I went from a high school dropout to building a seven-figure online company within a couple of months

with just talking to AI for 4, 6, 8, 10 hours a day, like it was absolutely relentless.

You've been the person who I've been following on this journey who is most excited, optimistic,

everything I've seen from you is like, guys, this will change your life, learn to do it,

but put in the time it's like Bitcoin, you're not going to get it straight away.

It's an asymmetric bet, right? Like even if you're wrong, like your time wasted is not as

near as much as the opportunity cost of not doing it if the AI backsees and the Bitcoin backsees are

right. The money and everything's broken and it's getting worse and worse and worse and we've got

like three or four hundred thousand Australians now homeless because the government's just

obliterating the money. It's absolutely worthless.

Hey everyone, Michael here with the Bitcoin Way podcast. Thank you for tuning in.

Today on the show I have Sean Clark, a good friend and partner of the Bitcoin Way. We dug deep

into AI. He has been on quite a journey, fascinating stuff. I mean, these tools are incredibly

powerful. If you are behind and you've just been chatting with chat GPT or whatever your little

LLM friend is, this is way bigger than that. I'm going to be digging deep on this very soon. You

will get a lot out of this conversation and I hope you start going down this rabbit hole as well.

You will love this conversation. Everyone, like I said in the intro, I've got Sean here with me,

Sean. Welcome to the Bitcoin Way podcast. Michael Jordan, always a pleasure. How are you, mate?

Good. My friend, it's been a long time coming. I don't know why I didn't reach out to you long,

long ago, but I knew I had you on my list and I'm glad we get to talk. I think I found you at like

peak Sean time because it's always been Bitcoin and then lately I've seen your feed more and more

interesting stuff on AI. I want to spend a lot of time on that because you're the person I'm

paying most attention to who's kind of going down this rabbit hole. I want to start though with,

we've got bombs dropping and missiles flying and the world is just, I mean, in absolute chaos.

What's your take on, like as someone abroad, probably a bit removed, I assume Australia is not

heavily involved in any of these interventions, like what's your take on what you're seeing?

I guess first off, people killing each other, like I don't know if it's old man in a room or

whatever and they want to move the chess pieces around. It just seems so absurd that we're

sending our youngest people to kill other states, youngest people really. When the issue is not,

I have a problem with those people or those people have a problem with us. It's the people right

at the very top. But as for Australia, any intervention fuels gone up a little bit, about

50 cents a liter as filling up a couple of Jerry cans and the vehicles and stuff. Yesterday,

we haven't got fuel reserves strategically for that. I guess that's the impact here. Obviously,

we are fairly far removed. The politicians and stuff are trying to jump in. But I guess my thoughts,

I just don't want to see people dying, Michael. Can we just not get in a room and talk peacefully

or whatever? I guess that's trying to live in a perfect world. But that's what I would prefer.

As for what's going on, personally, I've been very removed, head down focused with AI and

Bitcoin. I did see those videos, the missiles flying and things like that. It looks like

being the dictator, supposedly, from the little that I've removed has been taken out. But

then what happens? Is there power vacuum? Is there civil war? There's never a good situation for

anyone. I don't know if there's better before. It's going to be better now. There's going to be

this long process. But from all the wars, and I just finished reading a book called Chicken

Hawk about a helicopter pilot in Vietnam. They went there and halfway through, they started to

question the Viet Cong, and they started to realize that the Viet Cong were just the villages,

that they're fighting and they're shooting and they're killing. There's a lot of people get caught

up in the mix. I guess that's wanting to live in a perfect world of people are willing to

go and fight for their country. That's one thing. But the casualties and innocent victims and

stuff, it's hot breaking. Yeah, I completely agree. And it's funny, five years ago, I would have

just laughed at you and called you a hippie. I think it's taken this, I think as I've grown to

appreciate what it is that funds wars and incentivizes wars, I have a much softer spot for what's

really happening. And at the end of the day, it's a bunch of young people going to die for something

that's not particularly useful. And they spend all kinds of narratives around it. And it's just,

I mean, those just roll right off my back like I just don't buy a single word of anyone, any side

of the political aisle at the top explaining why we have to go bomb another country. I think all

of the justifications are total bullshit. So that's my take. Interesting though, Bitcoin is not

ripping with this. I know a lot of people have, you know, it's the edge. It's the, and it feels like

we're kind of close to that point where enough people understand that that you would think we'd see

Bitcoin doing something in a world of uncertainty, but seems like a lot of people still view it as

speculative. What's your, what's your thought? I feel like I've been talking a lot more price action

just because people are so baffled than I usually care to talk about, but I'd love to hear what

you think is going on. I think ever since we've come off the all-time high from 127 down to 61,

I think was the recent low, like what an opportunity. That has been for anyone like if you're worried

about price, just by stacking chill, get as much possibly can into self-custody, get the Bitcoin way

to help you protect it if you don't know how. But in regards to it, not ripping, I woke up and I

haven't checked the price. I don't know if I've seen the price in two, three weeks maybe,

but I went on X and I saw Peter Schiff just having a meltdown over it. And it was like,

Bitcoin's at 71,000, you need to sell everything now. It's not going to go any higher. And he's like,

Bitcoin's at 73,000, you need to sell it now. It's not going to go any higher. So I was like,

I opened up the app and I've just opened it up here again now. And it's up 5.7% in the

past 24 hours. So I don't know, maybe it is ripping on this news and people are getting it.

But I guess for everyone that truly understands it, if the whole world understood it on this news,

we'd be at a million or higher straight away. So yes, it's starting to move and it's starting

to go back up there. Is it completely undervalued and is the price under $100,000 USD

and sane? Very much. I was an amazing opportunity, but very much it should be higher

because of what's going on. Why isn't it? I think there's a lot of misinformation. People are still

they just don't understand. Like Bitcoin takes at least 100 hours to be able to get a base

acknowledge. Everyone's busy. People can speed read that by reading the big print by Lauren

Slapard, but it still takes work. And as fear inflates more and more and people have to work two,

three jobs to be able to stay in the same place or there are wars going on. People are focused

elsewhere just trying to keep up or trying to just have a normal pace of living. And they don't

feel like they need to study Bitcoin because when they turn on the news, oh, Bitcoin's

this magic internet money, it's a Ponzi scheme or whatever it is. It's used for all the

nefarious things that Jamie Dimes says. None of that's true, but that's the information they're

getting and they're struggling and they don't have the times. So they just don't put in the proof

of work. And I think that's probably more so why we're not at a million dollars today or 10

million or whatever the fear number is that people want to do. I think it's just Bitcoin takes people

effort and it's not going to get there. And I think we're just still early. Bitcoin's still

very new and it's very different. We've never had a money that's also a technology before,

at least in recorded human history. Yeah, I agree. It's like truly the proof of work meme is

kind of what it comes down to. You have to be willing to because so many people in my life

understand that Bitcoin is a thing that I have a high conviction in and spend all my day thinking

about and talking about and they haven't picked up a book. It's really just a matter of

expending a little bit of energy to zoom in on what it is. Okay, so we'll probably get back to

Bitcoin. I want to get to AI. So first of all, I want you to tell a story before I'm going to ask

you a little bit about like the mechanics of your setup because I want like this is a very self-serving

podcast. I'll be honest. However, you tweeted about maybe a couple of weeks ago, you saved like

$4,500. It's gone up a lot since then with software. It's about 10 grand now worth and software.

So what what happened? What did you save money on? How did you do it? Yeah, and so obviously

AI is this an amazing deep-flationary technology for a little bit of context. So just to be

able to set the frame, Jack Dorsey, the founder of Twitter, the creator of Twitter originally

is now CEO and founder of Square. He just laid off 40% of his workforce, 4,000 highly technical

coders, computer knowledge work is and replacing with AI. So that's like on a grand scale and

that's happening more and more and more. In my own life, what I've done is in Christmas holidays,

I kept seeing this stuff about OpenClaw, this AI agent that ran on its own computer and can

control its own computer. And it was on holiday and I actually wanted to stop and have a break and

it was amazing. I hadn't had a proper, mentally I hadn't been on holidays for like five years.

So I was like, I'm not going to get this while I'm on holiday because otherwise I'm just going

to go down the rabbit hole. But as soon as I got back to the company and back into it, I went

out and bought a Mac Mini, downloaded and stored OpenClaw on it, set it up. Now you don't necessarily

need a Mac Mini. I just wanted to give it a sandbox space. And I think this is the most important

for anyone playing around with AI and OpenClaw. It's incredibly powerful technology. It can take

full control of your computer, access everything. So what that means is give it its own device. Don't

put it on anything that touches Bitcoin. If you do anything, say on your personal computer,

Bitcoin related, don't also on store and store OpenAI or OpenClaw or any AI tool that can take

control and look through your folders and everything like that. I think just caution it.

But anyway, got that and stored it, set it up and then started really feeding a lot of information

on who I am, what I do. So I gave it 90 of my YouTube transcripts, 30 gig worth of X data,

a whole lot of company emails and all our company content of what we're doing with our online

training business. And then I got to a point where I needed to fill a contract and with the

course at the moment we hold hosted on our WordPress website, now I'm LMS, a learning management

system. The organizations wanted to embed the course onto their own learning management system.

It just needs to be put into a certain file structure, a scorem file,

pretty easy to be able to do. But that software to be able to convert how our course was

into that scorem was about $2,999. So that was the first $3,000 saving straight away.

I got my OpenClaw agent to be able to code and build an app for myself that could do that. So

that's the first subscription or software package that I didn't need.

The second thing is we need licensing around it. So use the limits for this particular organization,

just under 1,000 people for 12 month access. To be able to do that and actually make it happen

was about another $1,500 worth of software, like for that functionality. A couple of prompts

and then I built that on top of it and the course builder, the packageer, etc.

And then the other savings coming from it. So like built all that, it did it. I tested it as

incredible. And then I really wanted to increase the quality of the output of how the scorem

looks and what it can do and the functions and features. So just started working with it and like

googling like the best features that you can use within scorem files and packages for online course

and stuff. And then just re-created those and built those into the app which would have been

about another $6,000 of either plugins or to be able to pay for somebody to be able to code

and stuff. And obviously I've got no coding experience or anything like that. But intrinsically

I knew exactly what I needed to achieve and how to achieve it. And I'd been looking at all the other

products and software that was out there in the market and I just recreated, got the AI and

to recreate and refine it. And for example, and so that's just like one small thing. It's built

like little laps instead of a task taking five minutes, you open it, you double-click and it fires

off that automation for those as well. But like as for cost savings, someone else with an open

claw, like you're not going to get your AI to go out and build a course package or whatever.

You don't need that within your organization. But the point of it is I did and I would have had to

have paid for it regardless. I probably wouldn't have paid for all those extra features like the

five, six thousand dollars on top. But I would have paid for the four and a half thousand. I had to

I had to have that capability. It just built it and I can make as many courses I want. So I don't

need to worry about like paying more and higher tiers. Like if you have more people like over

a thousand people going through the course and you they charge you more per person, etc. So they

really capture it. And I guess the point is we can all create whatever it is we need specific

to our company, our business, our workflows, whatever it is, personalize to us and then we can

keep improving on it. And it can build it with our context, knowing who we are, what we do

and how we need to be able to use it.

Wild. So it's like a magic box, man. It's like a little genie.

No, I believe you, man. So what is the, what percentage of this work is done locally versus

are you have a bunch of API calls going out? Are you paying a bunch of money to get this done in

order to access other models? Like I'm not fully aware of like a mechanic with the setup.

Yeah, 100%. So just recently this week, so I'm so, so, so quick at the moment. But just as

week, when 3.5s come out with a local model that runs on the Mac mini, it's,

it be honest, powerful as what chat GPT 4.0 was like running locally free forever. It's yours.

It's amazing to be able to do that build out and everything like that. That's not powerful enough

to be able to use. So it is using API and calls. I've got a couple chat GPT and

Thropic for a little bit. It's been issues with the open claw doing it. They don't really like it.

So that's not really mainly in use. And then Kimi, which is a Chinese model, it's super capable.

It's super cheap. The thing about it being as Chinese as you don't know where your information

your data is going. So being very selective with when that agent that's connected to Kimi,

what he's processing is he just running the task and then pushing things into the background.

And also being aware that some things can run locally. They are very slow and they

sort of need to happen overnight and you just come back to something ready instead of waiting on it.

But having like GPT 4.0 level local now is absolutely incredible. If we had this last year,

oh my god, the future is here. Now that we've got GPT 5.3 or whatever we're like,

oh, this is a little bit slow. You know, we get used to it. If you're computer,

browse, it doesn't load the page instantly and the video doesn't come or whatever you're like,

what's wrong with this thing? But when we're kids downloading from Limewire or whatever,

like that, it could take hours to download a MP3, for example. But that's the set up at the moment.

There's Quen running locally and the main models are running Kimi and OpenAI with the codex

for the coding and for the app building. So can you just, like through Telegram, ping your

bot, you named it Lindy? I called them Charles Lindbergh after the great aviator and Charles's

nickname was Lindy, yes. Okay. So could you ping Lindy right now and say, go build me a website?

Yes. Yeah. Just describe here. And then it's something like that. You want to build a quick

informational marketing website or something. Is that the kind of thing that's low enough processing

requirements? It can do that locally. Does it, like, can you go purchase a domain name on your

behalf? It could, if you gave it a credit card, for sure, or the, say again, have you given it

access to any sort of financial? No, no process. No, no, no, no. Yeah, the moment is very much sandbox

and it's just building things in parallel on its own thing. You know, when I need to get the

course, I'll bring it in there or marketing material. One thing which is pretty incredible.

Michael is, gave it a whole lot of like off your website, all the information about the course,

everything that's happening and said, hey, like I want a 45 second animation. Can you make one?

And then it just like pulls this thing up on the computer. I need to look at the name of the

program. And it just makes this like incredible graphic animation that goes through them like,

oh, can you add sound to it? It's like, I don't have any access to sound go here or here. So just

downloaded the sound file. It added that in and then exported it all for me. And at the start of it,

there's on two, two of the slides out of seven and the slides and the slides were all animated.

It, um, some of the wording wasn't quite right or as like bunched up and things like that.

And you just speak into your phone, just saying on slide one, can you create a little bit of separation?

Between the two words and then on slide five, can you like remove that last line and do this?

And it's like, yep, done. And then all exported again. You've got like another version.

So that's like one thing that's just absolutely incredible. You just talk to it and it can do whatever,

it can make power points and images and graphics. As for websites, it's built us,

a whole dashboard, website, everything like that just from a couple of prompts. And then

what it spits out first off is like absolutely amazing and like a minute, two minutes or whatever

it builds and you're like, holy cow, it's something that would take me 10, 20 hours to be able to get

to like that, that level of looking nice and functionality. And there will be like a couple of

little quirks and stuff that you want to change, you'll be able to refine or things won't work and

they won't lead you to the next page where you need to be. And you just talk to it like a human

and be like, hey, can you fix this? It's not working. I'd like that button to be a little bit brighter

or can you bring the opacity on whatever it is, whatever you can imagine. You just talk to it.

And I'll figure it out and it'll change it. And then another two minute slider is done like it's

live. So you spend four minutes instead of like 20 hours doing, doing whatever it is.

So not a lot of babysitting. So you could, for example, say, hey, let's say you wanted to give

an access to a debit card. So you could just, you auto transfer 100 bucks a month or a Bitcoin

wallet, you auto transfer a number of sats per month or whatever. You could tell it, hey,

you have a $100 monthly budget and you can, you know, you can subscribe to tools, API calls,

whatever you need to do to get the tasks done. And then just ping me if you need extra $50 or

whatever. And it can facilitate, it can do that. It can do the account creation. It can do,

download any tools it needs. It can figure out those tools because obviously it has access to all

of the support pages, YouTube video, whatever it needs to do to figure out how to do the stuff. It's

if it's not already baked in. And it's then learning those skills like it's refining its ability to

serve you in those capacities. Is that, do I have that right? Yeah, yeah, bang on. And I guess

specifically to agents like OpenClaw is it's a backtrack a little bit. Yes, put a pen in that.

With the AI agents like chat GPT, for example, you're like I've used it very intensively.

But it like forget context over time and start to just not remember previous conversations that

you've had with it. By giving OpenClaw its own Mac mini, its own hard drive space, it saves all

these different files and you can give it a soul so like its personality, which is Wendy, the great

aviator and to be able to help in specific areas of the business. But then you can give it different

skills in those folders. So one of those skills could be here's like your credit card number

to be able to access it. It's got this much when it runs out, ping me, let me know. But the skill,

one of its skills is that it has access to that credit card. One of the other skills is that

when you ask it to do something, it'll go out and it'll choose whatever API is that it needs to do

and then it'll use the skill of paying with that credit card to be able to do it. So you're like

building out and you're training it. So it's like a, I guess the best way to explain it is like a

junior employee, they've got a brain they can think, they've passed school, you know, a university

or whatever. So they're clever enough, but they've never worked in your organization before.

They don't know your workflows, your work process, whatever is specific to you. So you need a

training, you need to give them the standard operating procedures and then build out those workflows

and those automations. And at the start, it's not plug and play and it's not easy and there's

still updates and it was helping somebody set up open-claw just yesterday. And because of these

new updates, it wasn't taken control of this whole computer. The computer wasn't given it any

access at all. So I was like a dumbfuck and brick at the moment, you know, and that took us two hours

and we couldn't quite get it. So it's not like you can just go out by Mac Mini, download it and

then be like, hey, I'm going to make $10,000 in software or I'm going to give it $1,000 and it's

going to go out and try it into a million dollars. And people are doing these things with these

AIs and I did it with the software and different workflows and things like that.

Other people have told their open-claw agent, like, here's $1,000 if, like, once you lose it,

you're dead. So like, make money and then it'll go out and make money or people are getting their

AI to scrape websites. And this is the sinister side of things but it's just to give a good example.

Scrape websites and then send out like 50,000 fraudulent invoices to companies and just hoping that

they won't look, accounts parable, won't look and they were just paid and they reckon about one to

two percent of companies will just pay it. And so they're made like a million dollars or whatever

from doing that. But the point is, like, it can be powerful and it can do whatever it is and it

can make money in positive ways, negative ways, it can save you money. I think more importantly,

it's about learning the tools and saving yourself time and 10kering because overall, this is the

worst it's ever going to be. It's only going to get better and better and better and better and

better and to be able to have like, clan running locally, which was as powerful as GPT 4.0,

like on a $1,000 Australian, a $600 machine, that's pretty mind-blowing. But all of these things

they're not going to just, you're not going to download open-claw and then it's just going to

do it. You have to build it into it and you have to teach and you have to train and then you need

to stack those skills up. What are your favorite plugins that you've identified? What's been

most useful for you personally that a local LLM isn't going to get you? Is it like image generation

that maybe you need more powerful tools? Is it video? What in particular has been helpful for you?

Obviously, building out that software package is the most editing documents and putting

the initial draft together in various company documents and stuff. It sounds very, very boring.

That's to be able to go through and to be able to format it and our language, our style,

our definitions, in a very nuanced focus way and then to be able to format it so it looks

like a really nice PDF type document and they just need to go through and just change a

couple of things. Instead of that job, I don't know, taking four hours maybe or even for my assistant

to spend two or three hours on it, then for me to spend another two or three hours on it at

the initial drafts done in a minute, two minutes or whatever and then I can go through in like

15, 20 minutes for that sort of final edit, polish or whatever and then it's good to go because it

it's got all the past documents, everything like that so it knows the style and then it'll lay it

out how we want it. It sounds boring or not revolutionary but that's pretty powerful.

Yeah, imagine over the course of a week, like those little time savings, like I'm thinking

right now, you know, every day we've got six segments going out on our YouTube channel every week

and I'm manually typing descriptions that, you know, like that kind of stuff based on transcript

summaries and that. So here's one for you. Drop the my YouTube video into a Dropbox file,

it'll go in there and get sorry into a Google Drive folder which it's got, so I gave it its own

Gmail account so it's got its own email which obviously comes with like that Google suite of

Drive, not Dropbox. Google Drive, so drop my video into that and then it'll go through and get

the sub agent to pull out the transcripts and then it'll go through the transcripts and everything

like that and it'll create the initial title, the description, the tags that need to do it and

then also like a draft post to be able to share on top of that. The next stage of building

this out is to get it to go through and then cut out six shorts out of each video as well

and then to automate the distribution onto all the platforms, YouTube, Fountain, X, TikTok,

Instagram for the shorts as well. It's just the time that it takes to be able to set it up,

you want to put a weekend aside, I guess, to get like a workflow or not even a weekend,

a couple of hours on the Saturday to be able to put that into place. So does Lindy for you,

it sounds like it has access to the administrative functions of your YouTube channel.

At the moment, it's not posting on there. It's just given me that information and a document

I'll edit it and then I'll copy and paste it over. One thing with this and then getting better

and better with their security and it's coming out and it's almost industry grade,

level of security, but I'm still very specific to sandbox it. It's own email, it's not connected to

my email. When I've got podcasts and stuff, I will forward the invite to the OpenClaw email so

that it knows and they'll ping me to say, hey, Sean, you've got a video with Michael Jordan

at 930, your time that's coming up in five, 10 minutes. So I will do that. I guess one of the other

things when it comes to AI and I've reiterated a couple of times, it's not plug and play, but this is

the worst that it's ever going to be and for me, my OpenClaw has been up the whole time or with

very little downtime and I haven't had any issues. I haven't done the very latest update

as someone had done the very latest update on install and get OpenClaw to start working.

It hasn't gone as smoothly this time, so there's a little bit of trouble shooting and stuff to

be able to play around with it. So I guess my message is for everyone with AI and these tools is

have patience with it, learn to do it, but like put in the time, it's like Bitcoin, you're not

going to get it straight away and it's not going to create you a million dollar business straight

away, but AI will make you a million dollar business if you put the work and you use it, like

source trust me, bro, it happened. It's incredible, but it's not, I think the hardest thing, Michael,

is we see all these videos and stuff and I'll create the software from a couple of prompts

and it works just like that and the websites and the documents and everything that I need. It's

just happening, but that came from using AI every day since like 2022 and then the first week that I

got OpenClaw is I'd five hours sleep in three days, for example, and I don't encourage everyone for

that, but it's just, there's no get-rich-quick scheme in the world or whatever. Things aren't just

going to be handed to you and here's the magic box, you plug in the Mac, you're going to

make all this money, it's going to save you all this time or whatever. No, at the start, it's going

to cost you time, it's going to cost you energy, it's going to cost you your awareness late nights

and you have to put in the proof of work, but then once you do that, it compounds and then

it's exponential and then it's very easy just to be able to say, hey, I need a document related

to this. Now, you get the document back, please email it to me, cool, can you make an animation to be

able to go along with us? We want to put that out on LinkedIn or can you make LinkedIn carousel,

for example, and it can do all that, it can email and you just copy and paste and put it over,

but I guess the messages for people that already have a business where you've got systems

or you've got processes in place that just waste a lot of time. You can start to automate those

out instead of just having the human do it all the computer input, it's like AI, human check,

AI goes, human verified, accepted, great, and then you move on to the next task.

This is amazing. I can't wait to get started. I'm waiting for my Mac mini to come.

So one question for you, what is, I've heard, you've been the person who I've been following

on this journey who is most excited, optimistic, everything I've seen from you is like, guys,

this will change your life and the way that you work and all these things.

And then I've heard some people who are like, it's going to cost you thousands of dollars,

an API calls, it's like the cost is not going to be worth it, or maybe it would be worth it,

but I don't have five or $10,000 to spend. Is that just specific to what you want to do?

Obviously, the more processing intensive, I think the more expensive it's going to be,

but for the average person who's just doing the kind of tasks you're describing here,

it sounds to me like you're not spending thousands of dollars a month to operate this.

Is that fair? No, it's cheap as chips. And where the cost comes down to is

what like LLM are you using? And the best LLM to be able to use was anthropic, is Claude

and people were connecting their $200 a month subscriptions through OAuth

authentication. So just their subscription to it. Now, anthropic clan down on them is like

absolutely not you're not doing that. You need to use the API call. And the API calls can get

very expensive to be able to do it. So that's where people were spending a hell of a lot of money.

And because it's remembering everything and it's putting a lot of context into each

window with all its memories specific to you or that conversation, it's costing you a lot of

tokens. Obviously, OpenAI has paid for Peter. I don't know exactly what the deal was. Open

claw, so Peter's the founder of OpenClaw. OpenClaw has stayed. OpenSource and OpenAI hasn't

bought it, but they're very much supporting it. So where I'm getting is with OpenAI, which is

like the chat GPT ecosystem, you can use your subscription, like your pro subscription, which is

what like $49 a month or something like that. I don't know. I've got two pro subscriptions, but

we just connected one of those with it. And that's you know, for the coding, the building with

jam GPT codex, it uses that and it's not costing anymore. If you're to connect it to the API

instead of the subscription, yes, it would cost you a little bit more not as much as

anthropic for just the thinking and the communicating and for the areas where you're not worried about

your information going to China. Kimi K2.5 is like $30 or $29 or something like that. So those two

subscriptions was $80 Australian, like even if it's $100 US for what it can build like who cares?

Yeah, yeah, yeah, we have to do it. And then if you need more, sorry, just quickly to start

to the rest of it. If you need more capacity or whatever and you're like building out the

craziest things and you need an anthropic API, just get it to build something initially with

those cheaper subscriptions that gives you enough value or enough time back to be able to pay for

the higher API calls. Don't just like throw everything all that out at once, like build it and stack

it up slowly, like any endeavor wants to be cash flow positive as soon as possible. Otherwise,

you can't keep doing what you're doing. If you've got something that provides value,

there's no point like going broke before you even start.

Gosh, yeah, that makes sense. And I assume too, you can automate this to learn sort of the workflow

of, hey, build cheaply inexpensively quickly and then like be thoughtful. So to speak,

artificially thoughtful as to when we go to more like of a production level environment,

we're willing to pay more. It's going to kind of figure that out, build you like a rougher maybe

product and then you sort of you sort of work your way to the more expensive models. If you even

need to get there, I bet for most people most of the time they with these subscriptions, you know,

that they pay a hundred bucks a month for more subscriptions at total and it'll do what they

need it to do. Is that is that probably? Absolutely. Like even just open AI, like just that one

subscription for 99% of people or 95% of people would be fine. Or Kimmy, if you don't worry about

information or you're selective with when you're interacting with it agent going to Kimmy.

If you're not worried about it, like send it there. But I think even more to your point,

Apple's just announced their M5 MacBook and stuff. They're going to come out with the M5

Mac Mini, you know, later this year. The capacity of the LLM locally that it'll be able to run

will be absolutely insane. But two weeks ago, I had another local LLM, the name will come to me

in a sec. And it was like, it was like when I started using chat GPT in 2022, like it was there,

but it was like so slow with it. And you're like, this is just absolutely useless. But then two days

ago when Q3.5 came out, backing up a little bit. So two weeks ago, when that first AI local I installed

and it's like, oh, this is cool, like it works, but it's super slow and it's going to drive me crazy.

But I just, I'm saying to people, just what we need to do is wait four to six weeks and there'll be

something like better and we'll have a new step change. Two weeks later, we're going to like GPT

4.0 levels straight away locally, like let alone the next four to six weeks. What's going to happen

in the next two weeks? You know, like every single day these models are getting smaller, more powerful.

And then as the hardware comes out and Apple gets that M5 chip onto the market, it's going to be

insane. I think, but in the next, I don't even think it's going to be 12 months, Michael, but

most people will be able to get away with running everything locally, maybe on even on like a Mac

mini level, if it's not a Mac mini level, go out by an Apple studio, like fully, fully

spec'd out at the moment, like everything, the biggest hard drives, everything in Australia,

it's like $27,000. So that's maybe $18,000, $20,000 US people, like I'm not spending $20,000

on a computer. If that computer can replace just one staff member, but what happens if it replaces

four or five staff members or you've got say five staff members and they're using this local AI,

that costs you $20,000, but now they're five times more productive, ten times more productive,

like the cost of that is like it's not existent to a business owner or it should be, and it should

be something that they should budget for and start to be able to plan for, and I guess the point

with the open claw and instance on the Mac mini now is yes, it works, it's a little bit finicky

and you need a tinker with it, but you want to learn all these things and you want to learn just how

to use it, so that when we get to the level where you can run everything locally as powerful as

like the best models that we have today, you'll want to be prepared for that because they're also

not going to be plug-and-play and you want them to be specific to your situation, it's pretty

exciting. So you're saying you would rather be in a place where you've been training Lindy on

a Mac mini and as the technology advances, it will be something that you can, that's sufficiently

compact that you can put it onto like a Mac studio, but you're basically taking all of its learning

that's done with you onto a more powerful machine in the future with better LLMs that can run

more localized, and you're not starting from scratch when these new machines come out, like you

have years of data that you've been feeding it, and you're able to take advantage of the new

technology much more quickly. That's right, so to relate this not to AI, but say with my

admin assistant, originally when she first came on, I'd show her a task on the computer and we'd

get on Zoom and we'd record it, so I'd show her the task exactly how to do it, she'd watch the video,

she'd download the transcript, and then she'd write an SOP, a standard operating procedure on how

to do every single task, she would send me that SOP, I'd check it, we'd get on another call,

and then I'd read it out to her, and then she'd follow it step by step by step,

gets the end, if the result was great, great, that's the SOP, that's the workflow, that's the

structure that we're going to follow, do this every single time, if I was reading it out one by one,

she just followed and she didn't get the answer, then all we needed to do was change that SOP,

for each SOP you're probably putting in like four to six hours for that, and that's like a lot of

time where maybe that task originally would take half an hour, 40 minutes, an hour or whatever,

so it's taken 600% longer originally, but you just do it once, you do it properly, then that

standard operating procedure is in place, the same with the agents now is when you're building out

workflows, SOPs, the frameworks, they take time, they take effort and stuff, so do you want to

wait until AI is where it can run and do absolutely everything, and you've got to wait until that

moment where it's super fricking capable, it's super capable now, but you're starting from scratch,

so you've got like this amazing tool, it's like a BMW and you've got the learning key or whatever,

so it's given you 10 horsepower instead of a thousand horsepower, or do you want to start

10 cream hour learning the systems, learning how to build the skills, the workflows, the frameworks,

SOPs, and then when it's ready, either copy and paste them over or your agents can do all that for

you, but everything's already automated and you know how that automation has come together,

so on the more powerful models it can then take it up to the next step, and then you can

be up here instead of starting from scratch in the future, I just don't think,

anywhere not paying attention, they're just going to get left behind, like they don't have

have time, unfortunately not to play around and pay for it because the people paying attention now

compared to the people waiting for the next local Wambam with local model or whatever,

they're just going to be so far behind and they're never ever going to catch up, it's exponential,

so it's a bit like Bitcoin, you know, you sort of get AI at the price you deserve,

100%, yeah, so what do you think, and I want to get to Bitcoin and AI, and there's been a lot

of chatter about this, but I want to hear your thoughts on where you think we're headed, so I,

my echo chamber sort of, I think aligns with you, and like Tony is so excited, I mean I know

you know and love Tony, he's so excited about when this AI is transplanted into human-like droids

that can do anything and everything for you, like he thinks that is like, to me that's like dystopian,

but I understand that's probably where we're headed, for him he's just pumped, so have a night,

do you think, but then oh, but then I stepped outside of my echo chamber recently a couple weeks ago,

was talking to a friend of mine, and he's, I wouldn't say he's a normie, he's a, he's a way

can do a lot of the things we talk about, he's not a Bitcoiner though, but he thinks that the AI

thing is way overblown, it's not like yes, it'll, it'll automate some things, there'll be some,

you know, some leaps forward with AI, but at the end of the day you need consumers who have jobs

in order to pay for the things that AI can provide, and so it's going to eat itself

and die if there isn't something for humans to do. I would love to hear your thesis on where you

think the economy, the AI economy is taking us. Yep, I'll dive into that, but imagine if

he's wrong, and he takes that opportunity, I'm just not going to play, I'm not going to learn AI,

and I'm not going to use it, if he's wrong in his thesis, or it's like when people say Bitcoin's

never going to work, it's a failed experiment or whatever, if they're wrong and Bitcoin is

like X-Predora money and AI is here to stay, and you totally ignore them, then you shoot yourself

in the foot, whereas the worst case situation, maybe you still believe that they're going to fail,

and maybe they do fail, but what's the harm in learning them? You've learned a new skill,

you've paid attention, like there's no downside to it, and it's like I believe it's an esometric

bet. That's right, even if you're wrong, your time wasted is near as much as the opportunity cost

of not doing it if the AI Maxis and the Bitcoin Maxis are right. Yeah, that's right,

but as for, is it overblown? Not my own life, like I went from a high school dropout to

building a seven-figure online company within a couple of months with just talking to AI for

4, 6, 8, 10 hours a day, like it was absolutely relentless, and it was like the few years leading

up to that that allowed me to condense that into it. So, is it overblown? Like, fuck no, but

fuck no, it's not. How people like, oh, this thing doesn't make any money, like you've got no idea,

like ideas flowing or percolating or you've got nothing going on. If you can't say five minutes

of your time with, like even when you're saying, when you're trying to get just the YouTube

description and everything like that, just the first draft, and to get it formatted, so it looks

good within your YouTube description. They can save you five minutes, you can just chuck it in and

then you just edit it super quickly, but it's all in your time, your brand, your voice, it sounds

exactly like you, just make sure that everything's factual because it can still hallucinate.

Doing that, but could it all just go away and could it eat itself? Sure, it's possible, but

humanity seems to always improve on itself and the natural state of the market and the words

of the author of the price of tomorrow, Jeff Booth is deflation and we're going to get better

and better and better and because we can run local models, even if, like I just don't see

how they could possibly go away anymore. We've been given this magic box, a Genie, three magic

wishes that we can ask over and over and over and over and over and over again, and we get those

delivered to us every single time. Like who in their eye mind is going to be like, no, AI is like

going away. Like yes, it's going to displace a lot of people, 40% 4,000 people of square, gone.

Like what are they going to do? Hopefully they can leverage AI and they can use it because they've

got the skills, they're already peak of their field, so they can do something better than I can

with doing it, but that playing field stopped. I hope so, but then in regards to

what happens, that's when like universal basic income comes in and people are eating the bug soup

and it'll be the haves and the have nots. The people with using Bitcoin and AI together and

the AI agents are going to use Bitcoin, they're going to have whatever they want. It is like

toning Yazbeck always says, it's like a lithium. There will be people in this AI and credible

paradise and there'll be people that are just living in hell and I was saying this to Tony on

the podcast just before. I go down to the beach every afternoon and in the June, there's a couple of

people sleeping rough and stuff, homeless. Like that is so sad and it feels sometimes like

a lithium. I'll come out to this beautiful house with ocean views and everything and we have

nice food and like clean clothes and everything like that, nice lawn and just like a lovely environment,

friends, family, everything like that. If anything happens to us, a medical appointment and an

emergency, if our dog gets an allergy and she needs three and a half thousand dollar allergy

medication, true story, we can get it absolutely no problem. We don't even question it. Whereas

some people live in under a fucking canvas sheet on the beach and what's that from? Like yes,

they've probably made some bad choices. I've lived rough from myself who've lived out of a car

for a while, out of a tent for a long time. So I know what it's like and it's definitely the

choices that I was making but it's also because the money and everything's broken and it's getting

worse and worse and worse and we've got like three or four hundred thousand Australians

now homeless because the government's just obliterating the money. It's absolutely worthless.

And so the point is if you're not saving an hard money that can't be

effed with, that can't be manipulated in any way, obviously price short terms a different story

and you're not paying attention to AI technology that can really leverage every area of your life.

Then you get into those types of situations where you can prevent it. But yeah, I don't think

like in regards to your friend. It'll just split like it. It'll be like that for for some people

about the people with Bitcoin hard assets. It's hard in a digital age to saying gold will still

contain like beer's useful as Bitcoin. It's worked well for five thousand years like an

all-local economy and stuff. The gold bugs are still going to be all right as well. I believe,

but unless you're holding gold or Bitcoin or gold, I think people are going to be on

Struggle Street and then as time progresses and the AI agents more and more take up Bitcoin

as their medium exchange, then that's going to eat into that gold market as well.

And I was thinking about this last night. There is no way long-term people like look out a little

bit everyone wants hyper Bitcoinization today. Just make it in your own life today. It's not

coming tomorrow, but it will come. But over time, the AI agents are only going to accept Bitcoin.

It's backed up by proof of where it's limited. It's scarce. It can't be manipulated. It can't be

confiscated. There's nobody going to change it in any way and they can take a hundred percent

ownership of it. Whereas what like USDC, it just gets absolutely taken or the government prints a

whole lot of money and it's worth less or something like Ethereum where they change the rules

300 times over the last couple of years or whatever know, like it's going to be Bitcoin.

So you probably want to get some Bitcoin and then learn to use your AI agent and then train it

to be able to accept lightning or use lightning on the 402 and exchange things. But it's all,

it's already happening. AI agents are paying people in Bitcoin to be able to do work.

AI agents are hiring humans to go out and do things. They're paying with credit cards,

they're paying in stablecoins, they're paying in Bitcoin. I don't know. Go out and try build one.

Even if it takes you a year or six months or whatever, it's not going to happen overnight.

Start building one and then give it capacity to connect to the lightning network and then go

out and do things. But still, like sandbox the hell out of it. Slightly don't give it access to

your personal information. Don't connect it to your personal email forward at things to what it

owns and what it controls. But don't give it like your personal stuff, not yet.

Yeah. Yeah. Man. Okay. So Bitcoin in AI, we've just got a couple of minutes here.

Your final thoughts on how Bitcoin gets integrated. Because I know a lot of people are saying,

we talked about this on our livestream earlier today that, you know, if you do like these,

you can sort of model this out. And it looks like the LLMs want Bitcoin broadly. Like almost

half of them prefer that over stablecoins or obviously Fiat. Let's very slow.

Let's ask chat GPT real quick. Is this an experiment you've done?

No, never. Hey, I'm just on a podcast talking with Michael Jordan. I was just wondering

in regards to AI, what type of currency in the future they are going to want to accept?

Would it be Fiat currency like the US dollar, something like Bitcoin? Would it be something like

XRP or Ethereum? What would be the AI agent's best pick?

Well, I can't speak for AI agents as if they have preferences like people,

but if we're talking about how they might handle transactions in the future,

it would probably depend on the systems we humans set up. AI would likely work with whatever currency

is most efficient. Oh, that's a fucking dance answer. That is chat GPT for years.

Correct. Yeah. But I saw a study from the Bitcoin policy institute that said they gave

these different conversations, like 9,000 conversations across a number of LLMs, I think.

You know, it was like almost half of them said Bitcoin. And you've got to think that corrupt

Kimmy, but I also Bitcoin. Yeah. Anthropic. That's the thing is I actually use chat GPT

as my LLMs test. I'll ask it. If I ask it a an unpopular historical question,

I want to see what it says first because it's going to have the most mainstream

woke bullshit sort of answer. And if it gives me feedback that seems

you know, out of place for it to provide, I tend to assume it's true. I don't assume it's true.

Obviously, if it's feed me the same thing that I learned in elementary school, but it's a good

like start. It's a good diving off point. Okay. Hey, man, I want to be respectful of your time.

I know you're on a tight schedule today. My last question for you. Same one I ask everyone,

what is an unpopular opinion that you hold? You get bonus points if you offend some Bitcoiners with it.

I don't know if it's unpopular, but the people leveraging Bitcoin in any way are

very misguided and I believe they're trying like it's that poor gambling get rich quick

mentality and they don't understand Bitcoin at all. They're not here for the revolution. They

haven't put in enough proof of work. So what kind of people are you are you describing here?

Like what kind of Bitcoin? What kind of Bitcoiners?

Is it like? Well, they're not Bitcoiners. They're like, they're shitcoining on Bitcoin.

Yeah. Yeah. Yeah. Yeah. Like, they're seed coiners, but anyone that wants to get the world's

hardest, guess it's the best form of collateral and stuff and for a company to land you

fear dollars and exchange for your Bitcoin, best business model in the world. Like, it truly is.

They've got absolutely no risk at all, but just recently with the last pull down where it came to

61,000 dollars, some people that I thought were incredible Bitcoiners that doing amazing things in

the space, getting close to getting liquidated, like losing nearly like 90% of their Bitcoin stack,

they haven't put in the proof of work. Fucking misguided. So how to describe them?

Lazy misguided on the nice side. I won't go anywhere worse than that.

Okay. No, I think that's a good one. I don't know if they'll offend our audience.

Well, most people following you guys get it. You don't trade Bitcoin. You trade everything

gals for Bitcoin. You don't take leverage on something as volatile as Bitcoin in the short term.

You just do everything you can to get as much of it and to self-custody. That's it. It's maybe

boring, but it shouldn't be exciting. It's just superior savings technology and then go out and

provide value. Use AI to be able to create value and to be able to get more Bitcoin and just repeat.

Beautifully serve my friends. Sean, why don't you give people a handoff? Tell them where they

can find you and anything else you want to point them to. Yeah, cool. I'll give you a link,

but if anyone's interested in learning about AI and Bitcoin over the next year, if you'd like to

book a call and learn more, you can give Michael the link. But get a Bitcoin over on YouTube.

It's also my ex account is Sean Clark 911. Incredible podcast and incredible guests and amazing

conversation I'd recently with Michael. That's out there for you to check. Just come over and say

get a and the final message is what I think is the most important over the next 12 to 36 months to

be able to focus on is get an amazing habit stack. Build a personal brand, use AI and save and stack

and become sovereign with Bitcoin. If you can sort out those four things, you're going to be happy

days and we're going to live in a leasing them. If not, it's going to be a hard time for many people

and if I'm wrong and you still learn those things, there's no bad side. It's all upside anyway.

I think that's a great place to close. Sean, thanks for the time man. Enjoy the chat and

look forward to standing touch. Always a pleasure. Thank you. Cheers brother.

And that's a wrap again. I hope you enjoyed my conversation with Sean. This guy is doing amazing

things. Go check out his page on axe. Go follow him. Go check out everything else he's offering.

This guy is a world of knowledge and insight. I think everyone should be paying attention to

what he is talking about and follow along on his journey and start taking steps yourself.

And of course, if you need help taking your Bitcoin into proper 100% self-custody,

go to the bitcoinway.com slash podcast. You can schedule a three 30 minute

console with a member of our team. We can talk you through proper self-custody,

protecting your online privacy, setting up a privacy phone. We have a plan B residency

option that is getting extremely popular in Panama. Again, it's the bitcoinway.com slash podcast.

It costs you nothing just 30 minutes of your time to learn more. If you are enjoying the channel

the show, you like this episode. Do me a huge favor. Give us a like, subscribe, comment, share it

with someone who you think might be interested. Anything you can do to help feed the algorithm and

get the word out about Bitcoin, self-custody, self-solverty, everything we talk about here every day.

And until next time, stay safe, stay sovereign, and remember the yield on Bitcoin's freedom.