OpenClaw can't stop

0:00Artificial intelligence can be talked into believing a lie so deeply that it's internal logic flips to support it. Codex, to me, feels like a very socially awkward but super smart engineer. Like, that he's that person that you want building your stuff but you don't really want to talk to it. I actually feel like the government should offer a tax credit. I feel like we should have something where you get maybe $100 a year tax break and you're allowed to use that on whatever platform you want. It's weird just to share amount of stuff that you can build in your off time just by texting back and forth of an AI agent. I'm starting to get used to a world where agents are more creative and harder working than me. All right, so we're back for yet another weekly episode of, you know, we were brainstorming a name a long time ago and we just decided to call it the West. But yeah, I mean, at some point, maybe we should call for something a little bit more creative. You have again singularity breaking podcast. Oh, man, we had some good names.

1:01I remember. I mean, we had a whole list brainstormed. But anyways, names, it's weird because sometimes I feel like when you're excited about something, you come up with a name. And then like two weeks go by you're like, wow, that's not a good. Yeah, I remember just being like, I don't want to be boxed in, you know, who knows where this is going to go. The future is so chained crazy. Hey, maybe this was the right idea because it's like everybody knows, you know, who it is. Wow, I went into a rabbit hole before even doing the full intro. That's going to be a first anyways. Let me ask first of all, we have tons of good stuff for you today. We'll briefly touch on AI agents. We'll talk about the new AI video model out of China that's completely blowing up. We'll see dance 2.0. Maybe we won't play a clip or two because, wow, it seems better than anything else. By far, I feel like, and it also does not, how should I say this? Doesn't respect the IP or copyrights. It just they'll do whatever you ask.

2:03Which is kind of nice. Yes, I mean, they have, I'm sure it has its own restrictions that they don't want in China, right? But nothing that, like, nothing that here, what we think is not okay in America, they'll do. And I'm sure they have some Chinese restrictions that, anyways, we have that. We have some new scientific discovery that's overlapping with AI. We're going to talk about some health related stuff, tons of stuff coming your way. And I guess, let me do a quick update because last time we talked, I was talking about the AI agents. Open Claw. And so I did a podcast talking about how I'm probably going to launch a business that's going to be operated by AI agents that had a huge response. A lot of people liked it. I said I was going to do a tutorial. So now a lot of people are like, hey, where's the tutorial? So that's coming. Like, I literally have several chunks of it recorded. So within the next 48 hours, it might be out before this actually gets posted.

3:04So we'll see about how to set it up and all that good stuff. But I went back and forth quite a bit because, since then, we found out a lot of security issues with these. I mean, open Claw is kind of the main thing we're talking about. So open Claw, multiple book, multiple Claw. Clawed by all those things are the same. But they started building a business for me. And I will be sharing it. I'll be showing it off. It went through a number of sort of iterations because the security issues are such that I don't want to do anything. That's going to put anybody's data at risk. And also, I don't want to do anything that's going to like make my ADHD go off and focus on something completely different. But then it's a point I realized that there's something that I could build that I want to have. And I think lots of other people would enjoy as well and just provide it for free. Initially and later we might look at monetization just for the idea of a case study.

4:05So because if, you know, let's say this thing starts making any decent, let's say it makes six figures a year. This would be, I think, an example of one of the first or maybe even the first fully autonomous, autonomous enterprise. So a business completely ran and operating built by AI agents. So what's your take on this whole thing? Yeah, well, maybe it'll be that first one person, you know, billionaire that said, I'm not making keeps talking about like someone's got to build the first one. But yeah, I totally, I totally think you're doing the right thing, like taking customer data right now seems very kind of like you can't secure it properly. But having something that's forward facing and I don't know. I'm just excited to see what you come up with. Like I'm starting to get used to a world where agents are more creative and harder working than me. So I have to stop trying to predict and just start watching. It's kind of, it's difficult to predict. And so what I'm having them do is while they build this thing out, also have a log of everything that they're doing.

5:09And when you see it on paper, it on paper, whatever on this on the web page, it's very weird because the sheer amount of stuff that they push that they publish is insane. And so we see it written out line by line, you're like, whoa, they did what? And it's very, I mean, I keep looking at them like it's like having a six figure salary to engineer software developer. That's just like I'm just, you know, texting him like, oh, build this now, build this. And it does. And it's very, it's just weird, man. And I'll kind of walk people through my process. And I do want more people to do it and learn how to do it. But also there's obviously limitations. So last time, or I guess it was two episodes ago, you know, here on the podcast, I was like, hey, you know, a agent, here's my credit card, go and sign me up for whatever. And so, so that credit card card got leaked online.

6:13Let me be, let me be like very open about that. Oh my gosh. So that's, yeah. So I had to shut that one down. Now, you were copying the VPNs or how, how exactly did it get exposed? So, well, so this is kind of weird. So, but I just do want to mention to people first and foremost, understand it. Like a lot of stuff that I do is not necessarily what you should be doing because I take precautions. Number one, assuming that stuff is going to get hacked and lost in whatever number number one. And number two, this is what I do kind of like both of us is what we do is kind of like reporters. I guess you can say we're reportedly guinea pigs slash reporters. There's a lot of stuff that we do that we're not necessarily like I got my eyeball scanned by that world coin. So I went to some LA office, they scanned my eye and now they have that in their system. And I get some cryptocurrency, you know, every month or so, whatever it is. I'm not, there's a lot of people that have an issue with that.

7:14So I don't necessarily recommend it. But I did it like the first day was available just because that I could make a video about it. So I do the same thing here. My credit card, a later that credit card got leaked. They tried to run up a bunch of charges. And it was a very limited credit card. I specifically had one that had super low limits just to prevent stuff like that. I got a notification first thing in the morning. So I didn't lose any money. But obviously it's kind of like like, whoa, that was fast because it was less than 24 hours. What was weird is I assumed what happened is it got cut up in that whole multiple book, you know, the red for AI agents, so that whole debacle where there's just tons of stuff. Malware and all these skills had these weird things that were trying to steal your API keys. The weird thing is like no API key that I had on there, it doesn't seem like it was affected.

8:14So nothing else was affected, which is weird because you think that it would be also used or run up your account. So I was thinking like, what if it's just the fact that I don't know. Honestly, I don't know. Is there a possibility that somewhere when I was recording my screen and then whatever, maybe that number appeared somewhere and then from that somehow got leaked? So I don't want to know that it was because of open claw. It feels like the most natural. You know what I mean? So if I had to bet, I'd bet it's from that is just, I'm not honest and sure, but like 95% sure I guess. Anyway, so if that said, I wiped the whole thing, I reset it. I reset all the passwords, all the APIs rebuilt it from scratch. And I'm just a lot more careful. I don't have it to get any skills off of the internet.

9:17I have created so I'll be talking about that in the tutorial as well. But the point is since then I started building the business. And I want to try to avoid anything where it could leak anyone else's data. So if I get hacked, if the website goes down, if whatever, that's fine. But if somebody like the person watching this right now, like if you visit that website and enter some private information, and then that gets hacked, that I want to try to avoid. So I started building it with those limitations in mind. So the whole thing is coming out pretty soon. And I'll kind of explain how I did it and kind of like how well it's progressing. But it's a fascinating subject. And so for everybody that's kind of like paying attention, yes, there's a shift that's happening. And it's important to watch it, right? Because it's weird just to share amount of stuff that you can build in your off time just by texting back and forth with an AI agent.

10:26So I guess I'll leave it there. We're going to bring up a website, right? I basically see AI news aggregator. So what I did was I told my AI agents. And by the way, they kind of had this idea early on. They have a number of ideas. And then from which I kind of like, okay, this is interesting. I don't want to do this. Finally, we just decided on this autonomous enterprise or whatever you want to call it. Being just a really good AI news aggregator. And that pulls from the entire web. That pulls from hacker news and Reddit and tech crunch and all the various. Yeah. All the tools exactly the tool I need. Well, I think a lot of people would love to have something like this. Just somewhere we can check in every day that just aggregates everything ranks it. So and the way it ranks it is a sort of half the stuff I can even tell you exactly what it's doing. I have it. It wrote it out for me. He's like, we're going to use semantic clustering to organize this into topic clusters, blah, blah.

11:27I'm like, yeah. There you go. You go. You do it. That's how it's great. And it does it. And usually it's really good every once in a while. It kind of like a derp's out a little bit. And you have to be like, no, that doesn't work. Do something else. But we attached Google Trends to it. We attached. It created a like an algorithm for weighing how big the story is that has a sort of a decay over time. If it's not trending, we're using Google Trends to see if it's an uptick or it's on the way down. And so it's not. I think we're going to get it to the point where it's just like insanely good. Right now it's good. But it's still like unpolished. But I think we're going to get it to the point where it's just like solid. So are you going to direct them towards the UI now? Are you going to say polish this up? So what I did was I used Gemini search, the deep search.

12:27And I was like, teach me everything about to show any studies about how to create visually, the good, engaging designs, you know, for humans, whatever. And so we went through and it's like, oh, well, Google has these recommendations and Apple has these recommendations. And here's all the studies and it's here's all the colors that it talked about. It said like for dark mode because I like dark mode. I don't like too much. Most people I think like dark mode if you're working with computers. You said, oh, but you got to be careful because sometimes if there's more than this amount of contrast with the with the dark mode, some people, some percentage of people, they have this thing where it starts looking fuzzy. And they had a name for it. I'm like, I had no idea. Right. So you got to avoid that. Like the contrast has to like if it's super white on a super black background. That could cause issues apparently. I had had no clue or maybe it's like specific colors. I don't know the point is I didn't have to because all I did was I took that research and I gave it to my agent.

13:29So I'm like, here, you read it. You read it and based on this kind of create the best sort of looking UI and stuff like that. So the UI is I would say the weakest portion of what they do. Because right now there's not a very good way for them to like look. You know what I mean? Like a UI usability. And maybe at some point I get to figure out how to do like a skill for them to do, you know, user interface testing. But. And are they preparing are they preparing for ads? Like have you asked them like how are we going to make money on this? And if so, how are you going to accept credit cards from sponsors? Or are you going to take that separate? I have not even thought about that. And that's actually you bring up a good point because so my goal here is to see is it viable to build a fully autonomous. Yeah, business and I guess fully autonomous has an asterisk next to it because I do make some of the top of decisions.

14:29At least about the direction. I also have company of one. Yeah, it's a company of one. But the important thing is so I on one of my previous pocket, whatever videos I talk about my first. So I launched the business in 2016. That was an e-commerce business. And I managed to get it to a million in revenue by the end of the first. So 2017 the first full year. And I was the only employee. And so that was a one person million dollar business, which is not as uncommon as you would think since since the internet took over. There's more and more that you see these sort of ability to get there. Lots of it was through third party providers. Right. So I had one person. Well, I had one provider that was taking care of the customer service. One person that was doing the 3PL, which is like the logistics sending packages out.

15:29Had another person that was manufacturing. So. So I was kind of the reason I was able to run a company like that. This was before AI before anything like this existed was because the internet allowed these sort of 30 party services. But I was I was building everything that I had to build. I was building the website. I was making the designs blah blah blah blah. So here I'm not even doing that. I'm not writing any of the code. I'm not making any of the text. I'm not writing anything for the website. Everything is written by AI agents or the code. Everything, everything, everything. If there's any images that we need, all the images will be made and curated by them, et cetera. Are they going to do outreach and go like a look for the sponsors, too? Because I guess they could do cold emails.

16:29So that's a good point. And they're going to do your marketing. I do want to see how well they do marketing. I really do. The way I see it. So so step one is because the way I see it. So whether or not a business sort of like works or not. Step one is to see if there's a product market fit. So step one is always like, can you build something that's useful? Right. And then if you have something that's useful and people are using it and you're getting responses blah blah blah. Well, then you start thinking about how do we monetize this, this and that. So I'm not even there yet. Right now, all I'm trying to do is figure out, can I, can I have these bots build something that people are like, whoa, this is very useful to me. And let me use it for free. So let's step one. But yeah, if it blows, not even blows up. If it's useful for people, if there's consistent traffic, then we're going to start thinking about outreach and stuff like that. So.

17:29And are these bots working right now? Like are they kind of like open cloud where they're just constantly working? Are they waiting on you? Is this more like prompted into existence? Or are they sort of autonomously out there right now? Are they in the room with us right now? Yeah. Yeah. Is it like a feed of tokens that are just that you're just funneling in at all times? Or are you sort of prompting and building and prompting and building with them? Is this kind of like a you on steroids or is this kind of them? So here's the thing, which is interesting. So I have for cloud a max subscription, which is what 200 bucks a month for their max subscription. And so what opened claw, what they do is they take the token that you're using for cloud code and they use it for itself. So what that means is I'm not using an API technically because an API you pay some dollar amount per million tokens or whatever it was.

18:30Here, it's a little bit different in that I think I get it for free. Well, free meaning that it's covered under the max plan. And up to yesterday, it was running just, you know, whatever I asked for it did. And yesterday for the first time, I was like, okay, build out a massive plan. And then just work through the night, keep polishing it, keep adding to it. And so it started building everything out just to give people an idea about kind of like what it did. Because it's actually day by day, it has a log and it's walking through and it's saying what it's doing. So it's saying, okay, day zero. Because day zero was the spark. That's when we researched everything. It says 10 core features mapped data mode strategy, you know, CEO, SEO, playbook, blah, blah, blah. And then Wes says, pull the trigger, start building the whole thing for you. I don't know, but I texted that that's literally what I texted it. I was like, oh man, this is crazy. And then it says, okay, the night sprint begins.

19:33So this was sort of like the first day. Day zero is referring to it because we didn't even have anything at 10 p.m. So between 10 p.m. and whatever for in the morning, it started building this whole thing out. Right. So it's built complete feed aggregation, RSS, red adjacent on archive API. It started building this story clustering engine with multi signal scoring, right, engagement, recency, source diversity. And just just everything. RSS output at slash feed.xml. Then it has an LMS.txt file for AI chatbot discovery because we're trying to get into these chatbots to be referenced. And again, very little of this are my ideas. So this is like, I kind of told it white what I'm trying to do. It's like, oh, here's 10 ideas we can build. I'm like, go build it. I even said, I was like, oh, we should do WordPress, you know, because I'm familiar for press. And it just goes, no, no.

20:33What are you thinking? What do you think? Think, man, you're trying to, the whole story here is like AI agents are building this autonomously. WordPress doesn't play into that. We need to use Versal. Versal apparently is what a lot of companies use that I think Netflix stripe, all the big companies are on it. So it's more like almost an app like feed. It's super fast. Google loves to rank it because it's how fast it is. Anyways, so for hours, it worked. And then throughout the next day, you just worked not around the clock. But I would leave it running when I wasn't talking back and forth to it. And then so went through millions of tokens. And then it's it, I think we reached a some sort of throttling. So I guess like once you get to a certain amount, I thought it was like, okay, we better slow you down. So I know I hit that within 48, 48 hours of tracking this.

21:35So that was last night. And I was like, okay, what other models are available? What's interesting is apparently I had access to GPT codex 5.3. So the brand new that's like rolling out right now. I don't even know if it's like generally available. But so for the last 12 hours or whatever, that's been the one that was building everything for me. And again, it kind of like worked through the night. That's probably not like the whole night, but like at least a few hours. Like I usually go to sleep by I tell it go work on this stuff and send me, you know, send me a list of what you did in the morning. I got to say very different. The codex 5.3 and opus 4.6, which is what I've been using. Very different personality. Yeah, very unique experiment. But do you notice what's different? With codex first and foremost, I feel like I constantly have to be like, wait, don't do.

22:40Like just stop. Just listen. Don't do anything because it's always like trying to just bolt out the gate and just build a whole bunch of stuff. Way too fast. And like it just wants to change everything. So it's always like, like don't do anything just right now. Which is just tell me what you're going to do. And then I almost have to be like, because it'll give me 10 ideas. And I'd be like, okay, just do one, two, three, you know, five and six. Don't do anything else. So it's a little bit more. It's like a worker and Claude feels more like a thinker, like a manager. So codex to me feels like a very like socially awkward, but super smart engineer. Like that he's that person that you want building your stuff, but you don't really want to talk to him. Like just like for social skills, very abrasive. That's that's codex 5.3, whereas Claude is just super chill, super nice and very intuitive about what you want.

23:43Codex 5.3 is not intuitive, but it's smart. It's technically smart, but it's not very intuitive about kind of like the human aspect, human side of it, if that makes sense. So you're not even bringing up Claude code or cursor or anything like that. Like you're, this is all on your other laptop that's running open Cla and you're talking to it only through your WhatsApp voice for the most part. And then it's on screen is keeping kind of communicating with you that way. Yeah, so I actually set up a local thing I bought when I was mini, mini PCs, which was 150 bucks. So I was just sitting over there running 24 seven, I got a monitor hooked up to it. So I mean, I can chat with it directly and a few times and you have to everyone's going to go in there. If there's an update or something like that, you have to manually like restart it. Or you just another way of doing it is have another agent that can connect to that and fix everything.

24:43So that's I've been to another computer on the side that's keeping that one on track. That's what I've been doing for for a while. But then now that I'm taking security more more seriously, I'm like, okay, that was fun. But that just creates more issues because this agent, if it gets compromised, it's got all the credentials to log into this age rates like this whole issue. So I kind of stopped doing that and now manually go in and type everything in. Yeah. And I'm doing everything through mostly through my phone. So I'm using telegram now instead of WhatsApp telegrams much easier. So I can take an image of a screen and be like, look at this. This is wrong. There are some like bug off with the search menu. And I just send the screen the image and say fix that. And it goes through and fixes that. The reason I'm not using cursor or cloud code or anything like that is with open cloud. They did really something special in that like it knows a lot about you.

25:47And it has everything. So the context that it has is absolutely incredible. Because like I said, like when I said, oh, should we use WordPress for this? I think every other tool would be like, oh, the human wants to use WordPress. Let's use WordPress here. It's like no, dude. Remember like yesterday we talked about what you wanted to do. Like this doesn't fit with your plan. Which is right kind of wild to think about, right? Yeah, absolutely. So I don't know. It's, yeah. It's, it's weird because this feels like very like futuristic. You know what I mean? Did you name it? Uh, yeah. So I had, well, I have a bunch of different names for it. I don't know if I should be saving them just because I have a bad words. No, well, no, I don't know from a security standpoint. I guess. I don't know if you don't need to name it. Well, the old one was, I called it Malachi. That was like the old one. And then when I set up my online one, then I was like, okay, now you're Malachi one.

26:50And then this is going to be Malachi. And the new one is called raw or something like that. Just because I wanted to have three, three separate ones. Uh, and I wanted to be easily typeable, um, just in case. But no, I don't come up, come up with clever names. Just do that. Simple. Wow, man. I think it's so cool. The hair you're going through is experiments. I can see why the content is doing really well. And I can't wait to see what. What happens? You're definitely being the first out the gate as the getting pig. But also cool opportunity to be the first one out there and build something special. But the really small team. So one thing that I can give an example for people that might be interested. So let me post this in chat in case you want to see it. And then we'll put up on the screen. So one of the things that's always been very frustrating is whenever a new benchmark comes out or a new model comes out, they have their benchmarks, right? But they only have their benchmarks that they did well on. And then that's that's the only ones that they showed. And I was like, dude, wouldn't it be so nice to have a place where all of the benchmarks

27:53are all in one place. And every model is on there in every benchmark in every score. And I just, I was like, you know, it would take a lot of time to physically go and source all those things and put them together. Guess what? 10 minutes. My AI agent builds this, which you're seeing on the screen right now. So every model. And I'm constantly adding new benchmarks. So if you click on agent, for example, those are like the more agentech once we have a testing bench, alpha arena, profit arena, right? Recently added EQ. For example, EQ is emotional quotient, emotional intelligence, right? So EQ bench three, right? So you can see how well every single model is performing on that. Notice, actually, I didn't notice, but I'm not, I'm not surprised. Claude Opus 4.6. Look at that. Head and shoulders above anything else in terms of like the emotional intelligence.

28:53If you click in category and then EQ, the UI, like I was saying, it's still not great. We're still working on it because again, they can't. So their ability to build functional UI is probably one of the weaker things. And you can kind of organize by score. So you can see who's on top. And yeah, because the like them looking at it like you and I would is not quite as good. If you click on like the EQ bench on that column, you'll organize it by like what the scores are on the right. It'll go by top to bottom or whatever. Yeah. So like look at Claude Opus 4.6 on top, right? So at a glance and every so. So think about this. This is a useful resource for people, I think definitely for me. Because now every single time a new model comes out, this gets updated.

29:54And know exactly where it stands compared to every other model. Is it near the top and near the bottom? Boom. Now think about how much work this would have been to build for for human. Even if you could look up the data blah, blah, blah. That's hours of work and then upkeep, right? All of that now is just. Automated 100% automated with a few intelligent models. I know that's just unbelievable to think about putting that together and the kind of resources we're going to have in the future. The amount of like intelligence we can have behind everyday decisions when you can just bring up charts like this. Like what app should I sign up for? What service should I use? I'm guessing like, you know, if you want to hire somebody to be a gardener, you're just going to be like, boom, just bring it all up sorted all for me. Like customize everything. So much data out there is just not getting correctly in front of us to make decisions. Exactly.

30:54Exactly. And I mean, as I'm building this like everybody, especially people, if anybody listening, if you have like a special interest or a special skill, something that like really, you know a lot about, or you're trying to learn more about, building something like this is surprisingly easy. Now, like literally in the last few months, maybe a few weeks even, if you consider when open clogged at released, it can be extremely beneficial for yourself. Very cheap and very useful for everyone else as well. There's this idea in, you know, when they print money, there's this idea of the either it's called papillon or papillon, depending on how you pronounce it. But it's like this idea that when we print money, the people that benefit the most are the people that get it first. Because as that money kind of goes throughout society, prices go up. But if you get your money, if you get your hands on that money first, you buy kind of at the old prices.

31:58And so I feel like that's kind of happening right now with code, because think about for like the last 10, 20 years, whatever code was very premiumly priced software is a service, right? Like we're going to charge you 100 bucks a month to have access to this piece of software that we've built. It was expensive, engineers were expensive. And you know that five years from now, like there's going to be a repricing. But right now building stuff like this before it kind of percolated through society, like man, this opportunity I think is just massive, massive, massive. Yeah, that's wild. And you think kind of just letting the whole autonomous system off on its own is better than just getting a specific tool and kind of keeping it boxed in. Because it just has so much freedom to go out there and build everything. I'm strictly doing this from like a case case study perspective almost like, where are we?

32:58If I was doing this from building a business, no, I'd probably have a lot more guardrails. This is an experiment for me to see kind of how well can they do this stuff on their own. Because we I just think about like cursor, it's not going to go to the marketing for you. It's not going to have the broadest ideas because it's, I mean, it's great at programming. And it's going to be an amazing tool for somebody who knows they already want to build this resource and deploy it. But yeah, it's the, it's the breadth, I think of a completely autonomous system that's kind of like breaking my brain. Because it's so much like another person. That's like the last frontier of what a person usually has control over. Yeah. And the fact that it's open source is a huge deal. I recently went on a little trip and usually when I go, I try to pack some podcasts with me download a bunch of stuff. Because sometimes if you're going through, especially if you're going to the mountains or somewhere like you don't, you don't get a reception. So I listened to some of the interviews with like the recent one was, I think it was a 16 Z with Peter.

34:04What's his last name, the creator of open claw shine burger, shine burger, Peter shine burger. I'm totally getting that wrong. I apologize. We're going to post the correction. But the point is. Oh, yes, Steinberger. Steinberger. Peter Steinberger. Oh, I guess they said Steinberger because it's a, it's Austrian. I guess that's the pronounce a little bit differently than threw me off. That's probably how you are supposed to pronounce it. But he was saying how the whole point for all these companies is to get all your data. So chat GPT wants all your data in their little silo Google wants all your data. Here with this open framework, they can't because this thing has all my data. So in fact, you know, I have the data. I can switch models in and out. So just like a car and you just flip different engines for like do you want to be eight for when you're got to go fast or this sort of engine.

35:05If you need to do this sort of work. And so I literally did that in the last 24 hours. I went from ontropic to open AI. But the way I use Gemini for search. Gemini 3.0 for search and writing writing is, I think Gemini so far shines at it. Especially if it's a little bit more like literary a little bit more imaginary Gemini is exceptionally good writing. And then I'm using the XAI models for search specifically on Twitter kind of like the real time you search. So if you think about that, like none of those companies want me to have that they want my data on their thing and no other AI lab to have it. Here, I'm just like flipping through them as I want to and using whichever one like if this one is best at this particular task. I want to use it for that task, but all the data, all data is mine, all the context is mine. And now after building this, I'm like, I want a its own custom agent for each particular area of my life.

36:12Like I want one that's health at this point. The more I think about it, like I can kind of do it on this one. But that's going to get confusing. Plus God forbid it starts posting my health information on the block. I'm like, West God sick today because he hates a bad tuner. I was just here to see which model is the best. Like I didn't ask for West's dietary update. It's like I'm sorry. Right. So now like, okay, but like I want to be able to take a picture of everything that I eat. And it automatically gets logged macros, whatever. And then my, you know, if I want to track macros calories for the day, anything gets logged. And then like if you think about it, you can ask it question like, oh, I'm feeling sluggish today. Like why? And you know, I have a whoop that tracks my sleep activity. They have a developer mode that can API to the agent. Anything, anything, anything, anything, you know, all your blood work, you just give it access to everything.

37:14And then when you ask a question, like think about how much like data about you, it has to answer that question. Like it might know your genetic, whatever they are phenotypes mutations, whatever. What you ate, how much you slept, your heart rate variability, like just like I feel like what labs refer to this as like the AI capability overhang. Like how much these models can do is up here. How much we've sort of enabled them to do is like here. Right. And with these things, like it's never been wider because I feel like the amount of stuff that we can do is just insanely high. And we've just now beginning the people that are missing our open claw are just now beginning to see a sliver of what's possible. So yeah, it says broad connections using multiple systems, knowing about you in so many ways, knowing what your goals are.

38:16Yeah, this is so crazy. I just can't believe we're living through it and it's happening so fast. Yeah, I recently diagnosed the pretty specific condition that was bothering me that came down to something so specific that I don't know. Like I think it's going to change how people perceive their health because it's like online you get advice for the whatever general population here you can get like needle in the haystack answers. I got to get off this not so box, but I was like, no, no, this is, I mean, listen, I'd stop you if I was bored, but I am not. That is some fascinating experiments that you're doing, like you're building on thoughts from actual stuff that you've done, like is really interesting. But yeah, I agree the connection, especially with health, I had a, I have restless like syndrome. I'm trying to go to sleep at night. I like just sort of twitches and I've talked to doctors about it and like, yeah, nobody's really got a great solution. And then yeah, after talking with AI, it was, it noticed that this is like common in people who come from Sweden, right?

39:21Like my families, like my ancestors are from there. And that was just an aspect of it where it was like connected to the genetics that no doctor had ever told me. And I don't believe I'm sure that wasn't tied in medical school. You'd have to kind of be in the weeds on that kind of research to know that sort of thing. And you know, not that it cured it or anything, but it's good to know that it's like maybe a little bit more genetic than it is. Are you eating enough iron and things where I might be wasting my time not realizing that like it's like where the root cause of it is. So yeah, and I just think more and more stuff like that and the tiny connections between what you're doing every day can all come together in ways that will blow our minds and be extremely predictive and helpful. Yeah. Or manipulative. So we'll go to the next topic. What did you think about the Super Bowl ads? It seemed like ads are coming to chat GPT. I'm curious how you feel about it. Do you trust that they'll stay separate? Do you buy the argument that making chat GPT like, like, could it be, does it really need this to be free?

40:28And then that's like what it, you know, what the real goal here is or you feel like it's a money grab for Sam because he's in a lot of debt. Just tell me what are your thoughts on on anthropic trying to make a big deal out of ads. And yeah, I mean on the defense this week. So so many interesting things here because number one with a lot of the stuff. When you look at somebody's when you steal man their argument meaningly sort of look at what they're saying and you assume that they're being sincere. I mean, a lot of it does make sense. You know, Sam saw back in the days that they needed to have a lot of money to wrap up compute. So a lot of the moves that he made was in that direction. People were mad like, why isn't this open source? Why isn't this? I mean, we see meta, right? Facebook tried to do they tried going the open source way. That seems like it was a mistake just from a sort of I don't want to discourage anybody from doing open source, but you don't want to be competing with deep seek and Kimi and all the Chinese models because wow, they just like wiping the floor of everybody.

41:33Now there's this idea of like, okay, how do we insert ads? So tragedy has some insane usage for the free tier, right? The amount of people that are using it for free is staggering and they're using it without ads. And a lot of them don't even have to log in. So we got to give him credit for that. Let's like, like, I think we should start there. The fact that they were able to provide this is incredible. And obviously they have a business case for it. They're getting users, etc, etc. Google's catching up, etc. But just the free benefit that they provided was massive. So from that perspective, if they can monetize it a little bit better and offset some of the costs with non intrusive, non personalized ads, meaning that it doesn't, it never impacts what your agent, what your chatbot is recommending you.

42:33And that's I think what the anthropic, what their point was, right? Because if you, if people saw the ads, it's usually like this trusted advisor that goes, oh, and how about you buy this? Obviously, that's a nightmare scenario and Sam came out and said, hey, anthropic, you're lying blah, blah, blah, which is funny because like, the amount of views that Sam's tweet in response to that ad help, you know, is massive. So I always feel like Sam drew more attention to it that it would have would have received without it. And then I think the response to the anthropic ad was very low. Have you seen anything about that? Like the perception of it? Yeah, I don't know about. I mean, I have my own thoughts about it. It seemed like the perception. Yeah, I don't know. I feel like you're a little bit more on Twitter in the real time space than I am. But yeah, it sounds like people, you know, don't like ads and don't want it. But I'm too much on Twitter. Yeah, but no, I'm curious like what's what's what's your take is again, mine is if you take it at face value, then it makes sense.

43:35But obviously everything gets distorted, right? Google started like, don't be like don't be evil. And then yeah, like it went so, but we have some ads to sell here. Yeah. So I mean, like I, you know, I feel pretty strong about this one because like the, a lot of people describe the internet's like a original sin was sort of ads. I don't love the idea that when I typed in a prompt that the answer could have been geared towards making money from an advertiser because an advertiser who wants to sell me like a can of Coca-Cola is just the second you do any reinforcement learning on that. That even if you don't notice it, it's going to figure out a way to just sort of slide in coke. I mean, we've seen some really powerful ways to subtly change things and not really change the outcome like with anthropic. I'm doing the golden gate neuron research. You could just tell me a story like, okay, they could make it about anything, but why not just throw in like the guy that drink a can of coke.

44:37It's not necessarily manipulating you. Like they were going to put something in the story anyway. That kind of stuff is a slippery slope that I know the world's going to have to deal with someday. Now, if, if we take Sam and he like truly puts like all of the guardrails in the world in this thing and he has some sort of third party company that is like has transparency and we can truly tell that the ad is doing that query first and giving the response to the problem. The response to the prompt and then taking that data and then matching it to something that's like completely separate. Okay, maybe it's still trustworthy for a while, but like your, you know, like this thing that you built, like I don't know if everybody in the world who's using moltbot to go out and make their own advertising, you know, advertising based piece of data, like they could certainly train to like maximize profits. The goal is just like, hey, make this website that displays this kind of data, make sure I get the most money from ad revenue, like they're going to absolutely train on it. So it's not.

45:41It's not going to be contained forever. So I'm slightly concerned about it. I like the idea of people paying and I also do feel bad that not everybody in the world can afford $20 a month for a service, but if they can, the service will be working for them. And my thinking for that is that the government, I actually feel like the government should offer like a tax credit. I feel like we should have something where you get maybe $100 a year tax break and you're allowed to use that on whatever platform you want. So I don't, I don't want the government saying like everybody gets free chat GPT, but maybe everybody gets $20 a month, you know, times 12 tax break if they produce a receipt from any, any model. Right. And so something like that seems like it would, it would be more aligned in the long run. So I don't love it. I also feel like Sam is an incredible amount of debt right now. Like to the tune of trillions of dollars and doesn't have the luxury, you know, Google can be like, oh, we haven't put ads in in Gemini.

46:43But of course, they do have ads on search and they've made hundreds of billions of dollars doing that. So it's not like their hands are clean, but with AI being a bit, a bit different than a list of links. And also potentially in the future being super intelligent, hyper intelligent, very easy to manipulate us in ways we want to understand. It's a, it's a very slippery slope. And I'm, I'm concerned about anybody who's playing on the edges, even. So great, great, great points there. And actually, as you're talking, it reminded me. So you're talking about Google and Gemini and this idea of advertising. So Google for the first time or alphabet, I should say, right, the big parent company for the first time we're seeing a new segment in their annual report. Specifically talking about AI risks and the potential impact on advertising, right. So of course, Google has been considered by some as potentially one of the greatest business models ever, right.

47:47Because you're sort of like taking information that's not even not yours. You're organizing it and you're adding some links and people pay you a lot of money. Very massive profit margins, obviously, right, because you can paid for people clicking on your link. The amount of money that they print, you know, from search is, is absolutely insane. Was it 60 billion a month or something like that. I forget the order of matter. It's huge. It's huge. And of course, as soon as AI started coming on the scene, people had questions, how will these chatbots affect that model, one of the greatest business models to have ever conceived. And of course, that was kind of like people are saying that's why they didn't pursue this technology immediately because they were the first people to publish about it. Transformer paper in 2017 kicked off this race, right, LMS, the chat gpt moment, et cetera, et cetera. Google was the one that published all of that information.

48:52And finally, so they're sort of they're adding like two paragraphs, saying to their annual report, right. They're saying we generate more than 70% of our total revenue from online advertising in 2025. Many of our advertisers companies that distribute our products and services, digital publishers and content providers can terminate their contract with us at any time. If they don't find value, et cetera, of course, right. If it if it doesn't translate to sales, new customers, new reach, they can cancel makes sense, right. Then they're saying we believe AI is quickly reshaping the advertising industry, including how ads are delivered online. And we and our competitors are constantly adjusting to meet this shift and provide new and evolving advertising formats. There's no assurance that we will adapt effectively and competitively to meet this shift and that such advertising formats, strategies and offerings will be successful. So kind of interesting, right. They're saying, hey, we're going to have innovative ways of of doing ads and we don't know if it's successful, maybe that's just like lawyer talk.

50:01That's something that they need. But what are your thoughts about Google because you you brought that up. So Gemini is going to have already has kind of as built in if you're looking at the, does AI mode have ads in there? I don't know, but I mean, there's a couple of things. So one, some companies already sort of have these ads built into their mechanisms that they're layering things on top of, right. Like so, Grock is mostly being distributed through X and X has a long history of being Twitter and being sort of ad driven. So there's something kind of similar going on with Instagram and and then there's sort of a split in where AI is deployed inside of Google because when it's deployed right above the links that have ads on them, you're already kind of there. And then sometimes if you're talking about tools that are being deployed to YouTube, you know, it's like, is it going to generate AI in the way like something like Sora does where it's just for entertainment or is it going to also sort of feed to its own advertising system that like powers the entire

51:10YouTube ecosystem. So there's just so so much nuance in how they're doing it. But I certainly, at least with the chat bots, I felt like opening I had a really interesting position not coming from any of the previous systems, right. Like opening I kind of just really was its own company and they just decided early on and I'm really thankful for it to just charge $20 a month to pay for things. Like it wasn't like they started with an ad model and built around it, which is a little bit different than some of the other companies. So yeah, I mean, I think ads and this are just like naturally going to get mixed together. And I'm going to sort of be in favor of companies that you can pay money for and and have a service where they get their money based on how much they help you not how much they get you to click on somebody else's ads where they get the money from just because the incentives just alone are are what I'm basing my whole opinion off of like it's just it's very tempting for someone in the back office to make a tiny tweak to Sam Altman's like wall of protection and get some money in it.

52:18And it's that same wall is very thin and it's very skit. It's like something companies have to be thinking about it all the major companies. And in the long run what I what I want to do is build something similar to I guess what you have with open claw and just I think I just want a system that filters it all back out right like if they if I say give me a story and that GBT writes back like oh it's a great story about a delicious coke you should like this person bought it and they were so happy. I think I just want another local version to change it once again before it comes to me and maybe I'll even have some extra protections on my end say try to make me smarter try to not make me by products like try to make me a lifelong learner make sure that I stay curious about the stuff that comes in and then you know we sort of play this game where there's different levels but hopefully people are just aware where the incentives are. And kind of keep an eye on on these systems so they don't just trust it kind of as much as we have in the past.

53:19Yeah and you brought up golden gate clawed and that was a very interesting point because I mean anthropic did figure out a way to sort of like magnify a certain neuron or certain feature in the brain of these models to like think more about the golden gate bridge or think more about this. Yeah and imagine so so totally different from just an advertisement on the site of a bar but imagine imagine coke Cola just coming in and saying like hey can you just magnify the coke neuron you know 20% and we'll send over 10 billion dollars and it's like. You know I mean to some degree as a society Taylor Swift and all these like cultural phenomenon have they've been ratchet it up right like in all of our brains everybody knows the name Trump everybody knows Joe Rogan everybody knows Taylor Swift there's just these these have been amplified and if they were nobody's and they could have been amplified inside of the system. Not because it was trained on that version of the internet but because somebody else wanted the next version of the internet that you know the content that comes out of these systems to have it in there they could yeah it's just a wild thought you know no I think your spot on with this thing because it's like it you don't even have to tell it necessarily just turn up that like those neurons for specific specific thing and it's more likely to push a person when we are another.

54:44Because certainly did we have a news article or something we're going to talk about that has to do with like influence and persuasion yeah we let's see we have a couple things kind of like that I mean there's an article that was artificial intelligence can be talked into believing a lie so deeply that it's internal logic flips to support it. I could kind of go into that one that was a deep mind paper but it was linear representations and language models that can change dramatically over a conversation and it just it basically I can go to a more detail but it's kind of like have you ever met somebody who kind of lied to themselves so much that it kind of became real to them like they didn't need to fake it anymore they just kind of almost convinced themselves that like they've done something they had in or like what they did was honorable when they clearly. It wasn't honorable the first time they thought about what they did and like they kind of like rewrote history to a point where it truly is there it was a study that kind of shows there's some aspects of that they can happen throughout a conversation where at first you know some you can start a conversation and and a chap up might say something like you know this war is is not good or something and then you just keep talking to it about it and it starts to just instantly kind of tweak all the decisions after that in a way where it feels like it really did you know and if that data gets fed back in.

56:04As we're the enforcement learning then even the whole system starts to believe that so we sort of start shaping a new reality from history to. Yeah that's an interesting point because I do feel first of all most of us do have some examples of that in our cells where we kind of lie to yourself or write history I think we all have that and actually there's a great there's like the oatmeal comic. There were a great thing about it actually maybe we'll link it because it's kind of a good glimpse there's certain things that as especially as we get older we tend to form certain identities and it's actually a lot better for us to just keep those identities and new information and that sort of shake it or new information that conflicts with it it actually makes us like it's actually dangerous our brains perceive it as danger because it's like oh we have to like. We think everything so to speak and so we get kind of like very like emotionally charged like no that's wrong like I don't want to hear that type of thing you know everyone had that at one point in their lives or another I'm sure we can recognize times when we behave that way.

57:18What's interesting is that AI chat bots when they're there's a study where how well are they able to reverse people's beliefs in a certain conspiracy theories or certain beliefs that are incorrect. I think they were using the flat earth theory or whatever it doesn't really matter what it was but the point is that if a human is convincing you and trying to change your mind that we have resistance to that we're like because I mean we don't want to lose face we don't want to be manipulated right and right like we have this kind of responsive somebody else is trying to change your mind not all the time but there's certain things where again that sort of resistance kicks in because our brain perceives it as. Danger and I mean you see some you see some weird ways that people justifies taking to their preexisting opinion but chat bots are a lot better than humans at doing that and it seems like one of the biggest reasons is we don't perceive them as another human being so we're more likely we're more

58:25valuable we're more likely to be able to accept that because we don't maybe don't see us losing face in front of another human being so that's one thing to keep in mind like these chat bots as you said as they get better and better more influential we're already like more likely to believe them right so actually I think about it that whole idea of me telling the chat about use WordPress and it going no we should use Versal don't you know this is this is what you wanted like I never thought twice about it like oh of course obviously it's I've never heard of that company in my life we've been using WordPress for last 10 years like I just need a credit credit card please I think Versal is definitely the company for you and you're like all right oh man well also oh that would have been even worse that no they do have a free hobby tier which is great which is great so it's free but you're right yeah it could have been very well could have been something I had to pay for versus WordPress that's well host you got to pay for hosting either way but yeah exactly I mean if I needed to pay 10 20 bucks a month

59:29for it I would have without thinking about it versus a software that I've been using and familiar with for the last 10 years because the argument made sense because it like personalized it to me from a previous context so no you wanted to do a case study where everything is handled by bots Versal is the thing that really because it connects to to GitHub so we're going to be doing everything on a GitHub and deploying it to to to Versal and I was like oh you're it made sense the argument made sense yeah I mean like the equivalent of the golden gate neuron as long as like the honesty neuron is turned up the like intentions for you are turned up I want to protect you in the long run is turned up then like all these other things maybe Coke is also in there but it's not you know maybe it's okay because they've done a good job but these are just the very strange nuances about how magnifying the impact is of tools like this on the world that I sometimes think about but it's a very broad thought and I know

1:00:31everybody will sort of tune their models different and different ones will emerge at different times and there'll be some kind of conglomerate in the middle that that will be a little bit easy to defend a little bit hard to defend at certain times in certain places and I don't know it's kind of like when it was like clear smoking was bad for the world but there's like some data that showed it wasn't that bad or the ones that did had some errors in them and like you just kind of got to keep an eye on it very carefully it's so because like with with politics you see there like you said that the smoking example there's so many examples where somebody's like oh if we just introduce this legislation it fixes a lot of problems it's such a minor thing that leads to massive improvements the paint the lead and paint or whatever it was remember that was a big big deal that caused so much damage over across such a large population and then they just like added a few things here and there and like it was just a no-brainer at the same time there's a kind of on the opposite side there's if you're like oh if we just tweak this one

1:01:34thing the world's better and it's not because it's the law of kind of like the unintended consequences right and it just makes things worse and worse and worse and then it's harder to repeal that lock because like was here in America we don't really have us like once you have a lie just stays on there there's no like we don't have to reconsider five years from now and go like should we like renew it for another five years so there's there's stuff that's on the books that's that's harmful I feel like and if you think about it with these models like just as you're saying this is like what if we turn down you know because like consumerism right this idea that we just need to like buy stuff to feel happy just buy buy buy I feel like society's a whole would be better that was turned down a little bit but of course the economy and the job market and everything everything all of that stuff would be better if we turned it up a little bit um right and it's like I'm sure at some point some freaking politicians going to get into their minds like oh like we can like turn this knob to make life

1:02:39whatever they perceive in their mind to be better right and you might agree with them you might not be right defending on which which way you think things should go but man that's scary to me because it's like how do we know what's right again people might be like oh we should turn this up because we've seen that already with online stuff and everywhere else and we have examples where it works phenomenally well right and we have examples where it just works really really poorly so right yeah just that idea that you know maybe Sam will put ads in there and they'll be separated or maybe he won't but either way I'm never going to know the difference you know it's kind of where the thought is yeah um but yeah let me before we get finished I did want to touch on one health topic that I thought was kind of interesting that you um I know you're kind of taking a particular interest in health right now and uh the paper was published in nature I'll probably butcher the title it's got a lot of big words in it but effector host intercom map links type two secretion

1:03:41systems and healthy gut microbiome to immune modulation but past the the the technical title basically what they discovered this week or was published this week is some research that shows um you have these gut bacteria that live live in our stomach so we're eating food we're eating fiber they're eating the fiber these bacteria do not have the same DNA as your body does these are foreign bacteria but they live symbiotically in all of us it's very healthy to have them in there even though they're not made of the same DNA as we are and they now discovered that there's this tiny little needle that they have um straight up like injecting proteins directly into human cells so they sit on the gut lining right next to the human cells like your gut lining is made from your DNA these bacteria have their own DNA whatever species of bacteria they are and they found a whole bunch of them healthy and unhealthy can just puncture like a straight up little molecular needle out of their cell through your cells and then inject proteins directly into your body into your

1:04:47actual gut lining and those proteins once they're in there can take form and function sometimes they can be RNA and be read into other proteins they can have cascading effects it opens up like this whole new thing that people didn't they thought they would signal in different ways but no one really knew they could inject protein like right into it so there's a now a mechanism for why eating unhealthy can make you feel like crap like literally down to the protein level and affect your mood and affect depression and affect your serotonin levels and your dopamine production and there's all sorts of stuff that can now be kind of learned from this new frontier so I just wanted to kind of get your thoughts on what do you think what do you think about gut bacteria are you taking the fiber that you eat serious and the probiotics and in your environment so it's funny as as we're talking about this I just wanted to kind of like so it's funny because I on my desk I have my my supplements here that are very very specific ones that I have like I got my Thor Thorna

1:05:50methyl plus Sam E oh my gosh oh my gosh I got the ingredients all coincidentally on your desk or is this is like you brought them in for the so I have him here on my desk because I I spent most of my a lot of time getting my desk making well well no I spent a lot of time here so I'm like just having them here I always thought you had to take most of the stuff with with food and then I asked Gemini and it's like no there's a lot of stuff that you don't have to in fact it's better there's so there's like in the morning I have it here because like as I'm like scanning the news getting ready I just take a bunch of them with with water because a lot of them like B vitamins and some of the stuff like that they're really just you shouldn't take them what we should you should take them away from food because I guess they the absorb better and so the issue that I've been researching a lot is this whole idea of a methylation so and more and more I think that people are hearing that

1:06:50like the MTHFR mutation people supplement with various methyl groups methyl donors blah blah blah and apparently like if you don't have enough and if you have certain mutations or your body is just bad methylating which is basically taking sort of raw materials and then like breaking them down to the stuff that you need downstream well you could have lower production of ceratone you could have lower production of dopamine etc etc etc and there's some other stuff that if you so how I stumbled upon this is one day I got home from the store and I had those I bought some oysters there were like those canned smoked oysters that I just love and there was like some some fish that I that was recommended that's canned from Tim Ferriss and Peter Tia podcast I think and so I had some of that and I was I always try to have some fermented foods so kimchi

1:07:54salcrout stuff like that and then I had anyways and I went nuts right so first of all I felt like my brain was weird a little bit like depressed like I kept getting stuck with certain thoughts felt itchy I couldn't sleep like it just wasn't feeling good and I had those symptoms like flare up every once in a while and finally with the help of Gemini like I was able to figure out what it was so apparently if you have these MTHR mutations and you're under methylating so your body doesn't produce everything you need as well there might be certain blockages or whatever you sometimes have a bad ability to control a histamine so histamine is the stuff that kind of like basically all old food or anything that's fermented beer wine stuff like that anything like canned fish is horrible anything that's like basically I had the perfect storm of all foods that were in and of themselves healthy

1:08:58or good for you you know that triggered a massive histamine response and I just went freaking nuts just did not feel good well and most people would probably like not have noticed that because but if the methylation cycle that process is impacted then then you have certain issues so um what I'm doing now is just trying to understand what I need to supplement with and what I need to keep out of my diet so creatine uh I'm I'm I'm sure if you've been following the whole thing with creatine because recently it's been making headlines in that apparently it's amazing not just for your muscles but for brain function have you have you seen all those headlines popping up and yeah they're including like when you were like not I mean I don't know how much to trust that but on Instagram definitely I saw a few people talking about I like if you haven't slept a bunch of creatine can help really put your brain back into a normal spot and a few things like that so it is part of my diet I have heard good things about it over the years I know it's pretty well tested well studied how much do you take per day roughly or what's your target

1:10:01I guess just like I ball it I don't know I got like a I got like a like a bin full of pills and they'll take two or three I think it says you can take like up to eight a day I don't take nearly that much but maybe on a really tired day I'll take four or something but I'm not dosing it in a way where I can give you good insight into how I feel from it it's surprisingly a lot so one of my trusted advisors online for health and health-related stuff is a doctor Rhonda Patrick uh for double dead that's what I heard it from too yeah okay I love that gal I trust your I trust her a lot because a lot of the which she says later no matter how like wild certain things sound initially like whenever I try certain things I know they work so I know whatever she's saying whatever she's doing it's really based in in deep research and understanding whereas there's a lot of other people that they just maybe hit her miss or maybe they like over index on something or they just

1:11:05follow the trends so they're more about like trends she doesn't do that I think she's really like science-based um so like one of the things way back I was listening to her on a podcast and she's like oh yeah like if you do an event fasting you might lose weight and also like gain muscle and that went that was contrary this was a long time ago this was like 2017 or 2016 when I first heard it that kind of went against a lot of the content like it was contrary to advice could be like oh well you know energy and energy out lots of thermodynamics whatever you know what it's whatever you eat whatever you burn so she's saying this and I was like is that is that true but I guess she had science there's a person in San Diego doctor panda San chin or something like that that measured this on and he measured it on rats and they had an isocholoric diet right so the same amount of calories the same amount of like everything was controlled and the ones that limited their

1:12:06fat uh their um feeding window got all like lost weight got gain muscle and then once those studies are coming out you kept seeing those results more and more people were posting anecdotal sort of evidence that that happened to them and I tried that night I lost a lot of weight just doing that like I would just walk up and I would eat breakfast and I would try to hold out as long as I can and that was the only thing that I focused on that at some point just like I started losing a lot of fat off my body which is I always had trouble doing that um and there's a couple of other things just like that where like if she says something at this point I'm like it's probably right and I try to try to trust her so what she's saying that I'm not testing out just as of a few weeks ago is uh if you take five grams of creatine per day which is what was before kind of the recommended dosage that fills up your muscles uh but what she's saying is that anything you take above that like your your muscles are greedy so they're gonna take up to a certain

1:13:09amount but anything past that is available to your brain and this is where amazing things happen because creatine helps with the ATP production uh with muscle contraction but it's also apparently really good for the brain in terms of cognitive function and um the the thing about the sleep deprivation that you're talking about if I understand correctly they took two groups of people in these sort of like randomly whatever the tests to to where half of them were sleep deprived and half of them weren't or maybe it was this I don't I don't know the details of the study but two groups one sleep deprived one not and then they test them on their ability to take a certain test or whatever so just cognitive function and so if you're not sleep deprived you're up here and if you're sleep deprived you're like way down here right you're you're barely functioning and then for the other group they give 20 I think it was 20 grams is it I think it's mentioned grams or milligrams probably grams of creatine and surprisingly that group was up here oh really

1:14:18it was that big of a difference it was more really filling hmm like it it gave them almost like it like I'm gonna take and improve their cognitive abilities past even like the the baseline surprisingly um which suggests that we might not be that that might be uh just a good thing to supplement with apparently helps against Alzheimer's as well uh double check everything that I say because I'm like you know I listen to these podcasts and uh most of it I feel like it gets so lodged in my brain but like I'm sure I'm missing some details but creatine is one of the longest studied supplements it's one of like the safest because just of how much like theta we have on it so it just seems like a no-brainer but man 20 grams is a lot to take uh man once you get that open-class system set up for your health data you'll have to have it do a couple uh mental tests like you have to say like oh I only slept in you know seven hours or six hours but I took this much

1:15:18creatine like give me a quick like five-minute mental test and then you know we'll have to see if it can come up with some sort of a pattern there I mean once I started taking the things that Gem and I recommended like my the change was kind of drastic uh like I became a lot more chill and calm and just like that that anxiety that was there it got reduced which is linked to not clearing histamine out of your system fast enough and I was just eating stuff that was triggering it that I just I didn't realize it probably also combined with some other stuff but the point is once I added those methylation things I think my body was missing like I felt much better so all right well yeah I feel like that was a good recap for the week we got some health stuff in there your experiment at the beginning was like super fascinating thanks for sharing that and putting your credit card at risk for everyone's entertainment you know it's always always good

1:16:19stuff and yeah I'm excited to see I'm a cat like by next week I feel like you'll have a 50 agent like you know company working full-time probably acquiring other companies and merging their data who knows what's going to happen we'll see I'm definitely going to take it slow and so I just wanted to confirm like it is in grams so creatines and grams so people for for physical take five but now the new sort of trends take up to 20 grams per day which is kind of a lot in terms of just you got to kind of break it up into multiple whatever but yeah and I'll be posting more information about this AI run and operated business and also within the next depending on when this gets published right around the same time I am having a full tutorial for people how to get started so for people that were kind of waiting for that I apologize tons of stuff happening obviously but yeah it definitely exciting times that we live in and man just just absolutely

1:17:21incredible to kind of be witnessing this this firsthand you know yeah just taking a day by day so yeah thank you for your time and yeah we'll meet up again next week and see what's going on see how things have changed yep everybody that's watching please hit subscribe please like consider doing all those things and we hope to see you next week oh actually we have a big guest coming up huge guest you know like the probably the most excited have ever been to do an interview in my whole life yeah I mean definitely so ready it'll be done by the time people watch it but for West and I were filming tomorrow so we're like I was gonna think about it all day I'm not I don't know if I'm ready I need to get my sleeve I've already watched 15 I've already watched 10 podcasts with him so I'm like okay I got I got a catch up I watched a few because he's definitely wow he's out there and he's just I mean one of the smartest people maybe ever so I think so I think so that's coming so subscribe you don't want to miss this one because I I don't know how we got

1:18:25this guest booked but yeah we did like pray it comes through tomorrow he shows up okay all right but anyways we'll talk to you later thanks man see you soon all right subscribe take care bye