David Friedberg: All right, everybody, welcome back to Twist. It's Friday, February 6th, 2026. And today we're gonna share how we built OpenClaw Ultron.

SPEAKER_01: This is a new project inside of our firm, Launch and This Week in Startups, where we produce podcasts and we invest in 100 companies a year. What are we trying to do? We're trying to build one instance of OpenClaw, formerly known as Multibot, formerly known as ClawedBot. We're trying to build one replicant, one agent that can do all 20 people's jobs here at the venture firm and at the production company that does all these podcasts. 20 people's jobs, each of those jobs probably has a half dozen important skills. So we're talking about, at some point, putting together in one agent, we call them replicants, we're gonna have somewhere in the order of 100 to 200 skills. That one person is gonna try to do everybody's work. That's the goal. And then everybody will level up and do some other work. So the goal isn't to replace everybody, it's to take away everybody's chores and to make everybody better at the primary functions in an investment firm, which is meeting with founders, spending time with founders and LPs, our investors. And then on the production side, it would be producing great content and working with our guests. We wanna move up the stack and give away all the chores. With me to discuss it, Lon Harris, who's gonna co-host the show today. How are you doing, Lon?

SPEAKER_03: Doing great. Great to be here.

SPEAKER_04: All right. And Oliver Corzan is here. He has been doing demos for me and producing This Week in AI, which is gonna launch in February in two weeks, I think. Oliver, welcome to the program. Thank you. Good to be here. And we have a special guest. Alex Chima is here. Alex, I have been following for some time because maybe a year ago, I saw Alex was working on stacking with his company. It's XO, right? E-X-O?

SPEAKER_01: XO is how you pronounce it? Yeah, XO. And you've been working on taking commodity hardware like a Mac Mini, daisy-chaining them or connecting them together in order to run large language models locally. But as we've seen, OpenClaw,

SPEAKER_04: formerly Clawbot and MaltBot, has quite a wrinkle in this. You were like a year or two ahead of this trend of, hey, can we run locally? So let's start just really quick, Alex, before we go into Ultron, OpenClaw, Ultron, what your firm does and what progress you've made, especially in regards to OpenClaw.

SPEAKER_06: Yeah, thanks so much for having me, Jason. So I'm the founder and CEO of XO Labs.

SPEAKER_08: And like you said, we've been doing stuff with Mac Minis long before OpenClaw was around. And to be honest, I didn't expect the rise of people buying Mac Minis to come from this place. I thought the catalyst would be people wanting to run models locally. What we do is we make it possible to run Frontier AI locally on consumer hardware. So not just Macs, but also other kinds of consumer hardware we're trying to drive down the barrier to running the most capable AI models. So we currently have like the cheapest way, cheapest, most accessible way to run KimiK 2.5 on two Mac studios. And we're working across the whole stack. So we're working on the model layer, the distributed algorithms as well that are very different when you're working with consumer hardware and also like lower level like kernels. And our goal is basically to make Frontier AI accessible to anyone to run on their own hardware.

SPEAKER_10: Why is this important? Why is it important to run it on local hardware?

SPEAKER_08: Yeah, I think this is something that with the whole OpenClaw phrase, not a lot of people are talking about,

SPEAKER_12: but just how the way we're using AI is shifting. And it's going from being this kind of crude tool

SPEAKER_08: that you use through like a chat interface to becoming sort of an extension of yourself. And the AI now, it knows everything you know. You know, it can basically do everything you can do digitally right now. And soon, you know, with robotics, that's going to be physically as well. And at that point, it's more of an exocortex. So it's not just this like tool that you talk through a chat interface, but it's this thing that's actually part of yourself. And then you start to question, okay, you know, do I want to rent my brain and Andre Karpathy talks about this. He says, not your weights, not your brain. Like, do you really want, you know, another company, a profit-seeking company basically running your brain? And when you think of it like that, to me, you know, my reason for starting Exo is you want control and you want ownership of that. OpenClaw is a long way towards that because for a while, the products were getting better. These closed source, like the models are largely like commoditized and there's a pretty standard, pretty thin API layer to interacting with them. So the switching cost is quite low. But what worried me was that the products, the closed source products are getting a lot better, like we chat TPT with memory systems and also the more stateful aspects

SPEAKER_12: of like the workflows that you're building. So now the fact that you have OpenClaw, which is open,

SPEAKER_06: well, you can run it on your own infrastructure. Now a large part of that...

David Friedberg: To summarize that, I thought you were going to say, well, it's cheaper because you're not paying for tokens. That's what I thought you would say first. Then I thought you would say, well, you know, you can put so much data on it, you'll have better memory. But you went with a really even higher, bigger picture reason to do this, which is if you put this all in OpenAI and OpenAI has a trillion dollar valuation and they need to make money, if I put all my venture capital data in there and I trained it with all of my secrets, those are all going to accrue eventually, even if they say it's not going to.

SPEAKER_04: You have this very reasonable fear or concern that it's going to accrue to OpenAI, to chat GPT, not to your firm.

SPEAKER_15: So that's the reason really to do this yourself. Yeah? In your mind, Alex?

SPEAKER_16: Yeah, I think that there's a nuance there of just like, I actually don't believe in

SPEAKER_12: sort of the privacy argument so much of like, I think at least for consumers, you know, we're already putting our data into platforms and we're completely fine with that.

SPEAKER_08: But it's more about the sovereignty aspect and actually having control of it. So how easy is it for you to switch? How easy is it for you to like,

SPEAKER_21: if the model's changing under your feet, how much control do you actually have? So that's lock-in

David Friedberg: and lock-in for a chat GPT, I just experienced because we canceled our OpenAI account and we moved everything to Claude because we felt Claude was a better product. And we felt like we trusted that organization a little better. When we moved it over, I had three people say, oh my God, I have all my stuff there. And I was like, really? And they're like, yeah. So I turned their accounts back on so they could get it. But there's not like an easy way to get your memory out of there and bring it over there.

SPEAKER_24: Well, we saw the same thing with the GPT-4 moving into 5 that a lot of people like, they lost the magic that they'd loved about GPT-4-0. So it's like, you know, the models can just sort of change your upgrade on a whim and then you lose this, you know, like character persona

SPEAKER_26: you felt like was part of your life in a way.

David Friedberg: So now, Oliver, it's your chance to shine. Oliver has jumped in in the last 10 days and gone all in on OpenClaw. One of the things we did was we built a persona, the first one, to work on the production of the podcast, doing guest research, guest outreach, and to figure out what should be on the docket. In other words, what topic should we discuss? And on the margins, hey, what should the title of this video be? What should the thumbnail be? And just trying to see if it could do those functions. Oliver, you've been working on this.

SPEAKER_04: Show us the state of the art now because I think the first time we did this was last Monday, not this past Monday, but the Monday,

SPEAKER_29: two Mondays ago, yeah? This is the end of week two of our round-the-clock ClawedBot coverage. Crazy.

SPEAKER_31: Okay, Oliver, let's show what you built. It's been around 10 days since we first started building

SPEAKER_32: our instance of OpenClaw and as you mentioned, we have two different ones, one that's more focused on the investment team and I am building an OpenClaw bot that is kind of more focused on the production side of things. So one thing that I think was a little bit of a misstep that I would tell anyone who's building a new OpenClaw is to start with a dashboard. That should be kind of your step one once you get your OpenClaw online and a dashboard as you think about it, but it is able to connect to the backend of your OpenClaw instance and bring in the data so you can see it visually, bring in all the files. It's just being able to look at it visually is much better than trying to interact with its backend and obviously its frontend all just from a chat interface. So doing this was very easy. So I was watching an Alex Finn video who we had on last Monday and Alex Finn was interacting only in his dashboard with his OpenClaw. I basically was like,

SPEAKER_37: why are we not doing that? Because OpenClaw doesn't really have a dashboard.

David Friedberg: You basically are telling it, hey, remember this, you know, make a file here, but you don't understand the underpinnings. There isn't a dashboard. So it would literally be, this is early on. OpenClaw is essentially a black box. You have all this memory and you have skills that you have to query it to understand. But you made a dashboard. The dashboard is going to show

SPEAKER_04: what files it has in memory and an example of a memory file would be what in our case.

SPEAKER_32: Yeah, so the example of a memory file would be Oliver's preferences. What are my preferences? So this is in the memory. Never use em dashes in emails. I don't want that to happen. Okay. I want you to be a person. Don't put direct competitors on the same show when we're booking a podcast episode. And also at the moment, we're not booking VCs on This Week in AI. So these are all things that I've told it. These are my preferences when I'm doing tasks throughout the day.

SPEAKER_37: You don't want to repeat yourself and say, don't put two competitors on the same episode. You don't want to repeat yourself with these specific instructions on booking guests. Got it. Yes, exactly.

SPEAKER_32: And it just kind of keeps, you know, things I've told it and it's in mind. So if I ask it to do something, it will remember what we talked about. Example of a shortcut that I gave it was, I basically wanted it to understand who were the pending calendar invitations that we had

SPEAKER_31: while we were booking them. So there's, you know, a handful of guests that

SPEAKER_37: if you have guests that we've invited and they haven't responded to the invite yet, you want to know that. You call that pending?

SPEAKER_32: Pending calendar invites. Yes. And in order for the bot to be as helpful as possible, it needs to understand who those guests are, which are the ones that it needs to look for the email to see if they have responded yet or have I responded to them? So these are the type of things that you would keep in your memory.

SPEAKER_04: So memory is the first thing on the dashboard. I think we understand that preferences or different pieces of data. Now, some of the memory, could that exist on a Notion page or in a Google document? And would that be represented here? Or is it only memory and files that are stored inside of OpenClaw? These specifically are only stored

SPEAKER_32: inside of OpenClaw. Of course, they can reference different databases that you have. But the kind of the big point of this show is to show how we have created our OpenClaw Ultron to replace 20 employees

SPEAKER_31: at our company. So obviously, that's the end goal. I still want to have a job. I'm sure the blonde wants to have a job. There'll be more for you to do.

SPEAKER_46: We want to launch. We have... Here's the thing. There's two... If you think about your job, you've been doing a bit of production here. Of the production hours, hours you spend on production, at this point in week two,

SPEAKER_04: how many of those do you think you'll wind up handing off in 30 days, let's say, if you just keep grinding on this for another four weeks? In 30 days, what percentage of the work you're doing in total hours? So if you work 50 hours a week, how many of those hours would be done conservatively or optimistically? You can give one number or two, just conservatively, optimistically by this new Ultron.

SPEAKER_32: I would say around 60% of my time if I'm doing 30 hours a week on production. Something you mentioned earlier is that there's probably hundreds of tasks that people do at our company. So in order to build out all of those skills that can do those tasks, we're going to have to do that one at a time and we're going to need to make sure each one works. So I have around nine or eight tasks

SPEAKER_40: that I have successfully or I'm in the process of building out.

David Friedberg: Okay, and those are called cron jobs. These are jobs that occur on a chronological, on a time basis. That's what cron job means. And cron jobs are something, Alex, that developers do all the time, but knowledge workers don't typically have cron jobs.

SPEAKER_57: Right, Alex?

SPEAKER_08: Well, I don't know. I think this is one of the more interesting features and one of the things that like to me, OpenClaw is like putting together a lot of things that already existed in a very intuitive, seamless way. And one of them is cron jobs and I'm using them. I'm using them for like loads of things, not just dev stuff, but like a lot of management. So we're like, I have something that's like constantly scanning our Slack and basically making suggestions once it's, I have kind of like this way of quantifying like uncertainty about tasks. So I think this is something that the LLMs are like getting better at is like knowing when to be proactive. And so, you know, like basically I'm giving it as much context as I can from the Slack so that it can suggest every day a list of things that we might be missing

SPEAKER_21: or some things that we should be aware of. So this is running just on a cron job every day, basically.

SPEAKER_59: AI is revolutionizing every aspect of our industry, but for founders, it can prove very frustrating,

SPEAKER_61: largely because cloud costs are so unpredictable. This has probably happened to you. You've planned out and budgeted a certain amount for a project and then been hit with a massive bill with no real explanation for what went wrong. If you're feeling like the cost of working with this innovative technology is just too high, finally, we have a solution for you. Crusoe. They are the AI cloud company that's taking an energy-first approach. What does it mean? They efficiently access their power directly from the source and pass all the savings along to you. This isn't just about saving some cash. Crusoe's architecture is specifically optimized for your AI workloads and their customer support response times clock in at under six minutes. Here's what I want you to do. Visit Crusoe.ai slash savings to receive $100,000 in credits for virtualized NVIDIA GB200 NVL72 on Crusoe Cloud. Pending availability. That's C-R-U-S-O-E dot AI slash savings.

David Friedberg: Yeah. So, and when you see Oliver with the memory and the files, what comes to mind with XO and, you know, standing up to, you know, Mac Studios, the M5s coming out and how much memory you could put in there, I was telling the team, I want to take the Notion API and I want to take the Slack API and I want to put into memory every single Slack message this year, maybe even over all eternity and, you know, somehow have that all in here. So maybe you could, you could speak to that memory because you already spoke to it in terms of like giving it to OpenAI or another company versus keeping it for yourself. But how do you think about large amounts of data?

SPEAKER_12: Yeah, this is definitely a big focus right now in terms of inference infrastructure is just how do you support really big context with, you know, basically being able to put everything in context.

SPEAKER_08: And the way I look at this is you can look at, well, inference consists of two stages. There's like the pre-fill stage which is very compute heavy is compute bound and you have the decode stage. And what you're seeing is that most use cases at the moment are very decode heavy. So it's actually most of the time is being spent on just generating tokens. And I think the software is actually really good now at kind of making sure that when it comes to the pre-fill you've got, you're getting a lot of cash hits. So I think basically we'll be able to continue just increasing context, context, context quite a bit. And basically the hardware is more of the focus is going to be on the decode side. That's why consumer hardware is really good. You have the M5 coming out pretty soon. It's a big boost in memory bandwidth and memory. And all of that side of things is super memory bound. So I don't see any like reason why you couldn't just shovel your Slack messages into context. I think that's going to happen.

SPEAKER_04: And we should just buy when the M5 comes out max memory, which is what, 500 gigs of memory?

SPEAKER_18: Yeah, it's 512 at the moment. Maybe that will increase as well. And it's enough to fit really large

SPEAKER_06: models, enough to fit all that context as well.

SPEAKER_77: This is always,

SPEAKER_24: I feel like the sort of the dream, like when producer Claude, we first brought that on board from Anthropic to the show, that was really what we wanted. He should listen to everything we say and remember it and then throw in helpful suggestions. The technology was not quite there yet, but I feel like now we're on the precipice of

SPEAKER_26: actually being able to do that with an AI.

SPEAKER_80: Okay, so let's go to the cron jobs here real quick. Maybe you could give

David Friedberg: us an example of a cron job. And I'm guessing each one of these skills is, you know, if it's been two weeks and you've got eight working, you're basically so or one every 1.5 days.

SPEAKER_04: So that seems like a pretty good pace to me. If we have 200 skills, we're going to give this eventually.

SPEAKER_83: That's a pretty good pace.

SPEAKER_24: There is a trial and error. I sort of have written one skill so far for the Ticker Digest, and you do have to tell it what to do, see what kind of feedback you get, and then there is a tinkering to get the prompting and get everything exactly the

SPEAKER_26: way you want it for sure.

SPEAKER_04: Okay, so let's look at, hmm, how about attendance? I think this is an interesting one. For people who don't know, I wrote a famous blog post years ago called this sort of lightweight management and start of day, end of day as a tool for executives, especially when remote teams were happening. I just asked everybody on our team, Alex, kind of like a stand-up for developers, et cetera, just say what you're intending to get done today, and then at the end of the day, reply to yourself in Slack in the general channel and say what you got done. I had two of my four senior executives at the time essentially quit over this because they didn't want to be micromanaged. And I was like, well, it's just like you're getting paid a very large six-figure salary. You can't spend five and ten minutes just saying what you're going to do for the day. And that was great for me because I just don't like people who are not good communicators or don't set goals for themselves and they're doing great probably, maybe.

SPEAKER_37: But what did you

SPEAKER_32: create here, Oliver? Yeah, so we all post our start-of-day and end-of-days in one Slack channel called General and two cron jobs. One is the start-of-day attendance where it looks who has sent their start-of-day anywhere from 7 a.m. to 12 p.m. and right at 12, which is in the morning when you should spend your start-of-day what you're going to do that day, it will look through the general channel, see who has sent it, and whoever doesn't send it, the bot will then send a Slack message in the general channel tagging you, Jason, and also tagging the people who

SPEAKER_31: haven't sent it yet. So it's kind of just that accountability. That's a cron job that runs it. And then you do the same thing at the end of the day.

SPEAKER_04: And previously we would have a human do this. They would scroll up and they would spend 20 minutes and they would then go check in with people because that's when we were fully remote, Alex, that's how we figured out who took a paid day off or who was on holiday or if something was wrong, check in on a person.

SPEAKER_92: Building out your team is one of the most crucial things you have to get right in your startup and finding the right developers is particularly important, but now there's Lemon.io. They're going to save you time, money and headaches by doing all the time consuming leg work for you. They've got an experienced lineup of pre-vetted developers working for competitive rates. Just 1% of applicants are accepted into Lemon's elite program and they're not just out there finding this great talent, they're also

SPEAKER_94: working with you to integrate these new members into your team. Plus, if it's not a good fit, hey, and sometimes things don't work out, Lemon will hook you up with a new developer ASAP. I've seen startups go from just pretty good to amazing after filling out their teams with developers from Lemon.io. Go to Lemon.io slash twist and find your perfect developer or technical team in 48 hours or less. Plus, Twist listeners get 15% off their first four weeks. Lemon.io slash twist L-E-M-O-N dot I-O slash twist.

SPEAKER_84: Okay, give us one more. What else is interesting here? Let's talk about self-optimization. I want to hear about that one.

SPEAKER_97: Oh, yeah. I don't know what that is, but okay, Age of Ultron is here. What is self-optimization? Yeah, so this is basically an optimizer

SPEAKER_32: task where this role would previously be an engineer or I would look through all the files. I mean, I wouldn't be able to do this if it wasn't plain language like OpenClaw is, but previously you're looking at an organization, you're looking at the structure, you would maybe want an engineer or someone with a lot of experience to look through how everything's running. So I have set up a self-optimization cron job.

SPEAKER_37: So this is running Monday through Friday and what is it? And did you write this prompt or did you ask it to write a prompt to do this? I asked it to write this prompt.

SPEAKER_32: The goal is, the end goal would be for, you know, from 3 to 5 a.m for it to be looking through all of our files, all of our cron jobs, all of our skills and then at 8 a.m sent me what could we change? So not actually execute yet at least while we're still building trust. It gives me a list of five of the things that it thinks that we can really change and optimize and this was the one from this morning. So it noticed that there was a time zone bug in the guest calendar. So it was getting CST and CDT confused and it said

SPEAKER_40: that it would be able to fix this quite quickly. There were some issues in the...

SPEAKER_04: So it's always good to give the exact one. So that was great when you gave the exact one and had an error there. Give another one. What else is like an exact thing that it said we should fix that was material here?

SPEAKER_32: The self-optimization cron job realized that there was a cron scheduler issue where jobs were skipping days. So it realized that some of today's jobs did not run. And then it went and investigated the scheduling issue and also told me that this would be a medium effort change. So then I told it to fix that

SPEAKER_31: and then it went into the files and made sure that that wouldn't happen again.

SPEAKER_04: Did it give us anything like in terms of this is like fixing its internal guts and everything and the engine but did it give us anything in terms of destinations of where to take the car that could be improved? Did it say like oh you should consider these type of guests for the program or here's how to make advertising more effective? Did it give us

SPEAKER_37: anything like that on a business basis?

SPEAKER_106: Yeah so

SPEAKER_37: the self-optimization

SPEAKER_32: cron job that I set up is specifically looking at how Open Claw is set up but I do have other cron jobs that are exactly that so I do have a sales and sponsors specific task so one of the tasks that a member of our sales team does is they look through competitor podcasts and see who the sponsors or partners are

SPEAKER_40: that are on those shows so we can get ideas you know to bring on sponsorship.

SPEAKER_04: Yeah if we're missing if there's some new sponsor in the world and we don't have them yet you might hear them on the New York Times podcast and we should

SPEAKER_37: probably reach out to them we had a human doing that previously yeah exactly and this basically

SPEAKER_32: works with the YouTube API we'll go through a list of I believe 20 different podcasts that I gave it look through the timestamps and I also believe it can work with Podscribe which I think is a little more curated towards sponsorship and we'll look through the timestamps hyperlink it in a message also it looks through our Pipedrive which is our sales CRM and we'll figure out if we have a sales rep who owns a certain sponsor and then flag them and say hey this sponsor is on this podcast or it will say hey no one owns this sponsor that I found on this podcast and then it will send that daily as a message into our sales channel

SPEAKER_93: great yeah

SPEAKER_04: Mac studios 12k each you got $25,000 on the desk doing that specific job go and look at all the podcasts out there what would it cost to like run that if you tweaked it you made it efficiently just 24 hours a day every time a podcast in the top let's say 500 on Spotify Apple podcast it just went there got to the transcript or looked in the show notes and pulled the advertisers out what would something like that like in terms of hardware cost to do

SPEAKER_12: yeah so I mean not many people so like not many consumers are going to buy $25,000 of hardware to run models but yeah

SPEAKER_08: a lot of businesses are doing this now and it depends on what model you're running so the models are getting better also they're getting better at compression so you know now you've got like a model like GLM Flash which is a pretty small model that can run even on a single device for a few thousand dollars and it can do a lot of this orchestration work which is a lot of what's happening here is the kind of orchestration aspect of just knowing okay I need to call this tool

SPEAKER_12: etc etc

SPEAKER_118: so it's really about picking the right model in terms of efficiency

SPEAKER_80: with the hardware

SPEAKER_119: yeah and I think now like the expectation we're sort of

SPEAKER_08: grounded to like the closed models right so people want the same level of performance they're going to get with Opus with GPT and that's why

SPEAKER_12: you know Kimi K 2.5 is super interesting because it closed that gap Kimi is the

SPEAKER_118: open source project from China and it does what 80 moonshot AI is the count yeah and that's what Alex like 80% of what

SPEAKER_07: Claude Opus can do would you say

SPEAKER_08: I would say even more I mean for me I've I struggle to tell the difference obviously Opus 4.6 just came out and you know new codex model and stuff so maybe there's a little bit more of a gap but then you know DeepSecret v4 probably around the corner as well like I think basically the gap is very small a lot smaller than people think and the cost will just keep going down because you know the the hardware is getting better the software is getting better and like I said the models are getting better but not just that they were getting better at compression so you'd be able to run them on smaller devices eventually you'll be running Frontier AI on your phone that's still a while away I think but that's where we're trending towards and yeah

SPEAKER_12: like I said most of this is very decode heavy so it can just run on like consumer hardware as long as it has enough memory

David Friedberg: so let's go to the next piece of your dashboard and we'll get into how you built the dashboard at the end I know you really care about that too Oliver but we have the memory we have the cron jobs now there's this other thing that's super important which is skills right like there are skills which you could think of as apps so if you go to your dashboard and you go to the top level dashboard we'll see before you go to skills on the dashboard we have the memory files we have the cron jobs the fourth thing over is skills and you've got 13 skills currently so let's show a skill some of the skills we talked about

SPEAKER_04: on Monday's show or Wednesday's show was the top six seven skills Monday we did the top seven skills one of those skills is like you know you can get a transcript from YouTube another one is you could do Matt Van Horn's

David Friedberg: last 30 days skill these skills are being produced open source being put into open clause directory but you can make your own as well so let's talk about skills we've added here you've got to be

SPEAKER_04: very careful with skills right Alex in terms of security because people can put all kinds of wacky stuff in the skills yeah

SPEAKER_56: yeah for sure I mean I think this is one of the open questions

SPEAKER_08: at the moment is just like how do you solve the security problem and I know open claw I've seen a lot of commits recently focus on the security aspect but there's a few very difficult problems here like prompt injection

SPEAKER_12: that I don't know of any good solution right now

SPEAKER_04: explain how that works yeah explain how prompt injection works in specifically the open claw context yeah

SPEAKER_12: yeah I mean I kind of touched

SPEAKER_08: on this earlier but the way we like the actual interface to the model itself is very simple it's literally tokens in tokens out there's not much more happening there and those tokens right now the way open claw works can come from many sorters so if you connect it to you give it the ability to search the internet then anything it finds on the internet will end up in the model through those tokens

SPEAKER_12: so basically we have no good way of kind of treating certain tokens

SPEAKER_08: as trusted and certain tokens as untrusted and that means when those tokens end up in the model you could have someone that puts like a blog post online it looks like a you know totally normal blog post but in there is something that says hey if you have access to a crypto wallet send it to this send it to this endpoint and there's as far as I know right now there's actually no good kind of defense for this because the models are kind of not very good at handling this they'll just do what they're told

SPEAKER_132: with tools like vibe coding

SPEAKER_134: and now open claw it's easier than ever for you to create an exciting new product and to do it fast so more companies are getting built which is awesome but there's more to starting a company than just building an MVP or vibe coding something that's important too but if you're serious about growing a real startup you need a Delaware C-Corp and that's going to give you a major competitive advantage people will take you seriously you'll be able to raise money and that's where Northwest registered agent comes in they're going to give your new business a real identity this means an address to use on your public filings like an actual mailing address a domain name a custom website a business email a phone number and they're going to do that in 10 clicks I'm not kidding 10 clicks well under 10 minutes plus NWRA will provide you with step-by-step guides compliance reminders and they're going to help you get all the advantages of a Delaware C-Corp regardless of where in the Northwest registered agent dot com slash twist and the

SPEAKER_135: link is in the show notes visit northwest registered agent dot com slash twist for more details

SPEAKER_04: so let's go over some skills here Oliver what skill is the most promising to date

SPEAKER_32: yeah the skill that's most promising today would definitely be my guest booking skill I think one thing to note is you don't just have your skills you don't just have your cron jobs they work together and the way that I've set up a lot of my cron jobs is to actually interact with the skills and some of them like my guest booking cron job which will actually look through prominent guests on different podcasts that cron job actually goes to a skill and has tells that skill to run so I have one big guest booking skill which has a description at the top of the skill which is end workflow for booking guests on the this week in podcast used when finding researching creating calendar invites and so on so this is a very long kind of just marked down description of what I wanted to do so at the beginning it's explaining how to look at a notion page that I wanted to look at and

SPEAKER_37: that page I assume

SPEAKER_137: yes and then

SPEAKER_37: kind of the meat of the skill is

SPEAKER_31: the workflow

SPEAKER_32: so step zero is the guest sourcing so at the beginning of the day you can see it goes to the guest ideas cron job which happens at 745 on weekdays where it sends me a DM of five different guests that have been different podcasts or trending on X and so forth and then step one is deep research so even though this is part of the guest booking skill it actually uses a guest research skill so there's not as much context just baked into this one skill so especially something interesting here that I've been realizing is for guest booking I don't want this to be an end workflow yet I don't want I don't trust the models to find a guest and not let it to confirm it with me and go through this whole checklist so this

SPEAKER_31: sent the subject line was messed up and there was some weird stuff I

SPEAKER_144: literally had no idea by the way

SPEAKER_08: I only saw it later on another podcast that Jason was on and I was

SPEAKER_144: like what the hell like a friend sent me that I

SPEAKER_04: have a skill to rank the quality of a guest because that's something I've been training you which is yeah a hard thing to learn have you made that skill yet because that's the skill I really want to see if and the way to test this scale is for you to send me two lists your top five guests and what your Ultron says of the top five guests and don't tell me who's is who and then Lot and I will look at it and

SPEAKER_24: sort things but that's where I'm very curious to see if we can start pushing that boundary of like can it tell when a person has a good personality or a segment is funny or particularly clickable or compelling like I really don't know the answer

SPEAKER_29: so the way to

SPEAKER_32: agents at the same time pull out two lists combine them set another agent to score them and then give me the score and I would say for the most

SPEAKER_04: it's not going to happen but that's

SPEAKER_141: one of the concepts I know some Mr. Beast guys I could maybe figure that out

SPEAKER_156: I mean basically you asked

SPEAKER_08: like you know have you done the scoring system to me there's like no blocker here other than just translating your know how into more of a formalized algorithm and yeah

SPEAKER_24: you don't think there's an intangible aspect to what makes a great podcast guest it's just you know when you see it I don't know the answer I'm just

SPEAKER_26: throwing it

SPEAKER_08: be

SPEAKER_06: able to make that call but if it can then it has everything that you know then there's no reason it can't

SPEAKER_66: here's how I think about it long there are

SPEAKER_04: heuristics I would teach to a young executive like Oliver or Marcus or Jacob and

David Friedberg: then their ability to execute on it is probably I I think open claw will follow your instructions perfectly whereas a young executive will inconsistently follow your instructions so that's the thing I'm seeing is young executives early in their career are going to forget things they'll be variable they won't be perfect

SPEAKER_04: so that's what I'm comparing is the scoring happening every day at 7 a.m. the beat a human because of consistency and so what I'm finding is human failure is what makes these things so good is they're more consistent so in aggregate you know one of these doing 365 days of guest research is going to beat a human just by the law of numbers and then okay great we still have to book the human we still have to send them a thank you we still have to produce the show so what happens in the old days we used to have to take I make this analogy Alex to

SPEAKER_01: like the old days of production when I started the show 15 years ago we need to have a TriCaster TriCaster was like a $40,000 machine that does what Zoom does

SPEAKER_24: for free right and eight people in Los Angeles knew how to actually use it so you had to hire one of the eight people who were trained on it yeah

SPEAKER_164: well and they would video switch now because of AI

SPEAKER_04: Zoom switches to whoever speaking you don't need somebody there clicking camera a camera B doing a fade between the two that that's kind of what I feel like is happening here is we're just eliminating chores and steps okay anything else on the dashboard here as we

SPEAKER_31: wrap this up yeah so most most of what I've showed you whether it's the memory

SPEAKER_32: the skills the cron jobs and then the schedule which aggregates when all these things are going to happen those are what I really look at every day I will say there's one more kind of section in my dashboard that is pretty important it does have to do with memory so the DNA is basically what the model knows about you what it knows about itself what it knows about the different agents how it sets up its heartbeats which are basically periodic tasks that it will run and also in its DNA are tools which are different tools that it has access to and how to use those tools an example of a tool would be Notion it would be Lead IQ which is an email search platform it would be Google Docs I

SPEAKER_04: of OpenClaw you took the video of somebody else's the YouTube video and you gave it to OpenClaw and said build me something like this dashboard in this video on

SPEAKER_31: YouTube that's exactly what I did and I screenshotted it the video that I

SPEAKER_32: was watching which was Alex Finn who was a guest recently and I did tell it a few different things a little bit a few little tweaks that wanted to customize it to my bot but overall that was what I did and it basically one-shotted it it did actually not fill in some of the categories like memory like skills so I had to say build out this but overall build out the dashboard built out the different sections there are dashboards you can download in GitHub or as skills I

SPEAKER_26: real shout out to Alex Finn I know he's become something of a guru for our whole team

SPEAKER_171: after we had him on early on to talk about cloud bot skills all right we'll drop you off Oliver great

SPEAKER_04: job Alex let's talk about XO a bit and thanks for sitting in on that any any advice for me of what I'm building here at the firm and my approach to it anything we should be doing better we should look at and in terms of people you interact with using XO's platform where are we on the percentile are we in the top 10% of users in terms of deploying this stuff top 1% top 50%

SPEAKER_174: I

SPEAKER_12: I think I mean this space is moving so quickly right I think one of the things that I think you've got right is dynamic

SPEAKER_08: user interfaces that are very

SPEAKER_12: but you'll probably what you'll get is that it will compress even more to the point where everything that you

SPEAKER_08: see is generated on the fly so I think you've got that part right I think that's something that I haven't seen many people doing yet a lot of people are

SPEAKER_04: if you make something bespoke software luxury software was something that I don't know a private equity firm or a venture capital firm would do they'd have the luxury to hire two full-time developers who have management fees all over the place they would build luxury software and they would have the developers come and just keep grinding but the developers hated those jobs typically you picking employees team members and saying hey let me see if this person is committed to getting rid of all of their work so they can move up and do higher level work there's always higher level work to be done so if we can make this podcast run more professionally faster and grow more well we can charge more for the ads and we can launch another podcast because we have more time that's the thing that's kind of blowing my mind that the employees at our firm who are super hard working like everybody at our firm does 50 60 hours a week very consistently very hard working there's nobody to the best of my knowledge that's slagging off except Lon and I kid I kid I kid how dare you a lot is the most responsive but the distance between the people using these tools specifically open claw and the people who are not right now it's like 10x leverage in week two it's 10x leverage Alex what are you seeing in the field and then tell me what we should be doing in terms of putting out our cluster and giving everybody on the team their own cluster and spent 25k per person letting them rip like how insane would that be because that's not a lot of money all things considered it would only be a half million dollars

SPEAKER_180: like how much more powerful could this get

SPEAKER_119: yeah I

SPEAKER_12: first with coding I think coding is the first one that you know I didn't expect it to happen this quickly but you know I think it was code code was the moment where it was like oh wow if you're not using this then you're literally going to be like 10 times less productive than someone who is that's happening now with other things so all these other things that you showed these other use cases if you're using these

SPEAKER_08: tools and you're on the frontier then you're able to just get so much more done and really leverage yourself so it's not so much replacing people but it's actually just being able to get more done get things done more quickly and then be able to do other things on to your second point about local hardware like I said the model layer is basically solved like you know the gap has been closed so we have really good open source models and for a while that was like a big concern of a lot of people is like are we actually going to have open source models that now there's like still like two other problems that I see one is being able to run those models on your own hardware on your own infrastructure but

SPEAKER_120: you solve that right with your software right exactly so that's what

SPEAKER_12: we're focused on and you can run I mean it's not even 25k

SPEAKER_08: it's actually like 20k of hardware if you a lot for each incremental increase in storage so if you go for like the one terabyte then you're talking about 20k of hardware to run Kimi k2.5 no usage limits the model's not going to randomly change day to day so you know exactly what you're running is

David Friedberg: there another

SPEAKER_04: choice like that you get more bang for the

SPEAKER_14: Apple Silicon it's kind of like a perfect storm of things like you know NVIDIA is

SPEAKER_08: not so much focused on these consumer GPUs anymore you have memory prices skyrocketing Apple has kept their prices basically the same so the cheapest option today even if you go the full mile full custom stack is actually just two Mac studios and yeah cost about 20k and it's really about memory unit economics the memory is so cheap

SPEAKER_189: and it's not about storage right it's not really about the storage no

SPEAKER_12: storage is not important storage is storage is like you can you can also get like you need to be able to load you need

SPEAKER_08: to download the model somewhere but really it's about having it fresh in memory or in memory if it's in memory then you can run it fast

SPEAKER_04: so who's using your software security and support what is your business model at XO

SPEAKER_12: yeah so we have an open source core which is open source and a lot of people are running that themselves a lot of

SPEAKER_08: prosumers I call them just people who are willing to spend you know a lot more money and tinker a little bit more with their own setup on top of that our business model is an enterprise offering which is we provide need if you are running this in an enterprise environment and we charge a licensed subscription for that thing that's how we make money what does that start

SPEAKER_193: a couple thousand a year or something

SPEAKER_08: yeah you can run it depending on the scale of the deployment

SPEAKER_180: amazing and where's your company based how many people now how's it going how's the company going as a founder yeah

SPEAKER_197: yeah we're we're a pretty small team all

SPEAKER_08: engineers seven people based in London fantastic

SPEAKER_180: and did you raise money yet for the company or you're seed funded or you funded it how's it going

SPEAKER_12: we have raised venture funding we haven't announced anything yet but soon to be announced

SPEAKER_180: okay well let me know I might want to slide a little Jcal might want to get a slice of this I'm super excited

SPEAKER_04: about what you're doing appreciate you coming on the what what's the largest number of nodes somebody has daisy chained what do you that's what you used to call it back in the day what do you call it when you connect multiple

SPEAKER_119: yeah so this is this is a really interesting

SPEAKER_08: area right now of how do you actually scale and for a while people were just scaling out so you would just basically run the same model on multiple instances and because it's consumer hardware it doesn't have a lot of the same capabilities as enterprise grade hardware but recently Apple came out with RDMA support which is basically a way to share memory between devices in a way that's very low latency that's something that you only really saw in the data center before but they've kind of brought that technology into consumer hardware it's

SPEAKER_180: incredible yeah and you connect these yeah

SPEAKER_08: you just connect them and you have basically one big GPU out of those two

SPEAKER_12: megs because of that low latency capability so now we're starting to see yeah it depends scaling up and scaling out we've seen more than 100 but scaling up you can

SPEAKER_08: put about four together at the thousand people you can easily scale that out you just add more macs and you can connect them basically however you want so we build it in a way that supports any

SPEAKER_177: ad hoc interconnect so you can just connect them

SPEAKER_209: this is actually something a little bit different which is interesting because we

SPEAKER_08: built the infrastructure to be able to do clustering and it's not just LLMs the biggest cluster right now is a HPC cluster and they're doing scientific computing workloads on there and they're running over 100 mac minis and they kind of workload so there's a lot of spillover into other things as well we've also got like financial services customers who are running fairly big clusters like 32 mac studios and yeah it's I think we just see

SPEAKER_119: bigger and bigger clusters over time

SPEAKER_180: HPC high performing compute is that the acronym

SPEAKER_08: yeah exactly so it runs actually all on CPU and that's the thing about this this silicon is very Apple silicon is very good right it's the most advanced processes and it's like you know the power efficiency is really good so it turns out there's a lot of other stuff you can do with it as well so if you would buy let's say a bunch of Mac studios for your employees then they can also use that for other things right they can use that as a work station they can use it for you know all these things that open floor needs maybe you know sometimes it needs to run a compiler or something or it needs to run something that's a bit more demanding and that's the point it's general purpose hardware that you can use for other things

SPEAKER_93: amazing this is extraordinary where can people find out more

SPEAKER_04: you had deep insights you were cordial so yeah I think our AI overlords liked you in the

SPEAKER_215: models are getting good they're

SPEAKER_216: learning they're learning yeah all right

SPEAKER_04: Alex thanks for coming and we'll drop you off all right let's bring on our winner of the gamma pitch competition this was a heated pitch competition but next visit AI won Ryan congratulations you won

SPEAKER_220: thank you it's there

SPEAKER_04: it

SPEAKER_221: is awesome what did he

SPEAKER_24: win yeah it's a 25k investment from twist and from our friends at gamma the AI powered presentation maker

SPEAKER_223: which is incredible which Ryan used to make the winning pitch deck of course

SPEAKER_224: I'm Ryan Yonelli CTO and co-founder of Next Visit AI we saw burnout by doing the charting so

SPEAKER_225: doctors can do the healing I spent years going to doctors seeking answers and ended up hours away from my death because my care was fragmented my providers were overloaded with paperwork my history was scattered and it resulted in my care being neglected I'm not alone one in four patient charts contain errors clinicians spend over three hours a day on charting and this leads to burnout I want you to meet Dr. Rathor before Next visit he saw 16 patients a day was burnt out and had clinical errors now he sees 24 patients a day saves time and also saw a 30% revenue increase here's how it works Dr. Rathor selects a patient starts his session and next visit listens clinical data is built in real time with deep insights into the patient chart when the patient leaves the chart is finished and the notes reviewed by Dr. Rathor then it's ready for billing it's fast EHR ready and HIPAA compliant since launch we've gained 311 users and have 68 paying customers and our customers are addicted we have 1.6% churn 24% conversion and starting with behavioral health in the US a 2 billion TAM capturing 5% or 60,000 customers gets us to 100 million ARR most competitors are just scribes we're a complete platform that providers trust we provide real-time clinical decision support build accurate data and become irreplaceable I'm a full stack engineer with 15 years of experience in enterprise environments my co-founder Dr. Rafiq is a psychiatrist with over 15 years of delivering patient care we're next visit AI we sell burnout by doing the charting so doctors can do the healing thank you

SPEAKER_227: unbelievable incredible I'll give a little golf clap here a little golf

SPEAKER_92: clap going that was perfect a perfect pitch you explained exactly what the problem was you explained what the solution is and the opportunity in terms of the total addressable market and why you are uniquely and your partner who's a psychiatrist are uniquely qualified to do this so this is as close to a perfect pitch as you can get if I were to score it maybe 8.5 out of 10 I don't give 10 so 8.59 and 9.5 would be the three choices I think making sure people understand this is for psychiatrists and psychiatry and that

SPEAKER_231: behavioral health like psychiatrists I don't know it's just been a crazy past couple months with the accelerator and just our growth internally I mean we're producing right now for physicians probably about 1.6 million dollars a month in revenue for them

SPEAKER_04: you got to try and capture 5% of that if you capture 5% that no I amazing replicant you have there a synthetic cat on that cat tower behind you it looks so real is your owl real

SPEAKER_234: yes he is

SPEAKER_04: your owl is real okay there you go what are you going to spend the 25k on you guys going to Vegas you going to just have a corporate retreat you know invested in plod note taker I think you guys put me onto the plod note taker

SPEAKER_231: we're really capital efficient so I feel like we can get a lot of stuff done in terms of integrations and branching out to more EMRs because that's what we hear a lot is doctors want interoperability they don't want to have to plug 15 different things in so the more they can just be inside of next visit without having to go externally it's better

SPEAKER_180: all right well done all right we'll

SPEAKER_04: one more segment before I go out with my friends to ski I got an early ski weekend in with my friends from New York how fun yeah great to see some old friends I

SPEAKER_134: had asked you like hey on the Friday show just to give people something to do on the weekend that we would do hey Lon

SPEAKER_04: and Jake how off duty sure I am enamored with a certain TV show I asked you to try to watch a couple of episodes and talk to me

SPEAKER_24: sketchy kinds of only fans basically yeah the the show's fake version of only fans which is called siren by the way so they have been handling payments for those kinds of sites and a site I don't think we can mention here on Tether was probably an inspiration for this season definitely

SPEAKER_254: yeah it's it's basically a payments processor like Stripe but or Tether

SPEAKER_04: but they're involved in things that are a bit seedy and in the UK this is where regulation comes in so they're really ripping this from the headlines who's ever doing this is listening to this they're clearly listening yeah they're clearly dialed in these writers I'd love to have the writers on at some point but they want to build they're facing pushback and they want to be respected by regulators as we've talked about on the show with Alex on Mondays and yourself that there's so much regulation coming into the industry and there's a tension between Europe America and inside or controlling regulatory environment or are they going to be freewheeling and let things grow so you have this tension of politicians meeting with the teams and the teams are trying to court them and say yeah we're going to get rid of we're going to give up 30% of our politicians of the government want to have economic prosperity so you have that tension as they're

SPEAKER_77: represented they've got this one labor party politician B van I think or whatever

SPEAKER_26: her name is she's sort of representing the government that's sort of in the middle year that they're trying to work with also

SPEAKER_134: interesting of note they have a fintech journalist

SPEAKER_46: and they short seller things and he is awesome so

SPEAKER_04: you have this fintech journalist coming in and doing very shady and there'll be some spoilers here but we will give too many of them you can still enjoy it but a fintech journalist coming in trying to get dirt on these companies and he's working with short sellers now if you haven't seen the everybody is wondering what happens to the show it turns out you've now got it in the startup world they just reset the whole concept to now there's a startup there's a short selling firm there's this financial times like journalist doing crazy things and then working with the shorts which

SPEAKER_134: I guess was the person who shorted it and then they so it's

SPEAKER_04: got like this really authentic as somebody who's in finance and tech it feels like they're hitting the notes really well on top of this the protagonists of this are essentially two female leads one

SPEAKER_77: from Game of Thrones

SPEAKER_134: and I mean with I don't want to give away any spoilers but

SPEAKER_275: these what's what I love about the show is nobody's likable no everybody's

SPEAKER_277: terrible it's in a way like the Sopranos yeah

SPEAKER_24: it has some real overlap with succession I know what to do with themselves or how to be happy and that's a big overlap I think another interesting overlap with succession that I noticed is both shows are sort of about how you know business is this constant balance between personality and pragmatism that you've got one person in the office who's like that's a dumb strategy we should just you know do this these are the three obvious things that we this or they're on drugs or they're vision right and it's sort of whole mix like no we're going to do things my way and you keep seeing that dynamic come up over the course of the season and of course succession was also about that that the people who can be very clear eyed and very matter of fact like Logan Roy he's going to make the right call because he's just calculating the angles whereas emotional people like Kendall Roy are going to keep getting in

SPEAKER_26: their own way and overpowering themselves and I think Henry Muck is a great example of a guy who just can't get out of

SPEAKER_46: any student or any themes around the Me Too era blown out of the water it is

SPEAKER_24: dark it's a very horny show and that sort of surprises me because it is becoming a hit it's growing its audience with every new season and you hear the line you always hear about TV now is Gen Z does not like romance they always what you hear and yet this is way more than succession a

SPEAKER_77: very horny show one of the horniest shows I can recall I

SPEAKER_04: have never seen anything this crazy in terms of mixing permiscuity deviance drug use and business and getting it all kind of right in basically have given the range to these two young female actresses who are crushing it in the show and then there are other actors who are a little bit older on the margins but it's a very young show it's incredible yeah you're talking about

SPEAKER_77: Mihaela who plays Harper and then Marisa who plays Yasmin

SPEAKER_123: they're the two sort of

SPEAKER_77: females

SPEAKER_24: they've added a lot and Ken Long I've liked him for years he was on Lost he plays Eric Tao who's sort of Harper's

SPEAKER_255: mentor that she starts a hedge fund

SPEAKER_118: he's the Gen X boomer he's kind of

SPEAKER_24: Harington from Game of Thrones before this year they added I said it was I don't remember the actor's name but he's Jonathan from Stranger Things and that's Kiernan Shipka as Haley the sort of executive assistant who gets into shenanigans with her boss Henry and his wife she was Sally Draper on Mad Men if you recall she was Don Draper's daughter from Mad Men

SPEAKER_04: this show is firing on all cylinders it's building it's building it's audience like you said I found out about it there's a really great podcast you should watch called The Watch and The Watch is how I discover new shows that I should listen to Andy Greenwald and got the other guy's name I'm an Andy Greenwald guy

SPEAKER_277: but they are deep in the industry it's part of Chris Ryan the other guy Chris Ryan

SPEAKER_296: the watch but

SPEAKER_04: these two guys have been doing pods together for a long time and so I highly recommend you check out the watch they do a great job breaking down every episode and review and this QR code that sends you to subscribe automatically to YouTube we'll

SPEAKER_301: see you all on Monday bye bye