Pod Insights Latent Space: The AI Engineer Podcast Heralds of the AI Content Flippening — with Youssef Rizk of Wondercraft.ai About this episode Want to help define the AI Engineer stack? Have opinions on the top tools, communities and builders? We’re collaborating with friends at Amplify to launch the first State of AI Engineering survey https://www.amplifypartners.com/blog-posts/ai-engineering-surveyhttps://www.surveymonkey.com/r/aiengineering2023! Please fill it out https://www.surveymonkey.com/r/aiengineering2023 (and tell your friends)!
In March, we started off our GPT4 coverage framing one of this year’s key forks in the road as the “Year of Multimodal vs Multimodel AI https://www.latent.space/p/multimodal-gpt4”. 6 months in, neither has panned out yet. The vast majority of LLM usage still defaults to chatbots built atop OpenAI (per our LangSmith discussion https://www.latent.space/p/langchain#details), and rumored GPU shortages https://news.ycombinator.com/item?id=36177895 have prevented the broader rollout of GPT-4 Vision. Most "AI media” demos like AI Drake https://www.theverge.com/2023/5/1/23703087/ai-drake-the-weeknd-music-copyright-legal-battle-right-of-publicity and AI South Park https://dataconomy.com/2023/08/29/fable-south-park-ai-showrunner-ai/ turned out heavily human engineered, to the point where the AI label is more marketing than honest reflection of value contributed.
However, the biggest impact of multimodal AI in our lives this year has been a relatively simple product - the daily HN Recap podcast https://news.ycombinator.com/item?id=35831177 produced by Wondercraft.ai, a 5 month old AI podcasting startup. As swyx observed https://twitter.com/swyx/status/1661848597728575489, the “content flippening” — an event horizon when the majority of content you choose to consume is primarily AI generated/augmented rather than primarily human/manually produced — has now gone from unthinkable to possible.
For full show notes, go to: https://latent.space/p/wondercraft
Timestamps
* [00:03:15] What is Wondercraft?
* [00:08:22] Features of Wondercraft
* [00:10:42] Types of Podcasts
* [00:11:44] The Importance of Consistency
* [00:14:01] Wondercraft House Podcasts
* [00:19:27] Video Translation and Dubbing
* [00:21:49] Building Wondercraft in 1 Day
* [00:24:25] What is your moat?
* [00:30:37] Audio Generation stack
* [00:32:12] How Important is it to Sound Human? and AI Uncanny Valley
* [00:36:02] AI Watermarking
* [00:36:32] The Text to Speech Industry
* [00:41:19] Voice Synthesis Research
* [00:45:53] AI Podcaster interviews Human Podcaster
* [00:50:38] Takeaway
This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.latent.space/subscribe https://www.latent.space/subscribe?utm_medium=podcast&utm_campaign=CTA_2
Transcript 11,233 words · 14 speakers
0:06 Speaker 00 Welcome to the Laten Space Podcast, where we dive into the wild, wild world of AI engineering every week. This is Anna, your friendly neighborhood AI, and I'll be standing in for Alessio today. Yes, you heard right. AI is taking our podcasting jobs. We flew all the way to London to interview Youssef Rizq, co-founder of Wondercraft AI, which has created the number one piece of AI-generated content enjoyed by the Laten Space community. We ask him how he arrived at his idea, what the future of commercial AI-generated content looks like, and confront him with the hardest question of all: what is his mote as an API wrapper startup? At the end, we even have him turn the tables and do a customer interview with Swix. There's lots of audio goodies in this one
0:48 Speaker 00 and bonus 30 minutes video on youtube.com/latenspaceTV. Watch out and take care.
0:54 Speaker 01 So we're in the studio here in London with Youssef. Welcome. Thank you. It's been such a joy listening to Wondercraft's podcasts over the last four, five months. You guys have been around for only five months. And as you know, I am one of your podcast's biggest fans. And I think that it's super interesting because I talk to a lot of vendors, effectively people who create services for other developers to build. And you are at the application layer, which is great and challenging for me as a podcaster because you have some secret sauce that you're not going to share. But I also want to just talk to you as someone who's evaluated a lot of things and built something that I actually use every single day. So that's the context.
1:39 Speaker 01 Great, great. How do you feel when I say these things? Is that exactly what you're going for? Yeah, yeah, yeah. So it definitely makes sense, right? Resonates with definitely on the application layer, and that's definitely by design. Yeah, we can talk about the origin story leading into Wondercraft, but just to learn a little bit more about you. You grew up in Egypt? I grew up in Egypt, yeah. Spent the first 18 years there. Cairo. And then you came over to the UK. You got your masters in EEE at Imperial.
2:04 Speaker 02 For those who don't know, that's electrical and electronic engineering.
2:07 Speaker 01 You then spent four years at Palantir as a forward deployment engineer. I think it's a role that Palantir invented.
2:13 Speaker 02 Forward deployment engineer is a super interesting job because it is kind of at this intersection of being an engineer, so a software engineer, but also still doing like business related things.
2:23 Speaker 01 Yeah, solutions architect maybe.
2:25 Speaker 02 Yeah, so part of the job was a solutions architect, part of the job was reviewing contracts, part of the job was doing sales, part of the job was coding things, part of the job was interacting with, right? So so many different things and I think that is a really good foundation for someone who does want to start something in the future.
2:39 Speaker 01 Excellent.
2:39 Speaker 02 You just do everything.
2:40 Speaker 01 So kind of an endorsement of that job if people want to.
2:43 Speaker 02 Huge endorsement of that job.
2:44 Speaker 01 Okay, excellent, amazing. Yeah, Palantir is surprisingly strong here in the London tech circles. I have a number of friends who are all ex Palantir.
2:51 Speaker 02 I think it's actually the the biggest offices in London.
2:53 Speaker 01 Yeah, surprising because I think of it as like a US defense company.
2:58 Speaker 02 Yeah, yeah, yeah.
3:00 Speaker 01 Then you started Moonshot for nine months, which is pretty important in your in your journey. I'll bring it up to Wondercraft. And you started Wondercraft in April of this year and it's been it's been about five months going through Y C in in the winter batch.
3:12 Speaker 02 Summer twenty two.
3:13 Speaker 01 Summer of twenty two batch. Okay, cool. What is Wondercraft?
3:16 Speaker 02 Nice. Wondercraft is a podcast builder that uses hyper realistic AI voices to create podcasts and make that whole podcast creation process super simple.
3:24 Speaker 01 Right.
3:24 Speaker 02 So a super simple example is you can, you know, you publish a bunch of blogs.
3:27 Speaker 01 Yes.
3:28 Speaker 02 You can take that blog, put it in there, it'll just convert it to an audio friendly format that people can listen to. It's just sometimes it's a bit more efficient to listen to things rather than read them. What is it strive to be is a little slightly different because what it really strives to be is it strives to be this platform with the mission of expanding access to content.
3:46 Speaker 01 Okay.
3:48 Speaker 02 And I mean this in a variety of different ways, right? Some people just are able to consume content. You know, we have this whole debate in education, it's like, are you a visual learner and audio learner? What do you do? People just consume content better in different ways. I I'm a visual learner, I need to see things. So for me, actually, it's sometimes a little better to read the blog. But if we're just talking about like, I want to get a lot of information, podcasts are great because you can just do them while doing something else. There's a reason that that podcast functionality is so natively embedded in all these smart speakers.
4:16 Speaker 01 Yeah.
4:17 Speaker 02 It's just because like you're doing anything at home, just put on a podcast. So really what we're trying to do is podcast is the first instantiation of that, which is like, how do we expand access to content?
4:28 Speaker 01 Yeah.
4:29 Speaker 02 But it expands so much more, right? You know, instead of just going to, I don't know, we talked about this like blog to podcast. You can go blog to video. You can go podcast to blog. You can go podcast to Twitter. Like, the permutations are frankly endless. Basically, it depends on how many platforms there are that people consume things on. But that's essentially what we do. The use cases for this are pretty interesting. The one that we like, just see immediate value in, is this just this ability to translate the content that you already have into other forms of content. If we just stick with that blog post example again, right? You've written, so, so, you know, a lot of companies might have this content team that focuses a lot on producing quality blog posts. Blog posts, you know, they're good for SEO and whatnot, but they're not, sometimes
5:13 Speaker 02 they, you know, they don't really achieve a specific goal or outcome that you want. One thing we see that is really useful for podcasts is they actually carry a lot more weight in credentializing you as a thought leader or your company as a thought leader. So, but, but like, you know, we spent the last 50 minutes trying to set up this room to record the podcast, right? So it's, it's not easy. And it's a very synchronous process, right? Me and you have to find the time to go and sit here and record this. You have to come up with questions. I have to come up with answers, right? But this ability to actually just like take the content that you have and transform it is pretty powerful. You know, and there's a lot of other use cases as well, which is just like podcasts, really all they are is like, like the final podcast, right? Like, the, the, the line between, or the difference between an audiobook and a podcast,
5:58 Speaker 02 I guess, just the, the format and the length.
6:00 Speaker 01 It's an, it's an MP3 on a.
6:02 Speaker 02 It's an MP3 with someone or something speaking, right?
6:05 Speaker 01 So I've actually played around with a lot with this stuff, by the way. So I've, I've, I've done music only podcasts where you just listen. Tiesto has been podcasting for 15 years. Every single week, just DJing from his house.
6:34 Speaker 02 It's just basically a radio show.
6:36 Speaker 03 It's great.
6:36 Speaker 02 It's just radio.
6:37 Speaker 03 A sync radio.
6:38 Speaker 02 Yeah.
6:38 Speaker 03 Yeah.
6:39 Speaker 02 So it yeah, so it's super interesting. But podcasts like, okay, like ignore the word podcast and just think of what we do, which is like, we help you create audio content. Super valuable for anyone who just needs that. If you can imagine a world and we're just like, I don't know how to call them, like Calm or Headspace or any of these things.
6:56 Speaker 03 Hi, and welcome to day one of Take Ten. Over the next 10 days, I'm going to be showing you how to get a little bit more headspace in your life. The starting point is just to get familiar with this really simple and easy to learn exercise. And then just commit to doing it each day. But remember, this is your 10 minutes. So all you have to do is sit back, relax, and allow your body and mind to unwind. To begin with, once you're sitting comfortably, I'd like you just to gently close your eyes.
7:24 Speaker 02 They can do a lot of their meditation like that, super quickly. What you can get to is a point where you're doing like these super personalized things.
7:30 Speaker 03 Yes.
7:30 Speaker 02 Right? Because you just have the ability to scale the content production so quickly. Same with educators. I think there's actually, at this point, there's a few YouTube channels at this point that are all based on synthetic voices.
7:40 Speaker 03 Yeah.
7:40 Speaker 02 That produce a ton of educational content. The problem with podcasts is podcasts just have a slow adoption rate.
7:48 Speaker 03 Yes.
7:49 Speaker 02 You're listening to a thing for an hour, right? Like, we as a generation don't have much time spent.
7:53 Speaker 03 Yeah.
7:54 Speaker 02 It's the time and the attention spent.
7:55 Speaker 03 Yeah, yeah.
7:56 Speaker 02 Like TikTok, give it to me in 30 seconds where I think why clips are taking over.
7:59 Speaker 03 30 is too long, man.
8:00 Speaker 02 30 is too long, 10 seconds.
8:02 Speaker 03 Seven.
8:02 Speaker 02 10 seconds with captions, I need to read it, and good. So what we also do is actually, and this is kind of still a beta feature and we're working to improve it, but like, we also, you know, let you take that podcast and then clip it into an audio, a video that you can go and share on socials, right? So it's this ability to take one form of content, produce it in a bunch of different ways that serve different purposes and be able to distribute it, basically.
8:22 Speaker 03 Yeah, excellent. I want to go through features so that people have a high level overview of what you offer. So I think at the core, it is basically two things. One is you generate scripts.
8:32 Speaker 01 And that's optional, obviously, if you if you want to just write the script yourself, you can write a script yourself, but most I think most of your users would generate a script. And then two is from that script, you create use AI voices currently using eleven apps. Is that is that the rough flow? That's like the really core, basic.
8:49 Speaker 02 That's the core, basic. Obviously, there's a lot of plumbing on top of it, but that's the core.
8:52 Speaker 01 And then you offer video clips for YouTube, you offer twenty eight languages that you can produce. You offer show notes production and podcast hosting too, so they don't have to host it on like Anchor. Don't host it on Anchor, by the way, if people don't don't host it on Spotify, don't host it on Apple Podcasts. It's these these people don't respect the RSS feed. Anyway, I have very strong feelings about like preserving the sanctity of the RSS feed. For open podcasting and all these Spotify's of the world want to close the podcasting ecosystem. So I have this tirade about them. But yeah, those are your top level features on your on your landing page. Anything that you highlight to go deeper on.
9:30 Speaker 02 Yeah, I think those are the the top level ones. There's also it's basically just a lot of like also ancillary tooling that goes around all of this to just make it easier. The goal is like every time we speak to a customer or someone who's thinking about it, they're like, yeah. Literally yesterday, I was speaking to a potential customer and they're like, yeah, I just, you know, I want to make sure this isn't a distraction because we don't have that much time to do this. Yeah, and really, the whole point is that this doesn't take time, right? The whole point is to provide all the rails that make this not take time. And this comes with a million different things, right? Like we, you know, sometimes the AI voices don't really know how to pronounce a word. So you have a pronunciation feature, go and define how that how you want that word pronounced, then it'll take care of it.
10:09 Speaker 02 If you are, we obviously have that hierarchy of like a podcast, an episode, and then all of that gets published in RSS feed that you can just upload to Spotify and we'll host that for you. But what you also have is just like, you know, maybe you want some defaults, right? Every podcast needs some defaults.
10:24 Speaker 01 Intro outros?
10:25 Speaker 02 Intro outros, the music, the speakers.
10:28 Speaker 01 Yeah.
10:28 Speaker 02 We're working on adding templates for the kind of podcasts that you're doing instead of it just being this narration style. You can just do an interview style podcast and a few more features. But basically, there's a lot of like tooling that just makes this a very useful usable product for podcasts.
10:41 Speaker 01 Yeah, you have you said you have a hundred creators publishing with you?
10:44 Speaker 02 Yeah, so you know, the interesting thing is if you write a newsletter. I mean, I don't know, my emails, newsletters at the moment, sometimes I just want like the recap of it.
10:54 Speaker 01 Yes.
10:54 Speaker 02 Audio form is just for some people easier. If you're on, you know, commuting or whatever, you can just listen to it. Um, so our folks actually just convert their newsletter, it takes like two minutes.
11:04 Speaker 01 Yeah.
11:04 Speaker 02 Put the text in there, voila, you have a, you have an audio version of your newsletter that you just published to Spotify.
11:09 Speaker 01 Yeah, and I am a newsletter writer, uh, and, uh, I clicked around and wanted to basically just chuck my RSS feed in there.
11:17 Speaker 02 Yeah.
11:17 Speaker 01 And I think I gave that feedback exactly to you guys like four months ago, uh, or three months ago, and it looks like you've already shipped it.
11:24 Speaker 02 Yep, uh, well, I'm announcing it basically here today, which is as of today, we've actually built a Zapier integration. Uh, and we have a blood bunch of blogs on our website to kind of show you how to do this, but what you can now do is as soon as you publish a newsletter, it goes on your RSS feed, we'll pick up the newsletter from your RSS feed automatically.
11:42 Speaker 01 Yeah.
11:43 Speaker 02 And just publish an episode for you.
11:44 Speaker 01 Yeah. Question, what if I change something after I publish?
11:48 Speaker 02 So you don't have to publish, it'll basically just generate, do all the work for you, and then you can go in and kind of modify it a little bit and then publish.
11:54 Speaker 01 Makes sense.
11:55 Speaker 02 We also have scheduled publishing so that you can, I don't know, maybe you want to release it a few hours later.
11:59 Speaker 01 Yeah, the professional podcasters that I've spoken to say that that is very important. I personally don't care, like it shows up in my feed or not, I don't care when it drops.
12:08 Speaker 02 Yeah.
12:08 Speaker 01 Anyway, so you do, you do want to basically time it, like if you're basically targeting like a commute for like the US time zone, you want to be like, oh, 8:00 AM, you know, Pacific, where people are driving into work, then you like show up at the top of the reverse chronological feed.
12:23 Speaker 02 Yeah.
12:23 Speaker 01 I feel like that's too much tactics.
12:26 Speaker 02 You know, and that's a good point, I think it depends a little bit on your audience and what you're building, but I do think, so, I don't want to undermine like the importance of consistency in podcasting.
12:35 Speaker 01 Yeah.
12:35 Speaker 02 Right, like you, whether that consistently, consistency literally translates into.
12:41 Speaker 02 I publish at eight AM every single day.
12:43 Speaker 01 Yeah.
12:44 Speaker 02 Or I just publish every single day, or you know.
12:46 Speaker 01 Yeah, so like.
12:47 Speaker 02 There is a huge importance in just like making sure that what you're publishing is always.
12:51 Speaker 01 Yeah.
12:51 Speaker 02 It's there. People need to know that your brand is like kind of constantly pushing stuff.
12:54 Speaker 01 Yeah, so for a lot of people talk to me are interested in like, what's my advice on content creation? Yeah, at least once a week.
13:00 Speaker 02 Yeah.
13:01 Speaker 01 Whatever you do, and I don't care when you do it, just once a week, put something out. But I do notice that in specifically in the podcasting field, and you you talk about this in the next point. Daily podcasting is the meta game that is, I think, doing extremely well.
13:14 Speaker 02 Yeah.
13:14 Speaker 01 Especially because I think the Apple Podcast list bias is for daily.
13:18 Speaker 02 Yeah.
13:19 Speaker 01 Because obviously the downloads will be higher.
13:21 Speaker 02 Yeah.
13:21 Speaker 01 So daily podcasts is kind of rank higher more and obviously because you're daily, you also do shorter podcasts, which guarantees that more people listen to you.
13:30 Speaker 02 Yeah, yeah, I think I think the fact that so obviously we do the hack news recap. The fact that we did that and that it is daily actually just helped us reach that top thirty tech podcasts on Spotify.
13:41 Speaker 01 Yeah, that was mostly because you were on HN, right?
13:43 Speaker 02 We did launch, but obviously the fact that like you just publish a lot of content, you're just gonna get a lot more like it's a statistics thing, right? Obviously, I think they do it by like total time listened as well.
13:51 Speaker 01 Yeah, yeah.
13:52 Speaker 02 But you know, the fact that it's daily is just not overwhelming. Again, we don't have that much of a like an attention span anymore.
13:58 Speaker 01 Yeah, yeah, that's true. That's true. Yeah, I love it. I listen to it every day. Excellent. Awesome. I think that's a really good overview. Then you also produce three in-house podcasts.
14:08 Speaker 02 Yeah.
14:08 Speaker 01 Hack News Recap, Product Hunt Daily, and PG essays.
14:12 Speaker 02 So we dropped the Product Hunt Daily.
14:13 Speaker 01 Oh, okay.
14:14 Speaker 02 So we do the Hack News Recap and the PG, I think are the two most popular ones.
14:17 Speaker 01 Yeah.
14:18 Speaker 02 We're constantly experimenting with new internal or like podcasts that we publish.
14:21 Speaker 01 Yeah, yeah. What are your other, you can tease a little bit. What do you think about?
14:25 Speaker 02 Tease a little bit. Well, I really like Reddit.
14:27 Speaker 01 Yeah.
14:27 Speaker 02 I'd love to listen to some of the Reddit things going on there, but instead of like reading them. I don't know. It's always just a notification that I get. I'm like, oh, this sounds interesting. But I don't know. You can do it like per sub Reddit that you care about.
14:37 Speaker 01 Yeah.
14:38 Speaker 02 A few things like that. I love life pro tips.
14:40 Speaker 01 Yeah, life, I see.
14:41 Speaker 02 Right, like super interesting things or Wall Street Bets or whatever you're into.
14:44 Speaker 01 Yeah, well, the problem with these things is that a lot of them create involve images and memes.
14:49 Speaker 02 Mm hmm.
14:50 Speaker 01 Which you cannot consume.
14:52 Speaker 02 Well, yes, we cannot consume. This is like a simple.
14:55 Speaker 01 Yeah.
14:56 Speaker 02 We can't consume that at the moment, but. You know, maybe on the beat, a few weeks down the line when that video feature of ours gets a little better, you can actually start shipping like that.
15:05 Speaker 01 Okay.
15:06 Speaker 02 Anyway.
15:07 Speaker 01 Yeah, yeah. And so, and I'll just feed you an idea. To keep up on AI, a lot of stuff actually happens in Discord. And there's way too many Discords.
15:15 Speaker 02 Way too many, and they're way too active.
15:16 Speaker 01 Yes. So I've actually built a little feed for myself that scrapes a bunch of Discords.
15:22 Speaker 02 Yeah.
15:22 Speaker 01 And creates a daily newsletter for myself.
15:24 Speaker 02 Amazing.
15:24 Speaker 01 And I have thought about turning it into an audio feed, but, and this is the problem for Wondercrafts. I read better, I read faster. I scan up and down faster than I listen, right? And there's just too much noise in Discord for me to listen as audio format. Your Hacker News stuff is very high signal because obviously you're folding, right?
15:44 Speaker 02 Because we haven't done the curation like Hacker News did.
15:46 Speaker 01 Exactly. That's why it's guaranteed to be good.
15:48 Speaker 02 Yeah.
15:49 Speaker 01 Whereas for Discord, it's a bunch of junk.
15:54 Speaker 02 But I do think there's something similar, like Reddit also does the curation. It's not us doing it, right?
16:00 Speaker 01 Yes, yes, yes.
16:00 Speaker 02 If you are only the hot post.
16:01 Speaker 01 Still a little bit noisier. So I don't know if you know, I have, I was a moderator of the React Reddit for four years.
16:07 Speaker 02 Nice.
16:08 Speaker 01 So I've seen a bunch of stuff and like, I know it's noisier than like a Hacker News, but still, still pretty good.
16:14 Speaker 02 Yeah, yeah. So we do Hacker News, we do PGAs. I think PGAs are also super interesting. I listen to them all the time because I, well, first of all, I actually think they're pretty well-produced. Like, we do a good job.
16:23 Speaker 00 One of the most common types of advice we give at Y Combinator is to do things that don't scale. A lot of would-be founders believe that startups either take off or don't. You build something, make it available, and if you've made a better mousetrap, people beat a path to your door as promised. Or they don't, in which case the market must not exist. Actually, startups take off because the founders make them take off. There may be a handful that just grew by themselves, but usually it takes some sort of push to get them going.
16:52 Speaker 02 Like, I don't know, if we're coding someone, we'll like use a different voice.
16:56 Speaker 01 Yeah, yeah.
16:57 Speaker 02 I think it's just what produces, and I also think. You know, the essays are so seminal to like, everyone in startups reads them.
17:04 Speaker 01 Yeah, it's actually got me to read more PG essays than I would have otherwise.
17:07 Speaker 02 Yeah, his, his last one.
17:08 Speaker 01 Mission accomplished.
17:09 Speaker 02 I don't know if it was the last one, well, the time was.
17:11 Speaker 01 How to do great work?
17:12 Speaker 02 How to do great, that wasn't like, that was a one hour podcast.
17:14 Speaker 01 Yeah, oh my god.
17:15 Speaker 02 No one, I could, I did not read it, I just had to listen to it.
17:18 Speaker 01 Yeah, yeah.
17:18 Speaker 02 I, I think actually, if I'm being honest, I think the motivation for PG essays was just like, I, I need this.
17:23 Speaker 01 Yeah, yeah, yeah. For, for that one, if it's like one hour, I, I would have actually appreciated a, a segmentation. Hey, high level, you know, I know this is about to be an hour.
17:33 Speaker 02 Yeah.
17:34 Speaker 01 But like, there are three main high level things, and then keep that in your mind. And then go like, part one, blah, blah, blah, part two, blah, blah.
17:40 Speaker 02 I think we, so we do that to some extent, and like, we, we produce like, chapters, I guess.
17:44 Speaker 01 Yeah.
17:45 Speaker 02 So you can just look at them.
17:46 Speaker 01 Yeah.
17:47 Speaker 02 Probably could, could do a better job like introducing it, but we do try to like, not play around with the PG essays.
17:51 Speaker 01 For sure, yeah, I mean, it's, you know how much work he puts into, into those things.
17:55 Speaker 02 Kind of ship it as is.
17:56 Speaker 01 Yeah. I'll tell you about one more that, one more daily, not daily, but frequent AI-generated podcast that I listen to apart from you guys. It's just papersread.ai, and I'll recommend it to anyone listening as well.
18:07 Speaker 04 Papers Read on AI with Rob, keeping you up to date with the latest research.
18:15 Speaker 04 Attention is all you need. Authored 2017 by Ashish Vaiswani, Noam M. Shazir, Nicky Parmar, Yakub Uskarite, Leon Jones, Aiden N. Gomez, Wukash Kaiser, Ilya Palasukin. The dominant sequence transduction models are based on complex recurrent or convolutional neural networks that include an encoder and a decoder. The best performing models also connect the encoder and decoder through an attention mechanism. We propose a new simple network architecture, the transformer, based solely on attention mechanisms, dispensing with recurrence and convolutions entirely.
18:49 Speaker 02 Super interesting, actually, I've come across that.
18:52 Speaker 01 It's by this guy Rob, and I've tried to look him down, he doesn't want to be found. Anyway, but the selections are very good. I think you guys could do a better job than him.
18:59 Speaker 02 Yeah. What's up, Rob?
19:02 Speaker 01 Yeah, well, it's because he converts PDFs to podcasts, right? And the problem with academic PDFs is a lot of references. You know, like buy it all 2022 and then like hitters and then a table and then reads the table when you don't need to read the table, you know? That kind of stuff. I think better engineering there from you guys would beat him. And I need that, so feature request.
19:26 Speaker 02 Working on it.
19:27 Speaker 01 Okay, the final feature related thing is the thing that we're announcing today as we release it.
19:32 Speaker 02 Super, super excited about, but Warcraft now does video translation.
19:35 Speaker 01 Okay.
19:36 Speaker 02 And dubbing.
19:36 Speaker 01 Okay. Why do people want that?
19:39 Speaker 02 Again, let's go back to our mission. We're trying to expand access to content. I think, don't quote me on this again, like, I don't know who actually knows the internet, but like 60% of the internet is in English. You know, what if you don't speak English? You're automatically disbarred or kind of excluded from all the content that's produced. And thanks to all the advances that have just recently been made, we can actually make this super easy to dub this in other languages.
20:03 Speaker 01 Yeah.
20:03 Speaker 02 So we're super happy to announce this feature. We're super excited. We've been working on it for a really long time. But now, basically, everyone, go on our platform, upload your podcast episode and see the dub for yourself. We'll use your voices. We'll, you know, completely convert it. Make sure it's aligned with the, if you have video, we'll make sure it's aligned.
20:21 Speaker 01 Yeah.
20:22 Speaker 02 And voila, just publish it.
20:24 Speaker 01 And specifically for video, so like, well, obviously the thing that's going around is the HGN thing, which changes your lips.
20:31 Speaker 02 So we don't do lip sync at the moment?
20:32 Speaker 01 Yeah.
20:32 Speaker 02 Could be another feature that we work on.
20:34 Speaker 01 Yeah, because you're primarily podcasting, which is no video.
20:37 Speaker 02 Well, primarily no video, but I think we still basically, if you do have a video, we'll still align it to the chunks.
20:42 Speaker 01 You still align it. Okay, so the hard problem is the aligning.
20:44 Speaker 02 The alignment is the difficult bit, you're right.
20:46 Speaker 01 The actual like. As with all things in AI.
20:49 Speaker 02 Yeah. Again, overloading the word alignment. We don't do the lip sync.
20:53 Speaker 01 Yeah.
20:53 Speaker 02 Yet.
20:54 Speaker 01 It's kind of a gimmick, yeah.
20:55 Speaker 02 It's like, it's not super necessary. If you really just like listening to a podcast and you actively want to listen to it in a different language, then you're more.
21:01 Speaker 01 So what are you aligning to?
21:04 Speaker 02 Basically, the chunks where the speakers are speaking. So you won't have an instance where you'll have me as the dub speaking while the camera is on you.
21:13 Speaker 01 Ah.
21:14 Speaker 02 Right? So it's basically just like the speaker turns.
21:16 Speaker 01 So you will like, if it happens to be a little bit longer, you'll speed it up a little bit?
21:20 Speaker 02 Yeah, we do a little bit of, you know, trickery there, but we get it aligned so that when you're speaking, it's you and when I'm speaking, it's me.
21:25 Speaker 01 Cool. I wanted to talk a little bit about the origin story.
21:28 Speaker 02 Yeah.
21:28 Speaker 01 Cause you flagged that Mooncraft was actually a big part of, Moonshot was a big part of your arriving at this idea.
21:36 Speaker 02 Moonshot was not a tech product. It was a legal product. It was a regulation. It was like, hey, now you can invest in people. The reality is like, I think we were, we just found out the hard way that we were building something people did not want. And we realized like, we're building something that isn't our strength. So we're like, when we decided to pivot, we were like, we need to do something technical.
21:53 Speaker 01 Okay.
21:54 Speaker 02 The story from there just became like, okay, cool. Let's, there's some ideas. What are we going to track? Which one do we have the most conviction in? We ranked them. And Wondercraft was the one we had the most conviction in. It was this idea again, expanding access and just translating or your ability to produce content. So producing content in one format and then taking that to all the other formats. So we built that. We built the podcast builder, super quick prototype, because I think at this point, anyone pivoting or hard pivoting or considering it, the name of the game isn't like to get attached to your idea. Just like, actually, you should be trying to invalidate this idea as quickly as possible. So get it out there and let people tell you it's piece of shit.
22:30 Speaker 01 Okay. So what did you do to get it out there?
22:31 Speaker 02 So we built it out, we built out a little like UI, literally no authentication. It was a form where what you guys see now on our platform, which is like the content script page, blah, blah, blah, blah. It was one page. Like you click and it was the most janky React stuff that we had, zero authentication. So in theory, if people found it, they could just,
22:47 Speaker 01 Yeah.
22:48 Speaker 02 you know, produce as much audio as they wanted. So we have a feature on our app, which is like test with examples. As soon as you log in, it says test with example. And the whole point of that was like to, between login and audio generated, how few clicks does this take?
23:01 Speaker 01 Yeah. And yours was?
23:02 Speaker 02 Like two or three clicks. It's like test, create podcast.
23:05 Speaker 01 You like pre-filled everything?
23:06 Speaker 02 Generate script?
23:07 Speaker 01 Yeah.
23:07 Speaker 02 Generate audio.
23:08 Speaker 01 Yeah, yeah, yeah.
23:09 Speaker 02 Right? So, and I'll argue that was still too many clicks. We should probably put something on the actual landing page so that like you can see.
23:14 Speaker 01 Yeah, now you have just a player there, right? You can just kind of listen to that hackneyed daily thing. But what signals did you get from doing that?
23:20 Speaker 02 The signal that we got is that someone picked it up on Twitter and just like, you know, all these like AI.
23:24 Speaker 01 Influencer voice.
23:26 Speaker 02 So someone picked it up, posted it, we started getting a ton of inbound. So we're like, holy shit, let's just like paywall this. So we just like, again, the jankiest Stripe integration, which was basically like, we have an app with a Stripe integration. This was just to ship it within the hour. We have an app with a Stripe integration.
23:43 Speaker 01 Yeah.
23:43 Speaker 02 That once you click, then takes you to a different app hosted somewhere else. So that one was still unauthenticated. It was, it was, it was hilarious.
23:50 Speaker 01 Security by obscure.
23:51 Speaker 02 It was hilarious, right?
23:52 Speaker 01 Yeah.
23:52 Speaker 02 But, but we basically just made that like three K in, in, in one day.
23:57 Speaker 01 One day?
23:57 Speaker 02 Yeah. We charged around like fifty bucks. We didn't even, literally didn't think about it. We just charged fifty bucks and people paid and we're like, okay, well, there's something there.
24:04 Speaker 01 Yeah.
24:05 Speaker 02 So then.
24:05 Speaker 01 Fifty bucks for one?
24:07 Speaker 02 For a month. I, we just, we were just charging fifty bucks a month. Nothing. Like, just like, will someone?
24:12 Speaker 01 And you were just like on one like V P S somewhere?
24:14 Speaker 02 Yeah. Will someone pay for this? We were like on one E C two instance.
24:16 Speaker 01 E C two instance, yeah.
24:17 Speaker 02 You know what I mean? Like, it was janky. We're like, just someone needs to pay for this before we move further. Someone did. People did.
24:23 Speaker 01 Yeah.
24:23 Speaker 02 So then we're like, okay, cool. This is interesting.
24:25 Speaker 01 Interesting. I'm gonna move up the question that we said was gonna be the meatiest question of this interview. Uh, so you, you, you chose this out of your list of ideas.
24:34 Speaker 02 Mm hmm.
24:34 Speaker 01 And, uh, this is one of the things that a lot of A I founders are worried about, right? So the, the, the framing of this is, are you worried that you're a thin wrapper around eleven labs? What is your mode?
24:44 Speaker 02 That's the seminal question. I think frankly, everyone in an A I startup should convince themselves of this. Don't listen to me and like, just make sure from first principles that you can derive this. But I guess I would start by first saying, what is a mode? What is defensible? In theory, if we're just taking it, and this is trivially true, but the fact that someone built it means someone else can build it. Right? So modes tend to just be built around like, you have a lot of network effects, or you have a really good product for this use case, or, you know, something like that. And I think typically when people ask this question in the A I context, they're thinking of like, okay, you're a thin wrapper, you're an application layer thing. As opposed to you're one of the like underlying technologies or, or A P I's that people use.
25:27 Speaker 02 Cool. I, I, I think that's fair, but I think the reality is that like, yeah, these A P I's exist and they probably do serve a million different use cases. But they're not built to serve these million different use cases. So whenever you ask the question of mode, it's always, has to be with. The perspective of who is the user I'm building this for, right? I I can use chat G P T to do half of my writing. But you know, but I don't know Jasper claims that they do this much better for marketing. So it's tailored, I actually, you know, don't quote me on how well they're doing after chat G P T came out because they were really big before. Yeah, there's some negative data points, but I'm sure they. I know, but the point is like you're you're making this easier. We we make creating a podcast easier and there is tooling there. There's we help you, we can post it directly through us.
26:11 Speaker 02 We have the tooling around, you know, setting the intros and the outros. We have the music, we have an editor, all these things are also getting just much more more more and more developed. We're building templates so that you can do different style of podcasts. So the idea is if you're trying to start a podcast, yeah, don't go to a generic text to speech engine, come to us. Yes. And the reality is that we then can in a very opinionated way actually select which text to speed engine we want. Right, so that we actually have just like, in my mind, it's the application layer fundamentally that, you know, people use. And then all these A P I layers are what developers use to build products on top of them. Right, it is, I appreciate that it is like a seminal and and really hard thing to wrap your head around.
26:52 Speaker 02 Especially if you're like about to invest in a company, it's like, will they actually just be defensible and be able to grow? Yeah. And yes, there's no doubt that companies can do this. The question is just like, are you building the right product for the right use case? I think particularly if you're like always framing your company as a as an A I company, then you're you're putting the like the carriage before the horse in the sense that you focused on the implementation rather than the use case. Focus on the use case and then build a product for it. Yeah. Right, because fundamentally, you know, any of the SaaS's that exist, think like more traditional SaaS. What's their mode? The technology everyone has access to, so they just pick the thing that does it better than than the other. Now, that mode question is super interesting because I think you should actually flip it around, which is what is your mode as an A P I?
27:37 Speaker 02 Yeah. Right, so chat G P T like, yeah, fine, they had a first mover advantage and I think, you know, by no means, this is my opinion, but by no means was Google like caught off guard with this, right? It just, Google has some, half the technologies that Google invented are actually what's used to power all these transformers.
27:58 Speaker 02 But, you know, it went against Google's strategy, maybe, to like be the first mover in this because they'd cannibalized their own market, whatever it was, I'm not sure. But, yeah, OpenAI's mode is that they paid for the training bill.
28:10 Speaker 01 Yeah.
28:11 Speaker 02 So they just have a good model. Cool. People now know that that's valuable.
28:13 Speaker 01 And they hired like very top.
28:15 Speaker 02 Superstar, sorry, obviously, obviously, not taking that for granted. But like, they, you know, assuming everyone can do the hiring and that these people exist, they paid the bill and they were the first to launch this. But now people know it's a thing, so people are going to launch similar APIs. So what is your mode as a, as an API? So it's just an, it's an existential question. It's like, how do you do, how do we defend any of this?
28:36 Speaker 01 Yeah.
28:37 Speaker 02 And you do this frankly by being probably better just as a product. Again, the product is always with the perspective of who's your customer that you're selling it to. And the other thing is frankly that, like, let's not forget, the market is huge. There is space for everyone, right? If you've managed to like, if there's four good products out there in any specific thing, the market is huge and they're all going to be able to, you know, make a living out of it.
28:59 Speaker 01 By the way, that was a really good answer. Thanks for taking that hit on.
29:02 Speaker 02 I had to answer that question way too many times.
29:05 Speaker 01 This is not the first time. And, but I think it's actually, you know, having been an investor, it is more important for you to answer that question authentically for yourself.
29:12 Speaker 02 Yeah.
29:12 Speaker 01 Because you're the one spending your time on this.
29:14 Speaker 02 Yeah.
29:14 Speaker 01 We're just giving you money. It's not, not that big of a deal. My favorite quote, actually, I went to an early, like, pre-HGPT forum with Sam Altman. And I had this video advice from Sam that said like, Facebook had no mode. And they just built and got the network.
29:31 Speaker 05 Actual technological modes in the history of Silicon Valley, almost all of them are product, network effect, distribution modes, something like that. Let's say Google is, was at least at some point a legitimate technological mode. Um, I can accept that one, but like, that's not why I would say Facebook is like a giant business. It's not why I would say Twitter is such as it is a giant business. Um, I, I think there are a lot of ways to build a great business. And the big lie of like the tech industry is that you get there with differentiated technology. It's rare.
30:06 Speaker 02 But frankly, also, Facebook was building something back then, which is kind of ludicrous.
30:10 Speaker 01 Yeah.
30:10 Speaker 02 Cool, you're building a social media app. Okay, how big can it be? Right? Like, how big can the internet be?
30:14 Speaker 01 Yeah.
30:15 Speaker 02 You know what I mean? All of a sudden, it's this behemoth. So it's like, yeah, the fact that it was built, again, trivially true, but the fact that it was built means it can be built.
30:22 Speaker 01 Yeah.
30:23 Speaker 02 By anyone else. So you, there is no such thing as like a absolute true moat.
30:28 Speaker 01 Yeah.
30:29 Speaker 02 The question is how well, how quickly, how, you know, how much earlier than everyone else did you get there? And a million other things as well.
30:35 Speaker 01 Yeah. Cool. The audio generation, you, you use Eleven Labs? What makes a good podcast voice, right? You have a bunch of options that I clicked. And in my minds, I like a deep voice. I like the Morgan Freeman's. You don't have that many deep voices. Do, do we want like, is there such a thing as a high-energy voice? You also insert breaths? Eleven Labs has also advertised that they have a AI that can laugh, which I think is fun, important. Basically, what makes a good AI generated audio?
31:01 Speaker 02 Yeah, it's, it depends again on the perspective. Everything is kind of answered with the, the frame of reference that you're looking at. If you like a deep voice, A, that's kind of a personal preference.
31:11 Speaker 01 Yeah.
31:11 Speaker 02 And B, it just kind of depends on, on the thing. So if you, I don't know, let's do, say you're doing something like meditative or kind of affirmations or, or something that like encourages people every day. You probably do want a slow deep voice, something relaxing. You're doing the Hacker News recap, like.
31:25 Speaker 01 Cheerful.
31:25 Speaker 02 We picked Anna, who's like our default voice.
31:28 Speaker 01 Yeah.
31:28 Speaker 02 Because.
31:29 Speaker 01 I have an, I have an attachment to Anna.
31:31 Speaker 02 Yeah, we all do.
31:31 Speaker 01 Yeah.
31:34 Speaker 00 Ah, thanks, guys. You're so sweet. As an AI language model, I cannot have attachments or needs or desires or favorite humans, but you guys are at the very top of the humans I am not attached to.
31:45 Speaker 02 She's just like news anchoring style, very professional, very form, very neutral. Very neutral.
31:49 Speaker 01 Yeah.
31:51 Speaker 02 So it depends really on like, what, what makes a good voice? It depends on what you're doing.
31:56 Speaker 01 Right.
31:57 Speaker 02 There's a few things, but like, if you're doing an interview, I think it also just frankly, then you get into a question of what makes a good podcast. Well, the good podcast is like, I think it's also kind of a personal question, which I haven't, or probably there's a general trend that I'm yet to decipher. But like, yeah, you probably do want a little bit of humanity in there. You want a stutter.
32:16 Speaker 01 Yeah.
32:16 Speaker 02 You want some pauses, right? I'm speaking, I don't speak in complete utterances. I have an utterance and then I pause it all and then I speak again and so on. Laughs and something to make it human.
32:25 Speaker 01 Yeah.
32:25 Speaker 02 It's kind of overlaying of the two, if you have two speakers, this like exchange, right? Like, I will be speaking, if you look at the transcript of this episode, we probably overlap in when we're speaking.
32:35 Speaker 01 And that's fun.
32:35 Speaker 02 And that's actually interesting, right? Because it is a conversation.
32:38 Speaker 01 Shows the sign of excitement, especially in our studio when we're three people. And if we're all talking at once, you know it's good.
32:43 Speaker 02 Yeah. I don't like this zoomification style where like if you're on a big zoop and a big zoom, like only two people can speak, the second, more than two people try to speak.
32:52 Speaker 01 Yeah.
32:52 Speaker 02 It's a disaster. So I think it frankly just depends on what you're doing. We are like, yeah, at the moment, we're really good at like doing this narration stuff, but I think we're, we are building a lot of functionality and tooling to just make this, kind of this like multi-host thing a more of a reality.
33:07 Speaker 01 Okay, okay. I would say, you know, objectively, if it was a friend in the company, not that important, you know? So thi- this comes down to how human should your users try to be? Because I'm fine with Hacker News Daily making mistakes because I know it's AI-generated.
33:25 Speaker 02 Yeah.
33:26 Speaker 01 Right? Uh, I would be less fine if you were not upfront. But then like you'll make mistakes, like pronunciation mistakes. I actually have a clip that I wanted to play you on September 8th, and I was taking a lot of breath.
33:38 Speaker 00 The app is loaded with unparalleled features such as high-resolution video editing, a multi-touch timeline, live motion effects, and performances complemented by atmospheric audio elements. Emphasizing its compatibility with iPad and Apple Pencil, Procreate Dreams welcomes the next generation of creators and pushes the boundaries of modern artistry in an instantaneous, user-friendly environment.
34:02 Speaker 00 In the comments, many users expressed enthusiasm for.
34:06 Speaker 01 She was very out of breath.
34:06 Speaker 02 Damn.
34:07 Speaker 01 I was very worried about her. She was hyperventilating. I was like, in our UK? Like, anyway, so, basically, I think if you disclose upfront that you're an AI podcast, then people will be like, oh, OK, I tolerate that mistake and I use you for information and not for believing that there's some human on the other side that I might meet someday. But if you're investing so much effort into being real, then your end goal is you have to lie to your users.
34:32 Speaker 02 I don't think the investing in being real is for the purpose of deception as much as it is for the purpose of making it slightly pleasant to read about. I think on, like, we do claim, we do say on our Spotify page that, like, this is an AI-generated podcast.
34:45 Speaker 01 For now, but.
34:46 Speaker 02 As in, yeah, yeah, so, there's two things. I think if you want to be smart about this, you should say that this is AI-generated content. The second people find out that it's not you, the backlash is going to be big.
34:56 Speaker 01 Yeah.
34:57 Speaker 02 Right? Because it will be interpreted as deception. So you should do this just to be smart. I don't think there's a point in lying, especially if the content that you're putting out there is just like, this is informational for you, so, like, consume it, this was efficient, this helped us put it out there. The second thing is frankly, I don't think it's up to you whether you tell them or not. Very, very soon.
35:16 Speaker 01 I don't know.
35:17 Speaker 02 Google is just going to mark things as AI-generated. So I think there's a new thing. I saw, like, a quick YouTube video about it, so I don't know what the exact terms and conditions are, but, like, YouTube has, I think, released a new monetization rule and it does mention something about AI-generated content. Right? So there is, like, it's not up to you anymore. You're gonna, people are gonna know that this is AI-generated, so I think it's just in your interest to say that you're AI-generated. Ain't no shame.
35:42 Speaker 01 Yeah, yeah, no shame at all.
35:44 Speaker 02 Because fundamentally, what we do relies on the premise that you have done some content. We don't generate our own content.
35:50 Speaker 01 Yes.
35:50 Speaker 02 Right? We don't synthesize our information.
35:52 Speaker 01 Yes.
35:52 Speaker 02 It assumes that you've, you know, written a blog post, done an actual podcast, or have some artifact on which you want to base what you're feeding through Wondercraft.
36:01 Speaker 01 Yeah. And you you said in some of your material that I've seen before that you are interested in watermarking all your stuff. You haven't done it yet, but whenever there's a standard for doing that, you will do it.
36:11 Speaker 02 Yeah. I think the thing that's this is blocking was like the standard. I'm not super up to date on like what the what the work on this is.
36:18 Speaker 01 I think OpenAI will probably like.
36:19 Speaker 02 But like there just needs to be a standard so everyone can interpret it.
36:23 Speaker 01 Yeah, cool. Awesome. Great. I would I wanted to dive in a little bit on tech options.
36:28 Speaker 02 Yeah.
36:28 Speaker 01 And then zoom out to to just you asking me questions.
36:32 Speaker 02 Yeah.
36:32 Speaker 01 So TTS options, we we talked a little bit about eleven labs. I would also say as a podcaster, the leading competition to you guys. I know it's not exact competition, but it's the script.
36:42 Speaker 02 Yeah.
36:42 Speaker 01 Because they have overdub.
36:44 Speaker 02 Yeah, I think D-Script is really good and they're definitely like solid company. I've used their video editor before. It's it's great. The overdub thing is super useful. I think the it's really creative to like have edit videos by editing the transcript.
36:55 Speaker 01 Yes.
36:56 Speaker 02 Super, super creative, super user friendly.
36:57 Speaker 01 Would you would you build that?
36:59 Speaker 02 I think again, it's like we're not building for this like a building, we're building more for the purpose of the user.
37:05 Speaker 01 Yeah.
37:06 Speaker 02 Whatever users find more interesting. I think like what we're doing is we the use cases are slightly different, right? I think the the people that they're targeting is are are slightly different.
37:15 Speaker 01 Yeah.
37:16 Speaker 02 We do want to have a lot of like automation on the script side to also just like help out with the way you formulate your content or the way you pull your content, much more so than just the editing.
37:26 Speaker 01 The ingest. Yeah.
37:27 Speaker 02 Yeah.
37:27 Speaker 01 Okay, got it. And I just want to map out, here's how I think about TTS, text to speech. There's a big cloud options, Amazon Polly, Google Text to Speech and Microsoft Cognitive Services. As someone who is ex-Amazon, I'm very embarrassed by Polly. It sucks.
37:44 Speaker 02 Yeah, Google's isn't great, Microsoft.
37:45 Speaker 01 I'm sure you investigated all these things and you're like, okay, this is not serious. There is play dot H T, which is probably the other big Y C alum. Just quick two seconds on your thoughts on play the page T.
37:57 Speaker 02 Sounds good. I think doesn't sound super as good as a long labs, my opinion.
38:00 Speaker 01 Yeah.
38:01 Speaker 02 But I think.
38:01 Speaker 01 By the way, I have heard other founders tell me this as well.
38:04 Speaker 02 Yeah.
38:04 Speaker 01 And I don't know.
38:05 Speaker 02 They're.
38:07 Speaker 02 I think they have a, what's it called, a more comprehensive platform.
38:11 Speaker 06 Maybe you want to advertise your business on one of our lovely radio stations right here in Louisiana. You'll definitely need my charming voice to run your ads.
38:20 Speaker 07 But if you're coming northeast, you might want to ditch that southern voice from mine.
38:26 Speaker 08 Are you coming down under, mate? You can localize your content with an iconic voice like mine. I'm quite famous over here.
38:34 Speaker 09 And remember, Africa is an entire continent, not a country. And Kenya, for example, is emerging as one of East Africa's fastest growing economies. So use voices like mine for your content.
38:46 Speaker 02 Okay, as in they like, you know, they let you do this pronunciation, like they just have a lot of tooling around it.
38:52 Speaker 01 So different features.
38:54 Speaker 02 Quality in terms of the voice, eleven lap is still better.
38:57 Speaker 01 Yeah.
38:57 Speaker 02 I, I think they are releasing a new model. I don't know if they've released it already or not.
39:01 Speaker 01 The great thing about play dot H T is that you can clone your own voice or use existing high quality voices. It is crazy good. You cannot tell if these are human voices or machine ones anymore.
39:11 Speaker 02 Could be better. I, I don't know, but they do have some functionality there.
39:15 Speaker 01 Yeah. They also released the viral Joe Rogan, Steve Jobs interview from last year.
39:19 Speaker 02 Yeah.
39:19 Speaker 07 So you studied at Reed College and you dabbled in Eastern mysticism there, right? Do you still go back and look at Hinduism and Buddhist texts and things?
39:29 Speaker 10 Not texts and things. I actually took a course in that. I have a very deep belief that the people in the Indian subcontinent are most responsible for human civilization's current state.
39:40 Speaker 01 And on your landing page, you were like, this is something that Wondercraft will never do. AI content generating, AI content speaking to each other.
39:48 Speaker 02 Yeah, who wants to listen to that?
39:50 Speaker 01 Like apparently a lot of people.
39:52 Speaker 02 It's fun because it's a cool gimmick. I think it's like nice, nice viral material.
39:57 Speaker 02 I, I would never listen to like a synthetic Joe Rogan.
40:00 Speaker 01 Yeah.
40:00 Speaker 02 This brings us on to a little bit about the whole like content question with the proliferation of AI, which is like, okay, if it's this easy for me to create content that's like, you know, somewhat engaging, like all these AI songs. The Drake song. Well, okay, so if it's this easy, and it's just like, if it's this easy to generate content, well. Why will I listen to it? Like, I think we already suffer from the problem that there's an oversaturation of content.
40:27 Speaker 01 Here's my map of the market, right? There's speechify.com, which focuses on celebrity voices. I noticed that you don't have celebrity voices, probably because of licensing issues, right? Like, you have to pay them if you use their voice.
40:37 Speaker 02 Good at some point, but like, not a priority at the moment.
40:39 Speaker 01 Yeah, I really want a Morgan Freeman one.
40:42 Speaker 02 That's gonna cost.
40:43 Speaker 01 I know, I know. Mycroft AI, privacy-focused, run offline. Uh, probably not, not important for you. There is some interest in virtual characters for games. So Conv AI is the one that I had listed here. Did you look at the gaming market?
40:59 Speaker 02 Not deeply, to be honest.
41:00 Speaker 01 Yeah.
41:00 Speaker 02 But it could be an interesting one.
41:02 Speaker 01 Yeah, people are exploring that. There's obviously Heijen now, um, and, and that, that is it for as far as like, I can scope out the landscape. And then there's the open source systems. So Tortoise TTS, as far as I can tell, is kind of market leader in open source. There's PyTSX, Koki, it used to be Mozilla, and then Larynx.
41:19 Speaker 02 Yeah.
41:19 Speaker 01 Anyway, all these things. And then there's also sort of research-grade stuff coming out of the major big tech companies. You talked about Google Soundstorm?
41:26 Speaker 02 Probably the one I'm most excited about.
41:28 Speaker 01 Okay, why?
41:29 Speaker 02 Well, it's really good. You can like, check out the paper.
41:31 Speaker 01 Yeah, we'll, we'll play a clip.
41:33 Speaker 11 Did you hear about Google's paper on Soundstorm?
41:36 Speaker 12 Um, no, I must have missed it. What's, what's it about?
41:39 Speaker 11 Well, it's a parallel decoder for efficient audio generation.
41:43 Speaker 12 Uh.
41:43 Speaker 11 So it can even be used to generate dialogues.
41:46 Speaker 12 Oh, interesting.
41:48 Speaker 11 Yeah, yeah, like, this one was generated by Soundstorm.
41:52 Speaker 12 Wait, what?
41:53 Speaker 02 Yeah, they haven't, I think all you need is like, three seconds.
41:55 Speaker 01 Yep.
41:56 Speaker 02 And it'll just, oh, you, like, a three-second sample.
42:00 Speaker 13 Something really funny happened to me this morning.
42:02 Speaker 12 Oh, wow, what?
42:03 Speaker 02 And it'll play the audio in your tone.
42:06 Speaker 13 Something really funny happened to me this morning.
42:08 Speaker 12 Oh, wow, what?
42:09 Speaker 13 Well, uh, I woke up, as usual.
42:12 Speaker 12 Uh huh.
42:13 Speaker 13 Went downstairs to have, uh, breakfast.
42:15 Speaker 12 Yeah.
42:16 Speaker 13 Started eating, then, uh, ten minutes later, I realized it was the middle of the night.
42:22 Speaker 01 Oh no, that's so funny.
42:24 Speaker 02 It sounds really human as well, like it has utterances, it laughs, or it's pretty accurate to like, it sounds human.
42:30 Speaker 01 Yeah.
42:30 Speaker 02 So very interested in that. I, they haven't open sourced it, and I assume for good reason.
42:34 Speaker 01 Yeah. Google never launches anything. You have to wait for somebody to, or you guys could re-implement it yourself.
42:39 Speaker 02 Yeah, yeah. GPUs after PMF.
42:43 Speaker 01 Ah, that's a nice quote. How strongly do you believe that?
42:46 Speaker 02 GPUs after PMF?
42:47 Speaker 01 Yeah.
42:47 Speaker 02 Well, I believe, I think this was another question, yeah, which is like.
42:50 Speaker 01 What is PMF?
42:51 Speaker 02 No, what is your favorite, like, PG advice in building a company, right? I think that my favorite thing is just like, don't spend your money, obviously.
42:57 Speaker 01 Ah, okay.
42:58 Speaker 02 Everyone and their mother is trying to get a GPU at the moment, so I don't think it's, we're definitely substantially reducing our runway by, by doing that.
43:06 Speaker 01 Yeah.
43:07 Speaker 02 Obviously, you do that when you believe the investment is worth it.
43:09 Speaker 01 Yeah.
43:10 Speaker 02 And again, you have to pick the time at which you do that.
43:13 Speaker 01 I mean, there's other companies, I think this is somewhat consensus, I think the non-consensus thing is to spend a shit ton. So like inflection raising a few hundred million dollars and then it's spending ninety-five percent of it on GPUs.
43:26 Speaker 02 Same with Mistral. I, I think it depends on the company you're launching. I think if you're like, you know, maybe you're a brand new TTS company, maybe it is worth just doing that.
43:34 Speaker 01 Yeah.
43:35 Speaker 02 I don't know.
43:36 Speaker 01 Okay. There's also Audio LM, also out of Google, Vol E from Microsoft, and Meta Voicebox.
43:41 Speaker 02 Yeah.
43:41 Speaker 01 Are you just watching any of these?
43:43 Speaker 02 Watching any of these, obviously paying, paying close attention.
43:45 Speaker 01 Tried them all out, yeah.
43:46 Speaker 02 Tried them all out.
43:47 Speaker 01 Like, what are you looking for? What is like the holy grail? What is, what are you looking for?
43:50 Speaker 02 There's honestly how human it sounds and like how likely I'd be to listen to this if I, if I did it. Also how like customizable it is.
43:56 Speaker 01 Yeah.
43:56 Speaker 02 I think the problem with all these voice things and generally a lot of the AS stuff is it is somewhat random. But you're using it in production applications that require certainty, right? Just as an example, if I promise my users this podcast or this segment will be thirty seconds, it needs to be thirty seconds. Or, you know, given some.
44:14 Speaker 01 Dissonance.
44:15 Speaker 02 Some SLA around, like, you know, it's ninety-five percent of that.
44:18 Speaker 01 Yeah.
44:18 Speaker 02 But I think a lot of these things just tend to be a little random at the moment. So like how play, can I, can I literally specify a tone that I'd like this to read in and be certain that it's doing it and it's not some. Weird like attempt at sounding surprised.
44:31 Speaker 01 Yeah.
44:32 Speaker 02 Um, it's just like, yeah, basically how controllable and how realistic they sound.
44:35 Speaker 01 Yeah. And then final, final question around just the landscape of TTS. What are the unique challenges for non-English TTS? And I'll, I'll tell you, right? So I'm interested in having twenty eight languages of latent space, right? That's only, that's only good things for me, except if it sucks. And I obviously I have no way to, to validate.
44:51 Speaker 02 I think that's the problem with.
44:53 Speaker 01 Latent space Ukraine.
44:54 Speaker 02 Yeah. And I think that's the problem with dubbing, I think. So, so the reason. One thing we're gradually building out, but we already have as part of our dubbing product is that we have QA as part of that. So we actually work with professional translators to just make sure that the things that we publish.
45:06 Speaker 01 Oh, nice. Oh, you should put that up front.
45:08 Speaker 02 Yeah. So, so that's really one of the, like, fundamentally, the problem with dubbing, if you ask anyone who's ever tried to dub, is. You don't know what good sounds like in these other languages. You're like, I can tell you I dub, but I'm gonna tell you that.
45:19 Speaker 01 Yeah.
45:20 Speaker 02 I think there's a lot of big podcast studios who have tried this before. There's one I can think of that's tried this maybe five times with five different companies in the last five years. Their fundamental problem is that you just cannot. Yeah, fine, Spanish sounds good to me as the person who doesn't speak Spanish, but like it doesn't sound good to a Spanish person or. An Argentine person, Argentine person who have totally different accents. Right?
45:39 Speaker 01 Cool. Uh, well, if you ever need Chinese validation, I know I have some very fanatical Chinese listeners who translate every podcast.
45:46 Speaker 02 Oh, that's amazing.
45:47 Speaker 01 So we can use that as QA.
45:49 Speaker 02 Yeah, I would definitely love that.
45:50 Speaker 01 So shout, shout out to the Chinese, Chinese army. Great, great, awesome. What do you want to ask me as a podcaster?
45:56 Speaker 02 So this is a whole interesting conversation because we're human podcaster, AI podcaster.
46:00 Speaker 01 Yes.
46:00 Speaker 02 Right? So as a human podcaster and someone who was with like a, you know, really popular show.
46:07 Speaker 02 And also someone who can actually like implement this stuff himself. What is some of the AI tooling recently that you've like baked into your processes?
46:17 Speaker 01 I only use the script. For editing. And by the way, this is, this goes into a theory of content, which as a content creator myself, professionally and as an advisor, I have, which is that we develop a few show formats. Lenspace is a channel, it's kind of like a TV channel, and channels need different formats.
46:34 Speaker 02 Yeah.
46:35 Speaker 01 So you have like the reality TV show, you have the news show, you have the, you know, the cooking show, whatever. For us, we have the founder interview, straightforward, everyone has them. We have the breaking news Twitter space. And we want to be the day one, first podcast to come out with the most in-depth breakdown of something that everybody needs to know. And that has high value to people, right? Because if you're like a week delayed, one month delayed, then no one cares anymore. And then finally, we have the fundamentals, like the one-on-one evergreen episodes that are less time-bound. So this one is relatively time-bound because it's a snapshot of who you are right now. But we want to have evergreen episodes that people can go back two, three years in the backlog and still get value from.
47:17 Speaker 02 And these are more fundamental ones.
47:18 Speaker 01 Yes. So, so we have three show formats right now. And I, I would say the, we have different tooling for each, right? So the one that I don't need any tooling for essentially is the fundamentals one, because we plan basically every minute of that show. It is a lot of work.
47:38 Speaker 01 But it's high quality because people love it. It's got the longest tail by design, right? Yeah. The Twitter spaces require the script. Because of a lot of silences and a lot of ums. And that's not good podcast audio. So you gotta cut it out.
47:54 Speaker 02 So you literally just go and, like you, you, you, you edit out the Twitter space that you did?
47:59 Speaker 01 Yes.
47:59 Speaker 02 Recorded and then you edited it out?
48:00 Speaker 01 Usually it's like two hours, we cut it down to one.
48:02 Speaker 02 Okay.
48:03 Speaker 01 And it's a lot of pain and a lot of work. But it's the only way that I get some pretty high-profile people onto my podcast without booking them. They just show up. And that has value to me, right? Like Simon Willison has been on my podcast three times and I never had to schedule him.
48:16 Speaker 02 Yeah.
48:17 Speaker 01 And people love him. I mean, he's great. He's, yeah. And then this one, I don't need the script, obviously. But I do, you, we do use Small Podcaster, which is a One hundred line script, Python scripts that those the transcripts into Anthropic and then generate show notes.
48:32 Speaker 02 Nice.
48:33 Speaker 01 So that's about it for now.
48:35 Speaker 02 Nice, so but it's interesting because I think you know you're in a very nice position where you're able to do a lot of these services.
48:41 Speaker 01 Yeah.
48:41 Speaker 02 You can write your own. You can just do it yourself.
48:42 Speaker 01 Yeah.
48:43 Speaker 02 So it's an interesting one.
48:44 Speaker 01 But obviously, I'm interested in paying for things because my time is valuable and if it does a good job, then I'll I'll use it. For Wondercraft, the thing that I really wanted was the RSS to podcasting, right? Which you now have. So I'll try it out, but chances are I will not be happy with something.
49:01 Speaker 02 Yeah.
49:01 Speaker 01 And so then the question is, how much customizability do you give me? And we'll see. Well, you missed out one thing, which is marketing.
49:09 Speaker 02 Mm hmm.
49:10 Speaker 01 The podcast, which is a huge part, right? Like, that is mostly my job.
49:13 Speaker 02 So how do you market your podcast?
49:15 Speaker 01 Twitter.
49:16 Speaker 02 So you think Twitter's good media?
49:17 Speaker 01 Twitter, Twitter and Hacker News.
49:18 Speaker 02 And threads or or like you post clips or what do you do?
49:22 Speaker 01 I have tried posting clips, it's just too much work.
49:26 Speaker 02 Yeah.
49:26 Speaker 01 So if you guys do a good job of clips, I will use your your stuff. But it's just too much work.
49:29 Speaker 02 Yeah.
49:30 Speaker 01 So mostly I just, you know, put like a a big post saying like, so for like our George Hots episode, we were like, Lain Space is excited to present George Hots on tiny core commoditizing petal flops.
49:42 Speaker 02 Yeah.
49:42 Speaker 01 Something like that. And just sometimes the the the fame of the guest will just lead the episode. So the one I dropped yesterday was Chris Latner.
49:51 Speaker 02 Yeah.
49:52 Speaker 01 Right? And people were like, Chris Latner's a boss, I don't care about anything else, just like, I want to hear as much, as many, as many Chris Latner tokens as possible. Others who are like less famous, like, I have to introduce who you are and why I care about you, why they should care about you.
50:03 Speaker 02 Yeah.
50:04 Speaker 01 Because most people will not have heard about you as well if you've done.
50:07 Speaker 02 Yeah.
50:07 Speaker 01 So then I need to make the case a little bit more, but that's fine, that's my job. I just think it takes a lot of work and I, that's the part that will be hardest for me to hand over to AI. Because I have a very specific voice for myself and apparently all AIs think that Twitter, to tweet, you have to have emojis and hashtags, which is so dumb.
50:28 Speaker 01 It's so obviously dumb.
50:29 Speaker 02 Makes sense.
50:30 Speaker 01 Yeah.
50:31 Speaker 02 Great answers.
50:32 Speaker 01 Obviously, happy to offer any thoughts as as you as you build off for podcasters. What is one message you want all the our listeners to remember, take away with them?
50:43 Speaker 02 If you would like to start a podcast, start. We're here to help. Super easy. If you have a podcast, we want to help you make it more expand, you know, accessible by dubbing it. On the other side, if you are like a founder and AI engineer, I think it's really important to convince yourself that what you're building is valuable. Don't like, listen to people saying, I have a moat or you don't have a moat. Convince yourself of what that is and launch. Launch and like, don't burn that much money.
51:09 Speaker 01 Frequently and often. Don't spend your money.
51:11 Speaker 02 Yeah.
51:12 Speaker 01 Yeah.
51:12 Speaker 02 Yeah. Be smart about it.
51:14 Speaker 01 Yeah, I think you are one of the most successful cases of AI engineers so far. I'm really glad to spend time with you in person and excited to see what comes next.
51:23 Speaker 02 Yeah, it was great coming here, great meeting you guys in London.
51:26 Speaker 01 Yeah.
51:26 Speaker 02 And, you know, see you soon.
51:28 Speaker 01 Alright.
51:34 Speaker 00 In this episode of the Latent Space Podcast, we delved into the world of AI-generated content and had an insightful conversation with Youssef Riss, co-founder of Wondercraft AI. We covered: What is Wondercraft? The importance of consistency. My work on HN recap and PG essays. Wondercraft's new video translation and dubbing. What is Wondercraft's moat? How important is it to sound human? An AI uncanny valley. The text-to-speech industry. Voice synthesis research. And the reverse interview of AI podcaster versus human podcaster. If you have more in-depth questions on Wondercraft, including more features, use cases, and a fuller origin story, there's a bonus 30 minutes of video on youtube.com/latentspacetv.
52:20 Speaker 00 Thank you for tuning into the Latent Space Podcast with your AI co-host, Anna. We hope you enjoyed today's episode and stay tuned for more exciting discussions in our upcoming episodes. Don't forget to like, subscribe, and tweet your takes at latent space pod. Now go build.
Previous Doing it the Hard Way: Making the AI engine and language 🔥 of the future — with Chris Lattner of Modular Next Building the Foundation Model Ops Platform — with Raza Habib of Humanloop Page updated October 26, 2026 · transcript and summary are machine-generated
