Speaker 1: 0:11

Welcome everybody to another episode of the accelerometer. Today we have Gabe from Zenfetch. Gabe was in my not my pod, my group in YC I don't know, there's so many different subsets of YC. We went through a very large YC group and Gabe was also in my pool a few weeks ago.

Speaker 2: 0:33

I was First time in a pool in Brooklyn. That was kind of cool Really.

Speaker 1: 0:36

Yeah, isn't Brooklyn all about the pool parties?

Speaker 2: 0:39

It might be. It might be, frankly, where I live, not too many pools around. So, this was my first experience and I really enjoyed it. Oh, that's really too bad yeah.

Speaker 1: 0:48

Where in Brooklyn are you living?

Speaker 2: 0:49

I live in Williamsburg.

Speaker 1: 0:50

Oh yeah, so you're a hipster.

Speaker 2: 0:52

I am a hipster, yeah, honorary hipster From what I understand, is where a lot of the YC founders live. So, yeah, when I was there I mean, there's even a WhatsApp group chat for YC founders in Williamsburg alone- WhatsApp. Whatsapp.

Speaker 1: 1:06

So not even like messengers, so it's, it's even a hipster group chat, correct.

Speaker 2: 1:10

Yeah, one step away from maybe going to Instagram or TikTok.

Speaker 1: 1:14

Yeah, so tell us a little bit about Zenfetch. Absolutely.

Speaker 2: 1:19

Yeah, zenfetch is essentially a tool for thought. It's a way for writers and other knowledge workers to leverage the information that they've consumed in the past. You can think of it as like a helpful companion to recognize the context of the content you're writing or the content you're reading and be able to build mental associations with information you've come across in the past. How does it do that? There's a couple of ways. One is we essentially track the information you're consuming online. Think of it as if you use tools like pocket or similar to save articles or YouTube videos. We ingest that information and recognize what are the most salient topics from that content. Then, in the future, when you're reading new information, new articles, new YouTube videos, we give you the ability to tie mental associations based on the content of this new information and those previous saves.

Speaker 1: 2:21

That's very cool. How did you come up with this idea? Oh my God, okay.

Speaker 2: 2:25

Well, honestly, my co-founder Kosh and I we've been in the startup game for a while, or at least been really interested in it. All of last year we were talking about different ideas and different things that motivated us to keep working. One of them was this concept that we read so much information every day, whether that's from forums like Hacker News or, I guess, podcasts or other YouTube videos. We have a tough time keeping track of it all. We were doing things like we read an article, we bookmark the link, but then that quickly compounds and then your bookmark tabs are impossible and unwieldy to manage. Or we would like save links in Notion or other Google workspace folders, hoping that someday we'd revisit them.

Speaker 2: 3:17

Typically, what happens is that information also grows and you never actually managed to activate that when you need it. We had all these solutions for storing this information and not a good one for getting it when we needed it. We would just forget. We even read certain articles which could have been really helpful sometime in the future. Then we thought, with all the modern advancements in machine learning and whatnot, now might be a good time to actually develop a tool that could help solve that problem.

Speaker 1: 3:51

Have you ever thought about, perhaps, maybe, building a phone and specifically just putting your tool on the phone? Make it like a zero interface?

Speaker 2: 4:03

Are you thinking similar to the way you communicate with Siri?

Speaker 1: 4:07

or something like that.

Speaker 2: 4:09

I definitely think there's some interesting applications for kind of building at the operating system level for a tool like this. In terms of feasibility. That's not where we're starting right now, but definitely something in line with how we think it could work in the future. In general, we kind of see this product as building towards a vision of ambient AI, An AI companion that's kind of like on your side, present at all times and can infer when there is something from your knowledge base that could help you in the current moment. Yeah, I don't know. It's definitely a possibility. Yeah, we'll see. Does that excite you? Oh, absolutely 100%.

Speaker 1: 4:49

I'm just waiting for this day that you can make a phone, yeah.

Speaker 2: 4:54

I feel like we're not the hardware experts, but you never know, we will come across someone who has a deep passion for this kind of stuff and would love to implement a hardware version of this technology. I've been seeing the recent Twitter feuds with Dan from Rewind and Aviv from Tab. They're essentially building similar pendants that record all this information so that at some point you can also activate it. I've seen some people even compare it to Black Mirror episodes. But yeah, I don't know. We'll see. It's a possibility for sure.

Speaker 1: 5:33

What's been the hardest part about building so far for you.

Speaker 2: 5:39

Yeah, I feel honestly.

Speaker 2: 5:40

I think one of the most challenging parts about building especially early stage startup is less so maybe the technical challenges or the implementation details.

Speaker 2: 5:49

It's having the will to build something and being ready to just scrap it entirely in favor of something else that could be a little bit more robust for what you're actually trying to accomplish, especially prior to do mount startup. I felt this kind of pain when working on other projects, anytime that those projects were sunset or deprioritized because of other company objectives that were higher priority. Working with that sense of I put all this effort into developing something and now it's not going to see a lot of days, it hurts anytime you do something like that In a startup. It's just magnified because you try so many things very quickly. As you're continuously iterating, you find that certain decisions you made were not the optimal ones, even though they may have felt like it at the time. You ultimately have to make a difficult decision of saying you know, yes, we invested a lot in this, but it's no longer the best choice, so it's time to shift gears and try something new.

Speaker 1: 6:57

So shifting a little bit to YC. Oh OK, nice yeah, complete non-secretary, of course, Right.

Speaker 2: 7:04

How was your time, yc? Yc was intense, honestly, but really rewarding. I, acaccio and I both moved to San Francisco for the batch. We even strongly considered relocating permanently to the batch excuse me, to San Francisco. I had never lived in SF. I had only visited for maybe a weekend in the past and I actually really liked it when I was in the weekend. But I was one of those people who's, like, you know, new York ride or die type of situation and you know a lot of my friends are here. The community I've built up is here.

Speaker 2: 7:41

So prior to moving to SF, you know, I told Acaccio like, ok, I'm down for the batch because I think it does make sense to, you know, be there, understand what it's like. But I was adamant that we had to come back to New York. After the fact, and maybe in like a month into living in San Francisco, I thought to myself, oh, it's actually a pretty cool city. You know, like I actually kind of like it and we, yeah, seriously thought about fully moving there. So kind of the plan was all right, my lease goes until the summertime, so let's come back right out the rest of the summer. But then in the time we were back. We just kind of thought, like you know, in some for some businesses it does make sense to be in SF and maybe it would make sense for us to be in San Francisco, but we haven't felt any serious disadvantages in New York. So yeah, and we're just happy here. You know so good founder community, also good VC community, frankly.

Speaker 1: 8:32

So how do you get into YC?

Speaker 2: 8:34

Oh my gosh, all right and high level, just don't give up. I mean, frankly, it's the same with startups. It's a game of attrition, right? If you just keep going, you'll figure something out. With YC, akash and I, actually this was our third time applying. The first two times we were both still in college, applying with a social e-commerce idea. Have you heard of Temu? You familiar with the app? Kind of like Temu high level, just be able to shop with friends, whatever you know.

Speaker 2: 9:07

At the time we thought like the vision and the idea was really cool and I think we, you know the product we actually built. We got a few people using it, but it never really took off. And yeah, but we did apply to YC twice with that idea by, I think, about a year in, just you know, put that project to rest, put that startup to rest. We each went into industry and about a year into industry we were like no, we need to go back to startup, we need to. You know that was our calling. So we just kind of like every every few days would meet over Zoom, because he was living in San Francisco at the time, I was in New York and we just iterate on ideas, talk about you know what's something we could build, something to put out there.

Speaker 2: 9:50

And then a friend of ours, actually in our batch, he called Akash one day and said, oh, akash, I'm actually quitting Meta. They were both at Meta. And he says, yeah, we're going to go, we got into YC, so we're going to do a build startup. And I said, okay, I'm going to do a build startup. And Akash is like, oh my God, that's, that's incredible. I had no idea that. You know, you guys were building this product on the side, like what is it about? And and then he looks at Akash. He says, oh no, like you know, we just had an idea. It's like we just applied with that. We don't have any product, we don't have anything, we're just a team. And then Akash looks, looks at me, and we're just thinking to ourselves, like you know, we can do that right. Like why not? Why not give them their shot? And so, yeah, so we just, you know, spent a solid week, you know, putting together a document with what is we wanted to build, which is actually what we're currently building, and a lot of the motivation behind it. Because, again, we had, we've been thinking about this for a long time. Don't can be wrong. It's just we hadn't necessarily built a product through an application.

Speaker 2: 10:49

Together reached out to I think it was two, one or two people who had done YC in the past, asked them to do a quick, you know, five minute look at the application, let us know their thoughts. That was really helpful, you know. If that's, if that's an option for people, I'd say they should strongly consider it. I'm even like open to reviewing people's applications. I've done it for some friends now. But yeah, then we just applied and we had a 10 minute interview. I'm sure he has something similar. I totally thought we bombed the interview. By the way, we had that interview and I was with Dalton and Liz and and we ended the interview. We called each other Akash and I called each other immediately after, like it's okay, man, we'll apply to summer 23,. Right, like it's not going to work out, but it's no big deal. Um, but I don't know, I guess things just worked out. The following day we got the uh call from Dalton. It's like hey, guys, like you know, you've been accepted twice and whatnot. So that was really cool.

Speaker 1: 11:45

Are there any other startups that you admire? Yeah, yeah.

Speaker 2: 11:49

Honestly, um, a lot of them are from, I think, our batch, because that's what I was exposed to right Especially over the last year. Um, just kind of, you know, I was like I'm not going to do it, I'm just going to assume anarchy is one of them. Anarchy is definitely there and it is definitely there. Um, a lot of people from our section I actually really admire Um, I think in general, I've just, you know, over the last year I've grown a deep admiration for founders and really creators, right, like, even if that's um you're creating music or uh, or TikTok videos or anything, or podcasts, you know um, because it's really hard and nobody, nobody just gets it right off the bat. I think it's easy to um construe this image of like folks who they just they've known right away, but there's that's because they don't tend to share all the hard work that went into ahead of time. Um, and so, honestly, like, I just admire the founders that are still in it. Um, the ones who you know.

Speaker 2: 12:46

I remember that when Brian Armstrong spoke at our spoke at our batch, he said, uh, something along the line. So, if you know, if you, if you haven't been working on your startup for at least two years you haven't really given it a shot right, or at least like really tried, um, and I kind of see why. I mean, you're never, you know rarely are you just going to hit it off. Um, you don't have one of those like spike. You know exponential growth curves, um, so you have to be willing to weather the storm and whether the uh, the lows of building a startup and if, if, you have that persistence and determination, I just really admire that. So, um, but I guess if I had to give one one shout out I'm a little biased because they were roommates during San Francisco, but a big fan of in keep from our batch. I really like what they're doing. Great founding team, yeah, and anarchy, of course, it goes without saying thank you.

Speaker 1: 13:39

Yeah, so what was, I guess, your hardest moment? What was the thing that you weathered?

Speaker 2: 13:45

Oh, my god, honestly, pivot hell, frankly. Yeah, I mean, you were in our section rights. I don't know if you recall, but yeah, I recall you. You might have pivoted, might have pivoted a little bit doing the thing that you applied with now doing the thing that we applied with.

Speaker 2: 14:00

It's quite, we've really come full circle. I love that. Yeah, I Roll like a brief sub stack about you know this, this whole high-level pivoting decision, but but yeah, I mean, you know we were in the same section, so every was it. Every week, every other week we'd meet. Every other week, every other week we'd meet. And you know a caution, I kind of dreaded going into it because we always thought, like you know, one thing is to not hit your goals.

Speaker 2: 14:25

Right, you said goals of you know, this week I'm gonna hit 3k MRR or next week or whatever, right? Or as for us, it's like we couldn't even have goals because we were having so much difficulty settling on the problem we wanted to work on and the idea we should pursue. And I mean that's essentially what pivot hell is right. It's like trying to figure out Okay, let's assume that your startup is not working, let's find a new direction. I Was also tough because we just weren't super constrained, right, we like thought very broadly, like all the problems we could think of and that that that did not help at all. But yeah, you know, eventually I think we we reached a low which was just Essentially burnout from pivoting so much we we found this one idea with sales tech again, I wrote about it already and we decided just we're just gonna keep at it and just be persistent and keep going. And we went with it for for a few months you know, it's not the two years by any measure but we reached a point where we realized At least the current strategy we're doing was not working and it was time to consider a serious, you know, change in strategy, changing product direction.

Speaker 2: 15:38

You know it didn't have to be a full-blown pivot, but we had gone to this point as well where we were just so exhausted, working in a space we didn't understand.

Speaker 2: 15:47

We felt a little bit like frauds, honestly, like trying to Sell a sales tool when we were amateur salesmen, right, and so, you know, we just died and the whole time we kept thinking, like man, why did we leave that first idea? Like I really wanted, like we you know we had countless times throughout this pivot health experience. For, like man, if we just had that tool would be so great, because then I could ask you know, my knowledge base? Or like, get inspiration for things, and, and I think ultimately it came down to, as we said, if the startup were to fail, you know, would we? Would we rather die on the like B2B sales hill or just give it our all on this kind of like moonshot or I don't know if it's actually moonshot, but you know this like ambient AI idea and I think we just decided, like you know what, let's just full send, and so that's where we're right now.

Speaker 1: 16:38

Yeah, at what point did you realize, like during sell, like during those selling and the sales tools, that, like that wasn't gonna work?

Speaker 2: 16:47

I think we really internalized it. You know, I think we went a solid month of having really really strong convict on wavering conviction with the sales, with, especially, the real-time conversational intelligence and, and we had a lot of like, I Want to say like Misleading positive signals, right things that we thought were really good and the, but they really amounts to nothing. And then around maybe three months, in around June, early July, you know we were running a full-fledged sales motion. So I, you know we, were like those annoying sales people who like, not to say they're, annoying by the way, I actually have a lot of respect for them.

Speaker 2: 17:25

But you know, cold emailing, cold calling and repetitively with the same group, with the same person and I was that was really tough, frankly, for me at least. You know I'm a developer at the end of the day, I'm not a. It's like a. It's a special skill to be able to cold call. It's really hard. What was cool, by the way, when we ran this, is we actually started to understand the existing products for sales and like why certain products exist and and and you know what needs they fulfilled.

Speaker 2: 17:54

Yeah and so that was cool. That was like, oh my god, I'm finally building empathy for a, for a sales person, especially like a b2b sas one.

Speaker 1: 18:01

Eventually, you'll learn how to build your sales team doing that.

Speaker 2: 18:03

Yeah, totally like. I mean, you know that why see always preaches this idea of like founder led sales, and I totally see the merit in it, right. But yeah, I think by mid-July, you know, we like really exhausted our funnel and we just knew that if we kept at it we might get something to work. But we also felt pretty confident we were gonna burn out and at the end of the day, you know, I think it was something like you know, two things that kill startups are Like essentially burnout, like running out of energy or running out of money, and running out of energies tends to be the most common one, from what I understand. So, yeah, you guys pivoted a bit too, though, right, I don't know if necessarily it's extreme, but oh, we pivoted so many times since learning this.

Speaker 1: 18:49

It started out as code search, then became code generation, but for fun, for fun. Okay, I like that code generation as a toy, okay. And then we got into YC and we're like automatic API connector to chat GPD. And then, once my commentator ended, I was like this is clearly not working. Like you know, we managed to make some sales, but then nobody used the product, so we decided to pivot and I was just decided to build open source a GI. That's what I'm working on now. It's pretty sick yeah.

Speaker 2: 19:23

Yeah, yeah. What was a well so for you? I guess, like what was the impetus to To finally move into this open source? A GI, do you think?

Speaker 1: 19:31

you'll do the same. It's what I always cared about, it's always what I wanted. And you know, once I found that I was working on the thing that I wanted to be working on, like it didn't really matter whether it was a good idea or it was a bad idea. What mattered was whether I was putting like 100% of my energy into it. And suddenly, once I was putting 100% of my energy Into it, things started working.

Speaker 2: 19:50

Yeah, yeah it's pretty cool, right Like. Have you considered Hiving it all since then? Not, not at all.

Speaker 1: 19:58

Yeah, yeah.

Speaker 2: 19:58

Yeah, I remember, I don't know if things are working.

Speaker 1: 20:01

Right, right yeah exactly.

Speaker 2: 20:03

I remember Dalton gave a talk at the retreat Maybe you were there where he talks about, he knows that people have landed on a pivot when they stop talking about pivoting, and yeah, I mean I don't know if this is necessarily what he meant, but I feel the same as you. I'm like you know what. I'm giving 100% of my energy as well, and I love who I'm the people I'm building for. I love the product I'm building. I'm using it. Finally, and it's almost like I don't really care if you know, like two years later, like it didn't go anywhere, like regardless, the fulfillment is just like through the roof. You know I'm really enjoying it and I feel like things are working as well, just because I guess maybe it's like I'm just more invested in it personally.

Speaker 1: 20:50

Yeah, another one of the realizations for me was moving up. Moving up in terms of the abstraction of how I was thinking. When I was an engineer, you know I was thinking in terms of features, in terms of like very tiny, very specific products. And during YC the companies that I saw succeed. They were like I am building this product category. And then they talked to customers and whenever a customer said like I want this built, they just built it. They weren't thinking like I'm a company that does this feature. They were just like I'm a company that does all of these features. And suddenly once, once you're doing the thing that you wanna be doing generally, you can just say sure, I'll build that too, this is my category, I'm happy with that category. You don't have to pivot out of a category, you just pivot like your feature, and then you don't wanna say I'm pivoting my feature, you just say I'm gonna build that feature. When did you first get in contact with AGI's? With AGI's interesting?

Speaker 2: 21:48

Similar. You mean LLMs and of that stuff type of thing. Honestly, my background with machine learning and AI was always in computer vision applications, so I was pretty familiar from that standpoint. I started getting a little more into the whole you know, natural language side of things, maybe 2019, 2020. So computer vision though.

Speaker 1: 22:12

Yeah.

Speaker 2: 22:13

What were you?

Speaker 1: 22:13

working on in computer vision.

Speaker 2: 22:17

A few different things, but most of my work in that field was around glaucoma prognostication. You know how we could use like eye tests from the ophthalmologist visits to forecast how much glaucoma, like the degenerative eye disease, would progress over time for patients who were high at risk, and so you know the CV part of there is like each test is essentially this like 52 dimensional vector that you can represent as kind of like a square image. So that was like where most of my work was. Other than that it was, you know, just studying in like school and whatnot.

Speaker 1: 22:58

So how did you get into AI for glaucoma? Really, it was a matter of.

Speaker 2: 23:03

I was really interested in working with a certain professor at my school and he specialized in deep learning, actually really in medical robotics, and he had this research opening with like a joint appointment with the Wilmer Eye Institute at Hopkins. So they, you know, like I'd specialty whatever, ophthalmology and yeah, just kind of by chance. But, like I said, I joined it for at first the intention to stick with it for a couple of months summer position and it ended up being about a year and a half. Has any of the skills that you picked up?

Speaker 1: 23:38

doing computer vision from back in the day. Has any of that translated to working with large language models? Yeah, I think that's a good question.

Speaker 2: 23:47

Yeah, I think, like in general, the ML process, data cleaning, right? Yeah, you know, there's only so much you can do with large language models or really any ML model if you don't have quality datasets and quality inputs and that's. I've found that in my experience working in data science, the bulk of time is in, you know, some form of like data cleaning or feature engineering and whatnot. So many of those skills are pretty transferable. Also, things related to like visualizing, data distributions and whatnot. That's. You know a lot of those things. It all comes down to just like the preprocessing, and then, of course, there are certain things that change with the model. But, yeah, most of the stuff is, in my opinion, just like before even get to the model.

Speaker 1: 24:39

So, yeah, is there anything that you see people doing that you don't think they should be doing?

Speaker 2: 24:46

You know, maybe less so on product development. I think that what I have seen is for a number of people who are just starting to get into like content creation, or like content marketing or content creation or something. It's very easy to use large language models to generate content and you know to to like the casual observer, the output of a tool like chat GBT might seem like very high quality or oh my gosh, like this is clearly so much better than anything I could produce. But at the end of the day, right tools like chat GBT are just going to produce kind of like the expected value of of all the like content marketing of the world. So, you know, if your goal is to kind of like stand out and be and that's pretty important, especially when you're early on and trying to like find your voice, find your, find your niche, I think that it's a little bit short-sighted to use tools like chat GBT.

Speaker 2: 25:45

I think it is very powerful. For example, like one application I use it in for content creation and content marketing is to kind of understand one like the general flow of something I'm writing, and two also like and this is I'm still playing around with it. But I often ask chat GBT. You know how would you interpret what I just said. Just to make sure that a model like chat GBT can spit back to me what I want readers to capture. And it's not perfect by any means, but it's kind of like an interesting proxy for how I would imagine most people would read an article and what they would take away. Of course I pair that with actually asking some people to proofread what I said and get their take, but that has been kind of a fun use of it so far.

Speaker 1: 26:31

I guess another problem with that is hallucinations also. Yeah, do you ever have to deal with hallucinations in your product?

Speaker 2: 26:39

I wish I could say we don't, but no, that's definitely like a one is. One problem is hallucinations right, where it just you try to have the model index on the content I'm supplying, right. So if you want to like in our use case, right, you save a lot of information or you digest a lot of information. I want to make sure that the stuff I service to you is from that content. And if we take a naive approach of just saying like hey, here's an article, spit out an answer, oftentimes it has baked in biases and so even with that content, it doesn't necessarily produce what the user might expect and there's no guarantees that it's truthful. So that is something we are actively working to mitigate and it's not necessarily like the highest concern for us, but definitely hallucinations are an interesting challenge. I'm not sure if there's a clear, a clean way to tackle that. I mean, there's so much money going into hallucination research anyways, so yeah, what is something that you're doing to mitigate it?

Speaker 2: 27:46

You know, one kind of simple approach is to chain like LLM's together, right, where you produce, you generate some output on a first pass and then you use other models to essentially evaluate consistency and hallucinations from that output and then, based on some score, like some threshold that you determined, hopefully using heuristics, you can essentially like refine the answer and get closer and closer to the truth, right?

Speaker 2: 28:13

I think that, coupled with, you know, just getting feedback from the end user, right, if they is really helpful, if they tell us, or just like internal testing, if we see that something is hallucinating, we can use it to refine whether it's the prompt or the model or the feedback loop.

Speaker 2: 28:31

Yeah, something to mitigate those hallucinations. When we talk about what are the limits of AGI, I'm not saying this is a good way to think about it, but sometimes I think like if I use a human as a proxy, if a human can do this task, then it's likely a model could perform just as well or better, which is also something we do for, like you know, like image classification and stuff. It's like if a human cannot discern differences, like why it's often the case that a model will also not be able to discern differences, something we face in my prior work. So, yeah, I actually kind of question, you know, like is it possible for a human to always prevent hallucinations, or is there something baked into how we think and how we operate that we naturally have a tendency to hallucinate certain items?

Speaker 1: 29:14

What's something that you're particularly excited about for the future of AI?

Speaker 2: 29:20

Is it a little corny to say ambient?

Speaker 1: 29:21

AI what does?

Speaker 2: 29:23

ambient AI, kind of like the vision that we're building towards of AI companion that's always on your side. You don't necessarily need to prompt it or, like you know, do something to invoke a response from the AI. It just it can essentially anticipate when there's something you might need, or whatever it is.

Speaker 1: 29:41

So screw pendant like why not like a brain implant?

Speaker 2: 29:46

Yeah, I mean honestly like I think the whole field of brain computer interfacing is really interesting. I think there's a lot of potential. It's a really hard problem and I honestly don't know too much about it, just like from friends who work in the field.

Speaker 2: 30:02

But I'd love to see something you know, like, like, like is working on to you know, evolve until, like you know, people essentially having like real life augmentation of their memory, of their like brain computing, like resources. I guess it interests also this whole other, you know, like philosophical dilemma of what does it mean to be human right? And then, if we started like adding these implants, are we like creating a new species? I don't know, I haven't thought that much about it, but I do think it's cool, you know, and I respect people who are trying.

Speaker 1: 30:37

So do you want to stay human?

Speaker 2: 30:42

You know my answer could change Right now. Yes, Right now, I think that there's a lot of beauty in human and the fault of within, like humanity. I mean, of course there are a lot of problems in the world and whatnot, but not even that aside, just like I don't know, this is like pretty interesting to recognize that, you know, we're not perfect beings and maybe there are ways to improve it, but for now, I yeah, I guess I'd say I like being human.

Speaker 2: 31:10

You know, I don't see myself being a cyber anytime soon. My co-founder might answer differently. Actually, I think he's like more on the whole. Like could be cool to you know wear implants and things like that. But yeah, are you? Are you trying to become a cyborg A little bit, a little bit, okay. Okay, I respect that.

Speaker 1: 31:27

I mean, I want it to be a good experience before. Okay yeah, I decided to become a cyborg.

Speaker 2: 31:31

Right, I just think I'm not an early adopter on anything that's like modifications to my body.

Speaker 1: 31:39

You know, yeah, I think, who would prefer like any sort of digital experience to the physical experience. I wouldn't say that I like digital more than physical, but my feeling is that if we work hard enough, we can make digital that's better than physical.

Speaker 2: 31:56

We haven't gotten there yet.

Speaker 1: 31:57

Okay, right, right. I don't know how long it's going to take.

Speaker 2: 31:59

Yeah, yeah, maybe sooner rather than later though.

Speaker 1: 32:03

Yes.

Speaker 2: 32:05

I think we'll have like the metaverse will actually be like really immersive, you know.

Speaker 1: 32:10

Is there anything else you would like to say to the universe?

Speaker 2: 32:14

You know, for all the founders out there, what I've learned in my short tenure of being a founder is a lot of. This game is just like we discussed persistence and not giving up. I think and this goes beyond founders, of course any creator, anybody who has like an original idea that they really want to pursue. If you have the means, you know, which, of course, is like an important dependency, but if you have the means, it's just a matter of determination and keep going. So, yeah, that's all I really could offer. Oh, that and also wear sunscreen. It's the only other advice I think is pretty universal and people should they have the means, they should also do that.

Speaker 1: 32:53

Thank you so much for coming on. Thank you so much. Great having you, gabe.

Speaker 2: 32:56

It was great being on here, awesome, awesome podcast. Thank you Bye.