Audio Files

New Recording 63

Kei

Kei

Apr 20, 2025 1:54 AM

Summary

  • 00:15-02:28: The conversation begins with greetings and updates on recent travels to London, Paris, and personal commitments on the East Coast. The speaker mentions returning to the Bay Area and highlights their current project focused on an AI writing system leveraging curated data sets, which they believe could outperform existing language models.

  • 02:28-04:14: The discussion shifts to the user journey with their product, Grasp. There's an emphasis on how users—primarily writers—organize highlights and notes from various sources before creating content, with plans to integrate AI for drafting and summarizing their writing.

  • 04:14-05:45: The speaker explains the planned features, including creating summaries of highlighted content for users, aiming to improve personal productivity and information management. They recognize the potential for their tool to appeal to a broader audience beyond just writers.

  • 05:45-09:56: There's a discussion on the difference between merely being a tool for writers and becoming an organizational productivity tool for everyone. The advisor suggests that the product could make information consumption easier on a larger scale.

  • 09:56-12:48: The speaker advocates for the value of highlighting and summarizing personalized content based on user preferences, making their summaries increasingly useful over time.

  • 12:48-16:59: The conversation touches on the organic growth of the product, with various methods of distribution such as guest posting and collaborations with YouTubers, highlighting strong user engagement and word-of-mouth referrals.

  • 16:59-17:36: The conversation concludes with plans to share the latest pitch deck with potential investors, affirming excitement over the project and encouraging further outreach for opportunities.

Transcript

Hi Kay, how are you? I'm good, how about you? I'm well, I'm well, thank you. I've been traveling for a bit and just dealing with some stuff. Feels good to be back home. Oh, actually I saw you were in London. Yeah, yeah, I was in London and Paris and yeah, yeah, yeah. So and then I was on the east coast for some personal commitments. I'm a trustee at Brown University, so I was there for a week, but it's good. Yes, welcome back to the Bay Area. Yeah,
of course, of course. How are things with you? Yeah, so things are going well and we are still like developing products and talking to users and currently like we see the user base is growing. Like this month we have around I think six to seven thousand monthly active users and it's growing 25 to 40 percent month over month at this moment and we are thinking about, and as I think mentioned in emails, but we are thinking about the AI writing system with curated data sets
and I think nowadays you see so many like people working on AI writing system, but I think our unique advantage could be like because we have curated data sets and then and as OpenAI said, like using like a curated data set that outperforms other current language models. So we are actually developing and testing out these things and so and then for this like a development, so we are actually currently like fundraising and speaking like around 1
million and so we actually talked to some potential investors and applied some accelerators and actually yesterday we got a kind of yes from one of the accelerators. Oh, that's brilliant. That is a launch accelerator run by Jason Calacanis, but we are... Run by who? Jason Calacanis? Yes, yep, but we are thinking about if we should take it or not, so that's yeah, but we have some like we get some yes at this moment, but yeah that's our
like a current snapshot. Got it. I don't think I have seen your latest demo video, but do you have walk me through kind of the user journey that you have in mind? Yes, so at this moment like user journey in our mind is like people like, by the way, 50% of our users are writers, so like writing newsletters, content or blogs and those people, so how they use at this moment is like they create information like highlight and leave notes on the web or
YouTube video or importing from Kindle, then they like organize everything on Grasp. Then when they need to write something, so they come back to Grasp, then look back at what they highlighted and left notes and they copy and paste or create the content on Grasp, then share on newsletters or blogs and with the AI-powered writing system and I think the user journey could be like, I think the first part would be the same, like creating information for themselves,
but later, so if they want to write something based on their created information and then our system can help them, like they can start writing something with their curated information so that they can just, how to say, oh what I, let's say user type, oh what I learned this week, the following, then they hit the AI like a help, then the AI takes care of the rest, and then they create the paragraphs, then if it looks good, then they can just
send, like how to say, send the newsletters or they can keep writing, like using the AI-powered system as just an idea and keep writing, then send it to the audience or that kind of like a user journey we are thinking about. Got it, got it, got it. Yeah, that totally makes sense and what are you doing for things like generating summaries? Generating summaries? Yeah, like you showed me a feature, right, one sec, let me, yes, yes, so yes, summary is like a, yes, like a weekly best
read and monthly best read and it's like for, also like one of the, I think the usage of it, like a use case of users, I think, like because some users like write a creation newsletter, let's say, like this week I found this information, these interesting articles or videos, so then some people just share the five links and with some comments and with the audience, so I think that summary function is for those people, like they just, you know, highlight and
leave notes, then when they need to summarize, like their creation or the newsletter that week, we can just help them creating the, create a weekly best read for them so that it helps them create the weekly newsletter or monthly newsletter, but that kind of summary we create and help them. Correct, got it, got it, and this is also a very powerful, like personal productivity use case, right? Yes, yes. Can you, like, to me, this is like the biggest
thing that you can highlight in how you think about, how you think about Blast, so, you know, it is one thing to say that you're going to help writers, okay, but there is a bigger market for helping people, more people, be more productive, okay, and what you are, and a lot of us have, you know, trouble remembering, like the things that we read or maybe we want quick summaries of what we have read so that, you know, obviously, as you point out, you can share it, anybody can do,
you know, essentially share, I often make, like, a list of books that I've read or articles that I have read, so I am wondering if the angle that you should pursue should be, like, make sense of your world more effectively. More effective, right? Well, yeah, basically, so, you know, for example, let me share my, you know, I'll tell you, like, the things that I get, okay, so I read, like, Scott Galloway's
newsletter sometimes, okay, and, you know, if I were to now go to, oh, interesting, you turned it off on Gmail? Yes, just in case for... Because it is personal, yeah, yeah, that's fine, let's see if I can see this on the web, yeah, is there a way for me to, like, I mean, I can obviously highlight here, but is there a way to summarize this? Ah, you cannot at this moment, but I think we can have a summarization function here,
like, summarizing the content and what this page says, or what this video says, and because we have a YouTube highlight function, so we get the transcript of the YouTube video, in case of YouTube video. Ah, it's not a YouTube video, these are all just text articles. Yes, what I mean is, like, it can be, it can work for also web articles, but also YouTube video, both, like, both things, I think. Got it, got it, got it, got it. Anyway, so I am, you know,
um, for sure, I think you should talk to a few more people, okay? Um, I am, um, and, you know, and, you know, I think your opportunity is, uh, is larger, uh, if you, like, are more than a web highlighter, uh, I think your opportunity is actually that of making just people more efficient with information, by using the power of AI, by using the power of large language models, at least, like, that is the broader vision, uh, that you can have. This way you help lots of people. Yes,
it is useful, and it is, it is, it is powerful. Um, I am, uh, definitely happy to share the deck with, uh, um, with Greylock. I can also ask my friend, uh, Sarah, uh, Sarah Guo, who started, uh, uh, a fund called Conviction, uh, whether she wants to talk to you. She is, uh, um, very deep in the large language model space, uh, and, uh, you know, she might, uh, she might want to talk to you as well. Thank you so much. Actually, I, I, I, I remember Sarah, uh, hosted an event last
week or two weeks ago, like AI, and I, I applied, but I got declined, but I wanted to go there, but yeah, thank you so much. Got it, got it, got it. Yeah, yeah, uh, but I think the angle that you should pursue it, uh, you should consider pursuing, it's up to you, um, is, um, if you, if you pitch this as a tool for, uh, writers and newsletter creators, it's necessarily a small market, but if you pitch this as a, uh, as an important tool for everybody, uh, then to me that
is a, it's a lot more powerful. I think there are going to be tools like this that, um, are definitely going to become, that are definitely going to become popular, uh, and, uh, um, I think you can create a lot of value. I see. Yeah. Thank you so much. Yes. And remember, you can also do things like, um, these large language models, uh, can also, uh, generate, uh, summaries. Um, we are also just, just so I, um, I'm transparent about it. We are also looking at, uh, large
language models for search, uh, and, uh, um, and, uh, you know, obviously it does not conflict with you in, in any way, but we think that is going to be, uh, a lot of power, uh, in basically our ability to consume information. So I think, uh, yeah, yeah. So I think that is a lot. Uh, so let me see. But the cool thing about them is that your, um, uh, these models can generate, uh, different summaries for, uh, uh, expert. Yes. Compared to, uh, you know, a, a regular person.
Mm-hmm. Yes. And, um, and so these are the kinds of things that you can, uh, you know, uh, that, that you can do that. I think, uh, uh, you know, that I, I think can be, uh, very powerful. Thank you. I think that's, I think one thing we were discussing, Kay and I were discussing, and that is like, uh, how many, like with GPT-3 or current language models, I think we can, anybody can summarize the content in, in the webpage, but with someone highlighted and left
notes in the content, I think we can summarize based on the highlighted part that reflects more, more, how to say, important part. What, what, what the person cares about, and you can also develop a personal model for the person so that you increasingly generate summaries that are more and more useful for the person. Yes, exactly. Yes. And I think we can follow, see what following people's summary and in mind,
and then see, ah, okay, so this is important and that should be fun. That sounds great. Yeah. It's amazing that you are number four on Harsha's list. Yes. Yes. Thank you. Nowadays, yes. Many people, like some people have started sharing about Graph on Twitter, YouTube, and blogs, and some YouTubers who has 300,000 followers or 1 million subscribers, those people started sharing about, hey, I found a graph on, on, on the web or,
or, or this is a great extension. And so we see like, just that kind of like, like, like word of mouth, like spread these days. That's amazing. Yeah. Thank you. And by the way, congratulations on, on like all the things like at NIVA. So, thank you. XYZ and touch on Web3 is interesting. And also one year anniversary at NIVA. And during July, right? Yeah, yeah, yeah, yeah, yeah. Thank you. It's, ah, it is, ah, definitely feels good.
And we are, like you, we are also very excited, ah, by, um, ah, just like, you know, the, ah, the power to create even, even better products. So I think it's, ah, it's, it's very cool. How are you getting distribution? It's all, it's mostly organic. Are you like, ah, ah, paying for ads? Not paying for us. So right now, so a little bit YouTube sponsor, so not big one, but small one. So like book YouTubers, because we have a Kindle import and export feature right now.
And so mainly like guesting post, guest posting and like, you know, organic and Twitter linked in. So yeah, we stop the advertisement right now. Got it. Got it. 50% from direct, ah, and 20% social, like Twitter, YouTube, ah, Twitter, YouTube, Facebook, and 15% organic search and 15% referrals from, ah, users blog, blog posts or, or, or somewhere. Got it. Got it. Got it. Um, this is, ah, yeah, this is, ah, super exciting. Um, so let me do, ah, let me do
both of those for you. I, um, I think David Packer is, ah, ah, the right person from, do you want me to, ah, the last update that I got from you was, ah, late September. Do you want to send me, um, a more recent update and I can share it with, ah, Sarah and David? Yes, we will share the recent update and, and yeah, thank you so much. Very good. Very good. I'm very excited for you. This is great. And, ah, yeah, I am, ah,
yeah, this is, this, this is very nice. Ah, it's, ah, it's an AD on tool that, ah, that I love. So I'm, I'm very excited to, ah, see you grow and get better. Thank you so much. So should I share the, the, the pitch deck is better or? Yeah, yeah, yeah. Ah, the, just the, the email with the link to the pitch deck, ah, it's up to you. Um, and then I will forward it with a note to both Sarah and, ah, then to David.
And, ah, if they want to come and I will update you, I'll follow up and make sure that they say yes or no. Thank you so much. Awesome. Thank you. Congratulations. Thanks. Take care. Take care. Take care. Okay. Thank you. Thank you. Bye. Bye.

Share This Audio Transcript & Summary 📚

New Recording 63

00:00
18:06