This Week in Startups - AI Demos: AI that interprets human emotion, Hugging Face Assistants, and more! | E1894

Episode Date: February 8, 2024

This Week in Startups is brought to you by…Squarespace. Turn your idea into a new website! Go to Squarespace.com/TWIST for a free trial. When you’re ready to launch, use offer code TWIST to save 1...0% off your first purchase of a website or domain.Gusto is easy online payroll, benefits, and HR built for modern small businesses. Get three months free when you run your first payroll at Gusto.com/twist.*Gusto pricing shown in ad is based on pricing prior to March 2025LinkedIn Jobs. A business is only as strong as its people, and every hire matters. Go to LinkedIn.com/TWIST to post your first job for free. Terms and conditions apply.*Todays show:Sunny Madra joins Jason to demo an AI that analyzes human expression (2:07), a generative gaming platform (23:26), Hugging Faces Assistants (39:12), and much more!*Timestamps:(0:00) Sunny Madra joins Jason(2:07) Hume AI - AI toolkit to measure, understand, and improve how technology affects human emotion(6:53) Squarespace - Use offer code TWIST to save 10% off your first purchase of a website or domain at https://Squarespace.com/twist(7:57) Hume AI demo continued(19:15) Gusto - Get three months free when you run your first payroll at http://gusto.com/twist(20:25) AI Bet Update - Small model deployed on iphone/android(23:26) FRVR AI - Generative AI to streamline game creation(34:16) LinkedIn Jobs - Post your first job for free at https://linkedin.com/twist(35:19) Calling GPTS with @(39:12) Hugging Face Assistants - LLM-powered assistants specialized in the Hugging Face ecosystem(45:05) Eleven Labs ChatGPT*Mentioned on the show:Hume: https://beta.hume.aiFRVR: https://beta.frvr.aiHugging Face Assistants: https://huggingface.co/chat/assistantsEleven Labs ChatGPT: https://chat.openai.com/g/g-h0lbLuFF1-elevenlabs-text-to-speechhttps://twitter.com/dnystedt/status/1754323506056266149https://www.youtube.com/watch?v=th3vzKTE0O8*Follow Sunny:X: https://twitter.com/sundeepLinkedIn: https://www.linkedin.com/in/sundeepmCheck out: https://www.definitive.io Follow Jason:X: ⁠https://twitter.com/jason⁠Instagram: ⁠https://www.instagram.com/jason⁠LinkedIn: ⁠https://www.linkedin.com/in/jasoncalacanis*Thank you to our partners:(6:53) Squarespace - Use offer code TWIST to save 10% off your first purchase of a website or domain at https://Squarespace.com/twist(19:15) Gusto - Get three months free when you run your first payroll at http://gusto.com/twist*Gusto pricing shown in ad is based on pricing prior to March 2025(34:16) LinkedIn Jobs - Post your first job for free at https://linkedin.com/twist*Check out the Launch Accelerator: https://launchaccelerator.co*Check out Founder University: https://www.founder.university*Subscribe to This Week in Startups on Apple: https://rb.gy/v19fcp*Great 2023 interviews: Steve Huffman, Brian Chesky, Aaron Levie, Sophia Amoruso, Reid Hoffman, Frank Slootman, Billy McFarland*Check out Jason’s suite of newsletters: https://substack.com/@calacanis*Follow TWiST:Substack: https://twistartups.substack.comTwitter: https://twitter.com/TWiStartupsYouTube: https://www.youtube.com/thisweekinInstagram: https://www.instagram.com/thisweekinstartupsTikTok: https://www.tiktok.com/@thisweekinstartups*Subscribe to the Founder University Podcast: https://www.founder.university/podcast

Transcript
Discussion (0)
Starting point is 00:00:00 11 labs, text to speech, GPD. Correct. He said, generate a story from me, then convert it to an audio book, two paragraphs. Yes. Wow. And then see, the first thing it did was it came back to me, so, well, what's the voice that you want to do it in? And I picked Jarvis.
Starting point is 00:00:15 And so then it did the story for me, and then it generated the audio. And if I go over here now, the story that was generated, I can play it. And let's listen. Metropolis beneath the shadow of towering skyscrapers lay a quaint little bookstores. that seemed almost forgotten by time. So that's Jarvis from Ironman. Avengers, Iron Man. Yeah, I'm not sure he made this debut.
Starting point is 00:00:38 Wow, that's wild. This week in startups is brought to you by Squarespace. Turn your idea into a new website. Go to Squarespace.com slash twists for a free trial. When you're ready to launch, use offer code Twist to save 10% off your first purchase of a website or domain. Gusto is easy online payroll. benefits, and HR built for modern small businesses. Get three months free when you run your first payroll at gusto.com slash twist and LinkedIn jobs. A business is only as strong as its people,
Starting point is 00:01:14 and every hire matters. Post your first job for free at LinkedIn.com slash twist. All right, everybody, welcome back to this week in startups. And I've got my good friend, my bestie, Sande Madra here. He goes by Sunny. His real name is Sunny. He's the co-founder of Definitive Intelligence. You can go take a look at their website, Definitive.io. They do all kinds of data analysis of private and public data with their AI-enhanced system, their platform. And we get together here every week, usually on Mondays, but we've both been traveling
Starting point is 00:01:48 to go over the latest updates in AI. The latest updates in AI. And we do that through demos. We've done over 100 demos. You can go see all of those demos at this week in Startups.com. A-I. Oh, that's it. Slash AI.
Starting point is 00:02:03 How are you doing, Sundee? I'm doing really good. I'm excited. You know, there is so much stuff happening, and we should just get right into it. Okay, give me the first one. Okay, the first one. So this one is super cool. It's called Hume, and I think that's how you pronounce it.
Starting point is 00:02:21 I'm going to just call it up here. H-U-M-E, H-M-E, I would think it's Hume. Yeah, yeah, exactly. And it's Hume.A.I. Got it. Hume. dot AI. And what they do is they basically predict call quality, attention, and mood using an
Starting point is 00:02:36 AI model. And so this is important. I think in a lot of customer support use cases or SDR use cases where you're trying to do, or even if you're doing a podcast and you want to basically get an understanding of how you think people are interpreting it. So I took this. It takes a few minutes to run. So I had to kind of do this beforehand.
Starting point is 00:02:55 but I took kind of a short clip of yours. And basically, I took one of their models and I had it classify it. And basically, you know, it kind of did the classification around, you know, whether you're self-confident or self-doubting just by looking at your video. And as you can see here, J-Cal, you will not be surprised. You came a cut 99.39% self-confident. Wow. You know what they say.
Starting point is 00:03:25 fake until you make it. I mean, if you believe it, you believe in yourself. I mean, so J-Cal, this is it, right? Like, we've got to send this in our poker group, but like, this is hilarious. Yeah. So literally, you took a 30 second clip of me from the podcast. Yes. And Hume AI determined I am 99.39% confident and 0.001%. In other words, I'm delusional to a level that this computer system, this AI, has finally unmassed exactly how delusional I am. I have so many thoughts on this, but the AI can't be wrong, can it? That's my first. I think we need a larger sample size.
Starting point is 00:04:12 Yeah. And so basically, it has a bunch of great models. Obviously, I tried this one, self-confident or self-doubting, but they have Parkinson's versus non-Parkinson's for health-related, alert versus drowsy, a tentative versus distracted. So you can imagine for the people that are still working from home. And if you layer this into like a Zoom or maybe even, you know, Vinnie's thing, which they renamed to Rumi now, imagine how powerful this could be. Like, hey, are your people even paying attention when they're on Zoom?
Starting point is 00:04:43 Yep. Yeah. And so there are a series of apps that do this for salespeople. They've been around for a while. What I like about this is you're uploading the video. You're not doing it in a real time, I take it. Yes. You're doing this as coaching for individuals.
Starting point is 00:04:59 So I kind of like this as a, in a way, it's kind of like having a speaking coach. You know, people will hire those to take out the ums and the aaz or, hey, make eye contact when you do this, look directly in your camera. So, you know, one of the things I have a problem with is, especially when I'm moderating, right, which you do. kind of the moderating job here, you're reading. So I wonder if when you're reading, and you're queuing things up, if that is distracted or if that screws with your confidence, right? Whereas when I'm commenting or like, you know, take the Olin podcast, I'm moderating, I have to like take a series of facts or I might be fact checking somebody or pulling up a chart. So I wonder like what impact that has on the sentiment analysis or these emotional analysis,
Starting point is 00:05:43 confident versus not confident, distracted versus non-distracted. But all of that's really good stuff. And yeah, you know, I think this is going to be built into Zoom eventually. So Zoom will probably give you some tips. Hey, you know what? You look a little sleepy. So this could be creepy and dystopian, but here we are. What do you think? Is it accurate? Let me start with that. Do you think this is accurate? Yeah, well, look, I only did like a couple of test things. But, you know, what I liked about it is exactly what you said. It allows you to kind of in an offline capacity go and address that. And I think especially with the growth in the creator economy, with so many more people trying to do, you know, podcasts or make videos, this is a great way to
Starting point is 00:06:28 get an analysis. Like I, you know, I watched this really good video by Colin and Samir. They were explaining like sort of the challenges they have a Twitter video versus the other platforms. And the amount of analysis that they do on a video, right, in terms of just the regular analytics of Where are people falling off and where are they coming in and all that? This is just another layer now because you can add that to your repertoire, right? And so, if your landing page looks terrible, I'm out. I'm going to just bounce. It's 2024.
Starting point is 00:06:58 There are no more excuses for an ugly website. Stop settling for OK and start using Squarespace and be awesome. Squarespace is an out-of-the-box solution that lets you build a beautiful website and engage your audience. And of course, it's sell anything you like. got an amazing drag-and-drop web designer. Gorgeous templates that are always optimized for mobile, where the majority of your users in all likelihood will be visiting your website. And you're going to get all the great advanced analytics like marketing, sales, etc.,
Starting point is 00:07:27 built in to Squarespace. You don't have to go buy third-party tools. You can also create an online store or start your own blog at a click of a button, create a subscription business for members-only content, and so much more. It's the simplest, the most effective, and the best-looking way for you to start a business online, and you can do it all. Squarespace.com slash twist for a free trial. That's Squarespace.com slash twist for a free trial.
Starting point is 00:07:49 And when you're ready to launch, you're going to get 10% off your first purchase of a website or domain at Squarespace.com slash twist. It's interesting how obsessed people get about these. It's why I like podcasting to a certain extent when you're not on video, because you can just focus on the words and just observe that in the conversation. But increasingly, people want to watch videos. And then people are obsessed with clips.
Starting point is 00:08:11 So we do clips of this show. This Weekend Startupes.com. Just double-click. The clips will always beat the actual original program, but the original program, you get people having a deeper understanding. So I kind of call it like drive-by, sampling, hors d'oeuvres versus sitting down
Starting point is 00:08:28 and doing the full tasting menu, as it were, right? So there are different experiences. Sometimes you just want to grab and go. I'm just going to grab a quick hit. I want to see one demo of This Weekend Startup's AI demos on Mondays. Oh, I just, I want to see Freeberg, on science. I want to see Chimoth do markets. I want to see Saks do Ukraine, Ukraine, Ukraine, whatever he's into at this moment in time. And, you know, that might fit, you know,
Starting point is 00:08:52 scratch your itch, so to speak. And so I think increasingly, you know, there are a group of people who only know certain podcasts from the clips. And they don't actually have the wherewithal, the attention, the desire, some combination of those things to consume a whole podcast. I think the people who consume whole podcasts are your really serious audience. and then you have the secondary audience, which doesn't make it invalid, but they just, you know, they're just kind of snackers, you know, they're grazing on the content, they're grazing.
Starting point is 00:09:21 And that's okay too. Sometimes you convert one to the other, right? Sometimes people just listen to the whole pop, but they think the clip is funny, but more than not, I think people go from like, watching a clip on TikTok or increasingly YouTube shorts, and then they find out, oh, there's just a whole podcast, maybe I should subscribe to that, yeah.
Starting point is 00:09:38 Yeah. And that's why I think there's a whole new genre of people who are like, why do a podcast? I'll just do clips all day. I'll just do three or four clips and then I hope that some hit. I find that you don't get the depth of podcasting with that right? And so it's kind of a missed opportunity
Starting point is 00:09:54 but it is valid to get a short people like shorts. Yeah. So, you know, I think the team's done a great job. They made it super easy. It's very usable. They have like a playground like you'd expect and so you know, I think folks should use it. Would you know what language model they're using or what visual model they're using for this. Maybe you tell us a little bit about the technology behind this
Starting point is 00:10:14 and how it works. They haven't shared a lot. So my my understanding by going through the website is that they've taken models and tuned them themselves, right? So they've taken models and given it or maybe even created the models that that wasn't super clear where they've said, here's a bunch of folks that are look to be confident. Here they look to be, you know, not confident. So this is the, this is the, I want to call it follow, but this is the next derivative of all this work that's happening in standard generative AI, people are looking at those techniques and applying it to even create their old models. If you were going to build this from scratch, how would you do it?
Starting point is 00:10:51 It all comes down to the training set, right? So I would use the same training philosophy that's used to build large language models. And I would basically have a bunch of video that is labeled, you know, labeled, you know, confident, non-confident, and then basically give it to the model to look at and learn. Right? That's the beauty of it. And then you'd have to take that original data and say, you know, this person's like an actor. Yeah. Like is the reason I'm coming across a super confident in that clip because I was actually confident
Starting point is 00:11:25 or because I've done 2,000 podcast episodes, I wonder. I wonder if it's a performance thing or it's actual reality. That's why I was wondering about the accuracy of this. So I would need to see more of this. I would need to put into. this, you know, like, let's just say, you know, the top talk, Bill Gurley's talk at all in Summit. You know, he was super confident. He was funny. He was pretty confident in his position because he had experienced it, right? And then, you know, see what it says because he also has a bit of
Starting point is 00:11:57 a southern axis from Texas. He's got a slower pace. He's very humble. So, you know, to set another way, you know, the training data could say maybe he's not confident. this is what I'm wondering, because he's humble, right? Or he's super confident because they're an actor, right? And so, yeah, I really need to get a lot more information on this to see if it's really, really working. What I would like to do here is I would like to run a test. This is how I would train it. I would have you give me a series of questions and answers.
Starting point is 00:12:33 Jason, what's two plus two? I'd say four. You say, well, how confident are that absolutely two plus two is four. And then you'd say, okay, what's the square root of this, you know, 20 digit number? Yeah. And I'd be like, you'd be like, that's not confident, you know, unsure. And then, or I would say, give this as the answer. And you would give me a piece of paper or on my screen, it would say, this is the wrong answer.
Starting point is 00:12:53 I made it up and you say it, but say it confidently. And I'd say it's super confident as a wrong answer. And we try to trick it, right? Yep. And so you could have it lie, right? Hey, what's your name? Can you say, my name's Vinnie? And I say, no, your name is Sandeep.
Starting point is 00:13:07 And you say, no, my name's Vinny. And we go back. back and forth and then we tell the language is lying. I wonder if I could tell lying. I mean, this is what's going to happen. Do you tell this story with Apple goggles? I'm not calling it vision. I'm calling it goggles because they look like ski goggles.
Starting point is 00:13:21 I mean, you put Apple goggles on and now I'm talking to you or I'm out in the real world and we're wearing this at a poker game. Would I actually know you're lying or bluffing? Interesting. Oh, what about that use case? You're wearing the, you don't need the Apple ones. You can just wear the meta ones. that just look like regular sunglasses.
Starting point is 00:13:41 Because all this needs is a video capture. Yeah, and then it just buzzes your ankle two times on your right ankle means not confident, lying. And the left side means confident, you know, probably has it. I mean, this is going to be really, really interesting. And I wonder if people, if you uploaded clips of a player, so like, let's say we play with Jason Kuhn or, you know, Phil Helmuth. And if you uploaded all their clips from when they played TV poker and then said, hey, they're bluffing. here. And hey, they have it here. Oh, that's your training data.
Starting point is 00:14:13 And that's your training data. And then see, then if you could actually take Jason Coon and break down his game in some way, or if he's in fact unreadable. You know, like top-to-lead poker players versus amateurs. It can be really interesting. Yeah. Yeah. Yeah. I figure out what the live reads are. So there are people who think they're good at live
Starting point is 00:14:31 reads, right? So this will be interesting. But this is going to totally change the game with those meta-glasses because those are also, You can buy them in like regular glasses, like not sunglasses mode. So you're going to have to tell people, hey, you can't wear those. I don't know if you saw my tweet the other day than what you think of this. But my position with Google Glass was, and I told this directly to Larry Page, like, you have to when this thing's recording or turned on, have a light. And it should have a light that if it's turned on its green. And if it's recording, it's red.
Starting point is 00:15:00 Audio or video. And, you know, I'm not big on regulation. You know, I'd rather the market dynamics, you know, sort of define this, but I don't know, what do you think of that? Like, if the standard was, you know, it'd have to be green, if on, red, if. Yeah. So the Snapchat glasses, they used to have a, like a kind of a circle, right? Yes.
Starting point is 00:15:21 Doing it, right? Oh, did it have a, did the circle go on when you're recording? Yeah, exactly. Yeah. And then the humane pin, I was just at a dinner with one of the co-founders, and then she had it. And it has the same thing. When it's recording, it's very clear that it's recording and it's active. How do you know it's recording?
Starting point is 00:15:36 Is it red? It has like a light on it. Yeah, like a very distinct. Blinking red light? It's not a blinking red light. This is what I think these guys are cowards. I'm not humane. They're probably cars too.
Starting point is 00:15:50 But I think these folks are cowards in the wearable space to put on a red light because they think, oh, it's going to be scary. I think it's the exact opposite. I don't think they're thinking straight. If these things are going to be adopted in society and people are not going to get punched in the face for covertly recording people who are kicked out of a bar like the glass holes did back in the day, the Google glass holes. Oh yeah.
Starting point is 00:16:12 I remember that. The woman got punched or something. I mean, I'm not for physical violence, but I am a realist. Like, if you came up to me with like, I put a recording camera in my face, now everybody's paparazzi, you see what happens when like people do that to celebrities over and over again. Like people snap. It's not the society we want to live in. Yeah.
Starting point is 00:16:29 So everybody should. Yeah, there's a human, you mean, you may imply it. Yeah. Apparently, it has a trust. When it's recording. Yeah, right, flashed. Trust light. I like the name trust light, but I don't know the color of the trust light.
Starting point is 00:16:41 Like on a video camera back in the day, remember those? Like those. That's what I'm basing it on. When you had a video camera, they put a flashing, they put the red light on it or a flashing red light. I give this one, I don't know how good it is. You know, our potential, I think it's obviously like this is incredible. I'm going to just withhold my judgment and just give it a B minus for now. You know, I can't tell right now from the end.
Starting point is 00:17:05 interface. I don't know how it's trained. So that could be, you know, me just not playing with enough. But I'm going to give it like a B minus right now. I think it's got potential. I want to really understand why when it, that fact that it's a 99.99% confident, that just makes me not believe it. So I would need an explanation of some incredible score like that. Okay.
Starting point is 00:17:26 All right. I give it a B. I think it's like early, you know, I was just pulling it up as we were doing it there. That model was only trained on 80 videos. which one could argue is like maybe not a big enough training set but maybe we can hear back from the team and see what they have to say. Yeah, explain to us what you're at. Yeah.
Starting point is 00:17:45 And then we had a lot of questions. Yeah. We had a lot of questions. Yeah. See right here is it trained on 80 videos. So it's like, so I pulled that one up. So right. You know what?
Starting point is 00:17:53 I'm going to give mine a B because I'm going to go from B minus to a B because they did explain that. And I was the reason I was kind of like a little down on this is because I didn't feel like when I looked at the airing phrase. originally that it gave me enough context to give, you know, like the score of the new. Yeah, because that was in the playground. If you go into the, yeah, I think it's the same thing. And, you know, there's a lot to learn from there.
Starting point is 00:18:14 And you should have them on. I like the idea, though, of also knowing combining this data with health data and knowing if you have early onset Parkinson's or Alzheimer's or you're slurring your words. Like, I think this technology, if somebody was on a Zoom call and was speaking, or if your phone was listening to you persistently, if you start speaking. slurring your words, if it knew, if it just gave you an alert, hey, you're slurring your words, are you drinking alcohol? Or do you think you might be having a heart attack? Or we noticed you didn't get, you only had three hours to sleep last night. You didn't get any REM sleep. Those are the three
Starting point is 00:18:48 reasons you could be slurring your words. And you just dismissed. Like, I get the loud alerts. You ever get the loud alerts when your kids scream or something? Yeah. So my girls like to scream when I, you know, play tag with him, whatever. So I'm constantly getting these alerts. Yeah. Yeah. That the decibels are going deaf from my girls. I'm playing tag. Awesome. But you could do that, right? You could just be like, hey, you might be having a hard attack.
Starting point is 00:19:10 There's a lot of potential there for someone to own this. Yeah, for sure. Great job. Listen, I know myself as a founder. There are things that I love doing. I love building products. I love hiring people. And there are things I hate.
Starting point is 00:19:22 Payroll, HR. So I use Gusto. Gusto is the best. Gusto's payroll and HR services make running a small business much easier because it was specifically designed for you, the small business owner. And payroll is something you did. definitely do not want to mess up. Oh, and I know it. Gusto will automatically calculate your
Starting point is 00:19:39 paychecks, do your payroll taxes, set up open enrollment, and that's not all. Gusto also handles onboarding, health insurance, the 401k, time tracking, commuter benefits, hey, people going back to work, offer letters, you want to get those right, and they even give you access to their HR experts. Gusto will let you focus on the most important stuff in your business, like getting product market fit, or bear hugging your customers and making sure they're happy. It's super easy to set up and get started. And if you're moving from another provider, Gusto, can transfer all your data for you easy, peasy, lemon squeezy folks.
Starting point is 00:20:10 So here's the best part. Because you're a Twist listener, you get three months free. Yes, that's right. Not one. Not two. Free, free months. All you have to do is go to gusto.com slash twist. G-U-S-T-O-com slash twist.
Starting point is 00:20:21 You must go to Gusto. Again, that's gusto. com slash twist. The next we're going to do, just a, it should be a quick short. This is going to be an update on one of our bets, which are also available at this week in start. up.com slash bets.
Starting point is 00:20:34 And just a quick short one, I want to get your reaction. So this one, this is an analyst. I follow like a chip analyst, Dan Nystet on X. The iPhone 16 with General Divi, I expect to launch debut in September with the A18 chip with, you know, the latest. So this is a little tease. And you know, you know how I know this is real? I'm just going to click this article here.
Starting point is 00:20:54 And it's like fully in like, you know, it's like some site from Taiwan, which means it's like, you know, this is a real deal. Like someone knows what's going. on there. So that's kind of a... I mean, that was a big portion of our league set and gadget were either executives who worked at the company or, you know, Shenzhen, Akihabra, Factory Express, where they just love the thing and they would just send us a picture from the line when they were making a star attack or some nonsense. And we would get it from them. This has like the whole, like some, you know, some kind of slide right out of some presentation talking about the different, you know, chips and all that stuff.
Starting point is 00:21:26 So we don't know, but what they're saying is there's going to be the A18 chip, which is the chips. that go in phones as opposed to the M-chips, which is a series of chips in silicon that Apple makes, that goes in laptops. And the A-18 is going to launch with the iPhone 16. With the iPhone 16, which is the next one. Yeah. And September.
Starting point is 00:21:50 Okay. And it will have, generative AI is going to debut, and we have a bet for July 1st. We do. If these models will be built in. into the phones. Correct. And so I'll just pull that up because we have the betts site here.
Starting point is 00:22:08 We now have the bet site, which is this week at startups.com slash bets that we're making. And I think our bet was that it would be announced or released. That's the question. We said deployed locally on phone. Models deployed locally on devices. I win this one. I'll win this one. Yeah.
Starting point is 00:22:26 I win this one. No doubt. Because you know what? The phone's not going to be in your hands until the fall. So I think I win this one. And so if we had made the bet, this is where, if you're a gambler, the definition of the bet matters. Because we didn't bet that it's announced. And then if you were going to bet it was announced, we'd say announced officially announced.
Starting point is 00:22:43 So this is announced. Yes. Unconfirmed. Yes. Announced by Apple. So there's three different bets we could have made. Leaked announced officially in your hand. And so, yeah, I win this one, I think.
Starting point is 00:22:56 Then we just need why you have to think for your bets. I gave a little like concessioner, I think, is going to be. I said it's going to be close. It's going to be close. I appreciate that. I appreciate that. Yeah. All right.
Starting point is 00:23:06 Okay. So I was just a quick update on our bets. Yeah. Yeah. But it's exciting. And these are $1,000 bets, by the way, folks. We're making $1,000 bets. And we settle up.
Starting point is 00:23:15 Somebody could sting, could sting. And I like to hear that history of how you guys were getting the scoops because that's kind of my scoop angle as well. Okay. All right. Okay. So this one is super exciting. I actually, it's an area that.
Starting point is 00:23:31 I'm passionate about personally. So this is a company called Forever AI. That's how you pronounce it, F-R-V-R-A-I is the website. Correct. It's forever without the vows. Exactly. And there are like a bunch of season game folks that have basically taken AI to create sort of these small casual games.
Starting point is 00:23:57 And so we're going to do like sort of a live demo here where, Yeah, here's a game, you know, that you can play. Oh, it's Gallup, basically. Yeah, exactly. And I say, oh, you know what, that's cool. So let me just die here. Okay, so what we're going to do is we're going to take this game and we're going to remix it.
Starting point is 00:24:13 And so I'm going to click this edit button here. I'm going to click remix. That's going to open it in their editor. And what I can do is I can say, make the ship shoot dual missiles. Right? And what I can do is, and I just hit. enter there. And what it's going to do is it's going to go through the code, write live, and this is doing it. And it will basically edit the game. Oh, my God. Yeah. And, um, and basically now you see.
Starting point is 00:24:47 What? And so, yeah. And then the other thing you can do is I can just go in here and I can say, let's just change this enemy and let's create a different enemy. So let's imagine this. And give me a description here. Yeah. Make it, uh, Donald Trump. 10. Okay. No, the content filter. Yeah,
Starting point is 00:25:05 I think the content filter is probably stop. Let's just try. Try doing Darth Vader. Okay. Make it look like Darth Vader. How about this? How about a make?
Starting point is 00:25:13 Try to see if we'll do like adjacent to Sith Lord head. How about that, right? Awesome. Yeah. Imagine. Now you click Imagine and see what happens. And then you click imagine.
Starting point is 00:25:23 And so it's going to do a generation here. And, And what's really amazing is, and they've got lots of different game styles here, from side scrollers to, you know, these shooters to space invaders to asteroids. But what this is going to do for game creation, it's like what we, the way I think about is what we saw happen with YouTube and video. And, you know, I believe video games have surpassed even, yeah. Wow.
Starting point is 00:25:54 It just made Kyle Ren and some very creative Darth Raider looking. I can make that. and you can see now so now I can play it again and now I have the Oh my God Isn't that wild? It's genius
Starting point is 00:26:04 Yeah and you can you can do even more sophisticated things than that but yeah and then you can basically publish this game so you go back here
Starting point is 00:26:12 and you publish it So it's democratized a number of things you've democratized game publishing and you've democratized writing code Yes
Starting point is 00:26:20 this is and you're making the analogy to how YouTube democratized owning a TV network A channel on YouTube is essentially like having a channel on your cable box. So Mr. Beast is like he's standing, you know, next to MSNBC and CNN and, you know, TNT, whatever, HBO. Absolutely brilliant.
Starting point is 00:26:42 Absolutely mind-blowing. Yeah. It's kind of getting here faster than I thought, if I'm being honest. We're in year two of large language models. The absolute creativity and insanity here is awesome. They did narrow it, right? So because you're editing an existing game that narrows, I think, what the AI has to do. So we didn't start with a blind piece of paper and say, I want to make a game that is like a Nintendo game that has a Sith Lord.
Starting point is 00:27:13 So it's not quite that. But it obviously is two jumps away from that. So you're editing existing codebase, I got to give this an A. I'm sorry for giving high scores of late. but this is literally mind-blowing to me. The stuff is getting so good. We're kind of in a world where it's really impressive. And look, you've been probably close.
Starting point is 00:27:36 Yeah, close. You've been in content creation from the beginning. And you've seen that transform. And you've definitely seen the video game industry. And this is going to fundamentally change it. So you think about Roblox, epic games, and everything that comes in and around it, I think we're going to see a massive explosion. And another prediction I'll make is, I think we'll start to see creators now as well,
Starting point is 00:28:01 where YouTube was the platform that was needed to highlight creators. And I think we're going to see video game creators. Because right now people create these games inside these platforms. We don't know about them. So we're going to see that as well. Someone's going to be known for creating really incredible games in that way, rather than the being companies, we'll see individuals doing it. Just like, you know, people have created great podcast, news channels, you know,
Starting point is 00:28:24 email newsletters, these things, you know, even Drudge Report, you know, as just a solo reporter, think about how he changed history with the Monica Lewinsky, Bill Clinton scandal, and, you know, other ones that he broke. You know, a solo artist, a solo creator, being able to create something just as one individual that generates a million dollars, $10 million, $100 million. You know, DJs doing that, right? There are DJs who make tracks, just them on a laptop with a, the Starlink connection, you know, on a boat, boom, just change the world.
Starting point is 00:28:58 Now, that might seem obvious now, but I can tell you, having grown up as the generation, Gen X is the generation that spans these pre-internet, post-internet, pre-online, post-a-online. Now you've got a generation that's going to be pre-AI, post-AI. Post-A. Wow, yeah. So just let's pause there for a second and recognize that our kids are going to remember. They'll be the generation that remembers the time before AI and the time after. AI. They would have submitted a school
Starting point is 00:29:26 report, at least my older child, would have submitted a report without AI, and now they're going to be doing it in college with AI. They're going to have done my younger kids who are seven, they're going to probably only remember submitting term papers or doing art
Starting point is 00:29:42 with AI. They're only going to remember an AI world. They probably won't remember the time before AI. And this is going to be like a whole different way of looking at the world. It's like being born in the year 2000. You don't know the world without the internet. You don't know the world without online. When I explained to people what it was like to get movie tickets. Yeah. Just the act of getting movie tickets and going and seeing a movie. So going to,
Starting point is 00:30:06 wanting to see a movie, required going to a movie theater. And it required going to a movie theater when the movie was playing, because there wasn't a virtual jukebox video player in the sky called Netflix or, you know, Disney Plus, where you could just get Star Wars anytime. If we wanted to see Star Wars, we had to wait for the summer when they re-released it. And it was so popular that every year, they would re-release Star Wars before Empire came out. When you had to open up the newspaper, find the time. Then you'd go, yeah, and you'd wait in line. No, you'd wait in line maybe even for that time, right?
Starting point is 00:30:40 Yeah. For that time. And you just got a ticket that was red, green, or blue. And it had a number on, and they just ripped it in half and handed you your stub back. You need to keep your stub so you get back into the theater if you went to the bathroom or something. And there was no arrange sheeting. And you couldn't watch it at home. And then VHS came out and you could watch it at home.
Starting point is 00:30:57 And we were like, whoa, we can watch Empire Strikes Back anytime we want. Like that's all of a sudden happening in this AI generation. It's going to be wild. Their expectation is that everything they touch. Travel agents are the other one. Travel agents are the other one. Remember? Yes.
Starting point is 00:31:12 If you wanted a ticket, explain to people, if you wanted to go somewhere in the world. If you wanted to go to Tokyo, what did you have to do? Or you wanted to go to Paris. Like you'd go to like the mall or a strip mall. and there would be like a business and you'd walk in and they'd be sitting on a desk and you'd sit down with them and then you'd tell them where you'd want to go and then basically they would be on a screen and they would maybe type something. They had a weird terminal. Exactly. And then they would give you a price and that would be it.
Starting point is 00:31:40 And then they print your tickets. They print your tickets and you needed to bring a giant envelope with you. I remember they had that red thing too because it'd make multiple carbon copies of it. and you'd have to then not lose your tickets. If you lost your ticket, you lost money and you could sell the ticket so I could take my ticket to France and hand it to you and sell it to you.
Starting point is 00:32:03 Then you can take that ticket to an airport and say, I have a ticket to go to Paris and they'd say, okay, when do you want to go? You'd hand the ticket that was paid for already and they would put you on a flight. It was bizarre. Yeah. Your name wasn't even in the system
Starting point is 00:32:19 as you on the flight necessarily. There was no system. There was no system. You bought five tickets. They printed them. They exchanged cash with you. You took these five magical pieces of paper and they let you on the flight with the paper. Yeah.
Starting point is 00:32:32 And if they oversold the flight- You watch home alone. You know, if you watch home alone doing Christmas because, yeah. They have paper tickets? It's bizarre. It's just absolutely insane. Sorry for two old guys having their minds blown.
Starting point is 00:32:45 Yeah, we're kind of like, oh my God, I remember back in the day. But think about everything you experience, Sande, for this generation, they're going to say, well, I don't like it that way. Change it. You know, that's a great movie, but I want the protagonist to be a woman, a boy, you know, an alligator. You know, and boop. Okay.
Starting point is 00:33:04 I give that an A. I give that an A. Yes. What do you give it? I'm an A plus. Like I really, whoa. I am a big believer in the video game economy and what's happened with it. And I've been a gamer for a very long time.
Starting point is 00:33:20 And so to see the ability for that ecosystem to explode is very exciting. I love age of empires, even in my older years. I love playing a real-time strategy. I started playing the Frost Giant game. Those are the people who made Starcraft, and they're making like a new real-time strategy is the RTS, is my category. Okay. Command and Conquer, Age of Empires, Starcraft, all fall into real-time strategy. It's, you know, like you build the civilization, you get collect resources, you then build weapons.
Starting point is 00:33:53 You mine gold. Yeah. Whatever it is. Yeah. Wood, meat, stone and gold in age of empires. Yeah. It's a lot of fun. My daughters are into it now, too.
Starting point is 00:34:03 So I'm literally going to buy three PC gaming laptops so we can all play Age of Empires together. And you have a land party again. We're literally going to have a land party. Absolutely. Yeah. If you want to build a great company, you need a great. team, it's that simple. And if you want to hire great team members, the first step is finding them. But where are they? You know where they are. LinkedIn jobs. That's the solution you've
Starting point is 00:34:26 been looking for. LinkedIn has more than one billion users, which is easily the largest professional network in the world, and you can land both active and passive job seekers. LinkedIn also knows that small businesses are wearing so many hats right now, and you might not have the time or resources to devote to hiring. So let LinkedIn make it automatic for you. Go post an open role on LinkedIn. Put that purple hiring ring on your profile. Critically important. Then, just start posting some interesting content. Talk about your day, talk about your career, maybe comment on a news story, whatever's trending in your industry. And then you watch the qualified candidates roll in. Sometimes from your network, sometimes friends of a friend,
Starting point is 00:35:00 and sometimes it's just those passive LinkedIn users reading their newsfeed, trying to get educated, and they see your hiring. So guess what? The first job listing is on us. And you can get that first free job listing at LinkedIn.com slash twist, LinkedIn.com slash TWIST, to post your first job for free. Terms and conditions do apply. We're going to go back to, you know, the folks that kind of kicked it all off. They've launched some new features recently.
Starting point is 00:35:26 Back to chat GPT. Okay, this is OpenAI's chat GPT. So here we are in the well-known interface. So what you can do now is you can say, you know, I'm thinking of flying to New York tomorrow. Thinking of flying to New York tomorrow. And you can just say at kayak. Oh.
Starting point is 00:35:45 And see, now it's that, right? Yeah. What are my options? She don't have to go hunt and pack to go find. That's exactly it. So within your command line, yeah. And so just give it allow here. And so within that, you can basically at mention all the different GPs that are in there.
Starting point is 00:36:08 Yep. And so and you don't even have to know if a GPT exists. You just be like at open table, at Redfin, at. Exactly. Amazon and just see what happens. Yes. Yeah. And I think this is really smart by them because you don't, you know, when this all started,
Starting point is 00:36:22 so we're just seeing evolution by them. Remember they kind of had these kind of apps. You had to add them. You had to plug them in yourself. Yeah, they were just not in line. Well, plugins initially, exactly. Then they became sort of GPTs. Now they're just within your command line.
Starting point is 00:36:35 So as long as you like recently used it or seen it and you can see it here and you can see the flights are coming up. Now it doesn't book it for you, but you can click book now. You off to kayak to book it. Well, hopefully that would launch the app if you had the kayak app on your phone. Or it should, well, because this was kayak, see, I want to have it have at United. Yes. You know, like I want to just do at United, at JetBlue, because those are the two I like most.
Starting point is 00:37:01 I like JetBlue men. Well, they need to make a GPT if they want to be. They need to make a GPT. Why don't you try that? Do a new one and see if there's an at United or. You have to kind of discover the GPT's person. The way they did it is like you have to have been through them. So I don't think there's like a United Airlines.
Starting point is 00:37:16 I don't see one in here. Yeah. Yeah. Yeah. So if somebody works at United Airlines, like, get on this. You got to, you know, this is like that whole race where people didn't have websites, then they didn't have apps. Now they don't have GPTs.
Starting point is 00:37:31 Now I got to like drag everybody kicking and screaming to add a GPT. But this is going to be incredible. You know, this is like the beginning of the super app like Uber, where you just keep adding features to it. And yeah, God bless. It's going to be awesome. So I give them, I don't know I get for that. I think it's like a B.
Starting point is 00:37:49 It's a B. It could be better. It's a good new feature, but it's going to make it exactly what you said. It's making it such that you can do more within your chat, GBT interface, which is smart by them. And going to continue to put them in front of a lot of folks in the race to be the Internet's apex aggregator. Yeah. I mean, listen, if you're the platform, you're the aggregator, and then you have a certain number of users, everybody has to build. it. Like I just said, there's no JetBlue.
Starting point is 00:38:16 You know, in the early days, JetBlue is like, oh, we have a web responsive website. Just use that. And I was like, but I want an app. You know, everybody's like, I like to have a pretty little button. Okay, I'll make an app and wrap our website in it. And then they're like, but your website doesn't work well in the app. So, okay, I'll make a responsive app. Yeah.
Starting point is 00:38:33 I'll make an actual app. Yeah. Anyway, I give this a B. Could be better, but it's a great. I'm kind of at the same. It's like, it's just a feature and it puts it in there. I think it's worth thinking about. Okay.
Starting point is 00:38:44 What would you add to it? If you were sitting there brainstorming, I had one thing, which is I want to be able to do it without having to add them first, right? And I want to be able to do multiple, right? That's, those are the two things that I really found. Like, you have to kind of have discovered them first and I don't want to do that. Right. And I want to use multiple. I want to say, you know, the ticket, then then open table and a few of them together.
Starting point is 00:39:04 Exactly. So we're both bees, solid bees. Yes. Yeah. Sandit, B for Jacob. Let's keep going. That's it. Okay.
Starting point is 00:39:11 So just on this. same track, what we're going to share is last Friday, I believe, or Thursday. Okay. So, you know, maybe we're like four or five days ago. Hugging Face launched
Starting point is 00:39:26 something called their assistants. And so what they're, what they've done is they've kind of created an open community to create their own GPT. So GPTs, so let's bring everyone up to speed. Three million were available
Starting point is 00:39:42 at launch. And so this was launched four or five days ago. They already have 3,000. Whoa, hugging face has their own GPs now. Correct. And the idea... What language are they written in? Like, how, you know, what's the standard here? I'm assuming it's some sort of open face, some sort of open source approach here, or you can write any code that you want. Exactly. And you can kind of plug it into, you know, they kind of follow the same template and framework as Open AI, but instead of having to build it using Open AI, you can kind of build it, which is whatever back can you want, and it becomes available in their ecosystem here. Is it portable? So, which is to say, can I take this one and use it at ChatGIPT?
Starting point is 00:40:24 I don't think so. I don't think they're portable right now. Yeah. So this is going to be, this is where somebody could, this is where Hugging Face could do a lot of damage. They could just say, we're going to make it open store standard. These apps are portable. So the chat cheap, what does chat GPT call them now? GPDs GPs GPs Do they own
Starting point is 00:40:43 the IP around GPTs GPT is a common word or do they own that Yeah, GPT is a technology right It's kind of like saying So they can't own that So
Starting point is 00:40:53 Portable GPs or making a GPT standard Would be a great thing For Hugging Face to do Yeah Okay, so you can see here When you create one You'd go kind of go through a similar process
Starting point is 00:41:05 And you know We did one of these When GPs first came out And then you can pick a few different models here that you want to have so disruptive yes
Starting point is 00:41:14 but what I would say here is look now you're just starting to see the ecosystem go in two different directions with GPTs you're limited to using their models here you can kind of pick between mixtral Lama Oh so pick the META one
Starting point is 00:41:28 I want to pick the meta one let's pick the meta yeah sure Lama okay this is Zuckerberg's open source and you saw his announcement with his stock ripping thank you yes yeah so let's name our
Starting point is 00:41:39 Let's name ours, healthy recipes. Okay. And the description is make healthy recipes. Well, this is the description. Okay. Make healthy recipes. Whatever. Yeah.
Starting point is 00:41:53 Like, this is not the instructions, right? Instructions are in the instruction. Right? You'll say, you'll act as a world-class chef. Okay. So you're giving the prompts. You'll act as a world-class chef. Yeah.
Starting point is 00:42:09 provides healthy recipes. I love it. Healthy recipes to your users. Oh, I got to lock in here. Okay, but I get it. You could put all the instructions and you could say, listen, we don't want to use those oils. Now people tell you don't use all these oils, right?
Starting point is 00:42:23 And then you can say, hey, we're not going to use these vegetables that have inflammation. You're not going to include these. You're not going to include these. You're going to keep sugars to natural sugars. And then boom, it just starts making your recipes. I love it. Exactly. Fantastic.
Starting point is 00:42:37 This is great. So, like, this is where, like, if you're going to build a startup in AI, you're going to need to have training data as a differentiator brand or the accruedment, the tools, the features around the AI, right? And so, yeah, okay. And I'm working on like 2.0 of my vision for the inside stuff I showed you. Yeah. Yeah. And I'm going to show it to you because I got a new team working on it. I switch teams.
Starting point is 00:43:02 Okay. Sometimes you got to switch teams, you know? Yeah, switch teams. The team did a great job on the first part, but I wanted to go with a different. of team to build it. And I realized, well, everybody's going to be able to summarize the news. That's like table stakes. Like it's not going to be a big deal.
Starting point is 00:43:17 So I came up with two ideas on top of that. That you can't replicate that have network effects. So anyway, I've been thinking through like, okay, all the cool little things you can do, summarizing, finding, whatever. That's all good. But there's another layer that you can't, which is the network, the brand, etc. So look for an inside. app coming soon. For this, this is another one where I'm going to give it a B plus to give it room
Starting point is 00:43:44 to grow. But yeah, it's totally fantastic. What are you given? I'm going to be a little bit, I'm going to be a little bit higher there because what I like about this is embracing the open source ecosystem. And so I'm going to put them at an A because what they're really doing is embracing all the different models, all the different companies that are putting models out there. And so that's going to lead to a much better ecosystem for us. And you can see. Open approach, it's much better for everyone. Could you say, I want to build like a two-layer, you know, app? I think they call these assistants.
Starting point is 00:44:19 Is there a power? So they call it assistance instead of GPDs. Could you build one that says, hey, step one, we're going to help you write a script using Lama, you know, you're a director, your screenplay writer. Step two is we're going to take that and we're going to make a movie out of it using. you know, instead of using the Lama model, what's the visual one? You use stable diffusion maybe. What's the video one?
Starting point is 00:44:45 Runway.com. Runway. Right. So like you've got to build it in two steps kind of situation. Really interesting. I'm going to give us a B plus. I think it's great. Yeah.
Starting point is 00:44:53 Okay. Awesome. Let's do one more and then we'll wrap the show. Another great one. Okay. Another great show. You can see all of our AI demos at this week in startups.com slash AI. Okay.
Starting point is 00:45:05 I'm going to go back to chat GPT for this. one as well. I thought this was pretty cool. And it's a use case, which is really, I think, is going to take off. It's using the 11 labs. We've had them on before. I think you guys even do a bunch of work with them. And so what I did here was, I said, generate a short story for me, then convert it into an audiobook, two paragraphs. And I did it using the 11 labs text to speech, GPD. 11 labs, text to speech GPT. Correct.
Starting point is 00:45:40 It said, generate a story for me, then convert it to an audio book, two paragraphs. Yes. Wow. And then see, the first thing it did was it came back to me, so, well, what's the voice that you want to do it in? And I picked Jarvis. And so then it created the story for me, and then it generated the audio. And if I go over here now, the story that was generated, I can play it.
Starting point is 00:46:00 And let's listen. Metropolis beneath the shadow of towering skyscrapers, lay a, a quaint little bookstore that seemed almost forgotten by time. So that's Jarvis from Iron Man. Avengers Iron Man. Yeah. I'm not sure he made his debut. Wow, that's wild.
Starting point is 00:46:17 Right. And so look at the number of things that have come together there. The story generation is coming from, you know, the model, right, GPT4. Then you have 11 labs doing their text to speech as well. And then think about what we're going to be able to do very, very, very. very quickly here. Yeah. It's crazy.
Starting point is 00:46:38 Yeah. And then the question becomes like, if I'm writing a novel, am I not going to use this? And then I'm going to start, I wonder if people are going to start saying, this is organic, 100% human, like literally the label, 100% human, no AI inside. Like I know this sounds crazy and dystopian. Yeah. But I think people are going to start.
Starting point is 00:46:57 Like you said your next book, your next book would say there's no AI. No AI, AI free. Like you know how some people are into organic, non-GMO, they put labels on food. I wonder if we're going to start labeling content. Like, this is organic, right? So I made this. That's impossible. Well, I mean, no, it's not.
Starting point is 00:47:16 I mean, people make acoustic music and they just say, like, imagine I said, this is one take acoustic music. I take out my guitar. I play the song. And you get to experience Tracy Chapman. I don't ever saw that at the Grammys. The Grammys, yeah. That was great, huh? I got a lot of feels from that.
Starting point is 00:47:33 Yeah. But, you know, like, whatever, 30 years later. I saw your tweet, right? Yeah, I used to listen to that Tracy Chapman out meant a lot to me because I was 17, 18 years old when it came out, a 17 years old, 1988. And I was going to Fordham. And I got, I had a CD Walkman, you know, and I would, you know, get a bunch of AA batteries.
Starting point is 00:47:54 I popped that in there. And that was like my guilty pleasure was listening to that album on the way to school, you know, Fordham. I went at night and I would commute. And then to see her, you know, she's kind of a, you know, doesn't perform, I guess, publicly or whatever. Not out of the spotlight completely. Yeah. I was like the first time in a long time that she's performed.
Starting point is 00:48:13 And she was just so powerful. And then you see the guy singing with her who had done the cover of it. And he's just in awe of her. Yeah. It's just like a great like multi-generational awesome moment. People were noticing a lot of things like she had no monitor, like just old school. Right. Yeah.
Starting point is 00:48:27 That's what I was noticing when she was playing. There's some very intricate finger picking in that song. Yeah. And I was watching for the finger picking and stuff like that. And I was like, you know, I think with this AI stuff, just like we're going to be like free range chicken as opposed to factory raised or no hormones, no GMO, whatever, you know, organic food, whatever. I think that's going to start happening where so much of what we consume is so processed content wise, the words, the music, etc. That people might say, you know what, I just want to see a human do this without the computer. just like you might want
Starting point is 00:49:03 some people might like to go to electronic music festival and watch somebody press the play button and then you know the DJs like moving a bunch of knobs I don't think they're actually changing anything I think they hit the play button in their set list everything is already done right I think it's kind of already programmed they're just jumping around and you know getting
Starting point is 00:49:18 you're kind of getting everybody hyped but it's kind of fun to see somebody actually do it as a human and say wow the human can do it too yeah it reminds me like classic classic cars like there's a Tesla which is incredible and super safe
Starting point is 00:49:35 drives you somewhere but then there's like you know but then there's like an old Corvette Mustang right yeah old Corvette yeah well yeah and then you think about those cars so I think that you can get in there and change the carburetor
Starting point is 00:49:47 you change the spark plugs you could you know adjust the timing a bit yeah yeah everything you can tweak it and then think of this all these cars now can drive themselves can be remotely controlled
Starting point is 00:49:59 I don't know if you've turned this on your Tesla, but you can turn the cameras on and it's like a remote camera system. Did you turn that on yet in the app? Yeah, I've seen that. Yeah, where you can turn it on and kind of look around the car. Yeah, I've seen that. Yeah. So, like, when your car is in your driveway, you can be in bed and you hear something outside,
Starting point is 00:50:16 you turn it on, you can see all around your car. So your car is like eight security cameras. It's pretty crazy. Yeah, and you can see, you can turn the inside camera on, except if the car is moving, then you can't turn it on. So, because some safety there. but I was just thinking if you don't want to be tracked
Starting point is 00:50:33 you could get in a Mustang leave your phone at home drive to Mexico and have a lot of cash and drive to Mexico and just disappear or you could have your phone and your car and you'll live stream it for the entire world
Starting point is 00:50:50 in HD. It's like the world has changed that much and this past year it's changed incredibly All right, so what do we give that? You basically created an audio book, soup to nuts. I mean, it's... In a prompt.
Starting point is 00:51:07 In a prompt. With Jarvis, who's like one of my favorite characters, right? Yeah, well, it's also an actor who probably should be getting paid. Well, I'm sure 11 is. Like, you know, they're... Eleven probably licensed it from him. Yeah, yeah. On a flat rate, I bet.
Starting point is 00:51:23 He should be getting paid per minute to public. When you hit... You should be able to list. to it, but when you publish it, you should have to pay like a dollar per minute or a dollar per hour or something. He should be getting some residual. I give that a B. Okay.
Starting point is 00:51:35 Right now, I would give it an A if it actually uploaded it to a store and let people buy it. Not necessarily audible, because they're going to not let you publish this stuff to audible. But if you could publish it to Distro Kid or whatever those third-party tools are like my friend Filmed, where then there was a store where you could buy the audiobook, that would be kind of cool. So, yeah, I give it a B.
Starting point is 00:51:55 It's incredible. It's extraordinary. Yeah. I'm kind of there like B plus. Like it's all stitched together a bit. And so the only thing it needs to go to like the next level, it needs to be totally seamless, publish, get on, you know, the thing is it's all inside the chat GPT ecosystem. I want that on my phone and I want it in my, you know, podcast or books out or whatever.
Starting point is 00:52:16 Let's do the cover. So it's like it's like it's getting there. But I give it a big. All right. Listen, been incredible. It's been incredible. Great episode. If you would like to be featured on This Weekend Startups AI edition, it's typically every Monday,
Starting point is 00:52:30 this week and startups.com slash AI to see all of the get the playlist and everything, see all these bets and everything we're doing. It's incredible. Yeah, just at mention us on the Twitter or the acts formerly known as Twitter. At Sundeepe at Jason. Yeah, I'm still at Jason. Unless my account's been banned for asking somebody who they are. You got pretty close there. Pretty close to my first Twitter man in 15 years. Your bestie jumped in for a second.
Starting point is 00:52:59 He jumped in, yeah. Elon jumped in. For people who don't know what we're talking about, we can we talk about it here at the end of the show. Yeah, of course, sure. So there's a, I'm not going to say the name of the account because I don't want there to be like pile on or I'm just trying to let this thing settle.
Starting point is 00:53:14 There are anonymous accounts on Twitter that have gotten very large. Some of them have been funny, like Goldman Sachs Elevator, right? And then some of them have been particularly hurtful to some people feel they've been hurtful. I'm not making any political judgments here, but like libs of TikTok is an account. And the libs of TikTok account,
Starting point is 00:53:32 I will talk about, because that's been a very public one. There is a term called doxing. Now, what do you consider doxing? I'm curious, before today, and this whole brouhaha that I got into on X, what did you consider doxing? Well, you know, I'm kind of polluted now
Starting point is 00:53:48 because I've read the thread. But like, so I'm going to try to go back. Yeah. Yeah. What I would have considered is basically highlighting someone's true, like, true identity and public information, like where they look like the whole doxing would be the whole, like everything. That's what I would have put it.
Starting point is 00:54:08 Got it. Okay. The package. Yeah. Yeah. Here's where you live. Here's where you work. Here's your name.
Starting point is 00:54:12 Yeah. Got it. Exactly. Just public information. Yeah. Right. Yeah. I've always considered doxing.
Starting point is 00:54:18 Like, here's somebody's location. And I understand that people. feel differently about this. We're going to need a new term. Doxing location, doxing identity. So I'm going to, for the purpose of this conversation, we're going to have, I'm going to say doxing location, doxing identity. Doxing location, we can all agree. No, Bueno. Dangerous. Don't do it. Now, doxing somebody's identity means if they use a pseudonym, they call themselves, you know, libs of TikTok or whatever that account is called. Does that person have the right to be anonymous, is the question?
Starting point is 00:54:52 What do you think on social media? Well, I don't know if they have the right to be anonymous. Yeah. But I think they're trying to be anonymous. Great. And, you know, it's on them to protect themselves, right? It's like most of these accounts, you know, even we've had some in our own circles, right, that were close to us. And then you realize this is one of our friends or something like that as well.
Starting point is 00:55:14 Fake chum up is coming to mind. Exactly, right? A parody account. Right. So there's parody accounts. Yeah. Yeah. And so I think the real challenge is once you're doing that, like, do you have the right to remain that way? No. But it's sort of, I guess, a double-edged sword that you get with. And so with great power comes. Yeah. So my position is, hey, with I simply asked, and I did it multiple times because people kept asking me to cover on the podcasts this one particular account's content. And I don't think your intent was.
Starting point is 00:55:51 like your intent wasn't like, who is this person, where is his address? It was like, sort of like, hey, who is this person? Who is this person? I just asked the question. Who is this person? It was with fake Chamath, right? Like, who is this person? Right.
Starting point is 00:56:02 And then I asked the person of someone, like, who are you? Yeah. Like, I'm curious because you're breaking, and to the person's credit, they seem to be breaking news or somebody's using them as a vehicle to break news. So to peek behind the curtain, when you have these large accounts, there can be, it could be anything. It could be an individual who is just a random person commenting. Great.
Starting point is 00:56:26 That's like one thing. It could be a team of people. It could be a political organization. It could be a foreign actor who wants to spread misinformation. It could be a useful idiot who is being given information by a foreign entity to spread it or by clever people who are using it to spread it. And so the question is, you know, these things can get very big. Hundreds of thousands of followers, perhaps millions. I don't know if any of them have millions yet.
Starting point is 00:56:56 And so I believe people should own their words. I believe it's fine for them to try to be anonymous. I believe it's fine for people to try to figure out who they are in that game of cat and mouse. You don't have the right to be anonymous and have a huge following. And then it's okay for platforms too to have different rules. It turns out the rule on X now for those formally. known as Twitter is you, I guess the rule is you can't reveal somebody's identity. Identity, yeah.
Starting point is 00:57:24 That's what was made clear by Elon, right? He said, you cannot do that. Elon said you can't reveal people's identity. And he wants to protect people because they don't want to lose their job. I disagree. I know this is incredible for people to think that two friends might disagree about something. But there's different platforms. Reddit has got pseudonyms too.
Starting point is 00:57:41 Twitter is going to be pseudonyms. I like people to own their words. And that doesn't mean I want to. you know, have people get outed or whatever. But I do think if you want to have a huge platform and you want to have impact in the world and cover things like politics or, you know, whatever, or do something full contact,
Starting point is 00:58:02 like the Libs of TikTok account is very controversial because it's amplifying to a very large level, things that are public already, that people maybe didn't expect that they would get brigaded when lives of TikTok highlights some live person saying something really stupid, and then a thousand MAGA people attack the person or whatever. So anyway, I think whistleblowers, pseudonym accounts,
Starting point is 00:58:29 and anonymous sources, they can all exist in the world. They don't have a problem with them existing. I also don't have a problem with whistleblowers being revealed. It's dangerous, but that's part of the choice you make as a whistleblower is that eventually you will be revealed. You know that. it's almost a certainty. pseudonyms, you're going to be revealed eventually if you get any kind of credibility or
Starting point is 00:58:50 growth. And I actually think that publications like journalists rely on anonymous sources too much. And I think if you, that's where I think you can get people to understand the point. We've all now seen anonymous sources are like a driving force of journalism now. And it's way too much. I mean, you just go to somebody and you're like, I'll give you anonymity if you say something crazy about this person. And then you don't know if it's a lie or not. And they gave the person anonymity. It feels profoundly unfair to a lot of people. And I think it's overused. Yeah.
Starting point is 00:59:22 Anyway, that's my position. I don't know what you think about what you're thinking of my position. What I like to know what Elon said is he at least put the stake in the ground and said, look, this is what we're going to do here. And so now he's clear. And yeah, and he's made it clear. And if you don't like that, I guess you don't have to use Twitter or X, right? And look, I sort of expect in other places that, you know, if I'm reading maybe traditional media, that they're not doing what you're saying, which it does happen, where if they have a source, they'll verify who it is and they don't allow the source. They may keep the source anonymous for the sake of the article, but they may go verify who the person is. Yeah, they will verify who it is, yes. Yeah.
Starting point is 00:59:58 Yeah. And so that's going to be the difference between X and maybe, well, not traditional. want to know how they do that. Other platforms, yeah. You know, let's say somebody was an employee of yours and they wanted to leak a piece of information out of project you're working on. The way they could prove that to a journalist is the journalist lets them contact them from their private email address.
Starting point is 01:00:18 And they say, okay, what's your LinkedIn page? And you say, oh, this is my LinkedIn page. So great. Put the word in your bio. Put a comma after that word, a typo. You know, put a typo there. Put two O's instead of 1O. And then tell me to refresh the page.
Starting point is 01:00:33 I'll refresh it. Okay, now take it out. Refresh. Okay. now they've confirmed it's you. Nobody else knows how they confirmed. And now you're off to the race. So hopefully journalists are doing that kind of thing where they try to
Starting point is 01:00:44 and then making sure that that LinkedIn page has existed for a long time and it is followed by other people in the company, whatever. But you know, then the question, then there's like all these edge cases that came up. What if a journalist outs the lives of TikTok account, which I think Taylor Lorenz did? Taylor Lorenz found out who the person was and then we're a story about it.
Starting point is 01:01:05 Recently, yeah. If she did that in the New York Times or Washington Post, but didn't do it on TikTok, does she get her X account ban? No, she didn't do it on X. She didn't do it on X, yeah. You didn't do it on X? Okay. Now, if she links to her story where she did it, does she get banned?
Starting point is 01:01:27 Slippery slope, man. I would say yes. In that case, I would say yes, because you link to it and so it's the same thing. Now, if you link to it after the story comes out and say, I disagree with it. they shouldn't have outed this person and you linked to the story. Did you dox them?
Starting point is 01:01:42 Yeah, I think if you link to a story where somebody got doxed, are you daching? I don't think so. I don't think I'm responsible at that point. No, because the story's already out. I think you got to know the original doxing.
Starting point is 01:01:53 Yes, the original docks. By the way, just a piece of advice, somebody who's had security accounts, don't run an anonymous account at scale if you have security concerns. Or turn this account off to start a new one with a different persona and then roll your account
Starting point is 01:02:05 so that, because you, the longer the footprints out there, the greater the chance they're going to get you, you know, and figure it out. And these people do have ways of figuring it out. They figured out who startup Jackson was,
Starting point is 01:02:16 remember that? Yeah, it was Parker. It was Parker. They figured that out. They worked very closely with you. Yeah. They figured it out, I think,
Starting point is 01:02:22 because they took his tone of voice and then they used AI to check another thread, I think? No, no, no, no. I think so. I think, or they used some sort of software to say, like, what words do this person use in common?
Starting point is 01:02:35 and it was, yeah. So anyway, I'm with people owning their words. That's a good little short discussion we had here about it. Okay. All right. Yes. So don't docks anybody. Don't obviously location docs anybody.
Starting point is 01:02:47 AI will figure it out, though. Maybe that's going to figure out who they are. If you keep trying to say, I just did this to use AI to figure out who they are. I could also protect you. You could say, hey, I want to say this. Yeah. But say it for me in a way that's not traceable and not in my voice. So there's a tip.
Starting point is 01:03:05 for people who want to stay anonymous is use AI to rewrite what you're writing. Yeah, that's right. You don't even know your own ticks. Okay, we'll see you all next time. Bye-bye. Hey, everybody. I talk to a lot of founders here on this week in startups and as an investor. And they tell me the same thing over and over again.
Starting point is 01:03:21 They want two things from me, more FaceTime and money. They want me to invest in their companies. And they want to spend time together. So we've been working here on a new meetup program. We call it Founder Fridays. And Founder Fridays are an event by, founders for founders. This is an event that is hosted in cities by people like you. If you're listening to This Week in Startups, you're a founder. So what are you going to do at Founder Fridays?
Starting point is 01:03:45 You're going to get together with other founders in your community. It could be four or five of you. It could be maybe up to 30 of you in a location. Pick a cafe, pick a co-working space. I like to go to a great Mexican joint or maybe a dim sum restaurant. You know, where you can do shared food, have a couple of cocktails maybe. You do it on a Friday. You get together and you host Now, why is it important for founders to get together? Shouldn't you be at home just focusing? Shouldn't you be in the office just focusing on your startup? Well, if you get together with other founders, true founders who are in the arena building like you are, you're going to get a lot of value from that because you can trade notes with that other founder about what's working at your startup and what's not
Starting point is 01:04:25 working. The truth is, if you're facing a problem, there are hundreds of founders out there who have probably solved it already. And instead of you banging your head against the wall, when you sit there and you talk to three or four founders, you're having some dim sum, you're splitting a casidia, some prajitas. Somebody say, oh, you know what? I had that same human resources problem. Oh, I had that same technical problem. Oh, I had that same marketing problem.
Starting point is 01:04:45 And they might tell you about a tool or a service that'll solve that problem for you. This happens over and over and over again when I do Founder Fridays with our portfolio companies. Now we're going to give you that same experience. But here's what I need you to do. I need you to host this in your city. So you're going to go to this week in startups.com slash meetups. That's it. And you'll see a landing page where you can sign up and you can say, I want to host in my city. Now, your city may already be hosting so you can just join that person. And what if you go to this event and you learn some go-to market strategy that 10-xes your growth? That might unlock funding. Or you might be talking to somebody and they say, hey, I'm a marketplace too. I'm not a competitive marketplace. Your marketplace is for use cars. My marketplace is for hairstylists, whatever your jam is, whatever you're working on. But they give you some technique that you didn't know about to increase your supply side or get more demand.
Starting point is 01:05:33 in your marketplace and you 10x your business. I see this happen all the time. And founders are like mutants, right? And I'm like Professor X here. I'm trying to put on Cerebro and find all the founder mutants in the world and then have you get together and do your own little meetup. And here's what you're not going to have to deal with. You're not going to have to deal with a bunch of service providers trying to sell you software or services. And you're not going to have to sit through a bunch of passive speakers. You can listen to this week and start up saying at the greatest speakers in the world on your own time. And you're not going to have to pay for a ticket to a conference or get on a plane or fly somewhere. No, this is about having an intimate experience with five, ten,
Starting point is 01:06:13 maybe two dozen other founders in your city. Please go to this week in startups.com slash meetups if you are a founder. This is four founders by founders only. If you are not a founder, this event is not for you. You can start your own meetup for lawyers, accountants, recruiters. This is for founders by founders. We vet everybody to make sure you're a founder. And if you host it, it's a non-commercial event. Our first founder Friday will start on February 2nd. So please mark your calendars.
Starting point is 01:06:42 And we're going to do these on a rolling basis. You can join an existing meetup if it's already occurring in your city. Or you and one or two other founders can start your own. We're using a wonderful piece of software that we've invested in called River. You can sign up for a River account just by going to this week in startups.com slash meetups. We've already got hosts and attendees lined up in San Francisco, New York City, Toronto, Los Angeles, Las Vegas, London, and even in India. So this is your chance to connect. And if you didn't hear your city named, you can start your city. Go to this week in startups.com slash meetups.

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