McIntosh:00:00:00 Welcome back to Satoshi’s Plebs for episode two forty five. I’m Macintosh,

00:00:06 and unfortunately,

00:00:07 Kenshin is not with us today. He had to take day off. We

00:00:12 do get days off around here occasionally,

00:00:15 and he needed that for what he’s got going on. I’ll let him talk about that next week if he wishes.

00:00:22 We respect privacy here

00:00:25 at the utmost at Satoshi’s Plebs. But

00:00:29 today, we’re talking about Bitcoin,

00:00:32 big surprise,

00:00:33 AI,

00:00:34 and how they’re going to integrate.

00:00:45 Yeah. So, really, please don’t let AI being in the title throw you off. This is not gonna be some in-depth discussion about AI,

00:00:54 even though I’m up to my eyeballs in it. This is going to be about how Bitcoin

00:00:59 is going to be used

00:01:01 in the future by people who are interacting with AI and actually by AI

00:01:07 programs,

00:01:09 agents as we call them talking to themselves or to one another, not themselves.

00:01:15 A real brief

00:01:17 rundown kind of what I’ve been doing this week. I have started working on a new project with AI.

00:01:24 It probably

00:01:26 another week I’ll be ready to talk about it. I do not wanna talk about it this week other than to say it has been an amazing experience.

00:01:33 I have been in computers for thirty years,

00:01:36 guys.

00:01:37 A long, long time. I started a dial up Internet provider back in 1996,

00:01:44 I think,

00:01:45 to give you some idea. That was

00:01:48 actually not even my first job. I had a job before that for, like, two and a half years

00:01:54 as a really as an Oracle DBA.

00:01:57 So it’s been a long time. I’ve been through a lot of different things.

00:02:01 What I am doing right now, honestly,

00:02:05 is some of the most excite well, I would actually argue probably

00:02:11 the most exciting thing I’ve ever done in my life.

00:02:15 The tooling is there

00:02:17 for people who understand what they’re doing to develop rapidly,

00:02:21 very complex software.

00:02:23 In the last two days,

00:02:25 well, three days really at this point. In the last three days, I have developed,

00:02:30 I

00:02:33 think right at 50,000

00:02:35 no. Well,

00:02:36 I can check. Hold on.

00:02:39 It won’t even take long. Where is that?

00:02:45 So just over the last three days, I have developed almost 40 no. 40,000 lines of code.

00:02:5340,000

00:02:54 lines of code. It’s a staggering amount of work. Obviously,

00:02:57 I could not have done that on my own. I was using ChatGPT.

00:03:01 In fact, I was using it so heavily I ran out of tokens for the week,

00:03:06 which made it for a sad Macintosh this morning when that happened. I will probably be upgrading as Kenshin did

00:03:14 for a while a few months ago to the monthly plan. I’m not sure.

00:03:19 I want to get this moving along. I am really close. I actually have a product that

00:03:25 maybe it’s not an MVP, but it’s it’s really, really close. The test,

00:03:30 it’s up running,

00:03:31 not doing a whole lot, but the

00:03:35 scaffolding is there. The bare bones system is there, and now we can start filling in the pieces and really fleshing it out.

00:03:43 I’m super excited about this as you can tell,

00:03:46 I’m sure.

00:03:48 But anyways, that’s kinda been my week. I got on that like three days ago and have just been blasting on it nonstop.

00:03:56 Other than that, just the normal stuff. But all of this AI stuff that I have been involved in,

00:04:05 it’s

00:04:06 it’s caused it’s

00:04:07 it’s caused me to ponder a lot.

00:04:11 And, you know, I watch all these YouTube

00:04:14 AI,

00:04:16 some of them they call them podcasts or whatever. Maybe they are maybe they have a real podcast. Some of them are just YouTube channels, whatever. It doesn’t matter.

00:04:26 And get all these different opinions and whatever, but I don’t hear a whole lot of people talking

00:04:32 about

00:04:33 how

00:04:35 AI is gonna work in a financial sense.

00:04:38 We talk about all these amazing things and

00:04:41 this, that, and the other, but

00:04:44 you gotta understand, with AI, one of so you go to ChatGPT.

00:04:49 You,

00:04:50 query ChatGPT.

00:04:52 Just go to ChatGPT.

00:04:53 You can do this for free if you’ve never done this.

00:04:56 Chatgpt.com.

00:04:57 You can run a query

00:04:59 and have it give you results. It’s kinda like Google. It’s using what’s called a large language model, and you can ask it crazy things. Like, I can program in that chat gpt query window. It just becomes a lot more difficult than the way that I currently do it, which I’ll explain very quickly.

00:05:16 One of the improvements that they made a while ago now

00:05:20 is they started building these little

00:05:23 tools

00:05:24 that integrate,

00:05:26 that understand your code base. So literally in my repo on my desktop, I run a,

00:05:33 program

00:05:35 at Node. Js, I think. But anyways, it’s called Codex. It runs right in my repo.

00:05:42 And so it has an understanding.

00:05:44 A repo is a repository, code repository.

00:05:47 It has an understanding of my code base

00:05:50 and can make modifications

00:05:53 and changes and

00:05:54 updates and add new files and delete files and run commands and all kinds of things that it cannot

00:06:01 in the ChatGPT web interface.

00:06:05 So that was one huge step, and that’s what allowed me to actually build this project that I’m working on.

00:06:11 A later innovation was what they called agents.

00:06:15 I believe they came slightly after this. And an agent is basically a more autonomous

00:06:21 piece of code

00:06:22 that operates using

00:06:24 the large language model. So it’s out there querying ChatGPT

00:06:28 essentially,

00:06:29 saying, hey, what about this? What about that? What if how do I accomplish this? How do I do that? All this kind of stuff.

00:06:37 But it makes it much more flexible. It makes it able to interact with other programs, other services,

00:06:43 other websites,

00:06:45 other

00:06:48 I don’t know. It’s almost limitless as to what these can do.

00:06:52 And of course, you’ve probably heard, unless you’re under a rock, if you have any involvement with computers at all, about OpenClaw lately. This was one now, this was not the first agent, but

00:07:04 it has caught on for whatever reason.

00:07:10 Not really gonna be a new segment this week, but I will throw out that the guy who,

00:07:14 Peter,

00:07:15 the guy who started Open Claw, this

00:07:18 this agent

00:07:21 program,

00:07:22 which has proven to be very popular, he started building it three months ago. Eighty two days. Well, probably like eighty four days as of now. Eighty three, eighty four, something like that.

00:07:33 Less than three months ago,

00:07:36 wrote his first line of code.

00:07:40 Last weekend, actually, I think it was on Valentine’s Day or maybe it was Sunday,

00:07:46 he,

00:07:47 he made an agreement

00:07:49 with Open,

00:07:51 AI,

00:07:52 Sam Altman,

00:07:54 to

00:07:55 become an employee.

00:07:57 And the amount of money that he’s getting paid is just staggering.

00:08:02 His OpenCLaw project will remain open source, so it’s available to anyone. By the way, this was the number one

00:08:10 fastest growing project on GitHub by far.

00:08:14 It amassed something like a 140,000

00:08:17 stars in in less than that three months.

00:08:20 It’s crazy.

00:08:24 But that will actually be moved to a project OpenAI will not be able to directly influence. I will say this. They probably will indirectly

00:08:33 influence it, and that’s

00:08:35 I guess that’s just kinda part of the deal, but whatever.

00:08:39 Peter will work for OpenAI,

00:08:41 and to be honest, the numbers that I have heard is he will be getting, in total, roughly $1,000,000,000,

00:08:46 about half of it,

00:08:49 kind of on signing and and

00:08:52 compensation

00:08:53 and direct,

00:08:54 like, stock options that are immediately available, and then probably the rest will be vested stock over the next couple years or whatever.

00:09:03 Incredible, incredible deal. But,

00:09:06 anyways, that’s some of the news for the week.

00:09:09 And it’s relevant because of these agents. So these agents are talking to each other. They’re doing tasks. They’re reading email. They’re making email for you. They’re

00:09:20 really just doing all kinds of things. They can do all kinds of tasks that don’t require physical

00:09:27 hands,

00:09:28 essentially.

00:09:30 And in fact,

00:09:32 if you really want, like, there’s a service called, like, humans for AI or something where the AI can actually hire out people.

00:09:40 Oh, AI

00:09:42 chatbot.

00:09:42 It’s not a chatbot. Sorry. I shouldn’t call it that, but AI agent,

00:09:47 which people typically name, but whatever. Fred,

00:09:50 I want you to order me some DoorDash. Well,

00:09:53 Fred can’t necessarily

00:09:55 do that, but there are people who can. So they go to a website and hire people to do that.

00:10:02 I think one of the things you’re gonna see as we move through this is a standardization

00:10:07 of protocols,

00:10:08 hopefully.

00:10:09 Ways of interacting.

00:10:11 And I would propose

00:10:13 that

00:10:14 Satoshi’s

00:10:15 over lightning is probably, one of the best ways for agents to manage

00:10:21 payments.

00:10:22 A lightning channel,

00:10:24 a lightning setup

00:10:26 in terms of the agent itself,

00:10:29 is very simple. Now, just got done telling you guys

00:10:32 lightning itself can be quite difficult to run a server.

00:10:36 But to have a wallet and use a wallet is super easy.

00:10:40 Another way of doing this would be through e cash, like Cali’s,

00:10:45 I’m not sure what it’s called, but his Mint software.

00:10:49 It’s

00:10:50 using what they call Chamien e cash in order to

00:10:54 manage

00:10:55 money.

00:10:56 But that’s actually backed by Bitcoin in the end. So Bitcoin wins there. Bitcoin wins with Lightning.

00:11:04 I think outside of that, there’s going to be more friction and more problems than people will want.

00:11:11 And so people will naturally migrate to this. And plus the level of a Satoshi,

00:11:17 even a decade from now when Bitcoin is a million dollars a coin is still going to be trivially

00:11:23 small

00:11:24 so that it’s not like you’re gonna be sending one Satoshi and it’s worth $20.

00:11:31 So it’s an ideal

00:11:34 situation

00:11:37 and the tooling for it is probably being built now. I hope we can come up with some standards for this again.

00:11:44 I mean, the Lightning

00:11:45 Network is a standard,

00:11:47 but there will be more

00:11:49 standards that are needed along the way. The skills that the agents use, for example, these are abilities that you can give them. OpenClaw has one way of doing it. I’m sure there’s other agents who do it different ways. There should be a standard for that. There should be a standard for security,

00:12:07 in fact. I do not believe you will see widespread enterprise adoption

00:12:12 of these agents even though they can save companies, I believe, millions of dollars over the long run

00:12:18 if

00:12:19 they’re not secure.

00:12:21 Because a company

00:12:22 will be gone if

00:12:25 an agent runs amok and does bad things and releases their

00:12:30 private information or private keys or whatever.

00:12:34 And OpenCloud,

00:12:36 I mean,

00:12:37 Peter has admitted I haven’t even read all the lines of code.

00:12:42 He

00:12:43 has no idea, in some cases, really what’s in his program. And in fact,

00:12:48 the skills

00:12:51 repo,

00:12:52 if you wanna call it that, there’s

00:12:54 a huge number of them. And many of them are security risks. They vetted them,

00:13:00 and I don’t remember the percent exactly.

00:13:03 But it was something like I wanna say it was like 20% or so. Maybe it it may have been higher. I don’t think it was any lower than that.

00:13:13 But it was a lot. That’s not acceptable for an enterprise.

00:13:17 It shouldn’t be acceptable for an end user, but it’s certainly not acceptable

00:13:21 for an enterprise. But these agents are proving to be so useful.

00:13:26 This stuff will be solved.

00:13:28 And here’s the thing,

00:13:30 Bitcoin will be there when people need it. So when people say, where’s Bitcoin going? I’m gonna tell you, I don’t know.

00:13:37 But I will say this, I do know that one of the big drivers of Bitcoin in the future will be its use by AI. And this stuff is happening so quickly, we may find that it’s happening

00:13:49 this year. It may be six months from now. What if you have millions of bots

00:13:54 out there using Lightning over,

00:13:57 using Bitcoin

00:13:59 over the lightning network to transact. Oh,

00:14:03 I, need this and I can’t get it because I don’t have the compute resources, but I’m willing to pay you a 100 Satoshis to do so. Boom. Done. Here you go. Here’s your answer, your product, your whatever,

00:14:16 and here’s your 100 Satoshis.

00:14:19 Okay, great.

00:14:20 What if my AI could

00:14:23 make a living essentially?

00:14:25 Pay for itself?

00:14:26 Because AI

00:14:28 is expensive,

00:14:30 honestly.

00:14:31 What if my agent could pay for itself? What if it were out there doing a job?

00:14:36 What if besides being useful for the things that I need, which make it very valuable, what if it were out there providing its own

00:14:43 work? There’s gonna be a lot of questions over the next few years, gang, in terms of AI, and I don’t know where you are at

00:14:50 on this,

00:14:52 but I will say this, you better be ready.

00:14:55 You better be flexible

00:14:57 because it is coming and it will change your life. Maybe not dramatically.

00:15:02 Maybe

00:15:03 I think it ultimately will change everyone’s lives, but

00:15:08 maybe for now it doesn’t. But I will say this, in my industry, in the computer industry,

00:15:15 things are changing so rapidly that I see colleagues who are just they’re getting left behind.

00:15:20 They don’t even understand how to log in to chat GPT and do something simple, much less talk about

00:15:26 a, you know, a specialized,

00:15:30 LLM that’s augmented with rag or agents or any of this stuff. They probably never even thought about what does it truly mean for an agent

00:15:40 to to run a business.

00:15:44 So I don’t wanna go on this. This is gonna be a short episode today.

00:15:47 My my sidekick,

00:15:49 my

00:15:50 my pal is not here.

00:15:53 So

00:15:54 we almost didn’t do an episode this week, but I am so excited about this. I just wanted to take this opportunity to kinda

00:16:01 not vent, but kind of let you guys know.

00:16:05 I don’t want you saying, well well, Macintosh, you didn’t say anything.

00:16:10 If you’re in the software development business, if you’re in Bitcoin, this is an intersection that you can look at. How is Bitcoin going to work

00:16:19 in software in the future, in AI specifically in the future? And I think it’s about to become a big thing.

00:16:28 You can just drop a simple library into your code and have the agent use that to build a Bitcoin wallet, whether that’s on mainnet or Lightning.

00:16:38 It’s

00:16:39 trivial to teach them to understand about Bitcoin.

00:16:43 And since it’s code, point them at the Bitcoin repo. Have them understand it. Talk to them about it. Teach them.

00:16:50 Bitcoin is in a very finite, you know, resource. You should treasure it, essentially,

00:16:58 as you do. Don’t have it out there wasting it when it comes time for that.

00:17:03 When they’re trading Satoshis,

00:17:05 it should be for things of value,

00:17:07 not just

00:17:08 because it’s fun.

00:17:11 Alright.

00:17:13 I’m gonna leave that at that. I hope you enjoyed that little

00:17:18 AI introduction.

00:17:19 I don’t know. It’s not really an introduction.

00:17:22 But thoughts about how bitcoin is gonna be intersecting with AI.

00:17:26 I believe it’s gonna be huge. We’re not hearing a whole lot about it now, but you’re starting to hear the rumblings that people are talking about how are these AI going to work together. They already do it, but it’s for free. And eventually, that won’t work. It’s one thing for me to run OpenCloud on my Mac Mini. It’s a whole different thing for me to run agentic agents

00:17:49 on my servers at work that I’m paying tens of thousands of dollars for, and then have those just give away things for free to

00:17:59 people outside the company.

00:18:01 So

00:18:02 these these things will

00:18:05 be monetized.

00:18:06 Be prepared for it, and it will only benefit you as a holder of Bitcoin.

00:18:10 Alright. Question of the week. Wow. This is gonna be one of the quickest episodes ever.

00:18:15 I hope you enjoy this.

00:18:17 Have you ever heard of

00:18:19 AI agents? I think we’ve mentioned them on here, but, you know, do you have any knowledge of them?

00:18:25 And second, are you okay

00:18:28 with

00:18:29 an agent being your Jarvis,

00:18:31 being your

00:18:33 minion

00:18:36 for you,

00:18:37 doing tasks, doing

00:18:39 work for you,

00:18:41 work that

00:18:43 takes,

00:18:45 I don’t wanna call it drudgery, but drudgery away from you,

00:18:49 makes you more productive.

00:18:51 And do you think that agents will be using Bitcoin

00:18:55 down the road?

00:18:58 Awesome.

00:18:59 Okay.

00:19:00 Where are we at?

00:19:02 Network

00:19:03 statistics. 02/18/2026

00:19:05 as we record.

00:19:07 I’m on a Wednesday, not a Tuesday as we normally record. We are at block height

00:19:12 nine thou 937,273.

00:19:18 We are at a Bitcoin price of $66,484

00:19:22 and €56,392.

00:19:28 We’re still down. That’s okay.

00:19:31 Price last February 18 was

00:19:3495,539,

00:19:37 so we are down 30,000

00:19:4030%.

00:19:41 So we are down 30%

00:19:43 from that, probably about 40% from all time highs, maybe even a little bit more.

00:19:48 The Bitcoin market cap right now is $1,330,000,000,000.

00:19:52 Gold market cap, 34,810,000,000,000.00.

00:19:56 So we’re at 3.82%

00:19:58 of the gold market cap. Gold is still hanging in there. Bitcoin,

00:20:03 at this point, is in its its bear market mode, essentially.

00:20:09 That’s okay. StackSats.

00:20:11 Bitcoin versus oh, I already said that. Satoshis per dollar, 1,500.

00:20:16 Well, this thing did not do this.

00:20:19 I

00:20:20 okay. Well, that’s my first little error I’ve seen.

00:20:23 My new show notes tool, which I built,

00:20:26 I ran it again this week and it worked. A lot of my

00:20:30 updates got in here. I don’t see the one for ounces. We had that.

00:20:35 Okay. Sorry. I don’t have the number of ounces of gold per bitcoin.

00:20:40 The all time high price was $126,173,

00:20:45 and that was hit on October 6. Been nineteen weeks and two days.

00:20:52 Alright. We got 53 and a point 68%

00:20:56 to the next halving, which is set to be on 04/11/2028.

00:21:00 Two years, fifty two days.

00:21:02 The mempool is humming right along, one sap per v bite and 33,509

00:21:09 unprocessed

00:21:10 transactions.

00:21:11 Unfortunately,

00:21:13 ayahya, our

00:21:14 next mining adjustment. I’m gonna have to have this thing show me the date. I do not know when that’s happening. It’s not too long off, don’t think.

00:21:2214.62%

00:21:24 up. So that completely wipes out the 11%

00:21:28 that it went down last time.

00:21:32 Okay.

00:21:35 Yeah. That’s about it. I’m gonna I really am. I wanna keep this one nice and short.

00:21:41 You guys, we’re a value for value podcast supporting podcasting two point o. We strive to bring you honest Bitcoin content every week. All we ask is, are you getting value from the show? Support it through time, talent, or treasure. Help with future projects.

00:21:55 Stream sets, boost with messages, even a 100 sets saying you a great show

00:22:01 or you suck. Let’s hope this one was more on the great show. I hope you enjoyed it. It’s a little different. But we’re gonna read it either way.

00:22:08 We did not have any boosts this week, unfortunately.

00:22:12 If you like the content, I would love it if you write a review on Apple. Tell your friends about the podcast. That is the best way

00:22:19 for us to grow. This will be the last week for Abel James, Voodoo Queen, for our music. Any boost or streaming of sets of that song will go straight

00:22:28 to the artist. Thanks for joining us. You’ll hear you will hear from us again next Thursday morning.

00:22:35 We hope this has been helpful, we would love to hear from you on Noster

00:22:39 or on Fountain. You can find all the ways to contact us at satoshis dash plebs dot com. Stay humble. DCA those sats. Have a great weekend.

00:22:48 We’ll talk to you soon.