#182 Check In: Outlier Database, Sherpa, and Other Reflections on the Journey

In this episode, we talk about where we're at with our apps, Jake's Outlier Database and Sherpa, the mobile app. We also reflect on years creating content, first time monetizing, and the journey thus far. 

Download Sherpa, the free AI-powered journaling app for athletes. Join the convo with Josh on Discord here.

Use the code "BJJHELP" for 50% off your first month on Jake's Outlier Database to study match footage, get links to resources, and more.

Use code “BJJHELP” at submeta.io to try your first month for only $8!

Transcript

SPEAKER_01 1:01

Hello?

SPEAKER_03 1:02

Dude, so I asked about topics and we were both like Claude, Claude Cut. I haven't thought about Jiu Jitsu in a while. So funny. I know you've got your trip coming up and still healing from injury, so you'll be hoping to get back on the mats, what, in maybe a month or so?

SPEAKER_04 1:20

Hopefully, uh, yeah, probably around there. Um, we get back from our trip like May 13th or something like that. Okay. So uh probably after that. And maybe then I'll be like more motivated to do like actual jujitsu stuff. But yeah, right now Claude is all I'm thinking about, and it sounds like you are as well.

SPEAKER_03 1:42

So yeah, it's it's amazing how many like friends or friends of friends are just like making apps and messing around with with vibe coding. It's pretty cool. Yeah, it's wild. What have you been using it for? Um, so I started using it to rebuild Cherpa on a new uh code base and then or like a new tech stack so that way we can release the new version and then start building a lot faster than before. And uh yeah, making a lot of good progress. We're actually gonna migrate, do our first migration, or we're gonna do our migration today of like we're gonna get like the old database, migrate it to the new one, and then me and Mike will join the new one on test flight. Apple's like beta testing thing, which is like our current build, make sure everything's good. Then we're gonna open that to a few people. I think we have like six people waiting, make sure they're all good, and if they're all good, we're gonna submit that test flight build to Apple, and then they will hopefully approve, and then it'll be uh available as an update, and then all our current users will update to that one. So, yeah, we migrate and then we have this overlap sync that runs every 30 minutes to pull any new entries from the old database to the new one, so that'll overlap for a little bit, and then we'll shut down the old one. So kicking it off today. I'm like a little nervous, but I'm pretty excited too.

SPEAKER_04 3:03

Nice. What how how like what's your workflow been like?

SPEAKER_03 3:08

Yeah, in the beginning, dude, I was just like I was just like one shot trying to like build stuff, and then I found out it wasn't the best. So then I learned about the Claude MD file. So then I started using that. Uh it's kind of like for people who don't know, like instructions basically it reads every time before executing stuff. And then I went online on like a like Claude chat and was like, what are the top vibe coding mistakes that people make for like scalability, blah, blah, blah? Had it spit out some rules, I added that to the MD file. So I was like, okay, at least I got a little bit of guardrails in. Then I was just kind of vibing my way through rebuilding all the screens, all that stuff. And then I found out about a couple skills, uh, simplify and review. Have you used those? They're pretty cool. So I would like do a bunch of stuff, like, okay, cool, it looks good. And then there's a skill you do forward slash simplify it.

unknown 4:01

Okay.

SPEAKER_03 4:01

And then Claude will like spin up three little agents, one for like reusability, one for efficiency, and I forget the other one, maybe like security or something, or simplicity, or like how elegant is the code solution versus like crazy stuff. So then simplify will come back and be like, oh, here's what I found. Like, here's a bunch of inefficiencies. I recommend you do this. All right, cool. Let's do that. I hit that, cleans it up. Then one more pass, there's a slash review skill, and it'll just review all the changes. Uh yeah. So that was like my next evolution. It's like, okay, it slowed me down a little bit, but hopefully it's all getting better. And meanwhile, I don't understand any of what it's doing. After that, my friend sent me a uh codex plugin, or he told me about it. So I plugged in codex, and codex can do what's called like an adversarial review, which is like it is trying to find like everything wrong with this thing. So after that whole process, now I do an adversarial review. It'll pick up a bunch of stuff, and then I'll send it to Claude, like, hey, what do you think of this review? And it'll be like, all right, yeah, actually, this is pretty legit. Let's fix this thing. So yeah, that's my workflow, and then I'll push it uh to GitHub. And I'm just working on the main branch, but Mike and I just started making PRs. So like I'll make a pull request to work off like a side branch, he'll review it, and then we'll merge it to the main thing. So yeah. Built the whole app, tested a bunch of times, test flight builds, whatever. And then we were building the migration script and like handling all the data transferring and stuff and this overlap thing. Now we're in a good place to like, all right, I think we can run it.

SPEAKER_04 5:39

So both you and I both you and I are still trying to figure out like all the coding language and kind of hell with it. Um but uh I'm curious, do you have like are you paying for API calls?

SPEAKER_03 5:53

And no. No, I'm just using the Max plan. It's like a hundred bucks a month. Right. Yeah.

SPEAKER_04 5:59

But you don't have like APIs on top of that.

SPEAKER_03 6:01

No, I have uh so what powers the insights feature is actually an old ChatGPT API that we set up a long time ago. So I just left it there. Maybe we'll switch to Claude in the future, but but you have to pay for that ChatGPT API. Yeah, yeah. And it's pretty cheap because right now we're not running that like anything too crazy. But uh yeah, yeah, I haven't dug in yet.

SPEAKER_04 6:23

So like from my understanding, when you plug in codecs to like review your code, yeah, um you have to like connect to the codecs like API, like you have to connect to OpenAI's API. Maybe that is what I did. Okay, I don't know. I don't know. Because okay, Mike, my my developer Mike was he did the same thing. Okay, and then he was like, but uh he was afraid it was gonna like rack up too big of an API bill. Oh interesting.

SPEAKER_03 6:54

Yeah. Let me see. So I have um so I connected it to Codex, but I believe Codex is there's like a free uh certain amount that you can use for free. And I've never gotten close to it, even with like the adversarial reviews. So it's kind of like a I view it like a free little code checker, but I don't use it like a crazy amount. Got it. Or enough, I guess, to trigger unless I'm getting my credit card bill somewhere that I don't notice. But I think I think it's okay.

SPEAKER_02 7:25

Gotcha. Yeah, I'm not really sure the way it works. Yeah. How are how are you set up? Um, I have do you know what VS Code is? Oh yes, I have that. Okay, I know the name of it.

SPEAKER_03 7:37

I don't really know what it does. I have it, it opens it sometimes and uses it, but yeah, yeah.

SPEAKER_04 7:44

So uh like from my understanding, it's where like Claude writes all the code. Um so basically I have VS Code and like all the stuff I'm working on is in like a folder. And then you open the terminal through VS Code so it has access to all the documents, and then you open Claude in the terminal in VS Code. And then it like you basically tell it to do something, and then it goes and changes all the code. Uh so you can like watch it, doesn't really mean anything to me, it's just like going crazy. But uh yeah, um, that's basically what I've been doing workflow-wise. And I've asked it like multiple times if there are some skills or some plugins that I can use, because there's like close to a hundred thousand plugins now for Claude.

SPEAKER_01 8:33

Yeah.

SPEAKER_04 8:33

And basically the thing it keeps reiterating with me is like what I'm working on is not like a crazy amount of stuff, and it's more niche in the sense that like I'm trying to train stuff on like jujitsu specific or like grappling specific stuff, and a lot of the plugins are more like generic business stuff or like generic website stuff. So basically what it said is like some of the memory stuff could be good down the line when it starts to get like out of control, or like the review type ones or stuff like that. Um, but basically it said when you add a plugin, it's gonna give a lot of information that's not gonna be used a lot, and the more information, the harder it is for the model to like sift through and be as reliable as it could be. So um basically it said like it's not worth it. Um the trade-off isn't isn't worth it for the scale at which I'm currently operating.

SPEAKER_03 9:35

So that's what I heard too. I heard the plug-ins and tool, like you can get kind of crazy with it. And the analogy I heard is like if you're like have a garage with a bunch of tools, it's like it lays out everything to do a job, and you're like, oh shit, which one do I? First is like, all right, let me just grab like a couple things I need to do each job or whatever. It's cluttered. Yeah, but you're working straight in the terminal. That's uh that's pretty gangster. I feel like that's like the way engineers have been working for years. I'm just working in the desktop app. It's gotcha. Yeah, I I don't know, the terminal is intimidating to me.

SPEAKER_04 10:08

It's so easy though, because like uh like I was running out of so the way we're currently set up is like I pay for the Claude subscription, and that's like me conversing with Claude, and then I pay for the API calls separately. Oh so basically when I don't know if you heard this, but like uh anthropic like changed the uh default setting to Claude to medium. Yes, dude, yes, I noticed and it was like drastically worse.

SPEAKER_03 10:47

Yeah, I didn't know what was happening. I was like, is Claude just super dumb like the last three days? And then Mike was like, no, you gotta turn the effort back into higher max.

SPEAKER_04 10:59

Yeah. So I wasn't running out on the$20 month plan. I wasn't running out of like me telling Claude stuff, I was racking up quite an API bill because I'm like training on a bunch of data, but uh I wasn't running out of like queries that I could use. Um but then I switched the effort to max. And then there's like the other thing that you can do where like Claude determines the amount of time to like think about a given thing based on the question that you ask, and you can switch that to like always answer to your the best of your ability. So I switched those two things. So effort always max and then always answer to the best of your ability. And then I ran out of stuff like so fast. So then I upgraded to the$100 a month one. Yeah. Um, and now I never run out even on the max and full thinking mode. Yeah. Um, but basically, like my developer Mike was like, you can use like, you know, commands in the terminal to go around Claude and not use the you know um tokens for like calling Claude to help you. But so I was trying to do that when I was on the$20 a month plan, but it was like pretty slow. Like I have to like, okay, what was this thing again? I like had a document with like all the commands and like trying to stay organized. And then once I just upgraded to the hundred dollar a month one and I wasn't running out. Now I just tell Claude, like, run this for me. And then she does it. So like I'm not really like doing anything fancy. Yeah. So it's basically probably very similar to what you're doing.

SPEAKER_03 13:00

I think it's very similar because every now and then I'll still open the terminal to do something, but yeah, now everything's just Claude. You you do it like that. You sent over. Remember that? Yeah, yeah, that meme was hilarious.

SPEAKER_04 13:12

Yeah, here's the five-step plan that you need to do. It's gonna take like two weeks, and it's like, bro, we're doing this tonight, and you're doing everything.

SPEAKER_03 13:20

Yeah, yeah, exactly. It's so funny. Oh, I was gonna ask, have you cursed at Claude yet?

SPEAKER_04 13:28

No, I didn't, but my uh developer Mike did. And uh this is actually pretty funny. So he basically like threatened it and said, like, this was like terrible work. And if you do this again, I'm gonna switch to uh like you know, a different company. Um, so like you'll be fired. And uh he said then he asked it a question of like, um, what's the like CSS color that we're using here? And it should have been like, you know, two seconds or whatever. It thought for like 17 minutes before answering.

SPEAKER_03 14:11

It was like, I can't get fired. I can't think of it.

SPEAKER_04 14:13

He was like, okay, maybe I went overboard. And then he saw this one thing that said, uh, Kwan doesn't tell you, but they keep track of any threats or like uh times that you've like cussed at it. No. Um and I was like, dang Mike, you're on the hit list now. He's like, I know.

SPEAKER_03 14:28

Damn, I'm on the hit. I'm not very patient. I I was telling my uh my Mike, I was like, damn, you engineers are built different because sometimes the debugging takes like hours, and it like sometimes it'll run in circles, and that's when I get kind of frustrated, like five hours in. I'm like, W2TF, you know, like question mark, exclamation mark. What are you doing? We've been at this for hours, like just to start fresh, do this again now, you know. Exclamation mark, and then it'll be like, I'm so sorry, you're right. Let me just start over. This must be frustrating. But um, dude, yeah, I I do feel like it's gonna remember though. Do you say thank you and stuff like that? Yeah, I try and say please and thank you.

SPEAKER_02 15:13

Yeah, yeah.

SPEAKER_04 15:15

Uh I apologize to it a lot too. Oh, you do? Sorry for all the back and forth. I know.

SPEAKER_03 15:21

Classic, classic Jake being so nice. Um, so yeah, what what what are you working on with it with the training and maybe you can give people a sneak peek into the new stuff?

SPEAKER_04 15:35

So my developer Mike is working on the app right now. Uh mobile app. Yeah, like the mobile app. Um, but that's gonna take a while to do. So like the initial thing was like I would start thinking about ways to implement AI into our website because at least the big problems that I see for it are like the navigation, like finding what you want to study, and then data entry as far as like clips, meaning like footage or resources, and then also now classes. Um so trying to use AI to help with those two primarily. And then from a selfish perspective, it would be like helping AI or using AI to help me with daily studies, um, to come up with ideas for daily studies, um, and also my like YouTube workflow for video ideas. And then another like user idea is to um give people like either daily studies or like topics that they're interested in, um, based on like what they've you know said that they're interested in or what they've been like searching on the site or whatever, um, as opposed to like giving everyone the same daily study. Like it might be more like curated, you know. Cool. Um so basically I kind of laid out all these ideas and Claude said that you should start with the customer-facing stuff first and then focus on your workflow after. So the first thing we did was um like right now when you do data entry, you have to decide whether it's footage or resource, and then you paste in the URL, and then you type in like the player names and the sport, and then you get taken to another page, and then that's where you enter like match general tags of like ABCC, you know, minus 66 kilograms, um, you know, whatever, like all the like general match info. Um, and then you take your notes and then you apply hashtags to those individual notes. So the first thing was like an intake process. So like now users can just paste the URL and then it goes, and then the AI basically like downloads the title, the description, um, the transcript, and then it sifts through all that, and then it auto-fills out the metadata. So it'll say, like, you know, this is a black belt match from IBJJF in the featherweight division, you know. Um, and then it will like ask you, like, does this look correct? You confirm it. So all that information is done for you and also determines if it's a footage or resource. Um, and then I got a little sidetracked because like the fancy thing these days is like the vision model stuff. Um so like I got a little went a little down that rabbit hole. But then um, I don't know if you heard, but like OpenAI pulled the plug on Sora, which is their like vision model thing. Um, because like a lot of people were using it to create like junk. Um and OpenAI was losing like millions and millions of dollars um because it was a part of their like you know lower price tier, and people were spending more than they were, you know, paying. Um so they shut it down. And in the meantime, anthropic focused mostly on like text-based stuff, which is very much lower cost, and it allowed them to get take a significant amount of market share. So the vision stuff, like having a model trying to like download the transcript of a match and take notes from just the transcript was not very helpful. Like, even the the notes weren't that good. Like it would get like a generic, like something important happened in this area, but it wasn't like anything useful. Um, so then I tried to use vision to like say, like, okay, the transcript narrowed it down to this time, so we don't have to run vision over the whole match, but like go analyze this and um do it. But like it was pretty expensive and it still wasn't like the best, you know. Um, so then I heard this story about OpenAI, and then I was like, maybe we just double down on text and we just focused on text instead of vision. So basically the way I set it up now is resources, meaning like people teaching stuff or like podcasts. These are the notes are very good from just a transcript. So basically now the user will plug in a URL, it'll autofill all the metadata for you for a resource, and then it will go take notes for you with the timestamp, and then it passes that note onto a hashtag maker, and then it applies hashtags at a high level, and then there's some that are like at like medium competence, and then the user can say, like, yes, add this one, no, add this one, and then um yeah, it's done. So basically you just like paste a you know 10 minute YouTube or like one of my I use it now for my YouTube videos. I just like paste a paste it in and then it writes all the notes for me, and then I just add or delete hashtags as needed or change the note as needed, and then um good to go. But then for footage, you have to manually type in the notes still, um, but then the hashtags get applied for you. Um so yeah, that's the intake slash like data entry process for clips right now. And currently I'm working on it for um like classes and blocks and stuff. Um, but yeah, that's like the first thing I did. Um lot more, but yeah, that's the first thing. I'll stop there.

SPEAKER_03 21:19

But yeah. Nice. That's awesome. No more uh it like what would have taken you like 10 zins and 10 hours could probably take you like one in one hour now or something like that.

SPEAKER_04 21:29

Yeah, and like the other big thing that's like super annoying is like, for example, the PGF events, there's like 40 matches on one URL. So every time you want to enter a match from there, in the current workflow, you'd have to like you finish one match, and then you would have to, you know, go back, put in the URL, put in the athlete names again, um, and then go back like to the video. But now the video is starting at the beginning, so you have to like remember the time that the matches. So it was like a very big hassle. Um, so another thing I added to it was it, if there's multiple matches in a given link, the model will basically list all the matches in the link and say, like, which ones do you want to enter? And then you can say like all of them or just like these two. And then when you're done with one match, it'll say, like, do you want to go to the next match? And you say yes. And then it it takes you to the next match and with all the metadata filled out already, and then you go. Um, so yeah.

SPEAKER_03 22:30

Nice day. That must be so satisfying to build something like that to save you so much time. It's uh it's awesome. That's cool. So wow, the the the outlier database now mean this means like the data and like um footage and resources are all getting put in at a much faster rate than previously.

SPEAKER_04 22:54

Yeah, I mean it's not uh like plugged in yet. Um so like the the issue is so Mike's working on the app right now.

SPEAKER_01 23:02

Yeah.

SPEAKER_04 23:03

And the initial plan, like I said, was to do all the customer facing stuff and then do my stuff. But uh Mike told me he was like, I'd rather just have it all at once. So like when I Do take a break from the app, then I can just do your thing and then um go back to the app. So basically I told Claude this, and then it was like, oh, well, then we might as well do some stuff to make your life easier first before we go like tackle the really hard problems. So do you want me to keep going? Or you Yeah, yeah.

SPEAKER_03 23:35

Well, it sounds like sounds like you got a few irons in the fire. You got like mobile app uh in the works, uh this piece, and maybe more making you more efficient and making the overall database just like uh more robust with data, I guess. Yeah. And then um and the coaching platform, I feel like. Are those like the three big categories of what you're working on, or are there some other fun things?

SPEAKER_04 23:58

Yeah, so the the next thing I worked on after the uh clips thing was basically I asked Claude, like, should I go tackle the data entry for the classes now? Um, or should we tackle the navigation for the clips and like just focus on clips right now, and then we'll focus on the data entry for the blocks and then the classes and then the navigation for the classes. Um, so that's what we decided to do is like do the navigation for the clips first. So like right now, I think the biggest thing is people don't know which hashtags to search for. So the first thing we did was that same like hashtag maker thing. Um so now it's applied to the search bar. So like instead of typing in like I want to see entries into K Guard that are not from Closed Guard or something like that. Um if you were to search that currently, you'd have to know to search K Guard entry, and then you'd have to search Close Guard and then click the Close Guard one to turn it red. So then it like filters out the closed guard ones. But now you can search with like open text, like show me entries from into K Guard that are not from Close Guard, and it will automatically apply K-guard entry without Closed Guard. Um, and then for queries like this, like open queries, very similar to Google, where like if you search something on Google, it'll give you like an AI summary at the top. Um so basically what we do is we give the AI summary at the top of the results, and then all the results are underneath it for those tags. And the idea behind that is like maybe you're interested, like it'll it'll help you filter down further, you know, like you can add tags and like our current we use Gemini currently, and we can't train the chatbot response. So it's kind of like a plug and play type thing. Whereas this I could train it. So I was like, you know, the when you give an example of a section after the section, say, like if you want to dive deeper, these hashtags would be appropriate. And then another section, these hashtags would be appropriate. So I'm very happy with the responses that it's giving. Um and then on top of that, we added this thing where when you're scrolling through the responses, you can click a button that says like show more like this. So, like if you see an example of something that you like, you can say show more like this, and then it'll throw that note along with the hashtags to the AI, and then it will go through our database and find more examples like what you just watched. Um and then on top of that, we're adding another button that says uh, like, you know, turn this into a class or you know, teach this, and then it'll you click that button and then it'll throw it to our like class AI thing. And then it will say, like, these classes already exist in the database. Um, do these make sense? Do you want like you want to use one of these? Do you want to um use one of these as like a template and you can edit it accordingly? Or do you want to just build one from scratch? Um, so yeah, like when you're doing your we wanted to link the coaching side to the clip side. So when you're researching clips as a coach um and you see one that you're like, I want to teach this, you can just immediately start creating it from there.

SPEAKER_03 27:29

So yeah, pretty wild. It's pretty crazy. Whoa. Yeah. Wow, that's cool. I almost wanted to say like outlier database like 3.0 or 4.0 or 5.0, but no, it's I feel like it's like hitting an inflection point with all this AI. I remember a couple of years ago you we were talking about how like if these models get better, does it benefit the product or take over the product? And like it's cool to see like how it's benefiting the products in so many ways.

SPEAKER_04 27:54

Yeah. Yeah. Yeah. And this is again just all tech stuff, no, no vision in it, which I think is keeps costs relatively low. Right.

SPEAKER_03 28:03

And then you have um the marketing, like the videos and the YouTube channel. Like, that's like a whole nother full-time job on top of all this.

SPEAKER_04 28:12

Yeah, I haven't made any YouTube videos in a little while. Yeah. But um, yeah, I think the idea is like make the product really good and then focus on marketing it more. Yeah. Um, but I've also been hurt, so I haven't yeah, it's just been like a nice little like almost like break from jujutsu to just like focus on the product more and just get it dialed in. But again, um, Mike, you you I think experienced this as well, where Mike heard that um sometimes Claude has been known to sacrifice security for the sake of performance. So Mike said even when I'm done with what I'm doing and I'm happy with it, it's probably gonna take him like a month to plug it in just to make sure everything's like good. And he's gonna have to change the UI a lot and all this stuff. So um, yeah, like I don't think it'll be ready anytime soon, or like you know, people won't have access to it um in like the super, super near future. But uh yeah, it's it's I'm can't wait.

SPEAKER_03 29:17

Yeah, yeah. Okay, so uh the stuff you're working on today, if they were to search, they they can't do open search just yet, or you're oh, but you're working on it. Got it, got it.

SPEAKER_04 29:26

Yeah, that's already done. So like all this stuff I told you about, other than the like, so we did the part of the class thing where it'll like recommend classes that already exist. Um, but now like currently I'm training the hashtag model to apply hashtags to classes.

SPEAKER_01 29:42

Yeah.

SPEAKER_04 29:42

Um, and then after that's done, then we'll edit it. Like we have the ability to edit an existing one.

SPEAKER_01 29:48

Yeah.

SPEAKER_04 29:49

Um, and then we'll also then then I'll move on to like building one from from scratch. Um, but it's pretty cool. Like basically, like when a user comes in, it'll like download their recent um classes to understand like this person likes to teach in CLA or they like to teach in drilling. This is the way they tend to like word things. So like if you go find a class from someone else, when you copy it over to your class and you want to edit it, you can ask it to edit to like just fit your you keep the content the same, but like fit my like structure that I like. Or you can say, like, I like this, but can you turn it into a game or something like that as opposed to a drill? So yeah, like the editing thing is pretty cool. And then after the editing thing, it'll be build build from scratch.

SPEAKER_03 30:36

Nice, yeah. Wow. Damn, that's crazy. Wow, seriously leveling up, but that's so cool. Um, I'm in a similar boat of like the uh making the that mindset of like making the product better before marketing too much, because like I was doing some I was looking up some of my metrics just to like get a grip on like historically what is like activation retention look like. And um yeah, it basically is like for Sherpa if people make it past their first week, it's pretty good. Like people stick around and come back, but it's but most people do not make it past that first week. So now my focus after we migrate everyone over, like we're on the new version is to build more stuff to make that first week experience way better because it's like, man, people are gonna start a new habit, you know. Even if it just takes like 15 seconds, 30 seconds before training, and then like a minute after training to do, it's still like a new habit, a new behavior to remember, and they have to like be invested in doing it, remember to do it, see the value in doing it, have a good experience the whole time. So, how do you like productize that whole experience to make it through that first week and hopefully they come back? Another thing that's tricky, it's like it's not like a daily behavior, you know, it's like people maybe train three times a week, four times, who knows? So, how do you make all this stuff more valuable and then timely like through that whole experience? So that that's kind of what I'm tackling next after um after we migrate, plus monetization, just to like give me and Mike some hope of like, okay, you know, like someone's finally paying for this, like we don't have to keep losing money on it for so long. So uh those are the next next two things. Like, hopefully, some of our power users uh will be willing to upgrade to get some of the cooler stuff.

SPEAKER_02 32:23

But yeah, exciting times.

SPEAKER_03 32:26

I know. Hopefully the migration goes well today, or else yeah. My goal is by the end of April, and I'm hoping we'll get there. Like, who knows? Maybe the test flight people and the beta test will find some new quirks or bugs and have to clean up some more stuff and re-migrate or whatever, but yeah, hoping hoping it all works out.

SPEAKER_04 32:46

The uh the thing I did um yesterday was um the daily study thing to help it with my daily studies. And um basically what it does now is it's so crazy that it's so impossible. But uh it takes the chat questions that people have asked um in the last like 14 days to get like an idea of what people are interested in, and then it tries to like highlight some trends in there, and then it goes and sifts through our last like 14 days of data that's been entered into the database to try and find some symmetry between those two. And then once it finds the symmetry, then it opens up to the whole database, and then it goes and finds examples to um like work with whatever it wants to teach. And then I said basically, like for the daily study, you can either dive deep into a given position or you can dive deep into a given like idea or grip or something like that and apply it to many different positions. So basically, it'll identify some like idea or theme that people tend to be interested in, and then um it'll go out and collect examples. And then if it was like something to do with like, I don't know, crossashi or something, it would be like I went out and found like eight examples from Krosashi, but I found like 15 examples from different positions. So instead of going deep on this one, we went wide and we're applying like the principle from the Krosashi thing to like many different positions, and then it runs the script every morning at 4 a.m. So basically I wake up and then it gives me three potential like daily studies, and then I can edit or um like approve them. Whoa. Yeah, it's so crazy.

SPEAKER_03 34:45

Whoa. So it's it like generates it for you. You just yeah, and you just make like the final call and like put it together. Oh, so that's is that part of your that's part of your daily process of um publishing the daily uh well, I just did it um yesterday, I made it.

SPEAKER_04 35:00

So this is my second day doing it. So I trained it all day like yesterday to give me like the right output that I like, um, with like the words I like at the top, and then like some examples and like a PS statement at the bottom, and then it has to give like a quiz um with some answers and everything. So yeah, I was training it um on that. So today was my first day doing it. Um and yeah, like it's it's pretty good. Like I have to still edit it a bit, but I imagine it'll um get better as time goes on. Yeah. But yes, it's so crazy.

SPEAKER_03 35:32

That's insane. Um, I'm curious, do you have something to see like what are overall what are the big trending searches in the database?

SPEAKER_04 35:42

So right now we just have like uh what hashtags people are searching for and the AI queries that people have. Um and that's basically the only way people can search in the database anyway. So um, yeah, again, we take like the past 14 days of searches is what we've been doing. Yeah. Um and send that to it. But yeah, that's the data we currently have.

SPEAKER_03 36:09

Got it. Yeah, I think it'd be kind of cool um to see like a trending, like a top three, you know, like how like Twitter or like whatever will have like trending things. It could be another like discovery route for people to go like search stuff while it's while it's like everyone's paying attention to it. Yeah, yeah, yeah. For sure. Interesting.

SPEAKER_04 36:26

Um yeah, for for a long time, like I was of the mindset, like we don't have to be first in this AI thing, but like we're just gonna gather as much data as possible to make it like super useful when we do get the AI tools. And now I feel like we're finally there, you know. Like it's like once we get these in, assuming they work well after Mike fixes all the security issues that I'm making. Um, like I think it would it's just gonna make it so much better, you know, and allow people to actually use the database the way I use it currently. Yeah. Um yeah.

SPEAKER_03 37:07

Um, the way the product has has evolved, has your thinking of like what the product is, what its value is, like who it's for, has that shifted at all, like as it's as it's grown over the years?

SPEAKER_04 37:20

Not really. Like, I feel like it's still kind of like the spot to go for for like match footage. Um and like mostly I think it's just kind of like pivoting that towards different avatars in the sense that like we started out by just doing like jujitsu people, um, and then kind of realized coaches like just absolutely love it. So then we made the coaching thing, but again, like the goal of the coaching thing is kind of the same thing, like be able to sift through a bunch of data, find the examples that you know fit what you're teaching, and then um making it easy for those coaches to like share it with their students. So, like I guess the end goal is to like get people to study outside of jujitsu. Um, and there's many ways to do that. There's so many platforms to do that. And probably the biggest advantage we have over any other one is the match footage as opposed to like just instructional type content. Yeah. So yeah, I think that's kind of the foundation of it. Um, and I feel like that's been pretty consistent the whole time. But yeah.

SPEAKER_03 38:32

Yeah, how about you? Yeah, I think in the beginning I was I was thinking it was gonna be more I was like very like um I thought people were gonna be journaling about like technical stuff a lot and like what should you work on next type of stuff. Right. And then once we thought a little bit more about like the daily check-in metrics of like how's your sleep, how's your soreness, how's your motivation levels? Um, it like zoomed out a little bit into maybe like trying to trying to give some metrics and like trends for people who don't use like Whoop or like Apple Health and like all those things. Now it's really easy just to answer a couple questions every day and it can track that along the way. Um, so I think it's like before it was like I thought maybe it was like a compass, like point you where do you want to train next, just by being helping them, helping people be more intentional and like reflecting what's going well, what's not going well. But now it's like hopefully gonna be a little more holistic of like, oh man, I'm like I've been working hard for three days, maybe it is better for recovery day. And how does that uh change things? Or maybe I'm like feeling really fresh and motivated and whatnot, and maybe this session I can work on something really hard or or whatever. So it's zooming out a little bit. But I I think still like the vision for yeah, I think it'll probably still evolve from there. Um yeah. But I think the core is still there of like taking just a few seconds before training, like, all right, what do I want to work on? And then just a few seconds after of like what's going well, what's not, what can I do next? Um uh Jonathan, who's been on the podcast, I like the way he's been using it. He uh he'll like set an intention for practice, he'll go train, he'll reflect what went well, what didn't go well, what what does he want to improve? And then that part of what didn't go well, he'll just set that as the next intention for the next training. So it becomes like this like really quick iterative cycle. Um and it reminds me of too of like Jared Wade, he's been on the podcast, and he said like after every training, he just tries to find one thing like he wants to fix, and the next training he'll just try to fix that one thing, and he just obsesses over those like tiny little improvement things. So um maybe there's a way to productize that. Like after you write what you improve, it's like, hey, do you want to set this as your next focus for your next session or something? So yeah. Um I'm hoping there's another.

SPEAKER_04 40:52

I feel like that's really good. Like you see what the like best use cases are, and then you don't like force them to do it, but you can like guide people in that direction.

SPEAKER_03 41:04

Yeah. Yeah, I feel like that's like what makes separates like good and great products is like the great ones, they have an opinion on what's like a good practice, and they guide you to make it easy to do that, versus like okay products are just like here's everything you can do. But if they can like that's just the crazy thing about like designing the thing, you know, like there's so many options, like how do you guide people's attention and like make things look easy and inviting and make it make sense to them to like yeah.

SPEAKER_04 41:35

All I'm gonna do is I'm curious when people do the journal entries, what types of intentions do they typically I don't know.

SPEAKER_03 41:41

Yeah, I need to talk to more users because I I can't like go and read all that stuff, you know. Yeah, that makes sense, yeah. Yeah, yeah. I've only talked to, I don't know, maybe 10, 15 people about it. And if you're listening and you want to please DM me, like I'm always curious what people do. Uh there the people I've talked to, some people it's like I think most people it's like a technical thing, like I want to work on passing today or something like that. Right. And then some people stick, uh think about Tammy, she's like one of our power users, she's been on the podcast too. She'll focus on one for a while, like she'll write the same intention for like a couple weeks. And she's asked, like, can I just like stick it there, you know? And like, so I'm reminded each time this is what I'll work on for a while.

SPEAKER_04 42:21

But uh me And is that intention more like is it like passing or is it like more open-ended?

SPEAKER_03 42:27

Uh I think it's a focus, something like passing, or like like I think one of her intentions was like, Yeah, I'm not gonna work on hand fighting at the hands and the feet anymore. Like, let me just go straight for knees and elbows. And every like every class she just kept paying attention to just going straight to knees and elbows, it's just skipping the hands, which is kind of cool. But uh, I commend her because that that takes like discipline and focus to work on something for a while, and I am like a shiny object like throughout this whole like journey of every week we talk about something new. But uh yeah. I'm I'm hoping maybe that shirt book can also make the journey just more enjoyable and fun and like interesting to look at, like to do the year in review or like month in review type of stuff of like, oh man, last summer was like your guard retention era because you hurt your neck or whatever, and like you've been working on guard, oh yeah, that was that was part of the journey. And oh, I got promoted here or like got my stripe there or whatever. Right, that could be kind of cool too. But who knows?

SPEAKER_02 43:27

Yeah, exciting times, yeah, yeah.

SPEAKER_03 43:29

I'm also like tempted to like try to get more people in other sports to use it, but everything I talk to, like on Claude and Chetchup, is like, no, just like focus on jujitsu for now and then we'll open it up later. But uh Yeah, yeah, yeah.

SPEAKER_04 43:42

Same. Like I'm very tempted to try and get uh like I have a decent amount of wrestling content now. Yeah, and if we could get like schools using it for their like coaching programs, I think that would be like amazing. But everything is like, you know, get it dialed in for jujitsu, maybe even specifically know gitsu, and then go to g jujutsu, and then yeah, go to wrestling. But yeah, like it is you just want to jump to the end.

SPEAKER_03 44:13

I know there's so much, there's like so much to do. Another thing I'm getting worried about is like because it's so much easier to like just ship out new features and stuff like that. I'm getting worried of like overloading, like I don't know, three months down the road with like any idea, just like send it, you know, like put it out there. So I want to be thoughtful of that. I'm also nervous of like so after we migrate people to the new app, I'll be working on the new app while people use it. So I gotta figure out how do I set up like a testing environment or like staging or something like that, or to like yeah, and how to not break stuff while people use it, basically. How have you figured that out? Because you've had live users like this whole time.

SPEAKER_04 44:54

Yeah, this is something that's out in my wheelhouse for sure. But I will tell you that um typically the way our thing works is Mike will work on something and then he sets up like a testing website. Um, and then only I get that link. And then it's just like a new Link that I do instead of our normal link. And then that has all the updates to it. And then I play around with it for a week or two and try and find as many bugs that I can. And then Mike fixes them all. And then when we feel like it's good, we send it out. But honestly, it's it's not the best like practice. Like we don't do it the best in the sense that like it's so strange that like everything will run well on Mike's computer. And then he pushes it to like the testing site and like stuff breaks. And then he's like, that was working just fine. And then he goes and fixes it. And then everything's working good on the testing site. And basically, me and him, we set up time. Like we know we're gonna be fixing mistakes for like three hours after we push this to the main site. Because like even though it works great now, when you push it, it's like now when I turn the hashtags red, the counts are all off. Like that was just when I was working. Um, so yeah, it's just like it's so strange how it just like breaks. And I think normal companies that have um, you know, active users have a very like staged approach where they like maybe release, like you said, like you release it to like five, six people, and then you slowly branch out. Um but uh yeah, we we haven't taken the time to slow down to set up that process. And basically, Mike said, like, is like I can do it. Um, I just can't do anything else like during that time. Yeah. And we were just like, yeah, screw it. Let's just fix fixed stuff in the fire. Right now, yeah.

SPEAKER_03 46:47

Um, not very professional, obviously, but I know I wish um like when I put out on Instagram and in the Discord, like, hey, we're gonna be pushing out the new version soon. Like, I need like five to ten new beta testers. I had like a few people reach out, but I I wish I had a way to like cultivate that group a little more of like who are like the Sherpa power users, the one who want to be engaged, the ones who want to give feedback and like also test new stuff. I wish I could like cultivate a better group um there, like uh to make the feedback like easier to give and to test stuff and like push into them first and they get first. Yeah, it's kind of cool for them, I think. If if like this is they're part of like your core people, um, because they get to try new stuff and give feedback and whatnot. But um, yeah, I need to get better at that, that for sure.

SPEAKER_01 47:39

Yeah.

SPEAKER_03 47:39

Yeah. Yeah. Yeah. Fun journey. It's uh yeah, I've been at it now. Um it'll be yeah, two years. I think this month is two years, and I feel like it's been like so slow, and now I'm hoping it just ticks up a little bit, hopefully.

SPEAKER_04 47:56

So you're monetizing after two years. Yeah. Hopefully.

SPEAKER_03 48:00

Hopefully, yeah, yeah. It's crazy. Yeah. How long did it take to monetize your um YouTube channel? Like from the very beginning.

SPEAKER_02 48:08

Oh, YouTube? Yeah. It took like nine months, I think.

SPEAKER_03 48:14

Okay, that's pretty good, right?

SPEAKER_04 48:16

It was it was the Gordon Ryan shout out. Oh, wow. Best day of my life. Um, yeah, I was getting like one to two subscribers a day. And then Gordon Ryan posted about my video, and that day I got 1,600 subscribers. And then I finally crossed the threshold that I could monetize my YouTube channel. Oh, okay, okay. Yeah. But I mean, you don't make much off YouTube. Like I make like$300 a month. Okay. Like it's not anything crazy.

SPEAKER_03 48:49

So well, when you first that first month of monetization, how much did you make? Do you remember?

SPEAKER_04 48:53

Oh, I you know, like YouTube's super glamorized, you know? Oh, yeah. It's like, oh yeah, you make so much money on YouTube. So I was like, I just need to get to the monetization thing and I'm set free, you know. I think I made like$24.

SPEAKER_02 49:07

Whoa. Yeah. After nine months. After nine months,$24. Yeah.

SPEAKER_04 49:13

Yeah. And that's like doing it all day every day. Yeah after nine months. Yeah. So for the time. That time period, I was just doing like two videos a week. Dang. Yeah. That's like all I did. Wow. And it sucked. I didn't even have like a microphone or anything.

SPEAKER_03 49:27

Because I wasn't making any money. Dude, wait, when you look back at the first nine months of like just grinding, making content, did you just love making the content? Like what kept you going for so long?

SPEAKER_04 49:39

Good question. I think I was more um like I had to quit my corporate job to move out here. And I really didn't want to get another job, you know? Yeah. So I think that was it more of like a scared motivation as opposed to like I love this motivation. But it was fun for sure. But I couldn't even train. So like it was all just watch as much as you possibly can. And yeah. Um, so yeah, it was a it was a grind. I don't know if I would do it again if I had to do it, but um yeah. It was it's a lot of a lot of work and not a lot of payoff um for a while. And like even when you do get it, it's like it's not that much, you know. Yeah, exactly. So you you definitely need like the it might be different for other industries, but like for jujitsu, the advertisers don't pay very much. So like you definitely need to have some sort of product that you funnel people towards. And that can be, you know, uh getting straight up paid for the amount of views that you get um by just like plugging in something, um, which I did for a little while. You can do it through affiliate marketing by selling other people's products, which I did for a while, and then you could also do it by building your own product, which I eventually got to. Um but yeah, obviously building your own product is substantially harder and it takes substantially more of your time. So, like my the growth of my YouTube channel, I think, has slowed down a bunch since I focused on building my own product because I mean you're only one person, you know.

SPEAKER_03 51:34

Yeah. Um, so the first product you launched, was it the uh half card passing instructional? Yeah.

SPEAKER_02 51:41

Yeah, I remember that.

SPEAKER_03 51:43

Yeah.

SPEAKER_04 51:43

That was when that was when I uh I spilled um tea on my computer and I broke it. So I couldn't make any YouTube videos. Um, so I used my wife's computer and just outlined this half-card passing instructional. Yeah. Um so yeah, that was the first like monetization thing I did. And I do remember I said like kind of the same thing, like I want to get feedback on this before I like push it out. So I was like, should I open it up to like 10 people, or 10 people would be gonna buy it? Should I do like, you know, five, should I do like 20? Like you're just guessing, you know, you have no idea. So I was like, eventually I was like, I'm just gonna do 50. Like, if 50 people buy it, that'd be amazing. And then I'm gonna close it, get feedback from those 50, revise it, and then open it up to everyone. And the 50 people bought it in like less than a day.

SPEAKER_03 52:34

I remember that day. It was so cool.

SPEAKER_04 52:36

Yeah, yeah. That was that was amazing. Um, so yeah, very grateful. But I think I honestly think that's like that's what happens when you make content for two years and don't ask for anything, really. Yeah.

SPEAKER_03 52:50

Um, so yeah. Yeah, that's incredible. Um, I was just journaling about this this morning, like reminding myself to enjoy this time, like this last month of two months of quad coding and like making this much progress and like getting to put this much time into it and having fun learning. Like, I'm like, am I in the good old days right now? You know? Yeah. And like, hold on, just like because it's just everything on my my mind just future, you know, like what all trying to figure out how to prioritize everything, like all the opportunity, whatever the costs. Um, and it's like, hold on, hold on, just remember, like, this might be the good old days right now. Do you do do you look back? Do you think now is like the good old day? Do you look back? There's like good parts of different phases, I guess.

SPEAKER_04 53:37

Yeah, definitely different, different phases. But I do think like the building phase is probably the most fun, you know. Yeah, like I don't necessarily like the marketing phase and like trying to push it phase, and I'm sure there are people that like that. Um, but I really like the like building it phase. Yeah. So I'm I'm having a lot of fun. This is this is awesome.

SPEAKER_03 54:05

It's so cool. Yeah.

SPEAKER_04 54:08

Yeah, because I know like in the future as it gets bigger and stuff, then there's like a whole like the business side and like managing all the taxes and all the it's just like so many things, you know? And it's like I just want to I don't know, watch jujitsu stuff. And it's like you get yeah, the bigger you get, the more problems you have that are not related to what got you those problems, you know. So like yeah, it is a little bit of a gift and a curse. But yeah, I do think like now is like the the fun time.

SPEAKER_03 54:41

That's awesome. That's so cool. Uh what was the hardest time? Was it the nine months?

SPEAKER_04 54:48

No, I don't I don't look back on that thinking it was hard for whatever reason, you know. Um yeah, I don't really know. I d uh I don't feel like it's been hard in the sense of like I have to push through it. Yeah. Um yeah, I've I've enjoyed it for sure. Um but I mean there there is like opportunity costs for sure. Um so like I would say I don't know if this is answering the question now, but like if you think about it, though that like two-year period I mean even three-year period before we launched the outlier database, I was not making like any money, and all of our money was coming from Sam. And during that time period, the stock market was like super down and it was like racing up. So if I would have just gotten a normal job, I would have been able to put a lot of money in there because we were living off of her salary. So if I just got a normal job, it was like even if I'm making more now than I had than I would be if I just got a normal job, money-wise, we would probably have more because I would be able to invest more during that time when the market was down. Right. So yeah, I mean, it's the the cost. And I mean, there's more to life than you know, money, obviously. And I feel like the lifestyle of creating your own thing is gonna have that opportunity cost upfront, but hopefully it leads to like a more enjoyable life down the road. So yeah. That's the way I try and look at it. But yeah, that's the only time I kind of was getting sad, I would say, or like second guessing myself, was when I wasn't making any money and I was like, I could be investing right now if I just got a normal job.

SPEAKER_03 56:48

Right. Right. Yeah. Yeah, it is it is very glamorized from the outside for sure. Gosh, to say the least. Yeah.

unknown 57:00

Yeah.

SPEAKER_04 57:00

How about you? Has it been a struggle?

SPEAKER_03 57:03

Yeah, I feel like the first two years, like I feel like right now is finally the point where it feels like like like the because like the first years, it was like I had my first co-founder who was just super busy with his uh regular job. I had the investor, um, and then we were hiring contractors who I had no idea how to work with engineering. Like we were just it was just this grind of like feeling like I couldn't make the progress that I wanted. It was like idea to execution was just a huge gap, like a huge time. And then Mike joined and was super excited. We had like started to make progress, but then both of our jobs, you know, sometimes I'm busy, sometimes he's really busy, he's got family, like opened up uh Corvus, like it's just not always in sync, you know, just with life. And so um I just was like I did and I couldn't code, and the coding tools weren't maybe I could have learned, but it was like I think the the learning curve was too steep for me, uh, to be honest. And now I feel like I've finally been able to unlock the the coding side, and so I get to be the bottleneck, which is like an awesome and difficult at the same time, uh, part of it, like falling asleep accepting like code suggestions or whatever. But um I feel more uh like I can run faster now. Um, and then my co-founder Mike doesn't have to be in all the nitty-gritty anymore. He can kind of zoom out and be big picture strategy and like architect stuff, which is which is cool. I think it suits our our our strengths and our time, like how much time each of us have. So yeah, it was hard for a couple years. Definitely went through lots of months of just like, well, I don't know if this thing's gonna make it, you know, like haven't shipped a new release in like months or or whatever, and having so much on the roadmap of like wanting to do. But uh yeah. And I was crunching some numbers yesterday and I was like, damn, it's still gonna be a long ass journey to like if our this could grow into a real sustainable business, like quit your job type of thing, do full time. I'm like, I don't know about that. It's gonna probably take years if things go well, you know. So it's still uh but I'm just trying to remember that's fun. And I'm on the army, like it's fun, even if at the end of the day it was just a hobby, like got to learn, got to ride this little AI wave and like be a part of it um in terms of learning or whatever. Who knows where it leads to in the future? But yeah, yeah, it's kind of my my perspective right now.

SPEAKER_04 59:32

But yeah, I think I think that's a good way. Like, I mean, I always at least try and think of things as far as like I'm just trying to make myself like as skillful and useful as possible. So yeah, just even if like this stuff didn't work out, at least I was a part of the AI wave and kind of like understand how the tools work and the limitations and what they do well, what they don't do well. So yeah, yeah, I think it's it's just fun. Like learning things is fun.

SPEAKER_01 1:00:00

Yeah.

SPEAKER_03 1:00:01

You know what's funny is uh one thing I noticed is like some of your projects came because you were forced to like do it because like the laptop with the T and then now with the injury and then light diving. So it's kind of funny how um that works out. Yeah, it is funny.

SPEAKER_04 1:00:17

It's like the that classic like parable of the I forget how it goes, but like something bad happens, and then you know, they're like, Well, I guess we'll see if it's actually bad or good, you know? Um I I don't remember the beginning of it, but like the ending of it is like some the guy breaks his leg, and then they're like, Oh my gosh, I'm so sorry you broke your leg, and then the guy's like good or bad, like we'll see. And then there's like a draft to the next week, and then he doesn't get taken to the draft because he had a broken leg. So I guess yeah, yeah.

SPEAKER_03 1:00:50

Um yeah. A lot of good lessons, good reflections. What a journey, man. What a journey. What are you focused on in Jiu Jitsu? Man, lately, like I think if I look at Sherpa, my last like five plus training sessions have been like just sweat, don't injure my neck. Right. Yeah, because I I uh retweaked it last week. Um, it's it's I think we were talking about it's like you get too excited, you know, you're you're feel like you're all healed up, and then uh kid put a guillotine on me, and instead of just tapping my way out, rolled in my back, like trying to fight it out. I feel great. Went through the rest of the round, did a couple more rounds after training. Man, it feels good. Wake up next morning, oh shit, my neck's so stiff. Then I was like, Oh, I didn't work out this week yet. Okay, I'm gonna go work out, worked out, got super stiff. Next morning woke up, unbelievable pain. Couldn't even like lift my head off the pillow without very sharp pain here. Had to like roll to my side, you know. And I was like, God damn it. I was so frustrated. Um, but then each week, each day progressed really, really well. So now I would say I'm like just regressed a little bit beyond the re before where I was prior to reinjury, and I just gotta be careful. Like so now I'm focused, like I'm basically each training, like don't go crazy. Like, you don't have to like I don't have to optimize my sessions for improving and skill and all this stuff and working around like just go show up and sweat, stay in like the grappling shape, and then yeah, get back, get back to it. So just just sweating basically.

SPEAKER_04 1:02:36

What is Andres been teaching?

SPEAKER_03 1:02:38

Um he's been prepping for ADCC. So ADCC West Coast trials this weekend, I think. We have Rana, um Ethan. Oh, I forget who else, but um yeah, so we did this last week. This last week was a lot of positional sparring, a lot of like start from here first to score type of games. A couple weeks before that, we're doing some body locking stuff. Um turtle stuff. And yeah, yeah, yeah. So West Coast trials should be interesting this weekend. Yeah, yeah. You haven't watched a lot do you have like a big list of stuff to watch on? Yeah.

SPEAKER_04 1:03:25

I haven't watched like the last couple weeks of PGF stuff, haven't watched the last couple fight pass things. Um we have trials coming up. I didn't watch the last W0 event um with DeAndre and did you see that match from there? Okay, yeah, yeah.

SPEAKER_03 1:03:43

It was it was so good.

SPEAKER_04 1:03:45

It was epic. Just been tunnel bajing. It's literally been like wake up and then just like see how much I can get done today. Yeah.

SPEAKER_03 1:03:51

It's addicting. Are you still tracking your hours of like the deep work on the it's probably too much now?

SPEAKER_04 1:03:58

Yeah. It's I when I started with the COD thing, I didn't have an issue with it. Yeah, I just I haven't done it for the past, I would say like week and a half. Yeah. But when it's more like, you know, get through the PGF event, I definitely set it so I'm not like getting sidetracked or whatever.

SPEAKER_03 1:04:17

Yeah, yeah. Same. It's like I'll forget to set it and I'll be like, I think I've put in already like five hours today. Right. So at least, you know.

SPEAKER_04 1:04:26

But it is a very cool hack. I I really enjoy that.

SPEAKER_03 1:04:29

Yeah, it puts in perspective like what you can control, you know, so clearly, it's right, so concretely. It's like, all right, can I get X number of hours of deep work on this project every day?

unknown 1:04:40

Right.

SPEAKER_04 1:04:41

And you think you're working hard until you do it, and you're like, dang, I didn't realize I was taking this many breaks.

SPEAKER_03 1:04:47

You know, yeah, yeah. And it's nice too, because it's like yeah, it's just something you look you can like mark down like I did, and hopefully it helps you be a little bit more content, even though for me it does not at all. Right. But yeah. Cool, cool. I uh don't know if anyone's still with us probably on this episode, but uh if they made it past the quad stuff, it it's probably cool to hear about your journey from the beginning and and all that. So yeah. Yeah, it's cool to hear about your journey. Thanks. Yeah, I feel like I'm a few years behind you, you know, just like similar yeah. Yeah, I'm in the not making money phase.

SPEAKER_04 1:05:29

Yeah, I feel like a lot of the startups they call it, uh Hormozy has this thing called the Rocky cutscene, where like Rocky's like training for the thing, and then there's like the cutscene, and then he's in the thing, and he's like, the cutscene in the movie lasts you know, three seconds or whatever. He's like, but in life it can last like five years. Yeah, like a lot of these startups, they don't make any money for the like the first five, ten years, and then some hits, and then now they're off to the races. But like, yeah, basically he's he said a lot of people are in the Rocky cutscene and expect it to go quick and it's not quick, you know.

SPEAKER_03 1:06:07

Right, right. Ups and downs. Yeah. Cool, cool. Well, uh, you head out here in a couple weeks.

SPEAKER_04 1:06:15

Yeah. Yeah, I forget the exact day, but end of April. Okay, cool.

SPEAKER_03 1:06:20

Well, we got episodes of Rod from Standard Jiu Jitsu coming out. I've got some coaches lined up, so I'll I'll do some episodes uh while you're traveling and and whatnot. And then um, yeah, yeah, we'll probably catch back up after your trip as you get back into jujitsu.

SPEAKER_04 1:06:35

Yeah, I can I can probably hop on next week if you do one with a coach or whatever, but after that, probably not.

SPEAKER_01 1:06:42

Yeah, yeah.

SPEAKER_02 1:06:42

All good. Cool, cool. Thanks, Jake. There, Josh.

SPEAKER_01 1:06:46

See ya.