NotebookLM

“If you're trying to learn about a topic, for example, pretty good is probably better than just generating a podcast because a podcast assumes like some certain understanding. Maybe it's, maybe, maybe it thinks you understand too much, too little. Whereas voice…”

“Credit where credit is due. I think Notebook LM is awesome and differentiated and helps me learn new stuff. I think it's great.”

“I think Notebook LM was actually the first look at this, and that was something truly new in, like, consumer AI audio.”

“So it started out as a protocol tailwind. There were four, five people working on it.”

“I think a lot of people know Notebook LM for just like, you know, you put in a source and you get a podcast out of it, but using it as a AI slide designer is super, super underrated”

“Deutsche Telekom, with John here, is, is, is, is, is, is creating some of the incredible new podcast experiences, and that came from, like, testing, Early models of turning a text into like a more notebook LM style of a podcast with incredible voices that you …”

“And then right below the top 10 was Notebook LM. At number 13. And I take your wow as a little bit of a surprise, which was a surprise to me too, and I think a bit of a narrative violation that they went viral almost a year ago. In September of 2024. But they'…”

“And what's been really cool is we've enabled a lot of sharing recently in the last couple of weeks. So people are creating notebooks and sharing them either with their friends or their colleagues at work or even in a class. And you're starting to get almost li…”

“So we really like things like perplexity for scraping and web search. We like Claude for mathematics. We're looking at something from Google called notebook LM right now to pull together massive language structures and to deliver an entire, entire audio file o…”

“Notebook LM came out, and that was the first real voice experience that put in those ums, ahs, pauses, other things that are so human-like”

“They invented this Transformers paper, and everything they put out doesn't work. Like Notebook LM, the great one, like it's already dead, right? The project. So it's like they're trying so hard, but they can't do it.”

“Notebook LM is one example where that sounded so human because they put in all of these things that to an AI might feel like an error, but to a human, it sounds like another human talking.”

“And then as soon as it went viral, most of the founding team of notebook LM quit to go do a startup.”

“One of the internal incubations we have is and I think we'll probably release it, I don't know, like in the next week or two. So it might be before this podcast is out. But it's called TLDR. And what it does is it takes all of the meetings that we have and the…”

“What I think was the great innovation of Notebook LM was that people generated monologues and what humans really respond to are dialogues.”

“Before we did the IO announcement in 23, we'd already done a lot of studies. And one of the first things that I realized was the first thing anybody ever typed was summarize the thing, right? Summarize the document. And it was like half like a test and half ju…”

“I just checked in on a Notebook LM Discord. 65,000 people.”

“I think, honestly, like our fastest way that we've been able to find out if like the servers are down or there's just an influx of people being like, it says system unable to answer, anybody else getting this? And I'm like, all right, let's go. And it actually…”

“We had this idea that you could highlight the text in your source passage, and then you could transform it. And nobody was really using it. And it was like a very complicated piece of our architecture, and it's very hard to continue supporting it in the contex…”

“We didn't actually start out thinking this would live in notebook, right? Like notebook was sort of, we, Built this demo out independently, tried out like a few different sort of sources that the main idea was like go from some sort of sources and transform it…”

“Like we work with the DeepMind audio folks pretty closely. So they're always cooking up new techniques to like get better, more human-like audio. And then Gemini, 1.5 is really, really good at absorbing long context. So we sort of like generally put those thin…”

“There is a transform that needs to happen. That is inherently editorial. And I think this is where like that two person persona, right? Dialogue model, they have takes on the material that you've presented. That's where it really sort of like brings the conten…”

“And then I had this realization of like, maybe Steven is the product. Maybe the work is to take Steven's expertise and bring it to like everyday people that could really benefit from this.”

“So if you have a PDF that's purely images, it will recognize it”

“Notebook.ln can handle up to 50 sources, 500,000 words each, like you're not going to be able to jam all of that into like the context window.”

“The audio model is definitely trying to mimic like certain human intonations and like sort of natural, like, you know, breathing and pauses and like laughter and things like that. But yeah, in generating like the text, we also have to sort of give signals on l…”

“So when you upload audio today, we just transcribe it. So it is quite lossy in the sense that like, we don't transcribe like the emotion from that as a source.”

“I think that's what Karpathy did, right? Like he has now a Spotify channel called histories of mysteries, which is basically like, he just took like interesting stuff from Wikipedia and made audio overviews out of it.”

“I think you just have to be really opinionated. I think that sometimes if you are, your intuition is just sharper and you can move a lot faster on the product because it's like, if you hold that bar high, right? Like if you think about like the iterative cycle…”

“That was a Steven catch. Like not give them names.”

“I mean, we have like a really, really good like character designer on our team.”

“I think one thing we find often is if there's just too much agreement between people, like that's not Fun to listen to. So there needs to be some sort of tension and build up, you know, withholding information, for example, like as you listen to a story unfold…”

“Interacting with a chatbot is sort of novel at first, but it's not interesting. Right. And it's like humans are what makes interacting with chatbots interesting. It's like, ha ha ha, I'm going to try to trick it. It's like, that's interesting. Spell strawberry…”

“Humor is contextual also, like super contextual is what we're realizing. So we're not prompting for humor, but we're prompting for maybe a lot of other things that are bringing out that humor.”

“I think at the same time, right, there are a lot of developers that are interested in using the same technology to build their own thing. We're going to look into that. How soon that's going to be ready, I can't really comment, but these are the things that li…”

“So I guess high level, like we're definitely working on adding more languages. That's like top priority. We're going to start small, but like theoretically we should be able to cover like most languages pretty soon.”

“We have some early signal that says it's a really good hook. But people stay for the other features.”

“We're actively, that's one of the things we're actively prioritizing. Actually, one of the interesting things is now we're like, why would anyone want to do that? Right? Like, what are the actual, like kind of going back to sort of having a strong POV about th…”

“I notebook LM is my favorite new AI product of late, but that's not the way I use it. I just use it as like a really nice little prebuilt rag gooey attached to all my data.”

“So in this case, this is based on what they think is the most appropriate thing to say at the end.”

“When we launched it, actually, when we announced Project Tailwind, I think we only had three. We had three engineers, it was myself, we had a designer, we had Steven, and that was it.”

“And so even between the period when we did the first IO last year to as recently as like last month, you know, we've only had about eight engineers or so.”

“The truly like the Gemini models are so powerful and we use Gemini 1.5 pro as sort of the base for notebook LM. And then we have a powerful voice model on top of it, an audio model on top of it.”

“The real secret sauce to what makes this really good is something we've built, which is a content studio.”

“Google has these Q three check-ins or like quarterly check-ins where we have to write performance reviews for ourselves and the amount of Googlers... That ping me that I don't even know. And they're like, wow, this was such a boost to my confidence just to upl…”

“We have about 60,000 people in it.”

“In labs, we have the environment to do it. We have the environment to move fast. We have far, far, far fewer processes, maybe even to a fault. Sometimes we'll go to meetings that are literally, you know, the product managers, the engineers, the designers all t…”

“Where I think in the beginning we saw educators, learners really loved the product, and they were a lot of our demographic. It's really now a big mix of, I would say educators and learners are still a big part of it, but really professionals, right? Huge inter…”

“It is now at a point where I have to hire a business development person because I was like, hey, I have to ship products, and I'm actually taking, like, customer calls every day now.”

“I imagined that in the future, You could have an AI editor surface, right? Fully remixable, any input, any output, right? And that, that to me is sort of like a really powerful sort of core nugget where if you imagine you could take anything, whether it's vide…”

“The mobile app to me is such a big gap in terms of the experience today, which I think is like understandable considering where we're at in like the product development cycle. But I think that's the next horizon is, hey, how different is the mobile experience …”

“I would say like the number one use case is actually a lot of students taking their study materials and wanting to turn it into an audio guide.”

“I mean, I think you might see in three, six months that the sort of Gemini core user experience is thrown out in favor of the notebook LM experience because it's just better for everyday consumers.”