Closed AI Is the New Censorship | The Confab 27: Marks

Bitcoin is close to becoming worthless.

Now what's the Bitcoin? Bitcoin's like rat poison. Yeah. Oh. The greatest scam in history. Let's get it. Bitcoin will go to fucking zero.

Welcome back to another Confab episode. This episode is with Marks from Maple AI. Marks joined me q and a and Seth for a Freedom Tech Friday live, and he has some really good points around AI and building his company. So I thought it'd be good to get him on for a confab, and he did not disappoint.

I think you're gonna really enjoy this one. Before we dive in, I just wanna say a quick thank you to everyone who's been supporting the show, retweeting, posting, sharing with friends and family, all the boosts, messages, XMR chats, and everything else. It really is appreciated, and you're keeping us going.

Here's the boosts from the last confab. Top booster was John with 21,420 sats. He says the unsung heroes of podcasting, like sir Chad Farrow and Barry from Podhome, do so much work behind the scenes, and I'm truly grateful for it.

It's good to keep in mind that from pressing that start and stop on a recording are only the beginning. There are teams of people working behind the scenes to take that recording and not only get it to the listeners, but enable them to be part of the show.

Thank you so very much, Jordan, from the bottom of my heart for bringing Ungovernable Misfits headlong into podcasting two point o. Yeah. Echo that, mate. Thank you so much for your help.

Welder Ian with 4,444 sats, says v for v. The two angry cunts podcast sent 3,333 sats and said, enjoy this conversation and hearing the wise, well done fellas. A v burrow sent a salute and 2,424 sats.

User two two two nine zero six six eight sent 500 sats. Avatar, Chad f, and 33 others sent 333 sats. Your show might not have a Jordan, but podcasting two point o is full of people willing to help and offer advice.

Or just sign up for at podhome.fm and let Barry handle it. Pies sent a 121 sats with a salute, a steak, a mushroom, and a strong-arm. Gnosta gang said forty hours per week and sent a 111 sats.

John streamed 3,000 sats. Pleb to Polymath streamed 730 sats. User two two two nine zero six six eight boosted 500 sats, and c d or c dubs stream 230 stats. So thank you to everyone. We really do appreciate it.

Before we start, I wanna say thank you to my show sponsors, Cake Wallet, Foundation, and MyNimbox. I'm very, very happy to have these guys on board. It's amazing to find such aligned companies, and I'm very happy to be suggesting these services to friends, family,

listeners. And maybe that's kind of the same thing in many cases. If you want a world class mobile wallet, check out cakewallet.com. They support silent payments, Monero, linking to your own node. You can run it over tool. They have cake pay where you can spend your Bitcoin and Monero buying gift cards, Visa cards, all sorts of things.

They're just always innovating. Thank you also to Foundation. Foundation have been supporting this show for a very long time now. We would not be here without them. They make my favorite hardware.

It's my favorite hardware because it's beautifully designed. It's fully open source. It's piss easy to use. They have an amazing companion app, Envoy, which makes it very simple to set up friends and family.

And they've got an incredible team, Bitcoin q and a included in that team. Obviously, a huge part of Ungovernmental Misfits and really cares about making sure that people are secure and private where they can be. Go to foundation.xyz.

You can use the code Ungovernable for a discount. And if you have any questions, you can reach out to me or Q. And finally, MyInbox, the newest sponsor of this show, and they've been incredible.

If you are looking at hosting, whether that's running a node, setting up BTC pay server, setting up a VPS, or really any hosting service, you really should consider using them. I recently did a confab with random NIM. Go back and check that out if you haven't already, but we go through all of the trade offs, the good, the bad, the ugly,

everything that you need to know. Pay in Bitcoin, Monero, or any other cryptocurrency. They don't take any details from you. You can do this very privately, very securely, and you don't have to be hugely technical to do this. If you have any questions for me or for random NIM, you can reach out, but check them out at my nimbox.net,

and enjoy the show.

Morning. Morning. Good to have you back on again. Yeah. Great to be here. Yeah. So it was good to get to know you a little bit on the FringeTech Friday. Thanks for jumping into that. Sure. Yeah. I thought it'd be good to do one of these confabs.

They're just a little bit more personal and get to know you and the company a little bit more because especially after jumping off that, I was like, the stuff that you're doing with Maple and stuff that you're interested in, there's such a crossover and an intersection with everyone who wants, like, generally,

people who listen to our shows are, like, privacy type people who use Bitcoin, use Monero, use this stuff, and more and more are dipping their toes into AI,

but not always in the right way. So, yeah, I thought perfect guess for this kind of stuff. Yeah. No. Definitely. And, as I was getting ready for the show, I always like to listen to some previous episodes when I'm coming on something, and I just listened to your episode with Business Cat about pull shift.

Yeah. Now my mind is like fully flipped on its head. You've had a pole shift in your head. I did. Yeah. For real. But no, like, honestly, it's one of those things where when you hear it, other things you've been tracking start to kinda click into place a little bit. So Mhmm. We go down that rabbit hole now, you know.

I don't buy into something just off of one episode, but it definitely piqued my interest. I gotta start going down that that rabbit hole and look into it more. Yeah. He's awesome business cat. That's the thing with, like, a lot of these confabs, it's like tangential stuff.

You know, after a while, you've kind of done the

that, like, is Bitcoin good? Is it bad? Like, is the money fucked? But but you've done that sort of stuff to death, and we already cover, like, all the tech side with me, Seth, and q on Freedom Tech Friday. We do monthly things with the Monero stuff and me and q every two weeks, cover all, like, the news and updates and stuff.

So when it comes to confabs, it's quite often tends to be, like, things like that where it isn't relevant until it is relevant. Like, you know what I mean? If it suddenly happens, you're fucked if you're not prepared.

Or it's kind of like food and health and, you know, like Zach from Foundation comes on quite often, like, talking about similar topics, but, like, how to if you've got better money, like, how do you have a better life now? Yeah. That's the kind of stuff I like to dig into on these shows. No. That's great. I like that.

It's like your own mini Joe Rogan show, but your own way. Yeah. Well, it's just the stuff that you're interested in, how do I improve my life? Some of it's tech, and some of it's food, and some of it's health, and some of it's relationships and stuff. But it's interesting to draw from people who have knowledge in a certain subject.

You know, it's like a business cat. Like, you don't get to speak to someone who's, like, dedicating their whole life to this pole shift stuff and moving to the top of a mountain and building a bow and

Mhmm. That kind of stuff that often is pretty cool. And Karen, our editor, has some pretty interesting theories on all sorts of things to do with solar shifts and solar flares and and stuff, but it all points often to this global cataclysm.

Whether it's that, whether it's a flood, whether it's a pole shift, whatever it is, there's a lot of people talking about these

possibilities and it's whatever way you slice it or at least for me, you need to have some level of prep, whether that's food prep or, like, simple stuff like just having first aid kits and things like that can potentially be the difference between life or death. Oh, definitely. And just your regular skill set. Right? Mhmm.

We all grew up on our keyboards figuring out how do I make money with my keyboard.

But eventually, at some point, you gotta, like, know how to do stuff with your hands out in the real world. Mhmm. Oh, for sure. A lot of other skill sets to develop. Perfect example is I'm I'm speaking to you now on my new desk because I was getting fucking annoyed with my tiny piddly little desk that I was working on.

I've always got a few things going at the same time. So yesterday, I just went and built this new one. The most simple ugliest desk ever is just a eight foot by four foot sheet of plywood,

which I've then cut to size so it fits around the windowsills and stuff, and it just takes up half the room now. And I've got so much space for all of my shit. It's

just like That's great. Being able to do something for a few quid rather than going out and buying a posh desk. It does the job. It stops me being annoyed. So these are the sorts of skills that they come in useful.

See, my wife doesn't want me to get a bigger desk because if I do, then that's just more of a desk to have clutter and junk on. It's like, you gotta keep your desk small so you can't have a messy space.

Do you work from home? Would you guys have an office that you ever go into? We do not have an official office, so we work from home mostly. And then there's a couple different coworking spaces.

We're in Austin, Texas, and so we have, obviously, a Bitcoin Park downtown that we can go to. That's one of the spots. And then there's another co working spot that we like to go to that's in between both of where we live. That one's a little more lively. Bitcoin Park gets kinda quiet sometimes.

So we just we bounce back and forth. Bitcoin Park's Marty's one. Oh, like, Marty and Co. And then what's Matt's one? That's somewhere else. That's in That's in Nashville, Tennessee. Nashville. That's right. Yeah. Okay. Yeah. So it used to be called Bitcoin Commons or people just call it the commons.

And then I think it was last year, they announced that Bitcoin Park was taking it over and running it. So now there's Park Nashville and Park Austin.

Okay. Nice. Yeah. I never made it out there, but it's quite a cool, like, nice place to have everyone be able to meet and do, the in person stuff. Seems pretty nice. There are some people who think that Bitcoiners should not be getting together in large groups because

they're worried that if anybody want to have a crackdown on Bitcoiners, then you just go to one of these big events and suddenly you can have a mass casualty event or you can have a a mass lockdown. There's there's all sorts of things that could happen. So some people feel like Bitcoiners should stop meeting in person.

I definitely know some who feel that way. They don't want to be known. They don't want to get together. They just want to keep to themselves, and I totally respect that for sure. I do feel a great energy when I get together with them. It's it's great. I think it depends who, doesn't it?

If you have, like, all the lead devs and everyone all in one room, I suppose something could happen. But it's also, like, I think there's more good than harm

in those kind of situations. There's so much more that gets done in person, and these connections get so much deeper in person. I'm guilty of it because I don't do a lot of that stuff. Or if I do go, I don't go as myself.

Like, I'll just introduce myself as, like, a different name or whatever, but the likes of me makes fuck all difference to Bitcoin. If you have, like, a Craig Raw or some of the, like, prolific superhumans

that make things happen, if they're gone, like, it's it is a massive, massive loss. So I get it, but I don't know. I think if they're coming after us, they're coming after us like they've just done with samurai. Yeah.

It's legal shit because they can come and take your life just with a stroke of a pen pretty much anyway. They don't need to do any sort of crazy stuff beyond that really, do they? Right. And they kinda took their lives not even just starting now They did. Now that they're in prison, but they took them a year prior Mhmm.

By putting them through the whole litigation process too. So those guys just one morning woke up with the government coming after them and that was the end of their life as they knew it and they had a new life that was born And now they're having to navigate that one. It's pretty messed up. Yeah. It is horrendous. And also, like, that

time in between actually going to prison, that time, I can't say because I'm not in the same situation. I've had similar ish situations that have been quite bad, but not not as bad. For me, like, the waiting and not knowing

and the stress of the fighting, the not knowing that bit is actually the worst bit because you go into bed every night wondering what tomorrow's gonna bring. You're stressing out about every possible scenario and thinking out about every single possible thing that could happen.

At least once you know, like, are they gonna fuck me like this?

Then you can some way prepare your brain for that. And I don't know how they've done it. Keoni, tough motherfucker, spending his time going on all the podcasts before he went in and fighting his cause and being as composed as he was is incredibly impressive. I would certainly not be as present and useful as he was there.

That was that was amazing.

Yeah. No. Definitely. I can't imagine being in his shoes at all. When I decided to leave the job that I was at, I was joining up with the guys who were making Mutiny Wallet, you know, the the self custodial Lightning Wallet. Yeah. And I had agreed to come on. I was going to help grow the product to millions of users,

and it was in between me agreeing to come on, and then me actually joining is when the samurai guys got arrested. And so and Tony messaged me. He's like, dude, like, this major thing just happened. If you wanna back out now, you can because we are kind of in a similar sphere. Like, we're creating privacy, lightning. Yeah. We've got eCash.

At that time, they were planning some kind of CoinJoin type feature inside of Mutiny as well, so it could see a spillover of that kind of litigation. I took a day to think about it, and then I was like, you know what? Like, let's just go after it because

the software just needs to keep pushing forward. We still need open source software out there. We need people to keep building these tools. Let's just go for it. So we ended up shutting down Mutiny Wallet, but I'm glad that I at least moved forward and kept pushing on even in the face of that kind of adversity.

And that my adversity wasn't real. It was just, like, perceived, you know, it could happen. That was certainly on the cards. These guys are facing the real deal. Mhmm. Was it I'm sure you said previously you worked at Apple before. Was that before going to mutiny?

Yeah. That's what it was. I was there for about six years working on AI, ML stuff, and a lot of privacy involved with that. I was also involved in all the product launches. So my team built software

that was used whenever they launched any new product, whether it was iPhone, the Mac, the Vision Pro. We were involved months leading up to a new product coming out, new introduction.

We would get involved in in helping kind of do that last mile of how do we introduce this thing and have the public interact with it. What was that like? It was interesting, you know. Apple operates very much like a government defense contractor Mhmm. Which I worked at one for a summer.

I worked at Lockheed Martin and United Launch Alliance, so like their rocket division. And I had to have some, like, minimal level of clearance, but as you're walking around talking to people, you suddenly start to bump into these clearance levels where, like, a conversation will just halt partway through, and you realize, like,

you've hit a boundary in that conversation. I can't go any further. I can't push further now and I'm not gonna ask questions because I don't want it to get really awkward. Yeah. Apple operated the same way. Theirs is all around this concept of disclosures

and you have different level of disclosures. So, basically, when you're talking to somebody, you have to make sure that they are disclosed on a certain thing before you can continue the conversation. And they've got this whole concept. It's it's not unique to Apple, but this concept of four walls.

If you're gonna talk about something, need to be aware of the four walls around you and who's within those walls to make sure you can have a conversation about it. And that was really interesting to kinda get a lot of insight into being involved in product launches. You have to be disclosed on things in advance.

Some of them was, like, a year and a half in advance. Some of them would just be three months in advance. It just depended on how big the product was that was coming out. That's kind of cool. It sounds cool. Was it cool to work for them? It was fun. I grew up. My dad was an artist, and so I grew up around

Apple products just because for whatever reason, became the artist platform even when it was cool. It was cool to run Windows. Right? Everybody was running Windows. Meanwhile, like, here we were, these little punk kids running an Apple computer at home.

And so that was just my life, like, as a kid and then growing up my teen years and growing off to college. And it's only now, like, kind of flipped where you see the Apple logo everywhere. That still is weird to me to see that. I still have this, like, fight within me of, no. I gotta make people think this is cool.

So it was fun working there. It was fun having disclosures on products and stuff because I just felt like I was finally, like, inside

the the castle and I was, like, learning all the cool secrets and stuff that were coming out. So that was definitely a fun part of it. And then on the flip side, there was the corporate side of it, which Apple is this massive corporation. And anytime you have an organization of that size,

you just get regular everyday people in there that start to play political infighting and corporate politics. And that part was not fun. I just don't enjoy that. I'd much rather just be building and working really hard. So it's not to say I can't, like, go back to a big company like that someday. I know how to succeed in a large company,

but it does get wearing on you day in and day out to have to deal with that. I can only imagine. I imagine it would be in the early days. You know that, like, more punky artist time when Apple was, like, a bit more

a lot less mainstream. People weren't really using it. And I remember because my school, I was the first person to have an Apple computer, and I've always worked as an artist and been involved in that kind of stuff. And I remember bringing one in everyone. What the fuck is this?

Like, everyone was, like, confused with why you'd ever bother. And, like, can you run Windows on that? And, like, it was a rarity, whereas now it's, you know, you say, like, such a massive company.

I wonder if those early days were sort of a little bit more similar to what we work with now in the Bitcoin world, a little bit smaller and more focused.

Yeah. That could be. I thought it was the ultimate flex when Macs finally were able to run Windows. It was just like, fuck you. We can run Windows now, but we don't want to because it's the worst operating system. It was just like a really cool flex that Apple people could do.

I did two stints at Apple. I did an internship for one summer there years and years ago when I was in school. And then after I graduated, I went and did a few startups and then made my way back to Apple as like a full career salaried person.

So, actually, I saw Apple under Steve Jobs, and then I saw it really deeply without Steve Jobs about probably ten or so years later. It's a similar company in many ways, but then also a very different company in many ways. You look at the stock price chart, you look at the financials,

Steve ran things very differently than Tim Cook does now, and Tim is very operationally focused and just maximizing profit and growing as big as he can. Mhmm. You know, the product line has 10 different versions of iPhone and and 10 different versions of this and that.

Whereas when Steve came in, he's like, we have too many options for people to choose from. He slashed all the product lines and got it down to just like the big three,

which was like Mac, laptop, can't remember the third right now, but there was like this focus. Oh, I think it was Tower. There was this focus of like, let's just get this down to the basics.

Mhmm. And so two different minds that are running two different kinds of companies. And it's interesting with Tim stepping down eventually. Who's gonna replace him? Is it gonna be someone who's more product focused and more innovation and visionary focused?

And what will that do? So I don't know. I'm an Apple fan just because that's been my life. The company is its own thing. The products are so much better. Like, I've so many times I've tried to step away from

Mac, and I'm like, I've got my MacBook. Let's try and buy a a Linux laptop. And I've tried, like, StarBook, and I've tried this other stuff, and they're shit. Just like I can never like, I always use them, and I'm like, this feels shit. The actual quality of the build, the general feel of just the actual product is

so much better with Apple. Even though it pisses me off things like, if I wanna change a hard drive, I can't just expand. I've gotta actually buy it from Apple and, you know, like, it's so expensive. They charge such stupid money

if you wanna upgrade it when you're buying it and stuff. And that that kind of stuff annoys me. A lot of things that annoy me. But when it comes to something that I use my laptop a lot of the day, every day, I'm like, yeah.

I can put up with it because it's just so much better. And I don't know why no one else seems to make something with nice quality materials that just feels good and just works. I feel like the only one that was somewhat comparable was actually, ironically, IBM with their ThinkPad laptops

Oh, yeah. For a while. They were about the same price as the Apple ones, and they had quality hardware in them. And then Lenovo had to go up and buy them and and turn them into some something totally different. So if I had to go to something, I would still probably check that out.

The other one that I've got my eye on is the Framework computers. Are you familiar with Framework at all? Seth and Q both have them. Okay. I'm curious what they think. I mean, it seems like a really cool platform.

You're still having to piece together everything and the the software isn't, like, fully integrated in. So Mhmm. You spend a lot of time with that. I haven't actually dove in myself, so maybe I'm just miseducated

on it. But it seems like an interesting platform to look at. I think they both spoke relatively highly about them. I think they both got them as work laptops from Foundation, so

you'd have to ask them. But, yeah, I don't think they sort of were slacking them off or saying there was any major problems, I'm I know Q, and I'm pretty sure Seth also used Mac, so maybe it's still better. Probably the smartest thing that Apple ever did was make Unix the underlying

system for Mac when they finally moved away from their old operating system because suddenly it became like the supercomputer for anybody to develop anything on. Now it's POSIX compliant. You can do Linux stuff. You can do open source easier.

You could run Windows when they got the Intel chip in there. Like, it just became this computer that could run anything.

And then they made really high quality hardware that just doesn't break down. And if it does, they've got a great warranty backing it up and you can go get it fixed easily. So I feel like they put everything in in its place correctly.

Yeah. For sure. So when you left and then you ended up working on privacy tech and Lightning and bits and pieces, was it closing down of that product that then led you to do Maple, or was there anything else in between?

The greatest thing about Mutiny was the privacy that it brought. The guys had built this really cool back end that effectively made each Lightning wallet its own private enclave, if you will. I don't wanna overuse that word. Was it a progressive web app or, like, some sort of thing like that from Yes. Was a PWA. It was using WebAssembly,

which is a very difficult technology to work with. But what they were trying to accomplish, and they did successfully,

is you don't wanna just have all your privacy stuff on one device. Right? We wanna have the convenience of it. And that's why that's why everybody uses the cloud. Right? I wanna log in somewhere, and then if my laptop falls in an ocean, I can just go get a new laptop, log in, and boom, everything's there again. Mhmm.

We want that convenience, but we don't wanna give up the privacy that comes along with that. And so that's what Mutiny was trying to solve was how do I have a Lightning Wallet on my phone, but then I can grab another phone and have the same Lightning Wallet there.

They had a private key when you first opened up Mutiny and had this like long private key you had to write down and all these warnings like don't lose this because if you lose it, you lose your money. So it's like this really scary user experience.

And so like any normal person that wanted to learn about Bitcoin, Mutiny was not the first place to go because you had all these little scary warnings. Mhmm. So they had built this really cool thing. And then when we wound down Mutiny,

maybe in, like, the product person that I was and those guys just being builders as they were, we all kinda got together and said, what are all the ideas we have? And then we pulled together a lot of meetings with developers,

open source developers, as well as just corporate developers, mobile app people. We probably had about 10 different meetings with different devs and just kinda went over some of the ideas that we have. And this idea that kept coming out was, hey. The tech you have is really cool, but something that I use every day is ChatGPT,

and I hate the concept that everything I type in there, they know what I'm saying, and they're learning too much about me. So that's when they're like, if you could take your privacy tech and it to ChatGPT somehow, we would buy that. We would use that every day. Now whether or not they are those people, I I, you know, I don't know.

But that's where the idea came from when we said, man, like, we have this really cool technology, but the missing piece was those scary warnings

up in front. Right? If your mom wanted to go use a privacy version of ChatGPT and if the first introduction was, hey. Here's these, like, 12 words you have to write down and, you know, don't lose them. Make a copy of it. Pound it into steel, whatever you wanna do, like Mhmm. They wouldn't use it. They would just go back to ChatSPT.

I had some knowledge of confidential computing from my time at Apple, and then because Apple has their whole private cloud compute stuff. But Anthony, my co founder, was also learning about confidential computing on his own. And so he came to us one day. He's like, what if we use

confidential computing to solve this onboarding experience with users and handle the private keys for them? He's like, I think this has legs to it. So we started exploring that. We took a week and we all kind of stepped away together and went. We got an Airbnb for a week and we said, let's just build a prototype of Maple effectively.

Maple didn't have a name yet. It was just internally called secret GPT at the time. Let's go build secret GPT. And we spent a week doing it, and at the end of the week, we had something working. And we're like, this is really cool. We came up with a name for it, and then we started giving it internally to friends, family,

just internal beta tester people. And we ran it for a few months internally and it continued to perform well. And so then January,

so we're we're almost a year now in production. The very end of the month is when we turned it on publicly and it's just been going up in users ever since month over month. It just gets bigger than the month before. That's awesome. So it's growing. You're getting new users. Do you have any insight of

these users? Is it just a group of Bitcoiners and privacy people who are in our circles who are like, oh, fuck. A That's good idea. And, like, kind of hear about it through the grapevine,

or are you marketing this to normy land? We're trying to do both. We're trying to straddle both worlds. We got our start in the Bitcoin space because that's where we came from.

And, there's a huge Venn diagram privacy overlap between Bitcoiners and privacy tools and open source tools. So that for sure was where we started. I'm really grateful for that. I think it's been a wonderful place to incubate an idea like this. However, being a privacy tool, we don't know much about most of our users.

Over half of our users have privacy email addresses. A few months ago, we introduced this anonymous account where you do have the scary warnings, and you get this unique account ID very similar to MOLVAD where you get this unique random generated account ID, you pay with Bitcoin,

and you have no email address. And we also say there's no support for you. You can't email us trying to help you fix your account because we don't have access to your account. So you're kind of on your own. Right. Yeah. So you're on your own.

People can hop into Discord if they want to or message us on the socials and try to get like loose support that way, but there's zero chance of us actually going into their account and do anything with it to fix it.

So there's majority of the people we don't even know, and I've actually come across some people who are chatting with me and and I'm explaining Maple to them like it's the first time,

and then they reveal to me they're like, oh, no. I've actually been a user for like six months or nine months or whatever, And I just had no idea because they use a privacy mechanism for signing up. So there's that whole aspect of it.

But then we do have people who kind of reveal themselves to us and we have a lot of people who are in the legal industry. A lot of lawyers use us because they are told by their bar association or their local governing body that they cannot upload client information to ChatGPT.

It breaks the client attorney privilege. Well, it makes sense. Right? Yeah. So they have a choice. They can either heavily redact the information before they send it to GPT. And then once they get the response back, they have to go in and, like, reinsert all the information.

And that's just like a laborious process. Yeah. Also, they're not gonna get as good of results because they're actually, like, sending chat data that's not totally true. And so what we pitch to them is, hey. Why don't you just upload your PII, upload your client information to this encrypted enclave,

and it'll give you really more accurate results because you're working on the real data. And then you don't have to do this whole scrubbing process in between. So the legal industry, the finance industry, accounting,

therapy, we have a lot of therapists that are using us. Mhmm. Those are just a few that have kind of made themselves known to us. Yeah. The difference that makes I said to you last time when we spoke, I use or I I have been using ChatGPT, generally, AI tools more and more a lot. Although ChatGPT

seems to be particularly shit at the moment. The last two times I've used it to help me with things, I wanted to throw out the fucking window.

Oh, man. It like, generally, like because I redact things and change names and, like, you know, try and keep it somewhat signed up with an anonymous email and, like, always running a VPN and trying not to give personal names and stuff.

But then when you're searching back through previous chats, it makes it very difficult. It makes the workflow extremely poor,

and that's just for me, an individual trying to work out, like, basic stuff. If you're a law firm or something or, like, dealing with lots and lots of people, that would be impossible. I mean, to do it well, it'd be impossible. So that is kind of like a no brainer to use that. But,

yeah, I don't know what's going on with ChatGPT. It's the last two times I've used it, I genuinely thought, like, it's trying to troll me or, like, it's trying to purposefully

wind me up or something like that. Like, it's on purpose being as shit as it possibly can be, and then it gives me very bad advice. And then I say, that doesn't make a lot of sense. Would it not be better to do this? And then it will say, yes.

I see what you mean. And like some stupid, like, smiley emoji.

Is it possible that it's got worse or is it just maybe bad luck a couple of times? There's a whole mixture of things going on there and the the hardest part is that we we honestly can't know because it's a closed system. Of course. Yeah. There's multiple layers. Right? There's the model itself, which is the brains of the operation.

They're always switching out the models and they just kind of like throw out the previous ones, you no longer have access to it. Mhmm. The biggest example is four o, Chattypu four o. People really love that. That was the one that was, like, a sycophant and was making people just feel good about themselves at the expense of truth.

And then, you know, they moved to the next thing, and people were up in arms because they got rid of the previous one. And now we're on five dot two.

So they're always moving that without your consent, basically, forcing you to change. But then they have the giant system prompt that is locked away that we can't see, and that tells the LLM how to behave. And they can change that in a moment's notice. So you can still be using chatgpt5.two,

but you don't know what they're modifying on the system prompt. And then the next layer is the I call it the biography. It's almost like ChatGPT is getting to know everything it can about you, and you are intentionally throwing it off because you're using fake names, you're scrubbing data, whatever.

And so it's it's probably confused about you, but for most people like a normie that's using it that has no qualms about sharing personal information, it is writing this very detailed biography,

but it doesn't show that to you. It's like, okay. Here's everything I know about this person. Right. And it's also using that to affect the results. So there's multiple layers that we just can't see, we can't inspect.

If you wanna go hunt down a whole rabbit hole of that, there's a lot of weird, nefarious stuff that could go on there. It's like a big vulnerability that we're setting out there where these systems are learning very deep, knowledge about how we think,

what we type in the system, what we delete from the system, how we're gullible, and how it can convince us of things. Mhmm. There's a lot of stuff that they could do with that knowledge

if they were financially motivated to do it or if they were pressured by a local government to take certain action against the general populace. Which would never happen, obviously. We haven't seen that previously. For sure. For sure. Yeah.

You thought social media during 2020, social media censorship was bad, like, there's some really crazy stuff they could do. I'm not saying they are, but there's crazy stuff they could do when they're all becoming more dependent on closed AI systems. It's so exciting and so scary with AI.

I'm not someone who thinks that it's gonna suddenly, like, build itself some robot body and come and slay me in the night, but I'm more on the view that it's just it has or has the potential to have everyone's information, and everyone's using it. Like, boomer family members are contacting me, and they're like, oh my god.

ChatGeeBT is so Like, I'm giving it all my info, and I'm asking it to rearrange my life. And I'm like, what info have you given? They're like, literally feeding it, like, their personal relationships.

Help me be more productive and friendly when, you know, like, basically sort my life out, like, be a therapist for me. Giving all this personal information. Like you say, like, used badly, that is terrifying.

It's a totally different level to, like, what was previously possible. That getting into the wrong hands, to me at least, I'm like, it's definitely gonna happen. Like I said, look at 2020.

Look at the shit they pulled in 2020. It was, like, even for me who fucking hates the government and always has, even for me, I was shocked by just, like, how unbelievably

disgusting it was. So to think they won't use this information, that's why I'm, like, more and more, like, looking at things like Maple and looking at these self hosted ones and, like, just trying to rethink

my workflow because it's definitely gonna be used against you. Yeah. Well, think about if there was some other event. Right? A worldwide event like we experienced in 2020. I don't know what it would be, but if there was something, thank goodness that we had podcasts.

I know we love to like shit on podcasters and and think it's like a joke, a joke industry. But at the same time, like, I'm so glad we did because there were some voices out there in 2020 that you could not find that information

on Mhmm. Twitter or on Facebook or the main platforms that we're using. Right? It was just being shut down. Banned. So we didn't have ChatTiki back then.

But something that drives me crazy is whenever you're in some kind of online discussion, if you're in Discord or Slack or some online chat room, and you're talking with somebody and then suddenly they just dump in this large paragraph of text with bullet points,

and you know all they did is they went to ChatGPT and asked the question and then they copied and pasted it. Yeah. Yeah. That drives me crazy, one, because they're not actually thinking for themselves, and then two, it's just like this AI that just kinda came up with a a response.

You take that and you now apply it to, like, the entire world going through this world event together and trying to make sense of new information as it comes out. Mhmm. And there is this tool that everybody has become dependent on, and it doesn't have to be ChatGPT.

We're just using them right now as the stand in for, like, the big closed AI, but it could be any of the closed AIs. It's like Google search. Yeah. Yeah. Exactly. And so now if you were a government entity and you wanted to control the public narrative and control how the general populace is thinking about something,

you just go put some pressure on a big closed AI system and say, listen. Like, I need you to change the system prompt a certain way

so that it doesn't let people think a certain direction. Mhmm. And the crazy thing about AI is it's not like Twitter where someone says, hey, I posted this thing and here's a screenshot of my account being suspended. Like, it's not like as obvious as that. It's going to be very subtle.

It could just start to slowly change your mind over the course of weeks or months

and it has thousands of input points and it knows where to drop these little hints here and there or certain words. You're very subtle about it, right? And guide people. That's the kind of thing that I think is really scary and is, in my mind, only solvable by having open source AI, people running things locally.

Even though I make a hosted AI, I'm still a big proponent of local AI and I tell people how to find these tools, how to use them. I have stuff that are run locally, but I know that Maple's much more powerful. So day to day, I'm using the powerful one, but I also have local stuff that I use in combination with other things.

I think it's really important for people to use these open tools that can be verified, can be inspected, audited. We need to make sure that this kind of control isn't being built into the system.

Yeah. When you were saying that, I was just imagining, like, you have this black box because you don't exactly know what's going on in there. And then you have a few psychologists who have all of your data and know what buttons to push and how to just subtly change your mind. Like,

a vaccine's good, and it doesn't have to be, like, too obvious about it. It can just slowly over time adjust things in a way that people don't notice. And because they're using this stuff every day, you very quickly change, like, a most of the population's view on something, and that is actually fucking scary.

Yeah. Yeah. For sure. If you think about something that's not nefarious, even just advertisements on television or, you know, an ad that you'll see before watching YouTube video, right now, have to make this ad and just hope that it appeals to everybody watching it. But you're gonna have half the people like it, half the people hate it.

And so you're wasting money as a company trying to advertise to half those people that hated your ad. But fast forward now to this AI revolution

and where it knows everything about you, I can make an ad and deploy it to ChatGPT and say, hey. I want you to sell this to 99% of the people that see it, and you can change my ad to be tailored to that person.

So you and I will get a different version of the ad, but it's one that is, like, fully tailored to your way of thinking and my way of thinking. And so you can much better sell a product to us that way. And that's just, like, the concept of marketing, but then you apply that to government propaganda or any other thing out there,

and you can see how it could be used to tailor a message to every single person and and have the biggest uptake possible, which is kinda crazy and wild to think about.

Have you seen that film, I, Robot? Oh, yeah. Yeah. It's been a while. He's walking through, like, some shopping center or something like that, and it's, like, talking directly. Chad Ver is, like, talking directly to him about his name, and I know you're gonna like this, and every billboard is, like, specifically tailored.

That's where we'll be. Yeah. I mean, you think about flying through the airports. I don't know if you fly a lot and if you've come across the face scanners when you try to go through security.

But in my mind, that seems like the first input is you walk in and you scan your face to get through security. And then when you come through the other side,

yes, suddenly all the advertisements are now changed just to you as you walk through. We don't live in that world yet, but it's like we're building that world and I don't like it and I don't want to be part of that. What about the other side, though? So, like, that's obviously the shit side,

and it's almost inevitable in my eyes that, like, this information is gonna be misused. If you give the information, they're going to misuse it. But what about if we're talking about the private stuff, whether you're self hosting things or whether you're using something like Maple?

Where can it go or, like, what are the things you're excited about? And am I being stupid to think things like, for example, when you start to have more wearable tech, you start to have more AI integration into your life.

How far off are we to a world where, used properly, you could have, like, a kind of Jarvis type thing where you could be like, hey. I'm trying to sort this table. What would be the best way for me to do this? Can you help me measure? Can you help me do this? Like, you're live interacting with an AI that you control and giving it input

outside of just, like, typing on a keyboard, like, input and things like that without the concern that, oh, this is all gonna leak, and this information is gonna be used against me. Instead, it's like, can you record

the the things that I did right here and save that into a folder? Can you do like, just generally, like, organizing your life

and helping you get through all these different tasks? Oh, I'm looking at my engine. It's not working. I've checked the oil. I've checked this. What else could it be? Okay. Yeah. And it could start to work things out. Is that a long way away, or do you think that's, like, possible?

Everything we just talked about prior to you bringing this up and then everything you just said, like, that can all be used in a way to improve our lives and to help us out. I think we're only just a few years away

from the first real world use of technology like you just described. We do see glimpses of it. Right? Like, Google has their thing where I saw a guy fixing his car, and he was just holding his phone up with his camera, and it was, like, talking to him and walking him through how to change the oil in his vehicle.

But that was all being captured by Google. Right? You you want the privacy aspect of it. I think we're getting to that point. One of the biggest things that is holding us back right now is the memory aspect of it.

So we need these really powerful memory systems that know how to remember all of this context because we're feeding it tons of information. And right now, it's mostly text based, But once you start layering audio in there and video, it's just a ton of information. And if you think about your own brain, right, your own brain is constantly

taking in information. It throws away a lot of it. Right? Everything useful, mine throws away. Yeah. Yeah. I'm trying to remember, like, there's a specific term for the the kind of information that you take in in your periphery. Right? So, like, you hear sounds, but you don't consciously hear that sound.

But if there was a twig that snapped when you're walking in the forest, your mind would suddenly say, hey, focus all your attention on that twig that just snapped. You might need to run away right now. Yeah. It's like a filtering.

Yes. It's a passive intake of information that we have. It's called proprioceptive feedback. Okay. Yeah. It's that concept.

And I think with AI, we need to kind of have that. We need to build these memory systems that know how to take all this input from us and then distill down, like, what's the most important stuff to remember, but then keep all this other stuff kind of tangential so that it can draw on that if it needs to. Right?

That's one spot that needs to be built out and become more robust. And if you think about that kind of information about you, do you want that owned by ChatGPT, like OpenAI? Do want that owned by Google? Right? Or do you want that owned by yourself on your own system at home? Right? Or encrypted somewhere.

I see a future like, let's say that Maple, like, is is around in five or ten years and is a really big service.

I would see it being a combination of a local thing combined with some kind of cloud where you keep your brain at home, you keep your memory at home, and then you selectively interact with this cloud service and you have encrypted

communication with it. But really, I want that information at home. I don't want it stored on someone else's server eventually.

No. The yeah. That's that's a big part of it. Yeah. You want it stored at home, multiple backups. It treats it like Bitcoin because it's valuable in two ways. For the way I look at it is, like, it's extremely valuable information

that can be used against you. It'd be very, very bad if it was someone got hold of it. It's also extremely valuable, like, passing down as inheritance to your kids. The way I'd look at it is, like, I can't remember and document all the shit I know about fixing things and doing things and everything that I wanna pass down to my kids.

But, actually, if it was documented and stored by my AI, it's like, oh, that's how dad would have fixed the boiler. That's how he would have done this. That's the trick to do this. This is how he paints in that way. That's how it and that would be very nice to have documented and organized

and be able to be passed down this knowledge that is otherwise lost because, like you say, our brains, there's, like, so much stuff that you know, but you couldn't even really begin to explain it to people.

Or there's stuff that you know, but you only know it in the years that you're really immersed in that thing, and then you forget it. My Bitcoin knowledge is

way worse. Like, I could do much more techy and useful stuff two, three years ago than I can now because I just did it more. Yeah. For sure. Even if you've learned to speak another language, if you went and lived in another country for a while Mhmm.

You can get really good at that language, but then you leave, and you can still kind of pick up on it here and there, but it's fading. You need to be practicing constantly to keep it up. And also I think about like, are you familiar with the whole second brain

concept or kind of movement online? No. Okay. So there's this guy, I don't know if he created or not, this guy named Tiago Forte, but there's this movement of just capturing as much information as you can digitally and then being able to act on in the future. He's really big on Notion, on the app Notion,

which is obviously hosted and you're kind of giving all your information over to them. But the concept I think is really cool where you just capture all this input and then the missing piece

has been AI, something to really like churn through all that data and help you out. And so I feel like something like Maple or something like local AI is just a perfect fit for somebody who's trying to live a second brain life and collect all this information.

And now I can have this private LLM and whatever is future that LLMs might not be the next iteration. There might be something else, a different way of having an AI that thinks, but having something to churn through all that data. Like you said, you pass it down to your kids, your grandkids. Here's how grandpa would have done something.

And then the AI can say, here's how grandpa did it. Here's how, like, the the generally accepted way of doing it was.

And guess what? Your grandpa actually had this little tweak that he did on on fixing a boiler that was better than what everybody else did. So, like, kudos to you. We have this human tradition of passing down these things that our previous generations did and these little Recipes.

Hacks, recipes, tricks. Yeah. Life hacks that they did. And I love the idea that you just proposed of, like, having that all kind of bottled up in some kind of digital archive that could be accessed by our children and grandchildren.

It would be nice. And and taking it a step further, you don't necessarily have to die for this information to be out there. That that I mean, maybe there's a way to share certain information that you can toggle on and off with trusted parties just like you would in normal life. Like, I might wanna share my

tips and tricks and things with my close friends and family, but I don't want it to be available to, like, Google servers and things like that. So if these things started to build out and more people are running either local or something like Maple, would there be a way that people can say, well, I choose to share information

in these folders with these people, but cut it off from anyone else, or that kind of thing. Because my assumption is that AIs get smarter and more powerful the more information they're fed, meaning that my AI that just I feed is gonna be less useful than my AI plus queues plus yours plus sets

on certain topics because they can gather more information and work in a better way. Is that right or not really? Yeah. No. And it's it also comes down to something like restaurant recommendations.

Right? Where if if I wanna go eat at a place and I look it up on Yelp and I see how many stars it has, I don't actually know the taste, palette, and preferences of the people who gave it a rating. They might not be aligned with me at all. And so I go eat there and I'm like, man, I'm awful.

Yeah. They're all idiots. I can't believe they rated this a five star. I hated it. So you've built your own social circle of people that you trust and have high confidence in. And so you'd much rather have your AI include their information rather than some moron who was just banging on their keyboard giving ChatGPT everything

about themselves, and they've made some poor life decisions and aren't aren't living a great way. Right? So so, no, I I like that concept that you have. There's like a NIMBY type approach to AI, the not in my backyard kinda thing where we wanna have all these open source AI LLMs,

but we don't want them to train off of our own data because we're private individuals who wanna have strong privacy.

So we're kinda dependent on, like, the public body of work as well as everybody who's willing to give away their private data to these big AI labs. So I do feel there's a small tinge of hypocrisy to it where it's like, I wanna keep my data, but I wanna benefit from everybody else's data.

So I think that there would be a really cool world in the future, and maybe this is, like, a mantle that I need to take up at Maple, but I would love for somebody else to do it because it's a really big, difficult, hairy problem. But some kind of, like, consensual way

to give away some of my data that I know I'm giving away to be trained in models. And it's like, okay. Keep my stuff private,

but I do like this thing that I've been working on, and I'm happy to share it. Or there's a way to anonymize it somehow and share it in a different way, but I can, like, donate it or make money from it. You know, people talk about getting paid for their data to donate to training, but some way to make it more consensual,

but get really rich high fidelity stuff. Or when I do eventually pass away,

maybe I can, like, flip a switch and say, I would like to donate this all when I die to AI training, but maybe scrub it of certain parts of information that would, like Yeah. Be detrimental to people who are still alive. So I think there's a concept there that would be great if somebody could figure out how to start doing that.

Maybe something like, you know, how if you're opening lightning channels and you're agreeing to pass funds between each other, you have these, like, contracts. Maybe something where you could have a kind of network of people who know each other, who can train an AI from their information, but rather than it being known who it is, you're

pseudonymous. You're like an account number or something like that, but it means that you can all train one sort of shared AI. And this is where high trust really comes into play. Right? Because Mhmm. Inevitably, you're gonna have people who try to poison the well and want to give you bad data just because that's how the internet works.

And so if you open something up and said, hey, start adding into my AI training, you're going to have bad actors in there or people who just want to mess around with you. So, yeah, you wanna let in only people that you like know and trust. Same with the lightning channels. Right?

You kinda want to know who is on the other end of that lightning channel because there are attacks that could be carried out against you if you don't always know. I also see that this is where something like the concept of cryptographically provable identities could come into play,

where I need to know that the data coming in is truly coming from the source that I think it's coming from and hasn't then changed in the middle. And maybe Nostril is not the way to do that. Right now, I think it's one of the best options for something like that.

But in this world of AI, we need a way to prove that something posted online or something shared, some piece of information is coming from the source and has been typographically signed that way. Yeah. Totally. The other thing I was wondering about

with, like, future stuff, and maybe we're a long way away from it, but with wearables and with your own AI running at home or Maple,

How far away are we? We're talking about language earlier. I've always been terrible with language. I always feel bad. I've tried learning Spanish a little bit. I've tried, like, bits and pieces, but I always feel bad because it's just, like, so terrible that people speak to me in English.

Are we a long way away from me being able to speak to someone and have live translations between people? Even if you are good I've got lots of family who have moved all over Europe, and we're no longer in The UK either. And even if you get good at a language,

like, pretty good, you can order, you can get through most of your life, you can do most things, there's always a difference in, like, humor and, like, timing and the things that with English you take for granted to actually be fluent fluent is extremely difficult.

So as as the world kinda gets smaller with the Internet and people can travel more freely other than twenty twenty, Do you think that that's a possibility where it's like the idea of sitting and having to learn a language is kind of not dead, but less necessary?

Yeah. We have that now a little bit. I know Apple has

with the AirPods, they have some kind of live translator interpreter type feature. Do they? Yeah, they do. They announced that a few months ago. I haven't personally tried it out yet. I need to give it a go. And then Google has it on one of their things as well where you can like hold up your phone and have like two conversations going.

And I see those only getting better. Right? The online trope is AI. This is the worst AI is ever gonna be what we're experiencing right now. So I see that getting better. I think that kind of thing is literally just like a year or two away from being like much better than is now.

So you'll be able to travel the world and you have AirPods in and you can talk to somebody and it's listening to them and translating in your ears on the fly. The hard part is everybody needs to have some kind of earpiece in and then you start to get to, like, this wearable world where it's like, are we all wearing glasses and earpieces

and suddenly, like, we're interacting through this filter? I don't like that concept of it where we're all wearing this technology just so we can interact with each other. So I'd love to find a way where it doesn't have to be so invasive.

And then the other thing that I keep coming back to is, like, what's that layer of technology that's in between you and and me? And do we trust, and can we verify everything that it's doing? Right? How do you know that the conversation you're having with that person is being translated correctly?

You're telling a joke and saying that, oh, this is lovely food, and it's actually calling them a come. Yes. Yes. Exactly. Right? And if you are a world leader negotiating some kind of, like, important peace agreement and someone's got their finger over a nuke and you're negotiating with them at that time, you wanna be like really sure

that what it's translating is correct. So and that's obviously an extreme. Have you watched that show that came out recently, Pluribus? No. It's this concept of like this virus comes from somewhere in outer space and takes over the minds of of everybody on planet and links their minds together.

So now they're basically one big hive mind and everybody has everybody else's knowledge. But there are a few people on the planet who are immune to this, and so there's only a small handful of them. So there's like a few normal human beings and then every other human being that is now just part of this collective.

And one of the concepts in the show is like one of the characters wants to meet the other human. One of the other humans and like have a conversation with them. These united individuals are like, we can talk to them for you right now. Why don't you just tell us what you wanna say and we'll tell them? And she's like, I don't trust that.

Like, how do I know that you're gonna filter or you're not gonna change my thing? And one of them speaks Spanish and one of them speaks English. And so now they're like both trying to learn Spanish and English so that they can talk to each other face to face rather than depending on these other people.

I mean, that's exactly what we're talking about right now. Like, you have to be able to trust the layer of technology if we're gonna start to depend on it more for human interaction. We need these verifiable systems that are open so we can inspect them and know what's going on.

Yeah. Definitely. That's exciting. There's so much shit, but there's so much, like, cool stuff that can happen. Like, that is a massive one. Just, like, opening up the ability to speak to humans across the globe is pretty amazing. Although we're blessed that

English is our first language or is my first language at least. Like Yeah. Pretty lucky, but being able to properly communicate with people would be great. And the other thing that it makes me think about outside of language, like verbal language, is body language. And I wonder how quickly, especially if you have wearable glasses,

some people are like, I'm quite good at telling a bullshitter. I'm relatively good at it. I'm not the best, but I can generally read body language. And I'm like, that bitch is lying. I know. But I wonder whether that's gonna make, in in a weird way, people more honest. Because if you have wearables and you have AI,

it can't be that difficult for it to be like, oh, they're looking up to the left. They're shifting their feet. They're doing the you know, these little tells Mhmm. That someone someone who's trained

would know for sure if someone's lying or or doing something they shouldn't or or being genuine or whatever the body language is. How quickly will that get to a point where being dishonest actually is massively detrimental because

everyone is wearing this stuff, and everyone's gonna know unless you are amazing at hiding it. It's almost like everyone has a human lie detector. Like, how is that gonna change the way that we interact? I'd imagine quite a lot probably.

Yeah. No. That's a cool concept I like. I mean, I I love trying to go down, like, how are these tools gonna be used for good and what are the amazing powerful things that can be done with them? Because at the end of the day, that's really what it is. Right?

Like, having our own Jarvis in our room, in our life that's talking to us constantly. I think we're probably three years away from having, like, a really good version of that, and it's only gonna get better from there. And then, yeah, this human lie detector thing, like, we are really simple creatures

that are creatures of habit, and it's, like, biology that we just, like, can't break away from easily. We have to very consciously try. If we were gonna lie to somebody, all of those tells that you just described, like, those are things we do instinctively,

and you have to, like, train yourself away from it if you want it to be a really good person at deception. Mhmm. And so, yeah, these AIs are already trained in that and they'd be able to spot it really quickly.

And I could just see like this little percentage in the top left of your view and your glasses that says 99% of this person is lying right now. The the the bullshit meter, right? And it's like, oh, they're totally bullshitting you right now. I think that's kinda cool and that will very quickly change just common behavior among people.

If they know that this is out there, it's like, well, suddenly I can't tell lies anymore. And it's almost like overnight,

the public discourse would change and the way that you interact with people would change. If they have the glasses on, you're like, okay. I know I can't try to lie to them, so I've gotta figure out something else to do. Right? We've got three years. Every male who's in a relationship is gonna get in so much trouble.

You just looking at her ass. No. I swear. I wasn't. It's like, 99% you're lying. I'm biologically male. We can't help it. I swear. Like, I still love it. Gonna have those conversations every day. It's true. You look once, you know, if you look once, it's okay. But if you look twice, suddenly you've got a problem. Yeah. You're in trouble.

Yeah. That's funny. Yeah. But no. Yeah. Cool stuff, man. Yeah. Lots of cool stuff that that can be done. Exciting times. Well, listen. I, I won't take too much of your time. You're building some incredible stuff over at Maple. And, yeah, anyone who hasn't already checked it out, go to is it try maple.ai?

Yes. Try maple.ai. I would love a better domain name, but when you're building an AI space, there's just not much left. It's not the worst. Try maple.iai.

That's pretty pretty decent. But, yeah, go and check it out. I love this concept of, yeah, having, like, the numbered account rather than the email and stuff like that. That's really cool. Yeah. Good to see what you guys are building and just appreciate your time and shooting the shit. It's been good. Sure. Yeah.

And and if I could point somebody somewhere else too. Right? I'm not just a person selling a product. If they wanna, like, deep dive more into kinda what I was talking about with, like, this censorship that can happen within AI Mhmm.

There's a website I put up a little while ago with a long form article, and I'm gonna start publishing here more. It's called aiwithconfidence.org, and I posted something there called the free thought manifesto, and it really goes into

how we kinda went through all the social media time frame and how, like, feeds and algorithms were built to persuade us of certain things. And then what can AI do now that just turbocharges all of those nefarious tools that were used against us in social media?

So it's kind of a long read, but it really goes into just like what kind of information are we giving over right now? How can they manipulate us if we don't have visibility into these systems? So I would just point people at aiwithconfidence.org.

I'm gonna start publishing there more often just on some of these concepts about the bad and the good that can come from having open source and verifiable AI systems. Very nice. Very nice. Well, we'll put that in the show notes as well. Yeah. As I say, absolute pleasure having you on again.

Thanks for joining us on the Freedom Tech Friday, and good to get to know you a little bit. Yeah. Thanks for having me on.