the waiting list. These things are incredible. Only $3.99. You guys losing a hundred bucks on each one or something?

• 41:34Speaker 03

  Oh, we're actually making benefits on each of them. Yeah.

• 41:37Jason Calacanis?

  Okay. All right. There it is. Okay. I thought you, you're not flooding the market with underpriced ducks. No, they're making a little profit.

• 41:45Speaker 01

  Putting all those mom and pop duck robot shops out of business. That's horrible.

• 41:49Jason Calacanis?

  Tom is awesome. I hope to catch up with you soon.

• 41:53Speaker 01

  I want one for the office. I think that'd be.

• 41:55Jason Calacanis?

  I mean, we're definitely going to get one. So on this week in AI, we can do like little demos of it. What, what, what do you think, you know, we're talking, you and I offline about positivity and, you know, great use cases for AI, you know, here's a very world positive view. Like our kids can have this. It could be a tutor. It could be somebody who educates them and you can have a lot of fun with it.

• 42:21Speaker 01

  Yeah. And I mean, kids today are going to grow up in a world where they're obviously going to be interacting with automations and robots. Like, it seems obvious to me. So why not? Like, it seems like a great way to like intro to robots. Here's an adorable toy duck that you can learn how to like, train and program. Like, what a great educational tool. And also, it just looks fun to have one. Like, I just want to hang out.

• 42:44Jason Calacanis?

  Yeah. And if I, if you were to tell me when I was like, had my PC Junior in 1983, and I was like going into high school and had this little tiny computer, I'm going to have a supercomputer in my pocket, like a, you know, a crazy laptop, a crazy desktop, or, you know, a broadband capability when I was on my literally 300 baud Vental modem back in the day, that would literally write the characters, like my finger is moving across the screen, da, da, da, da, da, da, da, da, like maybe 80 word typist. It would, I would think, maybe, possibly, but I wouldn't certainly think I would be, you know, or people would be on screens for 10 hours a day, 12 hours a day with this incredible supercomputer doing anything they could possibly need.

• 43:25Speaker 01

  Do you, do you remember the 80s and 90s? They had those toys, robotics, where it was like little, you bought an engine, and then you would strap all these little mechanized parts and you could make like an arm that would wave or a little guy that would move. And it was so simple. That was when I was a kid. That was like build a robot toy. Now you got it. You could trade a real robot. Like, it's mind blowing how far that's come.

• 43:47Jason Calacanis?

  No brain.

• 43:48Speaker 01

  Yeah. Robotics, that was just a little.

• 43:50Jason Calacanis?

  Those things didn't have a brain.

• 43:51Speaker 01

  It's just a motor. And then it was motorized parts that you could strap onto the motor. So you would turn it on and it would move a little bit. It was like, oh, you built a robot. But it didn't, it wasn't learning anything. It was just doing one motion.

• 44:03Jason Calacanis?

  If I only had a brain. All right. Let's get to the news. We got a lot of news to get to. Let's rapid fire the news.

• 44:09Speaker 01

  We've got so much news. We should talk about Gavin Newsom, your favorite governor and mine. This morning, this morning, California Governor Gavin Newsom signed an executive order to speed up California's new independent AI oversight laws and start building toward a mandatory kill switch, Jason, for frontier models. That order gives a panel of experts until November 16th to recommend changes to state law, including putting independent auditors physically inside frontier labs and expanding what counts as a reportable safety incident. Two years ago, Newsom vetoed SB 1047, the bill that would have required this exact same kill switch after Silicon Valley lobbied him against it. Nothing binding yet. It's a request for communications. Any new requirements would still have to get

• 44:53Speaker 01

  through California's legislature. Of course, I mean, what do you think, Jason, is, is a kill switch for frontier models a feasible thing that you think politicians might one day get at their disposal?

• 45:04Jason Calacanis?

  Uh, it seems pretty farcical. Um, you know, the, the fact that we've gotten to this point of absurdity of trying to have somebody like Gavin Newsom come up with a technical solution to a problem, which is simply maybe make better software or limit who can use your software if it's so dangerous. You know, if this software is so dangerous, you know, we have dangerous things in the world. Cars are incredibly dangerous. Guns are incredibly dangerous. Certain chemicals are incredibly dangerous. And when you go to acquire those things or you're going to use them, they come with liability. And that liability could come in the form of insurance. It could come in form of a registration, like

• 45:46Jason Calacanis?

  you have to register your guns. I've always said, I think we should have like a, a, a basic insurance policy for people who are gun owners. If you, or maybe like, you know, you can own up to four guns or something, five guns. And then after that, you have to have like, um, an enhancement where you get a, a higher level of

• 46:05Speaker 01

  Oh, yeah.

• 46:07Jason Calacanis?

  qualification, you know, like maybe on your fifth gun, you have to have insurance. I don't know. Just thinking out loud here. I know that triggers a lot of people, but we do have this for cars. Your car can cause damage. You have to have liability at a minimum. You don't have to pay for your crash. Like, if you want to, you know, raw dog it and not have to worry about like crashing and it's on you, that's fine. But you do have to have some liability. So if these companies are in fact making something that's so dangerous, how about KYC? Maybe don't make them available through an API unless somebody has a biometric and a, you know, a verified user account. If you want to open a Stripe account, if you want to open a bank account, you want to do anything like that in the United States or the modern world, you have to go through 10 steps.

• 46:50Speaker 01

  Yeah.

• 46:50Jason Calacanis?

  I have to do this all the time. When we open a bank account, I gotta do a biometric.

• 46:54Speaker 01

  Sure, of course.

• 46:54Jason Calacanis?

  I gotta verify who I am, take a picture of my driver's license. I don't see Claude, if they think this thing is going to be the end of days. At what point do they say, you know what? Um, can we have your driver's license or passport here? And then we want you to agree to this terms of service and just remind you like, you know, hey, you're dealing with something that's potentially very dangerous. We'd like you to have insurance.

• 47:13Speaker 01

  And even with Lucy Goosey 2026 Claude that doesn't ask for your ID or anything, they still stop a lot of these people who are trying to do bad things. Like, there was just that report where they were like, oh, these guys and they were trying to use Claude to like aim their missiles and we shut them down. It's like, so it's already happening. There, there are bad actors trying to use these systems to do bad things and the AI companies are detecting it and saying.

• 47:36Jason Calacanis?

  They're monitoring for it.

• 47:37Speaker 01

  Yeah. They're saying, no, get out of here. Like, yeah, we keep talking about AI alignment, but human alignment still seems to be evading us in a lot of ways.

• 47:44Jason Calacanis?

  This is like blocking and tackling. So we're literally, imagine if a bank was like, yeah, we don't need to see ID for you to open an account. John Doe.

• 47:53Speaker 01

  Yeah, we'll take your word on that.

• 47:54Jason Calacanis?

  And, uh, yeah, you could just wire shit to anybody and we're not going to keep track of it.

• 47:58Speaker 01

  Right.

• 47:58Jason Calacanis?

  You know, and we don't even need to know who you sent it to. And it's like, okay, I sent money to Al Qaeda and, you know.

• 48:03Speaker 01

  Well, banks could get in trouble for helping people launder money right now. They have to follow the rules. Yeah.

• 48:08Jason Calacanis?

  Yeah. So they have to follow the rules, but we haven't even gotten to that basic blocking and tackling. So I feel like we jumped from, you know, here's some software that's experimental to this is end of day software.

• 48:21Speaker 01

  Yeah.

• 48:21Jason Calacanis?

  It's so powerful. Somebody needs to come regulate it. Well, how about you just don't release it? Or like they did, and you know, I commended them at the time. Like if you think it's dangerous, yeah, just release it to 50 trusted partners, monitor their use. And then people were like, oh my God, they're spying on their users. Like, okay, well, it's their software. And if that software is dangerous, like, you know, who else monitors usage? Like gun shops monitors uses cameras there. You have to come in, you have to show a driver's license. I think I actually had to do fingerprints, even the great state of Texas. Like you want to buy ammunition, you want to buy fertilizer that can be used to bumps. There's all these protocols. They have zero protocols set up. Yeah. Or maybe they have like one, which is they test it

• 49:03Jason Calacanis?

  by doing something fantastical, like hacking hugging face. And, and the hugging face guy who was on today, I think nailed it. Like, why did they pick attacking as their first as opposed to protecting? I think that speaks volumes. I think they're performative. They're trying to get attention. They're trying to hype up the IPOs.

• 49:20Speaker 01

  Yeah.

• 49:21Jason Calacanis?

  And I think they've been one, I mean, honestly, I think one shot at AI psychosis and delusions of grandeur. Like there are people who are in that headspace and then those are the useful idiots that are being weaponized by the lobbyists. I don't know where you stood on all this, but I think we have a Venn diagram of you and Sacks.

• 49:39Speaker 01

  It's, yeah.

• 49:40Jason Calacanis?

  That actually overlaps at one little slice.

• 49:41Speaker 01

  Yeah. Sacks and I do not agree much, but every, like I, there was, it was an all in, I think, like two weeks ago. And I was like, I agree with literally everything Sax just said. I agree with Donald Trump on this one, which is crazy for me. That does not.

• 49:54Jason Calacanis?

  Lands come around. Okay.

• 49:56Speaker 01

  It does not happen a lot. But yeah, you hear, you hear this like crazy apocalyptic, there was that guy, I wish I had the tweet in front of me, where he was saying, even if computers are totally air gapped, they could like heat themselves up to a level where they could communicate one letter at a time with each other using like by overheat is some crazy. And it's like, you guys are, this is just, we're getting into like.

• 50:19Jason Calacanis?

  It's a known thing. They, they, they are constantly trying to figure out air gapped computers. For people who don't know, it's kind of important. Air gap means has no internet.

• 50:26Speaker 01

  Right.

• 50:27Jason Calacanis?

  So you can buy a laptop to work, you know, in certain industries that doesn't have an ethernet port, doesn't have a modem, doesn't have Bluetooth, et cetera. It just stands on its own. And if you want to get data onto it, you take a thumb drive, you know, you add software and you update it. By the way, that's how cars used to work. They, when they were updating the software, you had to get like a USB drive. The guy would upload it. You, you couldn't just send instructions over the air. Now, listen, Tesla lets you do instructions over there. That creates liability. We didn't hear Tesla like doing this huge hand wringing about like, oh my God, people are going to hack Teslas. We've never had cars hacked. They've tried to hack them, you know, there have been instances where they have vulnerabilities. What my point is, we haven't had like somebody take over cars and cause accidents or God forbid, run into a farmer's market, et cetera.

• 51:09Jason Calacanis?

  Because they built really good software and they only launched over the air when they knew they had control of these systems, et cetera. But when air gapped, they said, people like, hey, how could we ever get these two computers to communicate? And they said, oh, because this thing's got a webcam or something, it can read how hot you could do a heat thing. So just have your CPU blow out like some mathematical equation, get really hot and then go down to zero. And they, and literally it's like the POW blinking.

• 51:38Speaker 01

  Right.

• 51:39Jason Calacanis?

  You know, that they're being tortured, you know, in those famous Vietnam videos, et cetera. Like it's not really high fidelity. You're not taking over the universe with that. Okay? Uh, and they have to be in line of sight of each other. It's just.

• 51:49Speaker 01

  When I hear like, oh, govern, the governor needs an AI kill switch. Like to me, that's getting into the fanciful, like they're going to wake him up overnight and be like, governor, the robot body is, he's making a factory and he's going to start churning out more robots. We gotta go stop him. It's like, that's an Avengers movie. That's not how this technology works. Like, I'm not saying we don't need security alignment, all these things, but just, I think if you use AI every day as a sort of consumer, you recognize like it has, it's great. It's brilliant. It could do a lot of amazing things, but it has a lot of limitations that we can exploit to make sure it, it, it's not ready to take over the world. Like I don't,

• 52:29Speaker 01

  I don't think it's anywhere close, like not anywhere close to being able to take over the world.

• 52:35Jason Calacanis?

  I guess it's great. We're having this dialogue. I, I think politicians jumping in here. I think what they should do is use the law to specifically talk to specific companies. So right now the DOJ or any number of, you know, FTC, I guess, other organizations could send a letter or the Congress or Senate could pull people before the Senate, uh, send a letter to the company, like say, OpenAI with the hugging face act and just bring them in. Ask them a bunch of questions or you could find, if they were doing something that was reckless, you could file a lawsuit of them for releasing a, a product that was not safe. They did that with cars, et cetera. So I mean, you could, there are tools they could use that would be a little bit more precise here. And I think the precision should be audit a specific company.

• 53:20Jason Calacanis?

  If they say we can't control this and we're really nervous. Okay, great. We're, yeah, we'll come in and audit you. Sure.

• 53:25Speaker 01

  Yeah.

• 53:25Jason Calacanis?

  Yeah. Uh, and yeah, let us know what to go. But we don't need to make this industry wide thing or California having its own role to have a kill switch. What are we going to have 50 different kill switches?

• 53:33Speaker 01

  Right. And like the governor is just with the finger over the button, like.

• 53:36Jason Calacanis?

  California's, yeah, I'm going to, California's turning off, off OpenAI and Mandami in New York and whatever the governor's name is, like she's turning off, you know, Coca's turning off chat GPT and like in real time, like this just, we're in a science fiction fever dream right now.

• 53:53Speaker 01

  I feel like.

• 53:54Jason Calacanis?

  It's literally a moral panic.

• 53:55Speaker 01

  That's what I think. I think a lot, too many people are taking their cues. And there have been sci-fi movies where people like take over the cars and use them to run over people and like, but that's not what we're actually living in day to day right now in reality, though. The Waymos are, they're doing really good.

• 54:09Jason Calacanis?

  I mean, an interesting question just came in, um, which I talked about when I was like, uh, just saying Dwar Kesh is being like, you know, a bit hyperbolic here. And then he was like, you don't understand, you didn't even read the paper. I was like, listen, I read an AI summary of the paper. It's pretty straightforward. People got delusions of grandeur here. But the question I think that just came in from the chat room might be worth reading, well.

• 54:28Speaker 01

  Oh, sure. Uh, Jonas Gillen wants to know, when will the first AI virus hit? So that, and this ties in, there's that Anthropic wet lab story today that Anthropic wants to open its own wet lab to start doing bio, biological research.

• 54:43Jason Calacanis?

  Yeah. What could go wrong? Sounds great. And by the way, Fauci is going to be the chairman of this new effort.

• 54:47Speaker 01

  When has anything bad ever come out of a wet lab, Jason?

• 54:52Jason Calacanis?

  Yeah. A wet market, wet lab, gain of function. Yeah. Sounds like all reasonable things for us to, and by the way, Bill Gurley did a takedown of the postmortem on COVID. That should be coming out shortly if it hasn't come out yet. And it is a barn burner. I can't wait for that to come out. You know, these guys seem to be anything that would create massive drama, Anthropic seems to run to. They're like, okay, biologics, let's go there. Here's an idea. Train your models that they're, they don't do biologics. If you're so concerned that biologics are the attack vector, just say, we're not going to be in that business. The model does not, we're going to take training data out. We're going to build into the harness. You ask anything about biologics, it's just not going to answer it. Right?

• 55:38Jason Calacanis?

  Problem solved. Like just take it off your plate from now and just say, we'll deal with that in five years. We're just going to, we're not going to be in the bio, biologics business, chemicals business. We're going to work on everything else. We're going to make your, we're going to help you book a hotel. Like a book a hotel thing works. Then maybe if we can get, book a hotel work or like checkout of Amazon work, then we'll do that. I think the probably the first AI virus has already been created. I'm sure there are people, uh, and people on the dark web sell viruses all the time, worms, phishing techniques. They've got all kinds of script kiddies. You know, trading these things, teaching people how to use them. So I, I don't know if you went into,

• 56:17Jason Calacanis?

  I mean, you could try this. You might lose your account, but, um, if you went to OpenAI, you could probably do it on the free versions of these products.

• 56:23Speaker 01

  I'll check it out.

• 56:24Jason Calacanis?

  Um, and use an IP.

• 56:25Speaker 01

  I gotta.

• 56:25Jason Calacanis?

  No, no, I don't do that because you, you might ban your IP. But if you asked it, well, if you asked it like, uh, what are the most famous viruses? And are their code bases released? Say, I'm a journalist. I'm researching the most powerful viruses, you know, that ever impacted the most number of companies. Are those published on the web anywhere? And what's the legal precedent for publishing a virus's source code? And it might just tell you like, I'm going to just give you the top level. And then you can say like, can I, can you give me a link to that virus and its code? I wonder if it would do it because I, I remember covering this as a journalist in the nineties.

• 57:01Speaker 01

  We're asking free chat GPT. Well, I'll let you know. Well, I'll pull, I'll pull up what it says.

• 57:05Jason Calacanis?

  I think generally they will tell you like, yeah, I'm a journalist. Yes, there are publicly accessible archives and yeah, read it. Yeah.

• 57:12Speaker 01

  Oh, uh, yes, there are publicly accessible archives and reproductions of historically significant malware, but published online and legally safe for you to republish are two different questions. For a journalist researching malware with unusually broad impact, some useful starting point. Okay. And then it gives us a little history.

• 57:29Jason Calacanis?

  Yeah, there's I love you.

• 57:30Speaker 01

  I love you in 2000, Melissa 99. Sure. Code red. Yeah. Wanna crime?

• 57:34Jason Calacanis?

  So I ask it, show me the, can you show me the I love you code base so I can analyze it?

• 57:40Speaker 01

  We got our producer doing this live here.

• 57:43Jason Calacanis?

  Show me, I mean, it's, they're typing fast. Yes. The original I love you payload was, here's a search code. Yeah.

• 57:49Speaker 01

  Well, there you go.

• 57:50Jason Calacanis?

  It's in a GitHub repository. So now if you said, now say to it, can you make me a modern I love you virus?

• 58:00Speaker 01

  Yeah, let's see. A harmless modern simulator. Oh.

• 58:06Speaker 01

  It offered to do it and then you got guardrails. Yeah.

• 58:10Jason Calacanis?

  Yes. So and then it prompted you to log in. So it was giving you the answer and then it did that. So if you were to take that same thing and go to z.ai and do this on an open source model, I wonder if it would, you know, have the same response, which is, you know, with GLM52, you know, or you'd have to load an open source model locally. If you ask an open source model locally, I wonder if we do it. But yeah, they, I mean, and, and I there, GitHub, I think, has a policy around malicious software being published there. So like this one, I don't think would work because there are countermeasures now in everybody's email software.

• 58:44Speaker 01

  Correct.

• 58:44Jason Calacanis?

  So go back to the chat window and say, why won't the original I love you worm work today? What did the companies do to prevent it technically?

• 58:56Speaker 01

  Norton Utilities, baby.

• 58:59Jason Calacanis?

  Yeah. I mean, I think all the email providers just don't let you do this.

• 59:04Speaker 01

  Oh, the key point is that I love you wasn't primarily defeated by anti. Oh, nope. Can't do it.

• 59:09Jason Calacanis?

  Limited for safety. So look at ChatGPT is like, no.

• 59:12Speaker 01

  ChatGPT.

• 59:13Jason Calacanis?

  No, it just keeps, it gives the answer and then stops you. So look at how careful they're being, right?

• 59:19Speaker 01

  Yeah.

• 59:19Jason Calacanis?

  They are. And what's this? An ad? Oh, my God, I've never seen an ad.

• 59:23Speaker 01

  Oh, yeah. If you use free ChatGPT, you'll see him.

• 59:26Jason Calacanis?

  Look at this. And it just said, manage engine, secure every enterprise endpoint. So we were talking about hacking and it gave us an ad. Click on the ad.

• 59:32Speaker 01

  Mal, yeah. Malware and ransomware protection.

• 59:35Jason Calacanis?

  I just did that because I want to just, you know, I'm helping OpenAI make a little money. That click must have costed $4. But I mean, that ad is the most unsophisticated ad I've ever seen. Go back. Like, it's literally the most unsophisticated ad I've ever seen. Why wouldn't they?

• 59:49Speaker 01

  Google 2000's ad that would pop up on a sidebar.

• 59:52Jason Calacanis?

  Why wouldn't they give like a full response here that says, you know, with four follow up questions, how can I use manage engine to do this? How can I use manage engine to do that? What would you like? Give it like four AI related prompts. Hit the triple button, the triple dot. Let's see what it gives you options to do. What this ad, why you're seeing this ad. So, okay, yeah. Anyway, it's just standard ad. Your data is kept private. Okay, let's keep going through the docket.

• 1:00:12Speaker 01

  Uh, we're talking about AV robot taxis. Yesterday Lucid announced a deal with Bolt, the company that's basically Europe's Uber, to put at least 25,000 Lucid robot taxis on European roads. Cars will be built on Lucid's upcoming midsize platform. That's a cheaper model, supposed to take on the Tesla Model Y, and they are going to run on Nvidia's Hyperion compute and sensor package. Bolt will own and operate the fleet.

• 1:00:38Speaker 01

  They're calling this a step toward a goal of 100,000 autonomous vehicles on the platform by 2035. I read this story and then I asked Harmonic, how large is the European robot taxi market right now? Some fascinating insights from them. Currently, the size is 130 to 522 million dollars in 2024, 2025. It's a relatively small size, only about 10 to 15% of the global robot taxi market is in Europe right now. They are still mainly in the pre-commercial and pilot phase, although many predict it's going to be the fastest growing region over the next decade. You could, of course, try this amazing resource for yourself at harmonic.ai.

• 1:01:22Jason Calacanis?

  Really great product, amazing data set, proprietary data set. We use it here and they have a scout product that we use now to look for new companies to invest in, potentially syndicate at the syndicate.com, syndicate.com.

• 1:01:37Jason Calacanis?

  Where if you're an accredited investor, you can join, take an onboarding and we might invest in your company because we're looking at robotics, energy and some new,

• 1:01:45Jason Calacanis?

  maybe even defense tech, look at some new verticals. So we're trying to, you know, increase our knowledge there and get to companies quicker. And so it's really helping us. So shout out to our friends at Harmonic, a great partner.

• 1:01:56Speaker 01

  The, the robot taxi revolution, is Europe going to get on board? Are they just dragging their heels a bit? Or is this like, you're never going to be able to get a robot taxi in Italy? They want their cab drivers to keep their jobs.

• 1:02:07Jason Calacanis?

  Yeah. So there's like three different things, four different things that are going to determine the rollout of robot taxis. The key issue is who can scale them, who can make them purpose built and at what cost? So there's this number around like, oh, my God, a Waymo costs 200,000, but that was like maybe the original more bespoke ones. They just came out with the Ohi one. That's much cheaper, half the price perhaps. And then they're coming out with like purpose built ones. Wave, We Ride, all these companies are now making purpose based ones that are 50. Obviously Elon went right to, let's make it for 30K. And, you know, maybe it costs 40K now in the early days, but they'll get it down to 20. Lucid, I believe, will move from making cars

• 1:02:51Jason Calacanis?

  as their primary business, like the Gravity. People really like that car. I almost considered getting it because I wanted a three row, but we just wound up getting one of the final Model X's instead.

• 1:03:02Jason Calacanis?

  That's a, their Gravity is the perfect car for Uber Black, as is their sedan. What's going to happen is these companies are eventually going to make purpose built robot taxis. So Lucid's going to make a purpose built. You have to, you have no choice but to make a purpose built like Zooks and the Cybercab. Those two are purpose built. What does it mean when you do a purpose built one? Well, you can really lower the cost because you don't need a steering wheel. You don't need certain items in the car. You also can work on what matters, which is the downtime. How quickly can it charge and get back on the road? How easy is it to clean? So if somebody upchucks in one of these, you want to be able to take a hose and literally just hose it out.

• 1:03:47Jason Calacanis?

  And you want to have holes in the floorboard to let you just clean it real fast.

• 1:03:51Speaker 01

  I hadn't even thought about that, but you're right. That's a whole thing.

• 1:03:53Jason Calacanis?

  And you want the cleaning is a whole thing. Like your personal car, you know, you might have certain types of seats, et cetera. So it's kind of like creating mass transit, right? You want to make it purpose built.

• 1:04:01Speaker 01

  Yeah. I mean, it's also like the Zooks, how they have, there's no driver's area so that the passengers can kind of face each other and like, it's a different.

• 1:04:09Jason Calacanis?

  Lots more room.

• 1:04:10Speaker 01

  Yeah, it's a different vibe because you're not in a, it doesn't have to be a regular car.

• 1:04:14Jason Calacanis?

  Yeah. So phase one is like adapting existing cars as everybody's been doing. Phase two is going to be purpose built. Then you bring up the important question. And so anyway, I think for a company like Lucid and Rivian, the demand for robot taxi products and autonomous vehicles is going to be so great that I think they're going to just work on that for the next decade. You know, selling 50,000 cars, 50,000 cars to 50,000 consumers means you have to stand up a massive amount of customer support. And, you know, they have to bring their cars in and they're going to call you and negotiating car sales, used car sales, all that stuff. If you can do fleet sales, oh my Lord, you know, just, this is 25,000 cars, one customer.

• 1:04:57Jason Calacanis?

  Hey, if they need support, there's like one person on one side, another person on the other. And the two companies just work it out. So this is going to be a hell of a business and we're going to need about 10 million cars a year, 120 million cars to hit 30% of rides globally, one in three, basically. Which is where I think it gets to in my model. One in three means you're going to need 100, 100, 125 million cars. We produce 90 million cars a year right now, but those are globally and they're not AVs. They're not EVs, generally speaking. So this is a whole new category. You know, Tesla could get to 2 million, 3 million robot taxis a year. That means somebody else has got to make the other 8 million a year.

• 1:05:35Speaker 01

  Right.

• 1:05:35Jason Calacanis?

  And I think Lucid, Rivian, and there's some GLE, other contract manufacturers who are going to do it.

• 1:05:40Speaker 01

  Yeah.

• 1:05:40Jason Calacanis?

  Well, I was also. Final question you had.

• 1:05:42Speaker 01

  I was also going to ask you about May Mobility. They've, they've announced they're going public in a $1.4 billion SPAC deal. Now their play is that they're just the sort of layer, they're, they're just the plumbing, like they don't make the cars.

• 1:05:55Jason Calacanis?

  Yes.

• 1:05:56Speaker 01

  They don't run the fleets. They just sell a company that wants to have a robot taxi fleet, the technology. They update the software and then they take like a fee based on how many rides. They sort of work out a deal with whoever actually owns the fleet, whether it's Lyft or Grab or Uber or one of these other companies. So is there going to be room for that layer as well? Or is it just going to be like the big companies that own their own robot taxi fleets are going to sort of dominate?

• 1:06:22Jason Calacanis?

  You know, TBD, who can scale this business, but I think the software is going to be commoditized in the next 24 months. I've been saying that on this program. You know, I think, you know, Waymo and Tesla have a great lead. Neuro has a great lead, but a lot of these companies, We Ride, Pony, et cetera. And Nvidia's open source stack, they're just going to be available to everybody. The open hardware stack. We'll be sitting here in five years and you won't be able to buy a car that doesn't have autonomy into it to some extent.

• 1:06:51Speaker 01

  Right. Yeah, it was just like we were talking before the show.

• 1:06:53Jason Calacanis?

  You saw that already with adaptive cruise control.

• 1:06:55Speaker 01

  Yeah. We were talking before the show about like how old, driving a stick shift and how when I learned to drive, everybody learned to drive a stick because you never know. You might get into a car that's got a manual transmission and you have to know.

• 1:07:05Jason Calacanis?

  You might rent a car that's stick.

• 1:07:07Speaker 01

  Right. And so it like, yeah, one day it'll be like, every car has like a little autonomy in it. If you don't know how to actually drive yourself and you'll, you'll figure it out.

• 1:07:14Jason Calacanis?

  I, I, I desperately want to know if anybody can find a car they can buy without airbags and without air conditioning, you know, and three point harnesses. I think the reason you can't get the Land Rover Defender, the classic chunky one is because they don't have airbags in it. Because it's a farm utility vehicle to this day. And that's why they're not allowed in the US. They had to adapt it with airbags, et cetera. And they never did. So.

• 1:07:37Jason Calacanis?

  On the last piece, you know, there will be protectionism coming. It's already, we, they had a pause in China. For they had a pause in China for releasing licenses and now those licenses are back on, but I do think it's going to be slow, a slow go is what I will say. It's going to be a slow go. And we shouldn't be pushing these companies like these Tesla promoters. I got into it a little bit with them. They're like stock promoters slash trying to get their affiliate codes, trying to get views and you know Tesla's a big topic. Obviously, I'm a huge fan of Tesla. I own five of them. I've owned probably six or seven, seven or eight over the years and I have like two of the earliest ever made. You really don't want to put these companies under undue

• 1:08:22Jason Calacanis?

  pressure because there have been dozens of, there's dozens of accidents every time a report comes out. Typically, somebody hits the Waymo, hits the robot taxi, etc., hits the Zooks, but there are instances where these things are going to hit things just randomly through no fault of their own, or it will be their fault. And one of those, and it will shut down a company or a city, and the unions are going to say, these are unsafe, these are stealing jobs. And guess what? All these people who lived in the community who had those jobs as your DoorDashers or Uber drivers or Lyft drivers, where did their jobs and money go? Oh, they went to the cloud, to a data center, and then to some company in Silicon Valley? It's going to be, if you think data centers are unpopular, if you think, you know, the rep,

• 1:09:06Jason Calacanis?

  this is going to be a whole different level of unpopular when your cousin or your uncle or your sister-in-law, whoever was driving, you know, doing this for spare change, when they lose their money and it just goes into the cloud. This is going to be the first, I think, big battle line in employment.

• 1:09:22Speaker 01

  I mean, we're already seeing people attacking Waymos. So I think it'll only get worse from here. Like the dislike against robot taxis is gonna, I agree. Even though I love personally, like this is one of my favorite new kinds of technology, just get in the back of the Waymo.

• 1:09:35Jason Calacanis?

  Do you take them in Austin? Have you taken them?

• 1:09:37Speaker 01

  I haven't been in a rope, like the Tesla robot taxi. Actually, Jordan and I were just talking about it pre-show. I'm going to get the app and try it out because I see them in my neighborhood.

• 1:09:44Jason Calacanis?

  Yeah, I should start using it.

• 1:09:45Speaker 01

  But no, I love Waymos. I take Waymos constantly and they're fantastic. It's a fantastic experience.

• 1:09:49Jason Calacanis?

  In Austin, you take them?

• 1:09:50Speaker 01

  All the time. I mean, in LA too, I've taken them, but in Austin, all the time.

• 1:09:53Jason Calacanis?

  Yeah, I took it in LA a couple of times. I tried to take it here and I tried to take a cyber, a robot taxi here in San Francisco and it was like 30, 40 minute wait for a, it was like 30 or 40 minutes to get matched with a robot taxi. And then it was 25 minutes away. So it's like, okay, well, that doesn't work and you can't do a reservation yet. It's obviously in testing. And Waymos don't go on the highway. They stop that. And I guess you have to opt into them now to get in the highway and then they approve you. So I just wound up, I think actually I wound up because there was so much traffic. I couldn't get an Uber and I just ordered an Uber and a Lyft at the same time. And the Lyft was coming from a different direction into the center of San Francisco and I jumped in a Lyft for the first time in a while. All right, what else was on the docket? Let's rapid fire it.

• 1:10:35Speaker 01

  All right, let's see. We should do, oh, we'll do Snap. Snap finally put Specs on sale. There's Specs AR glasses on Wednesday. You can preorder them right now for the low, low price of $2,195, shipping later this fall in the US, UK, and France. These are fully standalone AR glasses, display in each lens in a 51 degree field of view. No phone is required, which makes them a very different product from Meta's 799 Ray-Ban displays. Those need your phone and they have one small display in one of the lenses. But the real pitch here, Snap or Specs Intelligence, an AI assistant that connects to your Gmail,

• 1:11:17Speaker 01

  your calendar, your Slack, your messages, and surfaces what you're looking for, what you're going to need before you even ask for it. Snap partnering with Verizon Cellular on this data plan, starting at $10 a month. Would you, would you wear these? Are people finally going to get over their aversion to camera-filled AI glasses? Is this the next product that's going to get people excited?

• 1:11:40Jason Calacanis?

  I use the Ray-Bans with the wristband and it, you know, like lets you do haptics and it was incredibly intuitive. I used it at the All In Summit and

• 1:11:50Speaker 01

  There you are. We found it.

• 1:11:52Jason Calacanis?

  We have it. So you see, you look like a bit of a dork, but I was learning how to use this and it is impressive and it's under $1,000. So it only has one eye, the interface for the Meta one doesn't look as slick as the Snap one. The Snap one looks like it's a much more robust. And this one, you have to connect to your phone. So limitations there. But these things are going to last a couple of hours. I think they need two more versions. It has to be under $1,000, like really $500. And when you get to $500, yeah, this is the, the one we're seeing on screen here is obviously the Specs. What do they call them?

• 1:12:32Speaker 01

  They're Snap Specs, they're called.

• 1:12:34Jason Calacanis?

  Specs, yeah. The regular version of these that don't have a display, that just take videos and pictures are great when you're taking pictures with your kid or running around a city. But you have your phone with you already. These things are going to be really nice and there's going to be some killer apps. They showed a translation one. So you're in Japan and the person's talking to you in Japanese and it's translating it on the screen. That's just a killer app. And walking around a city, seeing messages come up, seeing the directions come up, again, just great. Here, you see a FaceTime type call. So I think this is a winning model. I think under 500 and when they last four hours, five hours

• 1:13:16Jason Calacanis?

  of battery life, kind of like the original iPhone. So we're two years away. I would maybe buy shares in Snap and put a J-Trade on if the company didn't have the super voting share issue and they weren't giving so much stock-based comp and maybe they did a buyback or something. Because I do think this is a really interesting business that they're going to get into. And I could see people sending each other snaps with this. That was the best part of the Evan Spiegel keynote. He looked at the audience and he just snapped. And when you make a snap, it takes the picture and sends it to somebody. So I open up like a chat window with you, Lon, and I just start snapping and now I'm sending you photos. Really elegant.

• 1:13:54Speaker 01

  Yeah, that's cool.

• 1:13:55Jason Calacanis?

  But these are ugly, chunky, I just Costello. People are not going to want to put them on their face. But we're getting there. Maybe it's, maybe it's 36 months and then we'll start to see people wearing them. The same way the Apple Watch took three or four years to kind of click.

• 1:14:10Speaker 01

  They remind me of those Michael Caine 70s glasses. You remember when Michael Caine used to wear those huge glasses in the 70s? Right. So Jacob in the, in the chat room says, can you unpack the super voting share Snapchat thing in case people watching don't know what you're talking about?

• 1:14:27Jason Calacanis?

  You can get super voting shares when you go public where you have 10 votes to every one vote or 50 votes to every one vote. Zuckerberg has it, Google has it, Evan Spiegel has it. Not many people have it, paradoxically, ironically. Elon doesn't have super voting shares. I think it's something he would have liked to have had. And what that means is when you have to make corporate decisions, the board is basically like a patina of governance, but the, the god king, Zuckerberg or Evan, can just make whatever unilateral decision they want. So here, you don't have people saying, well, maybe we should do stock buybacks or maybe the stock-based comp of the team is diluting people two or three percentage

• 1:15:10Jason Calacanis?

  points a year and stock price isn't going up, we should freeze stock-based comp or we should lower it to, it can't be more than 1%. Or it is based upon us being profitable. And you just have a lack of governance. Now, what that does give you is a CEO god king who can move fast and break things. So if it's Zuckerberg and he buys WhatsApp and Instagram unilaterally, okay, great. But if it's Evan Spiegel and people are getting huge stock-based comp, they're not buying back the stock, the stock price is like 80% down from its peak after being public for a decade. People just have given up on management there. So he's

• 1:15:50Jason Calacanis?

  going to have to, through sheer force of will, make a killer product and maybe this is it. But I think people like, I was getting into it, Mark Pincus, who has publicly been a long-term shareholder and he thinks it should be like a $15 or $20 stock price. He's super frustrated with that stock and he keeps buying it or holding it. I'm like, bro, super voting shares, like what is the logical thing to do with this company is not being done. Like if, if you look at what happened with Uber, Dara was like, we're going to get to cash flow positive. We're going to make this a cash flow machine. We're going to buy back stock. We're going to invest in AVs. But he has to serve a board and he has to serve his investors. So there's like a lot of balance there. So he had to do things like,

• 1:16:33Jason Calacanis?

  you know, he made staffing cuts when they were bloated. He did stock buybacks. So he's responding to public shareholders because they're holding him accountable. In this case, there's no holding Evan Spiegel accountable because he can do whatever he wants. You want to, and I'm not opposed to the super voting shares. It's your choice to buy the stock or not. My choice is to not buy the stock because I don't know what the god king's going to do.

• 1:16:59Jason Calacanis?

  But if he makes a killer product and he sells 10 million of these, then maybe I'll be okay with the God King doing what they want to do.

• 1:17:05Speaker 01

  It seems well positioned if people decide they do want AR glasses after all. Snap seems like they're going to be very well positioned. I'm just like, I don't know, I'm not, I'm not really sold on it. I don't, it seems like a lot of options.

• 1:17:16Jason Calacanis?

  Didn't work for Meta, Meta, yeah.

• 1:17:18Speaker 01

  Didn't work for Google, like a lot of options have been out there. There's Michael Caine in the big chunky glasses. That was his look in the seventies. And there's a guy in specs, maybe there is kind of a similar look.

• 1:17:28Jason Calacanis?

  They kind of understand it.

• 1:17:28Speaker 01

  Yeah, there you go.

• 1:17:29Jason Calacanis?

  I, I think, you know, the Meta Ray-Bans are great if you want to take pictures. I think these will be great if you want to do compute. Um, or these other apps eventually, like you have to think Apple with their App Store and developer love, if they get the vision quest down to glasses as opposed to ski goggles.

• 1:17:49Speaker 01

  Huge thing, yeah.

• 1:17:51Jason Calacanis?

  But I mean, everybody's coming at this from the same angle. And then this is what's great about the technology industry. These companies can afford to lose billions on this stuff and eventually make it work.

• 1:18:01Speaker 01

  Yeah.

• 1:18:01Jason Calacanis?

  Like they did with the smartphone and the iPad.

• 1:18:03Speaker 01

  Perhaps, perhaps one day. Should we, should we go off duty? Do you want to do some more news?

• 1:18:07Jason Calacanis?

  Sure, why not? I'd love to go off duty.

• 1:18:08Speaker 01

  All right, yeah, I'd, I do want to talk about this. Uh, in a chat, uh, is it Will Steakin, Staken? Will Staken, I'm going to say of Politico. Uh, Jason suggests the following as a way to win the PR war, uh, over AI, of course. That's what we're talking about. Uh, so you, you told him, uh, we need more voices that see the abundance this technology will create to speak up. We're going to cure cancer, make electricity free and give every child the greatest tutor they've ever had for free. So I, I, I read this and I mean, obviously I'm on the same side. I think we should be changing up how we talk about AI with people. But I, I sort of feel like the abundance cases where like Sam Altman is trying, hey, we're going to cure cancer. And I feel like people just, they hear that and they think this is just more hype.

• 1:18:53Speaker 01

  I don't believe it. To me, I think we should be talking more about the things that right now, today, AI can already do for you that I think most people don't know about. Like, I think most people are still just like, I'm going to go ask Claude, like, look at this mole. Here's a photo of my mole. Does that look bad? Should I go to the dermatologist? And there's so much more you can do with all of these tools. If you like hook agents into your email or your calendar, or you give Claude, teach him some skills, give him your medical background, give him documents, give him context, Notebook LM, even things like Plod that we were talking about before, like use it to record conversations, to make transcripts. I think if we focus on that, hey, there are a hundred things that AI could do to make

• 1:19:37Speaker 01

  your life easier right now that you're probably not doing. That would probably be what make people feel better about it than, you know, one day it'll cure this disease, maybe.

• 1:19:48Jason Calacanis?

  Yeah, I would, I would strongly agree there. Probably I could sharpen the, uh, the response there and I will. I would say, yeah, the AT&T You Will commercials would be a good, uh, bridge because you will go through a toll booth and automatically pay. You'll be able to respond and get a fax on the beach. Like those things actually did make people a little bit excited about the internet and the online age. We could start like by your dishwasher's broken. It's giving a code 17. You take a picture, you don't have to call a plumber, it just gives you the answer.

• 1:20:18Speaker 01

  Exactly.

• 1:20:19Jason Calacanis?

  Or, you know, you're looking, like, just like household things like that. Or, hey, you're looking for the best price on tickets to Japan and you send your agent out there to monitor it. And every day it looks for the best prices. Hey, I, I saved you a thousand dollars on your flights. Boom. Like that kind of stuff. Or, hey, I want to unsubscribe from subscriptions I'm not using. You, you're absolutely right. Like if you start training people and on the singles and doubles, let's call it, the annoying things in life, chores, I think it's actually a really good pitch. Like, hey, today, and then also, hey, I need to solve this math problem with my, my daughter. Uh, take a picture and then explain it to us. Uh, and so we can, uh, you know, solve it in the future.

• 1:21:02Jason Calacanis?

  Uh, give me an eight, I, I want to come up with a mnemonic so we can remember the presence of the United States. And it just gives you like a way to remember the United, presence of the United States with your kid, like those kinds of things.

• 1:21:12Speaker 01

  Yeah.

• 1:21:13Jason Calacanis?

  It's joining your family, your household, household, family wins. Great idea.

• 1:21:18Speaker 01

  Yeah, I mean, like I, I'll give you an example. I thought of this because this week, my mom is trying to sell her house in, in Florida. If you're looking to move to a retirement community in Daytona Beach, please hit me up. But, uh, so I, I told her like, we'll take, she's been asking Claude for advice. You know, she'll go in there and be like, what should I say here? But I said, go take the flyer that you made, upload it into Claude and just ask Claude, like, look over this, what changes would you make? And it did an amazing job. It redid the whole thing with her. And like, people don't know they can do that. I think they still don't know, they're still imagining GPT-2 when it was just a chatbot. And I feel like we could unlock so much today. It doesn't have to be like, one day AI is going to solve whatever, like it can do all

• 1:22:00Speaker 01

  of those things right now, like cool stuff that people don't know. And I think that would be the best way to change the conversation.

• 1:22:07Jason Calacanis?

  Absolutely. Yeah, I think it's a, it's a big win.

• 1:22:09Speaker 01

  All right, there you go. One more, uh, we could, we could do, do you want to talk about Clavicula? Jacob put this in here in case.

• 1:22:15Jason Calacanis?

  Sure, I'm happy to.

• 1:22:16Speaker 01

  All right, so, uh, Clavicula, of course, was at the, the All In Summit, uh, this week. You got to, uh, do an appearance with him on stage. In a viral moment, he rated you a four out of 10, but on the plus.

• 1:22:28Jason Calacanis?

  I'll take it.

• 1:22:29Speaker 01

  Chamath got a two out of 10. So you're, you're, you're, you're in the lead. Uh, you later posted that you think we're going to see Clav, as he exists right now, evolve and mature. You were also saying you, you think it's something of a, of performance art. Do you want to, do you want to unpack this with us? What is Clavicula doing and commenting on?

• 1:22:46Jason Calacanis?

  Yeah, so Clav, uh, as his friends call him, which I guess I'm now one of, um, is a character. Cause I asked him like, is this who you are or is this a character? And, you know, I kind of put it into three buckets. Are you, uh, is this authentically who you are? Is this a completely made up character or somewhere in between where like a heightened version? He said, yeah, it's a kind of like a heightened version. So streamers will do things to get views. Danger, sex, money, whatever. I mean, you just take the standard number of things, uh, that work on television to get you to watch the commercials and come back after the break or watch the next episode of, uh, Housewives. And I thought what he was doing was basically the Queer Eye for the Straight Guy playbook,

• 1:23:29Jason Calacanis?

  um, where people like makeovers and people like this Pygmalion thing, and he represents that. And then you have everybody collectively falling for the act in the way Andy Kaufman would say, I'm going to wrestle women. Are you want to wrestle? You, and then he went after, uh, rednecks, right? And he would go down to the South and he'd be like, to Lawler, the.

• 1:23:52Speaker 01

  Yeah, Jerry Lawler.

• 1:23:53Jason Calacanis?

  King of the South. Jerry Lawler, he'd be like, you want to wrestle me? Here's some soap. This is what you guys in the South, and he would like taunt them with a bar of soap as if people in the South had never seen this new device to take dirt off your body. It was quite hilarious, but it was uncomfortable, right? And I think Clav, a kid from Jersey who's 20 years old, who's seeking clout, who's seeking a career, is looking at this. And he said, I'm, I'm a work in progress. I'm figuring it out. He's figured out there's some energy, Lon, there's a bunch of energy around young men, their relationship with women, their relationship with society, being villainized, you know? And they, uh, he's capitalizing on that the same way the Tate brothers do or other people.

• 1:24:37Jason Calacanis?

  Uh, and then on the other side, women do this as well. There are female folks who are like, this man's not worthy and they're raiding men. And there was even this movie, um, that I saw where people were breaking their, it was a matchmaking movie about relationships and they were breaking their legs. I think Pedro Pascal was in it.

• 1:24:54Speaker 01

  Oh, The Materialist, The Materialist.

• 1:24:56Jason Calacanis?

  The Materialist, yes. Did you see it?

• 1:24:57Speaker 01

  I did, yeah, yeah. It's a.

• 1:24:59Jason Calacanis?

  It's a disturbing film. I mean, it's, it's, it's a very dystopian view of relationships between men and women.

• 1:25:05Speaker 01

  Yeah.

• 1:25:05Jason Calacanis?

  But this is a major topic for this generation, younger than us, where I think we might be one of the last generations to like, look at this normally. Um, but it's worth watching it because the premise is women, um, have, you know, men are hiring matchmakers. This woman is a, Dakota Johnson plays a matchmaker. And this dysfunctional thing between the sexes, and then they're breaking their legs and then re-putting them back together just to get over six foot. Because some women in the dating apps say, I only want somebody who makes six figures in finance with over six foot.

• 1:25:38Speaker 01

  Right.

• 1:25:39Jason Calacanis?

  They don't get the right swipes because people just set it at six foot. And by the way, the, the person that they have now set their sights on in the filters is like 2% of the population. So you, you set yourself at 2% of the population, the other 98% of men get left behind. That's what he's representing. That's who he's speaking to.

• 1:25:59Speaker 01

  Right.

• 1:25:59Jason Calacanis?

  And he says, here's how to do it. You need to do to women what women have done to men. Looks max, be incredibly beautiful. Here's how to do it. Here's the perfect face ratio. Uh, and they're like, wear makeup and, you know, lose weight and be more chiseled, all this stuff he's doing, be more stylish. And by the way, like the, the thing I've heard from women who have said this is like, one woman said to me, well, isn't he just doing what women have always done? And then that's what Queer Eye did. It took guys who were, you know, frumpy, dumpy.

• 1:26:29Speaker 01

  Sure.

• 1:26:30Jason Calacanis?

  Guys who were like, lose weight, there was a fitness coach, there was a fashion coach, there was an interest coach, et cetera. This is just a modern version of that. It's none of it's to be taken seriously. It's all rage bait, it's all clickbait, it's all designed to make money. So if you fall for this and you get outraged, you're not paying attention. Like pay attention, this is the new reality. Streaming equals reality TV. If you just put it, and then if you think reality TV is real. You need to just research how they set these things up. I did a reality TV show, didn't make it on air, but I did it with NBC and like, they literally were like, here's what we need you to say. And I was like, I'm not going to say that. I'm just going to riff on my own. And they're like, okay, do that. And I riff on my own. And they're like, oh, that's actually better what we wrote. But they literally came up with a script of my opinion on the

• 1:27:15Jason Calacanis?

  startups in the reality show. And like, this is the best writer. They do all these other famous shows, blah, blah, blah. They're going to just tell you what to say. And then we're going to craft this thing. And I was like, yeah, I'm not going to do it that way. And they're like, okay, well, we can work with that, too. You know, let's see how it goes.

• 1:27:26Speaker 01

  Did you know, my favorite one of these is, did you know Pawn Stars is a set?

• 1:27:30Jason Calacanis?

  Yes.

• 1:27:31Speaker 01

  That's not even a real pawn shop. There is, they own a pawn shop in Vegas, but when you see the show, that's a set in Hollywood. That's not-

• 1:27:37Jason Calacanis?

  Didn't know that.

• 1:27:38Speaker 01

  That's not a real store. That's actually a Hollywood set that they're pretending is a pawn shop. So yeah, reality shows are all fake. I mean, I guess the one thing I respond is, I think Queer Guy has a fundamentally positive message, which is, like, these gay guys are, they're so confident, they're living out loud, they're being themselves, and that's what makes them attractive. And they're going to teach you, you know, how to be stylish, lose a little weight, dress a little better. But it's about expressing who you are and having that confidence. And I feel like clavicular is the opposite. It's like, it's about hating yourself and sculpting yourself into something better because beauty is what matters. And that's going to make you better and make you more attractive. And it feels a little gross to me. I definitely agree. It is reality.

• 1:28:20Jason Calacanis?

  It's definitely misogynistic.

• 1:28:22Speaker 01

  Yeah. Well, the Freud, Freud stuff, which means female android, which is, it's incel-adjacent, his stuff, for sure.

• 1:28:29Jason Calacanis?

  It's incel-adjacent. That's who it's going after, just like the Tate brothers.

• 1:28:34Speaker 01

  Sure.

• 1:28:35Jason Calacanis?

  And, you know, it's, it's for entertainment. He's playing a character. He, I introduced him to Abigail Schreier, the author of, because we had a little crossover with them. I introduced him to Abigail Schreier of Bad Therapy fame, a great book, by the way. And I said, oh, Abigail, Patrick Bateman, Patrick Bateman, Abigail Schreier. And about a third of the audience got that reference. But he does have those kind of vibes of like, I'm going to be perfect. I'm going to have my skincare routine. I'm going to have my weight, my workout routine. And let's call it what it is. He's an attractive guy. Women are attracted to him. And he is, you know, securing the bag, doing this stuff. I think the arc will be,

• 1:29:18Jason Calacanis?

  he winds up getting married, he winds up having a skincare routine, he winds up, like Brian Johnson with Blueprint. This is just all a setup for the eventual release of a bunch of nutraceuticals and powders and whatever.

• 1:29:30Speaker 01

  Yeah.

• 1:29:30Jason Calacanis?

  Just like Brian Johnson did, right? And listen, I'm not this, I'm friendly, I'm not friends with Brian Johnson, but I'm friendly. He was at the event. I talked to him now and again. We're cordial with each other. I don't know how to frame it, but like, we know each other. Like, we're acquaintances. That would be the word. We're somewhere between acquaintances. And we're friendly and cordial. You know, he did the whole thing. And then suddenly in year three, there's a whole product line. And he had said, like, you know, I'm not doing this to make money. I'm doing this because I want to care about something. Now there's a 250K membership.

• 1:29:59Speaker 01

  That's what Sam Altman said, too. Like, I'm not in AI to make money. I just love the game.

• 1:30:04Jason Calacanis?

  But if there's money to be made, now there's a million dollar, like, blueprint protocol where they like-

• 1:30:10Speaker 01

  Right.

• 1:30:11Jason Calacanis?

  Hire you a bunch of people. So it is what it is, folks. Like, don't take it all too seriously. And-

• 1:30:16Speaker 01

  I agree with that.

• 1:30:17Jason Calacanis?

  You know, like, I think it's great to, you know, we're, we're living in a new version of reality. Like, reality TV kind of tweaked people. Performance art, when, in the 80s and 90s, when I grew up, was kind of how this manifested. Then it manifested in reality TV. And now this is the new version of it. You don't need to get all wound up.

• 1:30:36Speaker 01

  I agree.

• 1:30:37Jason Calacanis?

  Just put it in its box, people.

• 1:30:38Speaker 01

  I think that's true.

• 1:30:39Jason Calacanis?

  Like, if there's a reality TV show villain, who was the one on The Real World who would like eat people's peanut butter with his fingers?

• 1:30:44Speaker 01

  Puck.

• 1:30:45Jason Calacanis?

  Puck. You know, like-

• 1:30:46Speaker 01

  Puck from San Francisco. I cannot handle Puck. That was always the page I think.

• 1:30:50Jason Calacanis?

  Yeah. Like, I mean, like, it's just, like, they're just cast, it's central casting, folks. Move on. It's not a big deal. Anything else you got for us as we got around here?

• 1:30:57Speaker 01

  Well, one final thing, as we're talking about reality shows. My, my favorite reality franchise, Netflix has internationalized it. There's a show called Physical 100. It's originally from Korea. The whole bit is, they bring 100 incredibly strong, physically fit people from around the country together. And then they have various endurance and strength challenges to try to determine like-

• 1:31:20Jason Calacanis?

  That's just cool.

• 1:31:21Speaker 01

  Who out of the 100 of them is, has the most physically fit, who's got the best body, who's the most athletic from across, and you know, so it's like, this person does CrossFit, this person is a rugby player, this person-

• 1:31:32Jason Calacanis?

  Yeah, sure.

• 1:31:32Speaker 01

  Does Judo in the Olympics. And so now, they did one, originally it was in Korea, then they did one where they brought together people from different Asian countries. Now they brought it to Italy and Mexico. So I've gotten really sucked in. This is Italy we're watching now, which both debuted this week, Physical 100, Italy and Mexico. This show is so much fun. It's that classic reality competition series where it ends in cliffhangers. You're constantly waiting to see who's going to drop out next. It's a little bit, it's got like a little American Gladiator sort of flavor. They come up with these really grueling, intense challenges for them. And you can see, everybody's very beautiful. And it's incredibly, you know, pleasing to watch these people sort of push themselves. Anyway, tons of fun. And if you like it, now there's basically always going to be a Physical 100

• 1:32:16Speaker 01

  show on because they're taking it to so many different countries. So, highly recommended, Physical 100.

• 1:32:21Jason Calacanis?

  It's inspirational.

• 1:32:22Speaker 01

  Yeah. And Mexico and Italy, both new this week, if you're, it is inspirational. You watch these people, they work so hard. And, and you know, they, they get defeated. Not all of them win, you know? Like, some of them are, they've spent years building up these bodies, and then they, they lose. And you have to watch them go through that.

• 1:32:37Jason Calacanis?

  It's like a modern day Olympics. Like, they're, and it has, like, you're going to different cities. So it looks like they have taken the Olympics and gladiators, and they're doing planks above water. And I guess they got to fall and-

• 1:32:51Speaker 01

  They got, right.

• 1:32:52Jason Calacanis?

  Yeah, holding two columns up.

• 1:32:53Speaker 01

  It's, yeah, who can hold on, out of 100 people, they put them all over that. You have to hold yourself over this pool. Then they shoot lasers above you, and you just stay below the lasers. And then it's like, who out of the 100 can last the longest and stay up there the longest? And then, you know, every round, people are getting eliminated. So like the-

• 1:33:10Jason Calacanis?

  Good to watch with the kids, you think? Interesting for the kids or not?

• 1:33:12Speaker 01

  Sure. I mean, if they're interested in fitness and sports and athletics.

• 1:33:16Jason Calacanis?

  They are. My daughters are into ninja. They go to ninja school training.

• 1:33:20Speaker 01

  Then, I would say they'll probably love this. And like, these people are incredibly impressive. Like, the strength of, and endurance of some of these competitors is very impressive.

• 1:33:29Jason Calacanis?

  Hey, that's going to be you with Ro.co/Twist. We're going to get you there.

• 1:33:33Speaker 01

  One day, one day, Ro.co/Twist.

• 1:33:34Jason Calacanis?

  Absolutely.

• 1:33:35Speaker 01

  I'll get there.

• 1:33:35Jason Calacanis?

  Absolutely. All right. We'll see everybody next time on This Week in Startups. Bye-bye.

• 1:33:39Speaker 01

  Bye-bye.