Tay Pattison: Hello, hello.

River Roberts: Ty Patterson. Okay.

Tay Pattison: So to date us here for a second, are the 22nd April, 2026. sitting in Bali. Mic check. you hear me? ⁓

River Roberts: Yeah, I can hear you loud and clear. It's almost as though we're at opposite ends of same cafe because ⁓ previous recording studio option is surrounded by construction noise, but I think we're making do with what we got.

Tay Pattison: as is the case often here.

River Roberts: Absolutely. Well, so for first episode of something that we're ⁓ of hatched, shall we get into it?

Tay Pattison: Let's jump into it. Maybe we can start with a... I think V2, ⁓ should record this so that we don't to do it on the fly every time. Or we feel about how ⁓ we can into how we want to do it on the fly. But can you give me ⁓ a short intro the great spin up? ⁓ And then I've got docker here for us. ⁓

River Roberts: Yeah, so awesome. Yeah, so the great spin up is exploration in working the ⁓ different tools that are upon us ⁓ and to work towards this kind of abundant sense of the future. But a very practical exploration on what it's like to actually work with these tools, what's working, what's not working. what's actually worth building. Yeah, and I don't know, my perspective also is this idea of spinning up is partly chaotic and partly like part of how evolution works in a sense and how despite all of this movement and spinning, we continue to spin up though I'm sure there's going to be many, many paths that don't necessarily lead to the best outcome. But yeah, I think just a really practical exploration in what it's like to be building different ventures and different tools with ⁓ in current age.

Tay Pattison: Yeah, cool. And I'll add to it. just that ⁓ so much talk, there's so much noise about building with AI and that you can do things like that and that there are these ⁓ tools. And both noticing ⁓ that right like, yeah, they're really good directionally. They're really good for some things. ⁓ There's also practical nature of...

River Roberts: Anything to add?

Tay Pattison: working with anything or working in anything that there these parts that are always difficult. There are these ⁓ of ⁓ that we learn through doing. ⁓

River Roberts: Hey.

Tay Pattison: And it's really easy ⁓ me at least to ⁓ about, ⁓ it's really cool, we can do this and we can do that ⁓ to actually try take a step back and go, well what is really like to do that? ⁓ And try

River Roberts: Peace.

Tay Pattison: Also use this as a way to connect with anybody else that is building with these things or is wanting to start to build with tools like this but doesn't really know to ⁓ those first steps. Maybe this can be an interface for ⁓ like sharing specific skills specific knowledge. and doing it in this format, in a podcast format, because we can do it relatively low barrier to entry and tools are moving so fast and everything's moving so fast that we need to be able to learn on the fly.

River Roberts: Totally, also, I think it's probably worth mentioning that ⁓ the idea of ⁓ great spin-up is also to invite lots of people, different from different building sectors to come in and have conversations us as well. So, looking forward to ⁓ that.

Tay Pattison: Cheers to it. ⁓ got a few things on the docket here. ⁓ start with a weekly POS. So are at? What's happening? What's from mind?

River Roberts: Great. Yeah, like what have we actually been building? What tools have we been working with? And I guess what are the things we're figuring out along the way? I'm kind of new to this whole idea of learning in public. So ⁓ I'm looking for ⁓ some from you. I can see you're a steps ahead. So I'm really appreciating that. But we've actually been, feel like We've been working really well together actually, and we've been producing a bunch of different things in the past months. this week specifically, I a bunch of time focusing on building out our first agent for our venture studio we're building. And so this ⁓ is a Hermes that runs ⁓ on a VPS, Hertzner runs through the telegram gateway. So that's how we communicate with it. that whole process of setting up and running with it and also building an LLM wiki or knowledge base. this agent actually has memory and that it doesn't rely on some of the big. the big open model. It doesn't rely on some of the big Claude models or open AI models. It's running predominantly on open source models. Gemma, there's a Nvidia, Nematron ⁓ open right now. ⁓ basically the way I've it up, so a prompt comes in from us then it gets routed to ⁓ one of these different ⁓ open models based on the context and how much context is in these message or if there's a vision aspect to it or if it's an audio aspect that routes kind of in a spot way. Yeah, a lot of different kind of aspects of this. Please, yeah.

Tay Pattison: flip this back just so I... So I understand. the communication, the way that we talk to this is through Telegram. We have a Telegram channel that chats back and forth with us. The VPS is a virtual private server. ⁓ Is right? ⁓ So essentially computer that sits somewhere else.

River Roberts: Yeah. Yeah. So the virtual private server is, yeah. Yeah.

Tay Pattison: I you got a little lag.

River Roberts: Yeah. Yeah. So the VPS lives. We definitely have a lag. Good. The VPS lives in this instance, it's a Nuremberg. So lives in Germany. It's always on. You pay a monthly subscription price for it. I think it's like eight to 10 euros a month. And that gives you...

Tay Pattison: ⁓ So we've got the virtual private server, we've got this computer that sits somewhere else. You got it. You got it.

River Roberts: we have eight gigs of RAM on it and we have 80 gigs of memory on it. So is the ability to actually download local models and run it specifically from the VPS. And then there's the option to ⁓ the models through ⁓ called open router. And then it runs through API keys that are in the cloud and doesn't rely on the local memory or RAM access. So that's what we started with. an open router system. And because it's on a VPS, it's always accessible. Like we don't need the computer to be open. Like if we were to be running this on an old MacBook or on an old phone or something, in order to interact with it, it would always have to be on. So the good thing about a VPS is you've always got access to it when you need it.

Tay Pattison: Nice, and then Hermes is, as I understand it, correct me if ⁓ getting this wrong, but is ⁓ ⁓ layer. So it's a way of orchestrating these models or orchestrating the way that our messages go to the models and come back. ⁓ And it's alternative to like OpenClaw. Is ⁓ that right? ⁓

River Roberts: Yeah, exactly. So, Hermes Assistant is a open source model that was released about two months ago. it's ⁓ adopted quite quickly as ⁓ alternative to OpenClaw largely because of its ability to have memory, like a working memory. And then... this guy, Andrej Kapathi, has come up with this idea of a LLM wiki, which gives it even longer term memory and kind of it builds itself at knowledge base, which it can access. So ⁓ our instance, we're it to have basically an agent to support us running a venture studio so we can actually ask it questions and it will have information specific to what we've kind of fed it over the months. So I think in terms of an agent that can support a lot of internal cases rather than an open AI or Claude instance that goes out to everything. We don't need necessary access to all the information. We can run that on our individual tools, but then actually having something that can act as an agent supports us operating our studio, this makes a lot of sense. Any other questions kind of got at this point or I could kind of talk about some the things that are broken so far? you

Tay Pattison: Now that's pretty clear. Yeah, let's talk about specifics. ⁓ ⁓ is one ⁓ that I realize is that, so I've been using, as far as looking at tools, I've been using Claude code a lot. And I a local dev environment. I have a folder that has all of the projects that I'm working on ⁓ on my computer. And when I'm using Claude code, it does retain quite a lot of this information about different projects. So I can reference different projects. does have some working memory, but it doesn't have a wiki and this is only on my machine. It's not across multiple. so we don't have a...

River Roberts: Hmm.

Tay Pattison: there's like kind of what I realize is ⁓ segments of this agentic workflow work in different ways but I think what's interesting about this that we're working with is that ⁓ potentially ⁓ context across sessions and devices and hopefully our studio.

River Roberts: Yeah, well, the other aspect to Hermes in particular is that it kind of has this self-improving or self-learning. aspect to it that the communities kind of really started to build around. So excited to keep exploring this and be interested to see where we kind of get in the weeks ahead of us. But just set this up, ⁓ two days ago and a of troubleshooting along the way, including, and this is second attempt, I've set one up for myself personally. And I found I have background or experience in terminal. I mean, I remember back in the day, like, you know, even before the internet was around, you know, you'd have these, these terminals, you'd have DOS, you'd have these big, you know, what was it? Nine and a half inch floppy disks, the B drive. And I was in the terminal then that was great, but I haven't really touched the terminal since I was probably 12 or something. So actually coming back into it to explore building agencies is been a really great spot, but there has been some really frustrating aspects of it. So for instance, when I was setting up homies on specifically the Hertzner VPS, there was an issue with even just getting access to server through SSH. So SSH is kind of like your website URL to get to your ⁓ VPS Hertzner, for whatever reason, out of the box, has a real with accessing it using a password. So you have kind of reboot the whole thing as soon as you install it to get it working again. And so it took a few cycles to get through that piece. It seems like such a... And out of the gates, like, what? Why are the functions so difficult? Should this kind of work? And the first step, you're kind of blocked. But it turns out that's quite a well-known bug, but hopefully it will sort itself out with the relatively soon. But there are actually some other VPS services that seems...

Tay Pattison: You

River Roberts: seem to have made it easier to start to work with these tools. And then the other troubleshooting that I've been focusing on this week is getting... So they call There's few aspects to ⁓ this So you've got the VPS which is in the server. You've got the ⁓ agent, is Hermes, which is the orchestration layer. And then you've got the models and the providers. So where does Hermes send each... requests for output. And then you've got a gateway. And so the gateway is how do you communicate with this agent and then out of the route. So in this instance, we're using Telegram as the gateway. And then within our Telegram, we've got a Telegram group with different topics. So the idea is to design each topic within Telegram with slightly different rules and slightly different functions and personalities for the agent. instance, the first ones I set up, Other than general was systems, topic or thread and a troubleshooting one, because particularly getting this thing set up is there's quite a lot of troubleshooting to get through. however, it's, it's responding well, it's responding well to you as well. you know, adding some security around ⁓ is really, limiting can communicate your agent. So I've only got it so that how Telegram IDs can communicate with it, which means it can't really get hacked, at least not in a way that seems particularly easy to do. But I haven't done a full security audit, but that seems to be one very easy thing to do. Yeah, what else? Is there anything else jumping out at you that is kind of interesting to double-click on and go into? Because I feel like there's... A lot of weeds I can go into, but I'm not sure how interesting or relevant this is at this stage.

Tay Pattison: Yeah, I think right now it's a watch this space. with the main note being that ⁓ set it up and we've only really, we really started using it. Like there's to me with a of things, ⁓ this is super interesting ⁓ to able ⁓ like intellectually go the exercise, ⁓ which is different how useful is it? Like, are we actually using this? How are we using it? What are we doing with it? Like the intention here is be able

River Roberts: To live.

Tay Pattison: to feed this model, feed this bot a bunch of things in it, be able to come back and say, hey, don't forget these things. Here's a weekly review, these are the statistics for each of the ventures. And these are all issues that we're solving in one way or another. It's just how are we using an agent to help us with some of these things. So I think right now it's a watch this space.

River Roberts: Yeah, I guess. Yeah, I think that's a really good focal point. this point of kind of setting up as like a CEO for us, someone that can really help us like operate the piece and stay across everything and also keep the two of us on task. And largely, we're going to be feeding it with tasks, but it also keeps track of like, what are the open questions? What the open things? Maybe what are the things we should reprioritize? Having, you know, like kill criteria for different ventures ⁓ a little bit less emotional. ⁓ this would be very unemotional. ⁓ so you're, although we can give it some personality, but yeah, the idea is to keep it very objective. Yeah. Although I will say it's already taught us a joke and a pun, which they weren't, they weren't that funny, but there was still something.

Tay Pattison: Yeah, yeah, objective. Yeah. soon. you remember that? ⁓

River Roberts: No, I don't remember the options all the way ahead. But I will also say, and I haven't told you this yet, but as part of study of these agents, they have what's called like a sole document or a sole markdown file. And with that, you can kind of give it its like ⁓ ⁓ I've actually embedded it with a little bit of influence from Buckminster Fuller, and this idea of trim tabs and always looking for what are the small things that have large impact. So Let's leave Hermes for but at least that's a bit of a start and let's see where we get to in the coming weeks. I'm happy to give updates on that.

Tay Pattison: do a check in next week ⁓ ⁓ a quick piece Hermes and I'm about like how we use this format to ⁓ continue ⁓ track these projects are going and maybe the check in next week is like what did we use for? What did it fail at? like what was less ⁓ helpful ⁓ than it was worth effort what was it like exceedingly helpful that.

River Roberts: Yeah, agreed.

Tay Pattison: sit up just a little bit because the rain is coming in.

River Roberts: I feel like I've just got a zoomed out perspective of you now. have you been working on this

Tay Pattison: You So as far as... yeah...

River Roberts: or what tools are as part of the weekly post like what are you on this week?

Tay Pattison: Weekly Pulse this week, been building an internal outreach tool, which we realized is essentially a CRM ⁓ that ⁓ custom what we need. And then is using various bits and pieces from projects that I've been working ⁓ on. for for this, I read about a podcast where you can go on Apple podcasts and you can see the RSS feed and you can pull in all the information from a particular person with an explanation of what their podcast is about. ⁓

River Roberts: Okay.

Tay Pattison: And with that in mind, I went, hmm, maybe we could build this into the ventures that we're doing and go, right, ⁓ are the people that I know? Who are the people that you know? Who are the people that are out there on the internet that would have significant overlap with the things that we're doing? And could click a button and have an agent go and pull in all of that context ⁓ and us a list of all these people might be interested in working with you. ⁓ And build in

River Roberts: Okay. Okay.

Tay Pattison: features that would be really ⁓ for our particular use case. What been finding is that the tools that we're using like these ⁓ AI Claude code etc ⁓ will give a lot more than is necessary.

River Roberts: you

Tay Pattison: like the actual ⁓ that we need. ⁓ quite a small subset of all of the stuff that you get given. Give ⁓ one ⁓ The other thing that I've noticed is we're working with AI are lots of like it's directionally correct ⁓ in lot of ways ⁓ but there are lots of gaps so another thing that we've been working on ⁓ is library and maybe you want to give a little bit of context on land library because you articulate this better than I do and then I can

River Roberts: You.

Tay Pattison: talk about the pipeline the of it. ⁓

River Roberts: Well, I'm certainly happy to get into some of the ventures in a bit more detail. However, I think it would be more interesting for everyone else, at least at this point, to better understand the tools that you're working with. So you mentioned code code. So as you've been building this week, what is your stack being? are you working with it? What are you finding are the limitations of the tool so far? And where are you finding that it's really exceeding?

Tay Pattison: Thank

River Roberts: And I'm also under the impression that Claude just dropped 4.7, so I'm not sure if you've played around and know there's much of a difference with that. ⁓

Tay Pattison: Yeah, a little bit. So... Mostly I'm working with Claw Code and in Terminal. And what that looks like is opening up a Terminal, ⁓ the command ⁓ ⁓ ⁓ already authenticated it so it's connected to my Claw Code instance. And I've given it access to the files on my computer ⁓ and also integrated with a bunch of tools. It's integrated with Versal for shipping products. It's integrated GitHub for pushing my files to GitHub to have a file history. There's been a leak Vasell, were a huge tech company.

River Roberts: I thought you were about a leak in the space, it looks like there's two types of leaks that we're talking about right now.

Tay Pattison: did it. Raindrop. actual actual rain on the computer hardware and water don't mix then within core code there's a couple of pieces there's a thing called climb that i've been using for a lot of the coding that I've been doing and that's something that you install and then run through terminal as well and what that looks like is in terminal typing Klein, it opens Klein in a browser and that gives you a Kanban style to work with prompts to the agent. I'll say, hey, I want to work on land library. you get the context for it? And code will go and find the file ⁓ go, OK, I understand the structure of this. I understand what.

River Roberts: you

Tay Pattison: authentication laser and then using Klein I can give it different tasks and it will go away and do those tasks and I can work on different pieces and the innovation of Klein is that it lets you work on different features and different parts of the code base and then when you're committing them ⁓ supposed to intelligently only pull through the bits that you've been working on and allow you to ⁓ on two different parts of

River Roberts: you you

Tay Pattison: same code base so that things don't break while you're working on it. ⁓ That's a pretty useful part of the stack. Another thing that I realized that we like we've been building things, we've been designing but

River Roberts: Okay.

Tay Pattison: without having a UI, UX background, it's kind of hard to understand how many different screens, say, there will be in a particular product. And so kind of a game changer this week has been using Lord Design. And I can give it a code base and this is just that claw.ai or slash design. I can give it a code base from my computer. I can give it a GitHub. think I can give it a screenshot and I can say, Hey, this is what I'm working with. Can you show me what we've already got? So go and kind of generate or regenerate what I've already got into a visual layout without having to do any of the backend wiring. And then I can say,

River Roberts: Good.

Tay Pattison: cool missing here what are the screens that we're implying exist but don't actually exist ⁓ can you go and ⁓ those in the same format as what we've already got so it go and fill in the gaps and then I can go and take that ⁓ and it or hand it off to Clawcode ⁓ or other coding agent I think One thing that I think we both started to notice is that when we're using different tools and we're different things off, there's this mismatch in wiring between different things. Maybe you want to talk about a couple of the design tools that we've been using and this handoff.

River Roberts: Yeah, mean, certainly happy to get into that as well. Before I do though, you mentioned that you run some of your queries locally on your machine. So I'm curious how you'd sort of structure that versus running it through an API and ⁓ you've learned maybe about token efficiency in terms being cost-effective and how the work could be saved.

Tay Pattison: Yeah, there's a bunch of stuff right now about Anthropic specifically changing expensive it is basically to use tokens or how tokens a specific model uses and to give a very ⁓ high level of this. There are different models that have different costs. Tropic right now have, ⁓ of, they until recently had three different models, Haiku, Sonnet ⁓ and Opus. Opus the most expensive. Sonnet being ⁓ a bit less expensive, but still quite good. Haiku I think is like one tenth the cost of Opus. This is ballpark I don't actually know off top of my head. But it also doesn't think as hard. This is all 4.6, like version 4.6. And they've just brought out Opus 4.7. And that tool, as far as I can see, is slower. And I'm not seeing a significant difference. Maybe it's doing a better job the first time around, but I'm not going into reading the code, ever. So I don't know.

River Roberts: Yeah.

Tay Pattison: I know I have a group chat with a bunch of people that are building and they're also raising questions about it. They're like look this is supposed to be better. Whether or not it's actually better is kind of a question mark. I believe that it is. The benchmarks say that it is. Yeah, there's this part of it, there's the different models. And then as far as like cost efficiency, I have a friend who's the CTO for a bank. and he's been using a lot of different agents to build and has, think, is very ⁓ with what he does and does a lot of research, is very rigorous. ⁓ he was saying that when you're on one of the plans, like the Max Plan for Claude, pay a monthly subscription and they give you a token allowance that is much more cost effective than just going through the API so API you pay per token per thousand tokens or million tokens so wherever I can I'm running things on my computer through the plan that I have rather than say I could set up an agent or could set up an API key in a project that we have and then if you and I program it to say when I click this button I want you to go and do this research and then take that data and put it in database so that it sits there and I can reference it or so that it's part of this larger database. where I can I'm doing that through the plan and it has the added benefit of can have more skills that live basically in files on my computer so I say hey can you use this skill to go and do deep research and previously I had to find the skill to say alright the deep research is search the topic on the web, find out who all of the experts in the field are, find out what their specific things are, and then go and make a plan based on that information, rather than ⁓ ⁓ would go and do this normally, which would be go make a plan on its own that is generated from all of this data, not the specific, ⁓ nuanced expert that I wanted to replace.

River Roberts: Yeah, interesting. It sounds quite similar to what I've been exploring with the architecture around Hermes and the knowledge base there. So perhaps we should do a little sprint and put our heads together on that. Yeah, I feel like just to keep us moving on the docket, it might be worth... What else is on the docket? We kind of put a few things on there.

Tay Pattison: Yeah.

River Roberts: I feel like there's lots of different ways we can go, however, we don't need to cover everything in our first call.

Tay Pattison: So we've got a few things here. We've got the specific things that we're working on this week so we've kind of delved that. We haven't touched on field kill.

River Roberts: Okay.

Tay Pattison: I think part of ⁓ that would be interesting is what's the difference between what we can do ⁓ what really moves the needle. We can quite easily design things, can quite easily build things, can quite easily ship things. Something that still holds true is that getting people to use this stuff and it's actually probably the harder bit. Harder is not the right word for it. It's a different skill set to just going and creating the thing. And then also on the docket we've got quick tool stack discoveries. So we've talked about core design, other ways we work, working in sprints, time boxing things.

River Roberts: Yeah, I think let's let's do a quick one on like what we've learned about how to actually work together with these tools kind of in the same room. I feel like that's that's really kind of

Tay Pattison: Yeah, idea.

River Roberts: And, and like, I mean, just, just to start, feels like whoever landed at so far is this idea of being in the same space and doing these 45 minutes sprints. And at the end of each sprint, you're a bit of show and tell. And just the other day, we were also like having at least defining what would success look like at the end of each sprint, and then, yeah, kind of. parallel working and giving little bits of feedback along the way where necessary, without complex shifting. And then also understanding in terms of sequence around kind a sequence of building, like design, ideation, research, build, and kind of figuring those things out so that we're not building something that's kind of half designed and then kind of trying to unscramble it halfway. But I know you're a big fan of building the rocket as it takes off. ⁓ Which I think obviously has its merits, but also like there's certain aspects where it can come in numb quite quickly. there anything else that's jumping out of like how we work together that might be insightful to others?

Tay Pattison: Ha ha. I think double tapping on a couple of these things would be useful. the sprints is something that we came to kind of organically and then it's been refined. So we were going and doing, you know, working blocks of four or five hours, sometimes longer, eight, nine hours, and we realised that we were getting burnt out by that. also, I think it's ⁓ A. like the nature of these LLMs or working with generative things is you can just easily go down rabbit holes. Oh, that's interesting. Oh, that needs to be fixed. Oh, this, it's very expansive in the way that the work goes. then timeboxing ourselves saying, all right, let's start to work in sprints. Let's give ourselves 45 minutes. And then like you said, oh, this is what success looks like. Well, at the end of this sprint, I would like to have

River Roberts: Mm-mm.

Tay Pattison: login and the sign in process all working so that somebody can come and create an account and be on this platform. think that has been a game changer in how we can move things along. building the rocket as it takes off. This was a big learning, it's like you design something, it off me. you would be like, how do you want this? I would say, ⁓ I want the code exported so that I can take the code. ⁓ but if there's any extra buttons in that code export, it's going to fuck up what I'm doing. And then would send something through and then ⁓ would ⁓ for another five minutes while I was implementing it. And you would go, ⁓ no, I have a new version. And would be like, shit, seriously? All now I'm going to. either next time wait another five minutes for the new version to come through or so we kind of shifted this to being alright finish the design get it so you're like mostly happy and then give it to me because once I implemented all of those buttons they need to go into different parts it and the implications of there's one more input field And this is something that I didn't realize until recently. One more input field requires routing to the database and writing database rules and reordering the schema or changing the type of input from, is going to be a multi-selected thing to a single selected thing. It changes it in the database. And so it's like a bunch of extra rewiring. And time you go and do this ⁓ with coding there's the potential for it to completely mess it up. like ⁓ to create another version ⁓ the ⁓ ⁓ instead of like updating the data scheme, it'll just write a whole new one and then you're like clicking the button. Why is the button not working? ⁓ cause it's connected to the old data model that was in ⁓ ⁓ of this that was an hour ago.

River Roberts: There's a line that's coming up for me, which is early mistakes cast long shadows. so I think how we design less long shadows? I think the other thing that we've learned in our sprints or our workflow is also being quite tight on the scope and not being, as we described, about things and having things kind creep past, which To be it's just really good design and build practice anyway, to not keep extending the build. And almost by having these 45 minute sprints keeps us within that frame quite tightly. And I don't know about you, but also since we've gone through more of this sprint approach, I feel like we've got healthy relationships with ourselves as well. And like what you were describing, these tools are very engaging and it's very easy to fall into these rabbit holes and kind of get taken away by them, but actually having a healthy relationship. with how to use the tools and your own sense of time. we're also very much looking for how do we build these tools and still have a healthy relationship with each other and ourselves and everyone around us. And I don't know about you, but I'm actually really excited by our ability to produce more but work significantly less. And the idea of just like hard charging 10 hours a day, every day of the week, although with these tools is definitely very enticing. I'm not sure it ultimately ends up in producing better results because you kind of end up spiraling down rather than spinning up. So I feel like that could be a nice thing to keep us grounded, so to speak. Let's shift to the next session because I think next on the docket was this idea of doing a bit of a deep dive on one thing or one or two things. And know what was on there was this idea of like ⁓ what we build or this area of like going to market. You some people talk about this being like distribution. And yeah, feel incredible in what be on the build. And it's been quite in input and getting people actually the tools that were built. And obviously we've built them, so we feel like they're valuable. But then when it actually meets the road and getting feedback, it's been a little bit slow. And I'm not sure about you, but I've also been a little reluctant to put too much energy into outreach because from my mind, the energy and the effort, as soon as you share this with like 10, 15, 50 people, all of a sudden you have much more inbounds to manage. if I share this with 50 friends, I kind of feel like I need to start replying to all 50 of them when they give something back. then, so there's been a slight kind of reluctance to let these out into the wild a little bit. I know with our first kind of two products that we've started to share so far, Land Library and Animates, it's been kind of a, do we share this with maybe 10 to 20 people and get feedback on that before we share it more broadly. And then within that, like focusing this week on actually building this idea of an outreach engine or of like helping us identify like who in our networks. And I know that you've built out this outreach engine using our LinkedIn data that we exported. Who in our networks that could be worthwhile. And this is kind of this idea of the trim type again, like how, ⁓ as venture studio founders, we focus on the first five, 10, 15, maybe 20 relationships that we share these products with, but then scale themselves almost like there's network effects embedded in every relationship that we help nurture. And that as founders, we're not then focusing on trying to get this to hundreds of thousands of people. We're more interested in the quality of that first circle of people that we have direct relationships with. And yeah, really really understanding the network effects of healthy relationships and healthy partnerships. yeah, I'm to see how that kind of plays out over the next few weeks as well.

Tay Pattison: Yeah, me too. It's also the... we do share these, at what point do you share them ⁓ what is the kind of expected relationship you're building with people ⁓ are... ⁓ potentially using them right because especially sharing it like with your inner circle with my inner circle like we about these relationships and we want to make sure that what we're sharing is good quality balancing that sharing it soon enough so that we're figuring out if this is something that people will actually use What also comes up to me you this book, The Mum Test?

River Roberts: I think so. if your mom can understand it.

Tay Pattison: This one is, your mom's probably gonna love anything that you make. And you go to people and you say, hey, I made this, what do you think? It's gonna be super rare for them to go, I don't really like that bit, ⁓ ⁓ not good enough. ⁓ So the whole idea is like presenting things in a way, or presenting things to people, or both, that don't you as, I've already built this thing, what do you think? ⁓ Your judgment this thing kind of reflects on me. ⁓ Like stepping away from and instead trying to get an objective view of how useful the thing is. So like having initial users ⁓ is really but also finding people that are genuinely wanting to solve this. And it comes back to like the pre-build things. Like, yeah, we can go and build all of this stuff. if it's not actually solving a specific problem, if it's not actually like a real ⁓ for somebody, then we're kind of ⁓ shortening ⁓ startup of build a thing, test it, ⁓ figure if it's something that people actually want. ⁓ To it just puts more ⁓ emphasis on...

River Roberts: Okay. Okay.

Tay Pattison: the upfront of research behind or at least spinning out a V0 or V1 so you get like ⁓ real feedback ⁓ whether or it's that's useful.

River Roberts: Yeah. So I found with my early research, my early outreach, it's all largely been on, ⁓ mean, it's actually all been on WhatsApp. Before I've shared something with them about, I've said, hey, I'm doing some research on this. I'm doing research on X topic. I'd love opinion. Here's three, three or four simple questions. ⁓ I'd love thoughts. Send ⁓ me a or a voice note. And my response rate for that has been 90%. And the feedback's been incredible. And then ⁓ after received the feedback and then shared with them the prototype and asked for a little bit of feedback on that, and that's kind of been part of the agreement, there hasn't been as much and ⁓ as quick response. also I think it's because

Tay Pattison: you

River Roberts: you know, once we send something that's actually kind of working or half working or in prototype form, you kind of see it, but it's a little bit harder to, to, um, provide maybe feedback with like the open questions, like the research questions that they've ever came really easy. And then it's like, what do think of this? Maybe that's part of the mum piece. Like, I don't want to be critical. Even though like I've been asking specifically like, Hey, I'd love brutal, honest feedback. know, any, any, any thoughts, there's no right or wrong. We're not going to get offended. We really want to know what we're building here and if it's worthwhile. But still that response has been tough. In fact, I had one friend just write back to me yesterday, goes, man, I've still been sitting with your land book prototype and I'm still not sure how I feel about it. I'm trying to get feedback to you, but it's not coming. I was like, even that was useful feedback. So yeah.

Tay Pattison: Yeah.

River Roberts: I'm finding that an interesting way to start the outreach. I'm not sure how effective it actually is at the end of the day. That adage of people don't really know ⁓ what want you to build, but they'll it when it's working. That's not it, but something like that. I don't know what Steve Jobs used to say.

Tay Pattison: ⁓ Yeah, I ⁓ to it says a lot about the time that we're in. Like, I've spent time this week actively not being on my phone, making sure that every time I'm on my phone there's a timer set ⁓ and I know what I'm doing there so that it's not like stealing my attention. ⁓

River Roberts: Hmm.

Tay Pattison: ⁓ And I feel that everybody is feeling this they're just maybe not as sensitive to it as I am or like maybe coping with in different ways of doing different things and realizing that everybody is with this noisy space ⁓ that is the internet that is

River Roberts: Yeah.

Tay Pattison: AI always asking you a question after you ask it a question that is social media taking every bit of attention that you have like that a noisy space and it's only going to get noisier I think and so

River Roberts: Yeah. love the irony that you're talking about noisy space as your space is getting noisy. ⁓

Tay Pattison: Yeah, me a hand using in the background. ⁓

River Roberts: Yeah, we should do a whole episode on how noisy the space is in a noisy space. That could be one of our deep topics. noise versus signal. feel like that's actually a really important piece to spend some time ⁓ and exploring and then figure out how to better build for as well. Yeah. Yeah, it's a really good point.

Tay Pattison: Ha ha ha ha. Yeah, coming back to these ⁓ tools as well, ⁓ that signal doesn't always look like what AI tells you it is. ⁓ Like, of tools are giving us answers. It doesn't it's actual answer. Saying, no, sorry, there's no significant signal here ⁓ is not in the mandate of most of tools, which is a real problem. ⁓

River Roberts: Hmm.

Tay Pattison: and I'll try and give it a concrete example, that if we go ⁓ and... I've a tool that builds, I've got a that builds a tool every night. It goes to Reddit, it looks through Reddit, it then runs it it looks for problems in subreddits and then it takes a list of those problems and it runs them all through ⁓ about six gates ⁓ try and weed whether or not it's a real problem worth building something for. ⁓

River Roberts: Do you have one of your gates called Bill? Is one of your gates called Bill? Because it's not, you should add a seventh gate called Bill.

Tay Pattison: they're build gates.

River Roberts: The Bill Gates filter? ⁓ Bill Gates. All right, all right.

Tay Pattison: That other. one of the bill gates but the point being like it might come back and say ⁓ yeah this significant this significant significant evidence here that bill gates ⁓ wants to build this bill gate

River Roberts: The Bill Gates filter, yeah, what would it be? you

Tay Pattison: And what's the evidence? I don't know. What's the thing? I don't know. But it came back with an answer. And I think that's a trap that we sometimes can fall into is just like taking any answer as, oh yeah, that's a good enough answer, rather than taking that responsibility of building the bill gates, bill gates.

River Roberts: the little Bill Gates, Bill Gates. no, these tools aren't particularly critical unless you prompt them to be. And yeah, they're very good at being, ⁓ that's a great idea. Keep going, keep exploring. I've found as part of, I don't call them but I've been running this kind of framework that ⁓ looks at a bunch different questions. And two of the ones I keep coming back to are this idea of red teaming. So be critical, try to figure out what's wrong with the thinking, what's wrong with this idea. And then the second one, which Elon Musk used quite a lot, is this idea of first principles. So how do we bring everything back to its simplest form? From a first principles perspective, is this even our value? Is making this going to solve something, or is this just more medicine-type thing that's keeping people sick in a way? And I like. always coming back to those points that feels quite valuable. then it's like, all right, cool. Now at least we've gone through those more critical lenses or gates. definitely need do these gates. This is brilliant. All right. but we'll go from there, but I'd actually, love, I'd love to do a deep dive on all the different kinds of lenses or gates or whatever you want to call them. ⁓ at later point, I think that would be a really nice kind of deep dive, to explore and the, and then also, how that kind of evolves over, you know, the next three, six months.

Tay Pattison: So this idea of lenses.

River Roberts: else on the kind of like sharing what we build or the go-to-market piece that you think is relevant at this particular point for this deep dive deep dive section of the docket?

Tay Pattison: actually a question for you or a ⁓ ⁓ is, right, let's work on the assumption that, that the space of and especially digital tools being created is going to get busier. And ⁓ the distribution mechanisms have traditionally been... SEO, how does someone use keywords to find something. Partnerships, ⁓ does somebody scale of mouth, sort of. of mouth is obviously a big and strong, the most effective distribution channel.

River Roberts: search.

Tay Pattison: Then you've got things like founder led marketing, you've got paid advertising, you've got underpaid advertising, all sorts of social, ⁓ social and paid social. Like there are all of these ways to get products into the hands of people that need to use them. Do you that there are any... ⁓

River Roberts: Hmm.

Tay Pattison: ways that will emerge or are emerging in this kind of agentic and will they have a bigger impact or will it just be like a slightly different version of these things that we have like I can see SEO like search for example is sort of morphing but it's morphing from Google search to agentic search and it's essentially the same underlying thing with a different rule set applied to it. Yeah, does that make sense? I feel like I implied the question but maybe didn't.

River Roberts: Okay. Okay. Yeah, yeah, I mean, we're seeing, no, absolutely. No, no, no, no, absolutely. I think we're seeing the early stages of how AI companies are starting to think of advertising, for instance, and tool discovery or product discovery through these tools. obviously there's been some big pushback. I'm not sure if you saw the Super Bowl commercials that Anthrocon put out. against open AI for the advertising campaigns. But you know, just brilliant, know, idea ⁓ asking your AI for relationship advice and then selling up selling you a Kuga dating app. ⁓

Tay Pattison: like you have mummy issues. Maybe you should go on.

River Roberts: ⁓ incredible. So, and don't know about you, but like, I've certainly noticed that, like, I'm finding a lot of tool discovery directly through, like, the main agents I'm using my research and ⁓ ideation are Akimi Aplexity Pro. And I kind of, you I ⁓ work across those in their frames. And yeah, like, I typically end up exploring what they suggest. I'm not like Googling anything anymore. And then it's what's coming up in different groups of conversations. I get the sense that idea of, you know, curated, small kind of WhatsApp groups and telegram are also a really good source because there's an initial kind of filter. There's an initial curation of the people that are there and then what's actually being shared. So I think that those streams will become a little bit more popular as well going forward. Yeah, I'm not sure SEO is, mean, yeah, I'm not sure how the space is going to emerge. I don't have like a broad perspective on it. But, know, I have in the past studied Marshall McLuhan, who's this kind of media theorist and he's one of his more famous quotes is that this idea of the medium being the message. So when the medium changes, how does that message change and then how different things communicated and ⁓ shared? So, yeah, I'm also getting a of a sense that there'll be a reversion to very small local instances as opposed to all these big global brands. And like what are the tools that are like locally important for you? There's quite a few experts that think that ⁓ won't exist in the future. We'll just have an agent that manages everything. So how many of these products or apps or tools are actually really important to market individually, as opposed to maybe actually connecting them to an agent prevents everything for you anyway? So you don't need to consider that. I'd say that's probably a very possible scenario. where we're no longer marketing to people anymore, where we're communicating to agents in a way so that they can make an informed decision. like, hey, is actually this a good fit based on the needs of who I'm building for? So yeah.

Tay Pattison: Yeah, interesting. We have this, tend to agree with you. And need to separate out what I would really like versus what think will actually happen. ⁓ And I there's some overlap there, but I see a in, ⁓ well, let me this in kind of like my thinking or at least ⁓ put the context ⁓ is that I humans ⁓

River Roberts: Hmm.

Tay Pattison: incredibly visually driven. brains, like than half of the neurons in your brain are visual. we're storytelling animals. And I think that anything that is specific to humans will have some visual element ⁓ or have some visual element and needs to have some storytelling element or some ⁓ kind connection to something that is longer that is a larger story. What would really like is to have a device that doesn't have a screen so it's not trying to steal my attention from a visual perspective, but that it's a tool that goes and does things. in that case, the like...

River Roberts: Hmm.

Tay Pattison: Agent that is on there that matters everything and has all these apps and does the things that the apps would do The way that it discovers that stuff is not how I would discover things and what interests that agent is not what would interest me and I think that my time horizon on this like 10 years out, but I would The pace at which everything is moving. I was short on that and say we're probably like ha whatever I think actually going to be the case. think things are going to move really quickly. ⁓ Like ⁓ see the start of... ⁓

River Roberts: Yeah, I was just wondering. Sorry, go.

Tay Pattison: So go ahead.

River Roberts: ⁓ little laggy this first episode, but we'll get there. No, just picturing, what does it look like? Are agents going to respond to, like, witty, instead of copy, like witty binary code or something? Like, is an agent going to be persuaded to explore one tool over another? Because it's got, like, ⁓ an aspect it that kind of speaks to its character ⁓ in a way. A little bit like how... we are kind of vulnerable, susceptible to a really beautiful with some catchy copy on it. you you look at the ⁓ campaigns in 60s and 70s, like, maybe not the 70s, the 60s, but they're just like so beautifully crafted in many ways. Yeah, I wonder what the the agentic layer of that might be. Or is it just like pure code efficiency? Like, yes, that's the piece.

Tay Pattison: Yeah. Cool. No, this is exactly what I was getting at as well, is that you look at, do they call the agentic media all the floors were on and that ⁓ they started making own language.

River Roberts: M-M-M-Micebook. Yeah,

Tay Pattison: Malt book yeah, okay, so on Malt book you got you got like these agentic means and then some of it is like They start using their own language or they start like making up their own language. It's like why are you translating? are these agents agent to agent ⁓ to all this trouble to translate into English or? Mandarin you know Russian any ⁓ when they Can just in a different way like meta set up these ⁓ agents that ended up pulling the plug on because got nervous about it because these two agents started talking to each other and then they started ⁓ essentially negotiate and they made up a language ⁓ when was decrypted was like, they were using balls and balloons, I ⁓ forget details, but basically like ⁓ they negotiating, these two agents were negotiating using some ⁓ made form of know bits and bobs and... ⁓

River Roberts: Balls and balloons, definitely male, definitely male birds.

Tay Pattison: things that... ⁓ There's also this whole, we won't jump into it right now, but there's this whole thing of encoded, the that the ⁓ creator their own personality into ⁓ something they create. So yeah, Balls Magic is definitely made by man. ⁓

River Roberts: I feel like that point leads us quite nicely to our next segment, which I'm very, very excited to explore, actually.

Tay Pattison: Me So ⁓ got a bunch of, ⁓ got a bunch of AMA questions here. And they are written by respective agents that we have been talking to for a So they kind of know quite a lot about us. And I've got, I haven't ⁓ read through either of these. a AMA questions tie and there's an AMA river.

River Roberts: Thank

Tay Pattison: How should we do this? like, I'll post your AMA questions in the chat here and then you'll be able to see them and we can just ⁓ out the ones that are interesting. ⁓

River Roberts: No, can actually pull it up on my phone, but I think AMA sections are always great, particularly in these kind of like weekly podcasts formats. And obviously considering this is our first episode, I think it's great that we have our agents helping us out with the AMAs. However, having said that, even once we get an audience, I think it would still be really nice to have at least part of the AMA asked from our agents. And even the different guests that we invite. We can invite them to ask the agents what questions they want asked. So I think this will be a really interesting as we continue to grow this side of it. I say let's pick two questions each from our own list and then one question from ⁓ each list.

Tay Pattison: Okay, that sounds good.

River Roberts: And and then if it's and then and then if and then either of us feel really compelled to also answer the question on the other side, I think let's let's be open to that as well.

Tay Pattison: Okay, sounds good. I need a second to read through.

River Roberts: All right, you want to key this up?

Tay Pattison: I've got different categories here. I've got ventures and building, nomadic life, inner work, money and strategy and wild cards. I'm gonna go with one from building. agent ⁓ alright gonna give you AMA questions, you like saucy and Would you like vanilla? ⁓ Or would like somewhere in between? And I asked. I asked for vanilla questions. All right. It's a little bit of a... pick a one at random. How do you feel about discipline following energy? Because two things seem to like they're in constant conflict.

River Roberts: Hmm.

Tay Pattison: is pretty relevant. I want to flow with whatever ⁓ I'm on day and I forget says maybe it's Mark Andreessen. And I've also that I'm king of ⁓ this...

River Roberts: attributing chords.

Tay Pattison: attribution. So thank you Mark if this is you, it was somebody else. ⁓ appreciate you anyway.

River Roberts: We can fix all that in the show notes.

Tay Pattison: We get run this at the end and have it in the show notes, the missed attributions and where they really came from.

River Roberts: think that sounds like a game show that we should do at some point.

Tay Pattison: We could have the eye. said. So how do I think about versus energy flow? ⁓ think they both have their place and I think it's on the individual too. ⁓ on you and I to ⁓ what works really well. Somebody I admire recently, we were talking about the gym and we in the gym together and ⁓ it about 3.30 p.m. and I was really grumpy. I saying, ⁓ I'm... I'm grumpy, it's the middle of the afternoon, this isn't my best time of day, and then we're like, you know what Ty? Being tired ⁓ not an important factor right now. Like here to do this specific thing, ⁓ and I started thinking about it like that, like yes, ⁓ follow greatest excitement, yes, ⁓ work things that really motivate you, but... It's helped me a lot to focus on discipline when it needs to happen. For example, if I know that I'm going to go and do exposure therapy like hot cold. Instead of negotiating every time and spending the energy on negotiating whether or not I'm gonna go into the cold It's like no I'm here and I'm gonna go in and I'm not even gonna think about whether or not I'm going into the cold plunge I'm just gonna go into the cold plunge and actually having that amount of discipline Jocko willing talks about this all the time like the discipline will set you free and I think in some areas being very disciplined really helps and allows you to flow with the energy so It's a bit of a both are important, but I think in the right place. And I think discipline, consistently up to do your best, working for number of hours every day and ⁓ doing it sprints like have been, like having these frameworks for discipline really help us be able to ⁓ yeah, flow.

River Roberts: coming up in response to that is this Bashar. And he about five different principles. I'm going to mention them all here. But he talks about do what really excites you, what really lights you up, and do it to the best of your ability. That piece like to the best of your ability. I feel like there is some discipline embedded in that. However, when there is like a real reluctance or resistance, that maybe that might be the limit of your actual ability, because I also look at like what we've been doing in the past few weeks and ⁓ could just keep crunching and burning through, but, ⁓ can be saying that's to the best of our ability. However, it's also a really natural. boundary that's actually healthy in many ways. So figuring out what is discipline and what is, is maybe quite, is kind of neglecting the reality of what's to the best of your ability, because I see a lot of people that are very disciplined and they're, you know, these hard charging types and you know, they work 70 hour weeks and, and I'm not sure that that is the healthiest way to go because in the whole, um, in the whole span of things. the relationships that suffer are huge. Yeah, it sounds like ⁓ some disco music skipped up at your end.

Tay Pattison: a party. Yeah. Absolutely. I think the difference though is the inputs versus the outputs and this is easier said than done but it's if I am disciplined about the inputs try to let go of the outputs then in I'm disciplined about the amount of time that I spend showing up consistently my best then ⁓ letting go ⁓ it generate this much money it will get this many is it will make this many ventures and focus on the things that I can control. I think that's probably where that line is, for me at least.

River Roberts: Let's move on. So I've got a question. I've actually got a question for both of us. The question is titled, The Bet. Each of you name one prediction about AI and building that you're willing to be publicly wrong about. The loser buys dinner. I'm not exactly sure who the loser buys dinner for entirely, what do you put in the table?

Tay Pattison: Yeah, what do you got?

River Roberts: It did also suggest binding it in Lisbon. So I'm not sure if it would be geographically tied to that. yeah, one prediction about AI and building that you'd be willing to be publicly wrong about. Maybe let's put a time frame on it. What do you think, like six months?

Tay Pattison: You Yeah, let's go six months and willing to be publicly wrong about, I think this ties back to the learning in public thing. I'm trying to let go of being publicly right or wrong about stuff and just publicly right or wrong about it. ⁓ I think it's very freeing.

River Roberts: Hahaha!

Tay Pattison: something that it's a party here ⁓ a prediction that I'm gonna make is that in the next months we will see a ⁓ be framed there will be companies ⁓ that have agent CEOs and I don't know how publicly the company will have an agentic CEO but it'll be for example a marketing agency that maybe has a person behind it but the actual thing that runs it will be an agent that agent makes all the decisions and asks the human to make and writes posts and orchestrates all of the ⁓ humans. ⁓

River Roberts: Yeah, I mean, that's actually a prediction that you don't even need to write because it's already here. already currently the place. There's actually entirely AI built organizations that have ⁓ basically ⁓ meat puppet ⁓ so that they can legally register in a place. And these ⁓ meat CEOs are just there for show. They don't do anything actually, other than just sign a document. So that's already here.

Tay Pattison: What are the... Do you have any examples? these working? ⁓

River Roberts: Yeah, yeah, they're working. I don't have them off the top of my head, but they talk about them the ⁓ podcast. Yeah. So definitely check that out. ⁓ right. ⁓ will.

Tay Pattison: there.

River Roberts: All right, so a prediction that I would like, so in six months, think that we, the two of us and this Animate's project we're building, will actually be able communicate with animals. So you'll be able to take an audio capture of what your cat is mewing or your dog is barking, ⁓ and you'll able to ⁓ communicate back it in a way that ⁓ people can't at the moment. I would love to be publicly wrong about that, but I'd also love to work having that possible. ⁓ know there are some apps that starting to already do that, but it seems to be a little bit more of gimmick than an actual piece. I think what we're building with Animates in terms of having the ⁓ breed base, all of the information around vet visits, all that stuff, then keying in the observations that owners make over a period of time will give it the context that it needs to really start to reform this communication layer. So yeah, that's my prediction.

Tay Pattison: thing that Dr. Dolittle

River Roberts: totally. But for everyone, not just a doctor, anyone will have capacity for that. and my, my excitement around that is, you know, so many people already treat their, their, ⁓ their they live with. I'm not even going to say pets, because I think pets is kind of demeaning to the relationship anyways, but the animals that they live with, they treat them like family. and kind what we've learned in our research with animates is that although they, think of them like family, they're the way that they treat them is, or the way they understand them is still a bit scattered and haphazard. And if we can provide a tool that allows people to be more towards the preventative side, particularly of healthcare with their pets, then this ends up being a really valuable piece. So yeah, I'm excited to ⁓ move resources energy in that direction. I'm super excited for that. And to like improve these family relationships and like such beautiful family relationships. you know, when that happens on a household scale, how does that happen ⁓ a more of a global scale across other species of animals as well? ⁓ super fascinated by, by that perspective. like for instance, you know, this, ⁓ this in Bali, a new bird's nest just arrived. And I would love to set up a little speaker to communicate, Hey bird, you're welcome here. You know, make yourself at home. I'm looking forward to meeting your kids. I can't wait to, ⁓ know, this kind of like, and then receive a voice message back from them, be like, hey, thank you so much for the warm welcome. ⁓ feel really at home here. We've our ⁓ first ⁓ after ⁓

Tay Pattison: You love that.

River Roberts: So in six months, that's what I'd love to be publicly wrong about, but I'd also love to be publicly right about it as well. right, what's the next question you got? think considering the techno music in the background, we've probably got one question each.

Tay Pattison: Alright, I got one for both of us. So, and it's a little bit of a lifestyle one, but since this is given our agents and also get to know us, it's what is your, what's your non-negotiable travel kit? The three to five you'd use.

River Roberts: Yeah

Tay Pattison: cherish the most when you travel.

River Roberts: Easy, easy, So three, right? Or three to five?

Tay Pattison: Three to five, up to you.

River Roberts: Okay. So, of mine is definitely a hammock and I just find having a hammock to travel with is just invaluable. ⁓ one of those, not one of those like hippie-dippie like, rock multicolored hammocks, but one of those are really nice, simple, green or one, one color. It's like a nice, nice fabric. I've been practicing quite a lot a, as a way, ⁓ as, as just space ⁓ to be more creative. 20 minutes a day, I think is such a beautiful rest on the body and then ⁓ on the mind. But then I'm fine. find it really quite a creative and grateful space to be. So having this, definitely one I always trouble with. I have these really nice traveling chopsticks that unscrew in the middle. So those like my traveling utensils. And then I've actually just upgraded, but a water bottle. I've got a ⁓ traveling thermos now. wasn't water bottle. And now it was a stainless steel water bottle and now I've made it a stainless steel thermos. So those are my three that jump out immediately.

Tay Pattison: like that. They're very river. I've got a little story ⁓ I kick this off because there's all the practical stuff. You need a passport, need a wallet, etc. My is that before I left ⁓ Portugal about months ago, I had a conversation with a friend. and he was saying that he was about to go on a trip and that the day before he left he couldn't find his wallet we laughed about it and it was so silly and then I went to leave and I couldn't find my wallet and I left Portugal without a wallet in November and ⁓ went Japan, Zealand, ⁓ Mexico, Brazil ⁓ and Without a wallet, no ID, no cards, been using Apple Pay and I've it through all of these places so wallet not My three things ⁓ are a ⁓ of kind, a ⁓ fold out knife for ⁓

River Roberts: But how do you travel with that? How do you even get through security? Or you got to check it in?

Tay Pattison: Most of the time I check it in. actually got a couple of like just... ⁓

River Roberts: you swallow it before you strip.

Tay Pattison: I've got a couple of OpenLs. And I used to have one like a... kind of old fold out knife that I bought in Pakistan a few years ago and I would travel with it in my carry on and it would always make it through security every time and I'd be like wow this is really amazing like I should not be allowed this on the flight and then one day somebody pulled me up on it and they were like hey you can't have this and this had happened a couple of times before and I've been like no well it's made it through before it's fine like I'm not gonna hurt anybody with it and they had let it through and then this one time this one person was like, no, sorry, you can't take this. So I've lost a couple of these, especially things that I like. But this is a, it's just such a useful tool. Carrying it around, like taking it for picnics, eating fruit, it's often useful for all sorts of things that I don't think I'm gonna need it for. having a camera, I have a camera that I travel with and a couple of disposable cameras that have been going for... maybe a few months now and what I like about that so much is that it gives me a specific lens to view travel through. So I'll go to a place and instead of just seeing the place or seeing people, I'll like really have an artistic lens for like the way that the light hits something. It's really nice. Or ⁓ the disposable cameras are really nice because I'll take them throughout multiple journeys

River Roberts: Hmm.

Tay Pattison: and then I'll go and the film ⁓ when it comes back it's like Christmas. It's like I get all of these snapshots from the year that's been.

River Roberts: Bye.

Tay Pattison: And the other one, and this is always a little bit too heavy, but I take it always anyway, is ⁓ a speaker. ⁓ speaker ⁓ just ⁓ having a dance in the kitchen in the morning, having some ⁓ of way to create a vibe or to like bring calm to situation, just like having a way to create your own atmosphere.

River Roberts: Hmm.

Tay Pattison: is a big one for me, yeah. I really enjoy it. And coupled with that is every month I make a playlist and it's the songs that I've most enjoyed that month. And then I can time travel so I can go, oh, I wonder what it felt like in August 2025. And I can go to August 2025 and play the playlist and it brings me back into like summer 2025 or

River Roberts: Hmm. ⁓ wow, incredible. That's so cool.

Tay Pattison: Christmas 23.

River Roberts: One of my favorite music apps is called Radio with like five O's at the end. you pick a country on the map and the decade and it plays music from that era basically. So I can already see your time travel version of that, but personalized. think that's brilliant. Just a really simple, like a really simple slider time travel. So cool.

Tay Pattison: be the kind of thing that Spotify, if anyone from Spotify is listening, I'd like that as a wrapped or some kind of feature of ⁓ were your most listened songs in this month, this month, this month. So you can pick a and go and generate a playlist from it. Pretty please.

River Roberts: I think we are.

Tay Pattison: Looks like we have lost connection...

River Roberts: All right, it's back.

Tay Pattison: Hey, there he is.

River Roberts: Just a little technical interlude. right. All right, last question. Well, at least for this one. But I'm actually really enjoying these questions. So this question is called the abundance paradox. definitely ⁓ both of us to answer. ⁓ talk about building towards an abundant future. But every builder I know ⁓ running on caffeine, anxiety, and the fear that someone else is shipping faster. Where does the abundance actually start? With tools or with you? Now, I find this question very interesting because I'm not sure how many builders my agent knows in that context. But I'm wondering what that brings to mind.

Tay Pattison: It says two things to me, and I think this is something interesting to highlight for anybody that's not working with agents as well, or not working with AI, is it's super clear to both of us immediately what the AI tendency is here in the question. ⁓ is just a little bit of an aside, but that builder I know ⁓ is very sounds ⁓ very, like it knows exactly what is right, which is not true. It's made up. very authoritative, exactly. And ⁓ I immediately pick up in questions that it asks me in writing, using verbose using actually, using all of these things. Anyway. What does abundance look like? I think it's... ⁓

River Roberts: Yeah, No, no, no. It's not what does abundance look like. It's like, talk about, you about building towards an abundant future, but every builder I know is running on caffeine anxiety and the fear that someone else is shipping faster. Where does abundance actually start from? With the tools or with you?

Tay Pattison: it stop. Yeah, starts with you. It a mindset, ⁓ but something that needs to be cultivated. And I that in humans it happens slowly. Specifically what comes to mind is there is ⁓ a... There's this ⁓ it's ⁓ Martex ⁓ idea is that technology moves, ⁓ evolves at an exponential rate, but move or learn linearly. So these systems evolving this, ⁓ we're moving linearly, there's inevitably a between what the technology can do and how fast ⁓ I, as an can keep up with it. I think it's my responsibility to create the conditions where I feel abundance and where I can use these as tools than being directed by them. it's things like ⁓ what it's ⁓ a specific skills is noticing ⁓ when being led down the rabbit hole ⁓ by AI that is just continually asking questions. noticing when I'm being trained by a social media algorithm to get a variable reward from swiping the ⁓ that I really help having some kind of routine some kind of grounding ⁓ so I can't get swept up in this stuff ⁓ and deciding

River Roberts: You're good for it. But like specifically been bringing you to more of a sense of abundance ⁓ more like thinking through that abundant lens?

Tay Pattison: Honestly, not feeling abundant. Feeling I don't have enough time. Feeling like ⁓ everything moving too fast. Has been the trigger to ⁓ make me to focus it. ⁓ What really gives sense of abundance ⁓ is... giving myself a pat on the back, giving you a high five, like going, hey, that was pretty good. Like benchmark this against what ⁓ it might've ⁓ six ago. wow, that's amazing what we did today. This morning I someone through a basically fully automated podcast set up from scheduling to post production ⁓ like creating all of stuff. They were like, wow, and that takes how long? 10 minutes per episode rather than like hours and hours. It's like We actually can do so much as as we set those criteria for I think that's what it is, like really ⁓ deciding ⁓ those What you find yourself? What comes up?

River Roberts: I feel like there's two modes of abundance that I feel actually know this, this three now that I think of a jump out immediately abundance in time, abundance in resources and abundance in relationships. And, um, yeah, I definitely like to move towards things that create abundance in all three of those. Um, like the idea of And there's, there's, there's, they often cross over a bit as well. So like one example I've, I've, I've kind of, um, done in the past and will continue to do in the future. One, one was actually born out of, um, a flat tire. So I remember getting a flat tire and at my immediate response, like, why did this happen to me? Like a flat tire, like I can't, and then I've, I've kind of flipped the script and said, oh, yeah, that's it's going to happen. But that's okay. And then, so. Got the car towed, got to the place to get it repaired. And then I paid for two flat tires to be repaired. And person, the place is like, what are you talking about? I'm like, ⁓ so is for me and the other one is for someone else, for you to gift. And just all of a sudden, like the relationship became abundant and our became abundant. And ⁓ then for rest of the day, they then had this like more abundant mindset. And presumably whoever received that gift had that kind of sense of abundance that was gifted to them as well. So just these small little acts and like I'm in the process now of a little farmer shack in the rice field here in Bali. you know, it's, it's a couple hundred bucks to rebuild it. But, you know, the guys I'm working with and just the, the farmers in the area, like it kind of creates this sense of abundance. And ⁓ my hope is. you know, the fact that I've supported in building this thing, it becomes kind of a space, a physical space that feels abundant already from the start. And the fact that it's in this rice field looking over what I would consider like one of the most abundant looking spaces there is, it's, quite incredible. So, yeah, I definitely think it's small little acts that you can do, but I guess you could also call those acts tools in a way. and I keep coming back to this idea that

Tay Pattison: you you

River Roberts: You know, man makes the tool, but tool makes the man. So it's this kind of recursive so to speak. And, you know, what of comes first and what informs next each stage. And, you know, it's kind of this evolving piece, which is another thing, another reason I'm excited to be building tools with you specifically, because I feel like these tools that we're building, you know, do allow abundance to emerge for a more people. ⁓ If it can be abundance in time, abundance in resources, and abundance in relationships, what a beautiful to be investing ⁓ towards. ⁓ very easy for me to get up in the morning and put energy towards this. ⁓ optimistic. It feels like it's moving in the right direction. And we just have such an exquisite set ⁓ tools that are emerging before us and to be able to help.

Tay Pattison: you

River Roberts: make those tools as they help make us as opposed to just being victims to the tools in a little sense. Like I look at the social media era and I feel like the tools are incredible. However, a lot of people have fallen victim to the tools I'd argue that entire social cultures and societies have fallen victim to the tools. The way the social media algorithm works, it spirals down instead of up. So yeah, that's where see the great opportunities ahead. with that being said, I feel like it's a pretty nice way to wrap up this first episode of The Great Spinoff. What do you reckon?

Tay Pattison: that is a fantastic ending yeah love it ⁓ first episode ⁓

River Roberts: Well, brother, thanks for your time. I so grateful to be building with you and I'm ⁓ also grateful to be learning in public with you and I'm excited to start to invite some other people into this space to learn from and ⁓ to with. So onwards and