# AI Accelerates Rust Adoption and Self-Hosting Trends

**Podcast:** The Changelog: Software Development, Open Source
**Published:** 2026-03-02

## Transcript

We can listen to changelogin friends.
Adam jamming people you know.
Changelogin friends.
It's your favorite ever show.
Well, it is called Finale and Friends for a Reason, my friends.
Today is uh the end of an era, Jared.
Are you sad about this?
Are you happy about this?
Are you are you like just beside yourself?
Can you believe it that next week you won't have to podcast?
Yeah, no.
You don't get to it, which which is it?
Don't get to have to.
It's both, I guess, in a certain extent.
Don't get to, don't have to.
There's something about having to do something that makes you not want to do it.
And then there's something about not being able to do something that makes you want to do it as well.
At the podcast itch.
Well, once you've had this itch for a while, it's kind of hard to stop the itch.
I get that.
It is.
It's so good.
It's fun to talk to people about the things we love, and our technology landscape has evolved and changed, I would say, dramatically over the arc of our relationship, I'd say even like gosh.
Holy cow.
I think uh we first met 2013, 2012.
I would say like end of 2012.
And then 2013, you started contributing.
Right.
And even then, like the the internet was so different.
It was the interests were so different.
Uh and it wasn't like we were besties right away.
We barely knew each other.
So like it took a while for us to to grow and to it had taken a little while.
I was just logging news uh back then, uh news items or whatever we called them back in the on the WordPress blog WordPress blog.
And then after a little while, hopped on the pod.
I just think we called it login.
I don't know what we called it, honestly.
I don't know.
I think you called me a changelogger.
I remember that's cool.
I want to be a change logger.
It was a term of endearment, obviously.
And then we had like semantic debates like, well, are we the changeloggers or are the people who are listening changelog?
It doesn't matter, but you know, these are the things you bike shed when you're trying to figure stuff out.
And things were different back then.
2012.
I mean, mostly it was like framework wars.
You know, the framework wars were cool.
React was out, uh, Ruby on Real wasn't as cool anymore.
I think React had launched pretty much early in your tenure.
Yes.
Um, I want to say what was the fork of Node IOJS.
That had happened uh like early, I would say, I can't remember when probably like 24.
Shout out to Michael Rogers, R.I.P.
Michael Rogers.
Yeah, I was just looking at some uh uh some messages with Michael recently.
And uh for those who don't know, uh Michael Rogers is uh actually I think he was he like helped Isaacs, I believe, if I can recall correctly, like on the early versions of MPM.
Like he was critical in its early, not that it gives MPM any credence this moment because there's some change there, but hey, it begins in great places, right?
Yeah.
But Michael Rogers was uh early contributor, great friend over the years.
Um came down with cancer this last year and uh shortly passed away.
Yeah, passed away like that.
Dear friend of ours when he was sure.
And he was instrumental in the IOJS fork.
And that's when we met him, right?
I remember that show.
Mm-hmm.
And we all hit it off.
That was a great episode.
It kind of served a little bit of a purpose to get that story out and and help, I think, to a certain extent, in a small way, Io and Node get back together and they resolved it.
I mean, it was a fork.
It was a hard fork/slash remerge in record time, probably for the history books.
I wonder.
There was the MERB Rails thing, but this was bigger that because it was more it was more social media.
It did, yes.
Murb was Murb Rails was probably like uh early changelog 2009, I bet.
2008, 2009.
Probably uh Penguin Days.
Yeah.
Yeah.
P-E-N-G-W-Y-N-N.
Uh yeah, that was good for IO and uh IOJS and Node to kind of come back.
I mean, it was Joyant, and I believe Scott, I can't recall his last name of the moment, was the CEO at the time, and we had him on the podcast discussing that change, and really just it was about the proper steward steward of that open source project.
And I think at the time it was so new to folks.
I was like, how do you actually do this in a way that represents the business?
Now Brian Cantro was on this podcast, he was CTO at the time.
Uh that happened.
He could probably speak to some of the uh technical details behind there.
Brian.
If you're listening, let's do an insider on the node I/O.
Uh if you have one, obviously, you probably do.
You have uh an insider on most things.
That's right.
Cool.
From D-Trace to ZFS to Sun to now uh full rack uh massive cloud compute that you own on premise.
So cool on side.
Serious C it has been a journey, man.
It's been a journey, obviously.
Uh I think it's bittersweet for me too, because you know, there's a lot about our relationship that I love.
There's a lot about podcasting with you, I just truly cherish.
And uh it's gonna be tough.
It's gonna be different, it's gonna be tough to do the show without you.
I look forward to it in positive ways because you have to, but also in in a lot of negative ways.
I I think that's uh that's one thing with a relationship that you truly cherish that it's hard to it's hard to walk away from and not do it with grace.
And that's what I wanted to do with us was just make sure that we walked away with grace and uh and dignity in the process because like I love you, man, and I love what we've done together.
And while the show isn't ending, uh you know it's it's uh it's just it's just challenging.
It's the end of a chapter and the start of a new one.
That's right.
I'm sure it'll continue to be great.
And it is bittersweet.
I mean, uh change is part of our name, change log.
Change logo.
That's the hardest name to like only say half of, right?
Yes.
Like some names, some brands you can sort of like uh give them a uh like a nickname in a way.
It's kind of hard to nickname changelog.
Yeah, it's either change or it's log, you know.
That's right.
And then both kind of trying to candle case it or snake case it or hyphen case it, or at least just told everybody uh gap in between change and log.
Is that our probably our biggest shared pet peeve of all time?
Is the uppercase L on a change log, you know, when someone's every time I see it, I'm just like they don't know us.
It's a tell, it's a smell, it's a telling you.
You don't understand how pedantic we are about such things, and you know it's a lowercase l guys.
Come on.
I I think to sweat the details, Jared.
You know this, like you have to sweat that kind of detail.
Like you have to be that level pedantic to to deliver quality at a clip that we have for so long.
You have to have that level of care and detail.
You have to right on.
And if you don't, it shows.
I mean, it's not that you can't do it, just it's gonna show the cracks in a show.
Yeah, yeah.
If you care, then you'll care, you know.
And uh and we certainly care, and that was what makes this hard.
Of course, it's gonna be it's gonna be good and bad, it's bittersweet, as you said.
Uh, I'm excited for a new chapter.
I don't know what comes next for me.
Uh I know in the short term, a little bit of a break will be nice from everything.
I'm planting a little tree farm.
Uh so getting my hands in the dirt, you know, kind of doing some of that analog stuff I've been talking about recently.
Um Back to the dirt.
Back to the dirt.
Ooh, that's a good name, too.
You know, I came from dirt, and I'm going back to the dirt.
Call me Joe Dirt.
Joe Dirt.
Yeah, man.
But we don't need to uh hover on it for too long.
I there's big news that surprised the heck out of me.
I was gonna speak of a crack since you mentioned cracks, or I think we mentioned cracks, somebody mentioned cracks.
I said it was this was I think that the Rust that the that the Rust world is cracking in a positive way.
Ladybird cracked.
Yeah.
And said, you know what?
Swift is actually not what we need.
Even after they did create careful evaluation early on, they cracked and they said, you know what, Rust is actually better.
And I think the bigger, more unique crack here, which is like I kind of expected in this world was the leveraging of AI to get there, which I thought was quite uh, I don't know.
I guess uh at some point it's gonna be everywhere, and it's gonna be expected and anticipated that AI is involved in everybody's day-to-day uh coding adventures.
So it's it's not surprising, but nice to see they cracked and came back to the the good land, which is Rust, which is Rust.
So a little bit of context here when Andreas Kling and Chris War Chris Wanstroth, who has a difficult name to say quickly, were on our pod uh probably last summer, maybe two summers ago.
I don't know exactly when it was, but it's when they announced uh the foundation around Lady Bird and all the stuff they were doing with the money raise and stuff.
Um, or not money raise, what do you saw?
Chris donating, just speaking plainly, that whole deal.
Uh, we asked them about that, and at the time, Andreas said, We're going Swift.
Like Ladybird is written in C, but they're gonna start porting certain portions of it over to Swift, and we released that as a short clip of a longer conversation, and it's become a quite a popular YouTube video where people like to go there and argue about whether or not that was right and why Rust is better and why Swift is better, and why didn't they go with Go?
I don't know.
There's all kinds of comments on that particular video.
And so I was very surprised to see this because Andreas didn't really pull any punches and saying not that he dislikes Rust, but that he didn't think it was a good fit for a browser because of how object-oriented the DOM is and how browsers work.
Take his word for it, not mine.
And how that Rust doesn't really have the facilities he thinks that he would want in order to do that.
And um he still hasn't really changed on that.
He actually mentions it in his blog post about like Ladybird adopting Rust that he still thinks that, but it's just a pragmatic choice.
Rust has the ecosystem, Rust has the momentum, Rust has a lot of other good things about it.
Security, of course, for a browser is imperative and it's so important to the Ladybird team that it was the pragmatic choice.
And that's a curveball.
I was just like, didn't see that one coming.
Yeah.
I think the way they proved it themselves was was less of a curveball when you get into the details.
So I think when you look at the details, uh, they start with lib.js, that's Ladybird's JavaScript engine.
Uh describes it.
Uh Andreas describes it quite well in terms of what it is, but it's got a great uh it's got extensive test coverage per Andreas, which I think that makes it a good starting point because if you've got great tests, it's pretty easy for I would say an LLM to say, okay, here's a bunch of tests that prove all the things.
Let's just use that as a great way to get to the other side here.
One thing he pointed out was there was the result was around 25,000 lines of Rust.
The entire port took around two weeks.
We had noted that the same work would have taken him about multiple months to do by hand, which is kind of obvious to anybody who's been using AI uh coding or agentic coding or an agent.
It's pretty obvious that what you can do in a day or two or a week or two with uh with a tool like that.
But 25,000 lines of Rust, the entire port took around two weeks, and the same work would have taken multiple months to do by hand.
But the the fact they start with a great project from the inside, like lib.js with great testing, that is what I think gives you confidence.
Like, okay, we've got a great code base with great test coverage.
That's a great start to pull in this transition.
Yes.
And it's is it safe to call Andreas a gray beard?
I know he has a beard.
I don't think it's necessarily gray.
The guy's been around.
And his title of this post was uh Ladybird Adopts Rust with help from AI.
And when I logged this in changelog news for Monday, I almost pulled that part out because I I searched for the whole thing with AI, and I was like, well, where is the AI?
I didn't get to that lib.js part fast enough.
He actually specifically names Claude Code and Codex for the translation.
I was looking for AI and I thought, was he just link baiting?
Is this link bait just saying with the help from AI because he wanted to get more attention?
As if he needs help with Ladybird.
It's already a very popular project.
And that being said, I actually did find the portion.
I was like, okay, this is legit.
It makes sense coming from him.
But the thing produced a lot of what he calls like C style Rust, which is totally fine for what they've done so far.
And it was just such a very good use of these tools, especially in the hands of an expert, to see how it made that which is intimidating, if not, I wouldn't call it impossible for him.
He certainly could have got it done.
But just an intimidating project, so much more approachable.
And that's really, I think the unlock for people of all skills, but even for somebody who could sit down and hand code every single line and pro and does a lot of that work and continues to and takes pride in his craft and all these things.
It's like there's huge unlocks here.
And that got me thinking about Rust in in general and maybe its benefit.
I know you've been picking some up a little bit with the help of Claude and other tools, AMP, whatever it is.
Rust's biggest drawback is its complexity and mental overhead, right?
It's hard to learn.
Let me just say that.
It's very terse.
Yes.
It requires a lot.
The compiler pisses you off.
Exactly.
Dead code sucks.
Yeah.
All the things.
You know, you know well.
Even before you put into prod, you know your issues.
That's the thing.
Right.
Compile time bugs better than runtime bugs.
Um, as all of our static friends will tell us.
Um still a fan of some dynamic languages, but I totally get it.
Ruby.
Yeah.
I'm just gonna Ruby for life.
I was writing Ruby before I started changelogin.
I'm gonna be writing Ruby after I leave changelog.
Ruby for life.
That's right.
So beautiful, so elegant, so amazing.
Yes, Elixir too.
So big fans of both of those.
Um I'm thinking that like it's such a win for language like Rust to get mass adoption because now we have tools that help us over all of those difficult parts.
You know, like eventually, I think if you stick with it, even if you're not hand coding, like you're you are code reviewing, you are uh instructing, you are testing, etc., I think you learn it over time.
But I think you learn it without any of the like showstoppers that certain people of stubbornness make it through, but a lot of us just throw up our hands and say, like, ah, this language is not for me.
So I think it really lowers the barrier to Rust, and I think makes Rust way more approachable to way more people.
What do you think?
As it should be, honestly, as it should be.
I think after using it, there's obviously in any language there's rewards, right?
And there's trade-offs, but I think largely the safety around Rust is just bar none.
It's efficiency, the way it handles memory is amazing.
You know, I'm not the super Rust nurse, I can't go into it.
I'd love to have that kind of person here to just go deep into it, but the memory efficiency with Rust is just uh the one thing with Go is you've got the garbage collector, you know.
And so if you have an interface or even like a network interface kind of thing that requires zero latency or very, very minimized latency, almost zero, then Rust is your choice over Go, because Go will introduce those steps, those stops with the garbage collector.
And they're small, but if the byte matters, then the byte matters.
It could be a database, for example.
That's a challenge.
One thing he says though is if you uh I'm in the uh the blog here, he said if you look at the code, you'll notice it has a strong quote translated from CIB.
That's because it is translated from C.
And so I wonder, you know, if that's a bad thing for the long-term code base, let's let's just like spitball this a little bit.
While there was a test coverage, while there was, you know, an existing code base, existing test coverage, why translate from C versus new implementation?
And I know that he was trying to solve a problem.
He was trying to prove the concept probably.
But if Rust truly is, and it seems like they're going to keep writing C alongside of Rust, is what he mentions later on the post.
So it's not like a straight cutoff, like this is not the winner, and they're still in like an interopt kind of scenario where they're doing both boundaries.
Uh, but just thinking about it, like, why not?
What would have been the process, do you think to not translate but rewrite it?
Take feature for feature.
How could you not translate it versus just say, here's this, examine this code base?
We're not translating it.
Here's the code, but here's the test coverage, here's all the things it's testing for, which can be intent-based, which is what you hear a lot, or what you will hear a lot in this new world worrying, which is like specs, intent, uh, you know, document driven, all those are synonymous with I understand what intent is, I understand what the user experience should be, or the DX should be, I understand what I'm trying to get to.
Here's where I'm at.
My intent is X, which is what he could have done here.
What do you think about an intent driven versus uh I guess ported?
Yeah, like a this this way, yeah, where it was translated from C.
Why not inspire by versus translate it?
That's a really good question.
I think that he could answer it better than I could, but my what I would surmise is kind of like when we talk about, you know, what's gonna change here around the change log, and the answer is like as little as possible.
You know, like I won't be here, but like everything else is gonna be pretty much try to stay the same because you don't want when you have big change, you don't want to have big radical change necessarily.
And I think in this case, when you have a big change being introduced into a code base, which is like here comes a new language, and you're proving that out and you're you're taking steps in that direction.
It's like let's do as little as we can, but still make progress.
And so as little change as possible, you're way more likely to get it replaced when you can give the tools the existing code and the tests.
Yeah.
And say, let's not rethink the architecture and everything.
Let's just, you know, here's our scaffold.
Let's pull out this part and let's put in that part and make sure the building still stands.
And let's call that a win, and that will help us decide if this is a direction that we want to go.
And yeah, we're gonna have some C look in Rust code at the end of the day, but we can slowly swap that out as we re as we do rethink and re rewrite, you know, subsections.
Just like you're gonna make little tweaks here and there, you know, to change log news as you go out making it yours.
And so, like, I think that that's probably the reason.
Probably he just got done way faster that way, too.
Because, you know, give Claude and Codecs as much as they can have, and they're gonna crank way faster than having to.
I mean, I guess that's an assumption.
Maybe they won't.
What do you think?
Would they will they do better with existing code than with none?
Uh you know, I don't know.
I think I would uh honestly, because the the cost is near zero or as close to zero as it's ever been to give it a try.
Uh I would say I would try, honestly, because a fresh new code base, and I honestly it might be something where uh, and I'm not sure how much uh experience with Rust Andreas has, but I think as he becomes more confident, the one thing that Rust does for you is it gives you confidence in what you're delivering because of the way it compiles, because of the way that you get past that compile time, it gives you a lot more confidence in the code you've actually written because you you squash so many possible bugs uh in the future with that.
I think a future endeavor of his might be to literally rewrite it from scratch with Rust, but I you know, I think that would only come if they give it a true this is the only language we're gonna use, this is the way forward, etc.
But that's not what they're doing.
I think that I would, you know, because the cost is near zero, I'll try.
But friends, I'm here with my good friend Chris Kelly over at Augment Code.
Chris, I'm a fan.
I use Augie on the daily.
It's one of my daily drivers.
Now I use Claude Code, I use Augment Augie, and I also use AMP code and others, but Augie, I keep going back to it.
And here's where I'm at.
I feel like not enough of our audience knows about Augment Code, not enough about Augie, the CLI.
It's amazing.
I love it.
What can you share?
Yeah, we often say Augment is the best coding assistant you've never heard of.
And that's both frustrating as someone that works there and it's like very proud of the work we've done, but also like inspiring.
Like we want to go and and sort of punch above our weight because like we aren't anthropic and we aren't open AI.
And so the quality of the product itself, you know, with our context engine, once you do touch it, people are like just blown away by that.
And so like that keeps me going every day.
So not to bear the lead here, but this is a paid spot.
You are sponsoring this show to get this awareness.
Now at the same time, we're selective, and I love to use your tool.
But there is in the world, so a lot of developers look at the space and they say, okay, well, how long can this work?
How long is this sustainable in the case of Cursor or Windsurf, or you pick the name and you think discounted tokens?
Help me shape a lens for our audience.
I think it's a lot of awareness, right?
Like um, Cursor got a lot of um publicity early on for like fast revenue growth, which well deserved.
I think, you know, frankly, some of the media got the gets the story wrong, and that like if I gave you a dollar fifty for every dollar you sent me, I'd be the fastest growing startup in the in the valley.
And so when you're selling discounted tokens, yes, of course you're gonna grow very fast, but all that money plus more goes to the model providers.
So I think the real story is the story of Anthropic and you know, being an API provider.
I think the market has just moved so fast, and there's so many pieces of competition out there that it's just hard to get noticed.
So, friends, I love augment code and I love using Augie, and I highly recommend you use it.
I love using Augie.
I can hand Augie a well-defined specification, a well-defined pep, as I call them in my world, an agent flow, and it executes flawlessly.
So the cool thing about August that I love most really is that context engine, and I can hand it a task and it can just churn away on my well-defined plan and just never bother me and accomplish the mission.
It is so cool leveraging the latest models, the context engine, and all the fun things behind the scenes in that awesome CLI.
So, yes, go try it out.
Augmentcode.com.
Right in the top there is a CLI icon, a terminal icon.
Click that, install it, and change your world.
It's gonna be awesome.
Augmentcode.com.
Uh, if you're thinking about contributing or saying, you know what?
I want to try this.
I want to port some of this code.
Not invited.
Okay.
Don't get sad.
Not invited.
Not invited.
It says it says uh we want to be deliberate about which parts are getting ported in which order.
So porting is managed by the core team.
Please coordinate essentially.
Don't do it and waste your time if it's not something they can't merge.
Right.
Paraphrasing that last sentence there.
But uh, you know, I think if you're interested in this port or if you're at least uh tantalized by the effects of Rust and you're seeing it daily in your newsfeed, and in this case, I saw it three times.
We haven't covered all three yet.
Uh, then I think go check out the code base and watch them in action and see what's happening.
You'll see what does it take to go from C to Rust and what's happening behind the scenes.
And I'm sure on their YouTube, they'll publish more content.
But I uh I'd like to see more of this because Rust is the language in that space of the future.
Just is.
You heard it here first.
He's calling just this.
Wow.
I'm just following the lead, you know, just in the stream.
Yeah, man.
Just seeing what happens and just saying it out loud.
No, I don't disagree with you.
It's it's overwhelming.
Uh its adoption and its benefits.
And um, I think he did.
I mean, I think C will be in this code base for years, if not in perpetuity.
I don't think they're going to like ever completely crush it.
And I think he said that even on the show.
It's like, yes, this is more like a direction.
There'll be parts that we exchange or other parts that might even make more sense to just leave alone.
Um my answer back to your why not just try, you know, since the cost is zero.
The only answer I'd have for that is like if it ain't broke, don't fix it.
Like there's other stuff that you can be working on.
Obviously, you can get multiple, multiple agents going, but uh, you know, the price of agents is not nothing.
And it's probably, I don't know, is it gonna go up?
Is it gonna go down?
It feels like so many of these.
There's a lot of blood and water right now between the agents at large, you know.
Right.
There's this whole like claw code subscription thing with open open code.
I don't know if you've been tracking that, but the way that yeah.
So Anthropic, I think, had an amazing idea with Claude Code subscription, which I'm I loved it.
Like as soon as Jose added ACP, is that what it's called?
Agent client protocol.
As soon as I could use my Claude Code subscription inside Tidewave, I was ready to dive in.
Whereas if I had to go out and get a separate source of funding and get an API keys and like spend API tokens, agent client protocol.
That was Zed's thing.
Yeah, Zed started it, and I think that Tidewave is using it in order to use Claude code from inside Tidewave versus a clawed API token.
And it's like I can piggyback that on top of my already existing monthly fixed limit as a guy who likes to know what he spend every month.
I like that.
Like it just makes it an easy adopter.
Same thing with like these new open claw style tools.
If we can just plug in our codecs, we're already paying whatever open AI, we're already paying anthropic, just plug them in, and then it gets rate limited as I have tokens available per month.
It's a fixed cost.
I feel good about that.
Well, anthropic does not feel good about that, so they've been changing kind of the rules around your Claude Code subscription and saying that you can't use it for third-party tools.
You have to use API tokens.
And this has a lot of people not too happy.
I'm in I'm one of them, but I also totally understand it because I mean they gotta get they gotta make some money, right?
They got these huge valuations, they gotta have revenue, right?
Yeah.
When you look at the landscape, this is this is at least where my lens is at.
And this is uh Google has money to burn, and uh they can they can operate at zero margin and profit for a long time.
Totally.
They got a cash cell, which is a different precisely, and so then you have anthropic, large investment, large capital raised, massive valuation.
You see the big number, but the spend is so astronomical.
And then you couple that with opening eye, the same, right?
The the same largest problems.
Yeah.
And you got two key players that can't lose, right?
That's the that's the game of business.
Like, don't lose.
This is the object, right?
Don't lose uh and play for as long as you can, right?
Play forever.
But when you have Google, and they're doing well, I mean, they're doing quite well.
They're very impressive.
They're I mean, just uh they're they're actually surprising me.
I didn't think they would, but they are.
Um, but when you have this kind of scenario where you've got such massive capital raised, and you've got Google Who can essentially operate at no profit for as long as they have to, they can bleed them dry to some degree.
So they have to win, they have to profit.
So when you look at this landscape, y'all, it's not by surprise that Anthropic and OpenAI are going for ways to make money.
And Google's giving away Notebook LM, giving away uh nano banana, giving away things, you know, in in a lot of ways, because they they're the ones with the moat to lose.
You know, the these other folks are essentially new.
That open AI is the incumbent when it comes to the first large language model that was largely usable by mass.
Yeah, right.
But Google's catching up, and they have the dollars to fund catching up, and they have the dollars to lose catching up, and uh flatten the market.
And so you've got those worlds and you gotta expect this.
Now, is it good business for open code?
No, I think our friends over at there, you know, that's not cool uh to do that.
I don't know what that means.
There's also this news about which is not in any of our notes, but this news of uh of what is it called?
Uh distilling the open models, distilling from you know uh a mass infiltration into anthropic and Claude and whatnot, like a massive amount.
This is the same thing.
And the world's kind of applauding this.
I'm not sure.
Yeah, sure.
Explain that what that means.
I didn't prepare for that.
Okay, it's not in our notes.
I could probably do a short version.
I think I know what it means, but I'm not uh totally up to it.
Um, which is that, and this is anthropic is saying that there are organizations abroad, I'll just leave it that vague, that have been basically creating tons of accounts and prompting Claude and extracting a whole bunch of stuff in order to train their models on the results.
And that's what you mean by distilling, I think, like taking the results and turning them into a new model.
And um, they're mad about it, which I understand why they are.
Am I explaining it right?
Is this what you're talking about?
I yes.
Okay, so let's see.
I grocked it really fast because honestly, Grok is the best for real-time news.
Yeah.
Uh let's see if they can actually help me out here.
So I'm gonna paraphrase a little bit of it, but the biggest recent story involving anthropic, this is me reading directly from Grok's response.
So forgive me if there's incorrections or a lack of correctness.
The biggest story involving anthropic, the company behind Claude Code, Claude AI models, and distillation broke just yesterday, which was February 23rd.
This is February 24th, 2026, as we record.
Anthropic publicly accused three major Chinese AI companies, Deep Seek, Moonshot AI, and MinMax, or sorry, minimax, of running large-scale coordinated distillation attacks on Claude to steal/slash improve their own models.
There you go.
Now the world will say uh our fellow friends out there on Twitter, slash X and other places will applaud this.
You know, I'm just not, I'm just not sure I want to applaud anything in this tumultuous world where where there's massive downness, I would say, in a world we want sort of massive upness.
Okay.
I don't I'm not sure I want to applaud this.
I'm sort of in the middle there because they would say that, well, anthropic did the same thing to other things to get to where they're at.
Right.
You know, how did their models get trained in the first place?
etc.
And I get that argument.
But that genie is out of the bottle, and and applauding somebody's downfall, I suppose, because of somebody else's upfall, especially in a coordinated distillation attack.
I mean, would you would you categorize it as an attack?
I suppose so.
They are categorizing as an attack.
I guess it depends on who you are.
I think they said like there was 12 million.
I saw this post from um 24,000 fake fraudulent accounts, if this is correct.
From they were going to slurp down results.
Generated over 16 million exchanges, which are conversationslash props with Claude.
Yeah.
So, like, are they, I mean, are they paying?
If they're paying, then it's like, I'm sure that's a guess that against the terms and they could suspend their accounts.
I don't know.
It's a tough one.
I think that obviously the existing players who have trained their models on the world's information, including proprietary content and the work of artists and the work of uh photographers and just et cetera, et cetera, and the work of coders.
Uh, in the case of some of these models trained on publicly available code, but not uh liberally licensed code, not merely liberally licensed code, and so they're stomping on GPL code, for instance, which we know has happened.
I feel like they there's recourse for them in the United States at least.
These are all American companies in the case of Microsoft, OpenAI, Anthropic, and Google.
They're all can be adjudicated in the court of law.
And you know, if there's any justice in in our court system, which sometimes you get it, sometimes you don't, uh, I think that will play out.
Like are and that and there has been results based on scanning books, buying books, etc.
Um, but when it comes to the world stage, I mean, what do you do when it's coming from a different country or different jurisdiction?
I don't know.
Yeah.
So it's a tough one.
It's murky.
Yeah, it's summarized at the bottom.
It says, in short, distillation itself is normal tech.
I mean, if you look at just a short short tangent, I believe, the way that Airbnb became so successful initially was because they leveraged off of a non-existent web-based scraping API against Craigslist.
Right.
They would find things that they could post.
I mean, that's this is in the history book.
So this is that's a version of distillation, right?
It's kind of like one person's scrappy startup is like another person's attack, you know?
Yeah.
It's like, well, we were just trying to figure it out, and we needed data, and so we scrape some data.
And who hasn't gone out and just scraped some websites for their own use?
I mean, we all have I mean, how do you yeah?
I mean, a hundred percent.
You gotta do it with taste.
Obviously, terms of service is a real thing.
Yes.
And that's what they say here is like doing it at massive scale against a competitor's paid API in violation of TOS terms of service, especially across borders.
This is what anthropic is calling out as a major problem right now.
I understand that, I guess.
It's who's got time to read those terms of service, okay?
Right.
We need the one person, uh just have your agent click on it, okay?
That's right.
All you need is just spin up an agent, call them, you know, clicker, clicker bot, agree bot, and have it do it for you now.
Yes.
Did you actually are you at fault then if you didn't actually click that acceptance?
Something else clicked it for you.
Time will tell.
Time will tell.
It's a weird world as things get increasingly uh agentic and increasingly like the snake eating its own tail.
You know, I think we'll start to see it unravel even more.
I don't know.
Well, that was not at all in our list of things to talk about today.
Well, that's a little bit more.
Well, besides a little bit what happens.
Yes.
Uh let's quickly mention two things more about Rust and move on to a couple other topics.
So I want to mention Ubuntu is also using Rust.
We're not gonna go into detail there, but there's a lot of cool things happening there.
Okay.
Uh, I think Rust has a recent survey out that I don't have notes on, but I want to mention.
There's like a new survey for it.
We'll put that in the show notes.
And then obviously there is some change in terms of compilers, registries, tooling for Python, JavaScript, et cetera.
Yes.
That's now being written in Rust.
And one of the most recent was OXC's JavaScript oxidation compiler, which is a collection of high performance JavaScript tools written in Rust.
Let me, okay, let's just put two, maybe a phrase together, say two words together.
High performance, I guess that is one words if you hyphenated it.
Yes.
Right.
High performance and written in Rust.
Those two phrases will forever, those, those that's the way to describe the next project you do in Rust.
If it is a tool for X written in Rust, it's going to be high performance written in Rust.
And if you just wrote it recently, then it's modern.
So put throw modern on there.
True.
Okay.
And now I mean that that dog hunts, you're selling.
You're selling right there.
So they wrote it.
I mean, they're reusing the same verbiage that uh Astral did with UV and others.
Like the Rust is high performance.
Rust is fast.
Rust is memory efficient.
And just one thing on C the one issue with C is how you have to manage memory in your mind as you're designing and writing the code.
And now if you're using an AI, then you're not writing much of the code to write it, but you still have to manage the memory manually, whereas Rust does it for you.
So that's kind of what makes it high performance is the developer can let go of that concern because the compiler's just that smart.
That was the design of Rust all along.
That was what what the they try to do writing it, but I wanted to mention that quickly on that front there.
You want to touch on that at all before we high tail the similarity?
It sounds like it should be.
It sounds like it should be, but you know who it is from is from Void Zero, which is EvanU's startup, Evenue of View and Vit fame.
And so he has his bona fides.
And so this is from people who know what they're doing.
Umxc.rs.
And so it's the foundation of modern JavaScript tooling.
And so what is it?
OXLint.
There's a linter.
So again, following a little bit in Astral's footsteps there, they started with Rough, which was their linter, I believe.
And then there's OX Fumped, as the gophers like to call it.
OX format.
But in in GoLand, they always like to say.
Yeah, fumped.
FMT.
So there's it's either OX Foomed or Oxfumped, you know, you decide.
Um, which is a prettier compatible formatter.
And then they have a parser, which of course I think these other things are probably built on, and a transformer and a resolver and a minifier.
So it's a bunch of subprojects with uh free and open source.
So this is cool.
This is cool.
How in the world do you build such amazing things on free and open source these days?
It is a conundrum of this new world.
Wow.
I'm so happy about it, but it's also very scary.
In the case of Void Zero and Totally makes sense.
That's what they're calling this.
OXC, yeah.
OXC.
Uh built on sponsorship.
So they have some silver sponsors on the website, bronze sponsors, and then individual backers.
And so, you know, they're going straight up sponsorship style.
And hopefully that works for them.
Yeah.
OX.
Look at this domain, Jerry.
We didn't even mention the domain yet.
Rust.
Rust.
Right.
That's the extension for Rust.
Uh that's a five domain right there.
That's a nice one.
Uh OXC dot rust slash sponsor.
If you are interested, they have details on why the sponsor who current sponsors are, which is seems to be plentiful from individual backers to uh corporate sponsors.
One of a recent sponsors of ours, so kudos to them is Miro.
Miro, actually, I should say.
They told me to say it.
It's hero.
I can't say see the word mi-r o and not think Miro.
It is the same.
Miro.
But it is actually Miro.
Miro, like hero.
Sounds like Hero.
So you all know.
Yeah.
So when you think Miro, I think Miro like Hero.
Okay.
There you go.
Just to be clear, not sponsored, but they have been a sponsor.
Right.
And so that's just baited into my our desire to say things correctly.
That's right.
That's right.
Let's go on-prem.
Let's go self-host.
I think I don't know what you're feeling, Liz, Jerry, but I feel like the the future.
This year, this last year was the was the crack in the earth to say on prem is where it's at.
I think as we uh drive more of our own internal tooling, as we start building more of our own SaaS replacements, let's just say.
Yeah.
Um, you know, you're gonna desire to self-host.
And now is the best time ever to get into home lab if you're not into that at all.
And I think the future is gonna be self-hosted slash on-prem.
And I'm excited about that.
I am so excited about that, as you can probably tell.
Well, I have a little secret I'm happy to reveal, which is that I bought a Mac mini.
I bought a Mac mini.
Oh, wow.
Yeah.
So was it for our friend OpenClaw?
Well, it's so maybe to a certain extent.
Um I've just been building more stuff, you know.
Like, it's just living in a world where you can go from idea to something working in minutes.
It's just hard not to just build kind of everything, not everything you think of, but just more things.
And you know, I just build stuff here on my laptop, and maybe I wouldn't source it, maybe I don't.
You know, it runs here.
Fine.
I end up building stuff that runs on schedules, and I'm just like, you know, my laptop closes.
It's not a server, that's not a server.
This is a work computer, you know, that moves places.
And I don't have all I have the only other computer I have in my entire house, besides laptops for my wife and kids, is a Raspberry Pi that's in the basement that's not plugged in.
And you know, I just three, right?
Raspberry Pi 3, if I recall.
Yeah, right.
Yeah.
And uh still capable.
Totally, depending on what you're doing.
Well then I thought it's just time for some some new hardware.
And you know, everyone was buying Mac minis.
I kinda got caught up, and I told you about my friend that emailed me after or texted me after the la the news about open claw.
And he was like, I just bought a Mac mini, and I'm just like, you know what?
I'm just gonna I'm gonna buy one.
I I I actually bought the exact specs that you said you would buy.
I think on the show that when we talked about it, yeah.
The second tier or the top tier.
It's the top tier.
Ooh.
It's the pro.
Yeah.
I did not go full RAM because I'm not gonna run much local.
I don't know.
I didn't want to spend a lot of RAM, especially when it's basically.
Here's the other problem.
I wanted to play with it, and when you go more RAM, it kicks it back like three weeks to ship.
Oh yes.
And I'm like, yeah, I mean, if I'm having tons of fun, I I don't know, maybe I'll buy another one or something and give this one a cell.
You can bolt on storage, you can't bolt on RAM.
Yeah.
And you could bolt on storage at Thunderbolt 5, I believe, in the Mac Mini.
I I could be wrong.
It could be Thunderbolt 4.
So you've got that full bandwidth.
Anyway, so I'm going on prem at my point.
I'm going on prem for whatever it is that I have I have nothing to run, but um I've been having some fun.
And I do, I did try out nano Claw.
I haven't set up OpenClaw, but I do want to know this weird world all these people are living in and try it for myself.
Zero Claw, actually.
Uh might be the one I set up first because that's rated and way less resource heavy.
I don't know.
But um I'm excited to just move some stuff off my laptop over to there and put it on a scheduler.
That got me thinking, like, you know, Max built-in scheduler is Launch D, which really doesn't have a good interface, you know.
And then just in the shower the other day, I was like, I should just build a GUI for it and be sweet.
You know, like I could do that in an afternoon with me and my friends.
Um, and then it'd be like a cool launch D GUI.
Otherwise, you can use cron, but it's a little bit different on Mac and things get weird, you know.
So yeah, I'm uh I'm all about it, I guess.
We'll see.
I'd like to see that Launch D GUI.
That's actually uh that's cool.
That's cool.
You want to let me name it?
I I I got stuck on a name.
I couldn't think of a good name.
That's why I didn't build it.
You know, you gotta have a good name first, and then you build the thing.
But uh, I don't know.
Yeah, I don't know.
I can nidle on that a little bit.
Well, friends, today's podcast is brought to you by our friends over at Squarespace.
Now I use Squarespace personally.
I love it.
We use it on one of our small businesses.
It's amazing.
But here's one that hits closer to home for some of us.
Maybe you got friends, family, yourself, maybe even.
Maybe you're selling content.
Maybe you do consulting.
Maybe you have the desire to do content and put a paywall in front of it and make some money from different ideas you want to share.
Well, that's the gap that Squarespace fills.
It's the all-in-one platform for turning what you know into a real business.
Not just a website, but a way to offer services, book clients, send professional invoices, and even get paid.
It's all in one place.
Two things worth knowing.
First, offering services.
If you do consulting, workshops, coaching, freelance, speaking, dev work.
Squarespace gives you scheduling, on-brand invoicing, online payments, email marketing.
All of it built right in.
No stitching together five SAS tools, just one dashboard, one way, no chasing payments in your inbox.
You set up your offerings, clients book and pay, and you get back to actual work.
Back to that AI project you're doing.
Second, selling content.
Got a course idea, maybe video tutorials, probably.
Second, maybe you're selling content.
Maybe you have a course idea, video tutorials, premium posts.
Squarespace lets you gate content behind a paywall, add value, deliver on that value, get a one-time fee or subscription, your call, and even recurring revenue from what you already know.
The platform handles the access control for you, the payments for you, the presentation for you.
And all you gotta do is bring the expertise and ideas.
Here's the next step.
Head to squarespace.com slash changelog for a free trial.
And when you're ready to launch, use our offer code changelog to save 10% off your first purchase of a website or a domain.
Again, squarespace.com slash changelog.
Get that free trial, use it, enjoy it, and have fun.
I did think about uh Bill O'Reilly.
So a little request from the audience: if you're a memeer, if you like the meme out there, uh take Bill O'Reilly when he says F it will do it live, and say F it will do it on prem.
Okay, because that's what that's what the world needs.
Is uh is on-prem and self-hosted is the way of the future.
Uh the cloud's still there and it's still great, but I think you know, I just miss you know, like a lot of folks touching things like rack and server.
Uh, like I've got uh off camera, you can't see it here.
I've got one motherboard there, one motherboard there, a case there, uh, an SSD MVME over there, a cooler right there.
You got some work to do, and lack of time, man.
Lack of time.
But I got things that I got two motherboards, two CPUs, a couple of things of build, and uh, I'm just enjoying that kind of stuff.
I love building as you already know.
Yeah.
But I say we'll miss it as if I miss it, but I don't actually miss it because I get to do it frequently.
I miss it in terms of time.
But I'm paraphrasing and mentioning this post on Reddit, this fella, own general 6755 one day ago.
Am I the only one who genuinely prefers on prem over the cloud?
Now, Jared, we use Fly.
We love Fly.
That's right.
Uh we're friends with render, we're friends with other folks as well.
I think there's, I'm not saying those folks are bad by any means.
But I think when your only answer, I think this is what own general 6755 is alluding to, and I'm not gonna say this whole entire phrase, all the all the details in his uh Reddit post on our DevOps.
But it's this fact that when you go into a new job, you go into a new position, or you go into a new scenario where you're stacking new infrastructure, it's largely APIs and services you're point you're kind of piecing together.
And as you know, the GUIs are dying and CLIs are winning in this new world.
And you have to navigate all these settings.
We've recently just did this with keys.
Uh, you know, yes.
I couldn't see the keys that I mean, it's okay.
What's up with that?
You know?
They're called account keys.
Right.
Why can't I not see the account keys you create, Jared?
Help us out, Cloudflare.
Does that make any sense?
What's up with you guys?
Uh Cloudflare.
That doesn't like make any sense.
It doesn't make any sense.
That is, I mean, like the fact that I couldn't see him and you could like they're both.
Same credentials, even like the same permissions, the same administrative level permissions, made no sense.
Okay.
No sense.
And so when you're in that world and you have to like, okay, sure, you have to have keys in on-prem.
You have to have SSH keys and permissions and things like that, user accounts, and you got GUIDS, GUI ID, you got all sorts of things, UUIDs and stuff.
Sure.
Sure.
You have those issues, but that's something you can control on-prem by just SSHing in and Linuxing, the way Linux works.
You know, when you miss this world when you have to build infrastructure on the cloud only, it just you just get to miss it.
So that's all I'll say.
You get to miss that kind of scenario where your only choice to build your new next big thing is cloud.
I say, no, that's not cool.
Self-host it first.
Uh in fact, tail scale funnel your way to success.
What's that?
Well, it's like a tunnel, but it's a funnel.
Okay.
So you could take something, let's say you got something uh running locally on a port, let's say port 8080.
You can uh tail scale serve and tail scale funnel that thing into essentially punching a hole, which actually doesn't punch a hole through the firewall because it's a mesh network and enable connectivity from the outside in and the inside out, for example.
And so you can actually run a service on your Mac mini like you've got, right?
Or on your ain't got time to build it yet, but I'm gonna assume kind of thing.
Okay.
I got other things.
Proxmox is over there, don't you worry, Proxmox is humming along and doing well.
My gosh.
I've never had more on Proxmox in my entire life.
Um, but you can use it there and build there and just tail scale your way to success, okay.
Look into it.
The next episode, actually, next week's episode, uh tomorrow.
Tomorrow's episode?
Soon, very soon.
Very tunnel when it's coming.
Check your feed if it's in there feed.
Go listen.
Very soon.
If not, it's coming very soon.
Uh, is talking about uh identity in tail scale.
What I didn't realize so much so when I did this pod, and it's a slight teaser for it, is how embedded identity is in Tail Scale's networking.
So you larger thing about tail scale is this glorified VPN.
And it is that.
Yeah.
But think of it like everywhere you go in the network, if you're authenticated, you are you.
You got things like OIDC and things like uh OIDC connectors and being able to just already be logged in to uh true NAS or already be logged into Proxmox or pick your internal service that can support OIDC because your tail scale network, your tail net, when you go about it, you are you, and so you are you as your identity as well, which I think is super super cool.
Anyways, on premise of the way.
That's all I'll say.
Tunnel and funnel.
I like it.
Tunnel and yeah, it's not a tunnel, and we didn't mention that net pod.
I was like, you know, I thought it was a misspelling because tunnel makes more sense, but I think Cloudflare D has tunnel while Tail Scale D has funnel.
And I don't think it's actually called Tailscale D, but it is Tail Scale.
I mean, that's effectively like what NGROC was always doing for us back in the day, right?
Which is like open up yeah, expose us now.
Does that require some a version of it?
A version requires you to port open on your on your firewall, whereas Tailscale does not.
Okay.
I did not know that.
Yeah, there's no port opening whatsoever with Tail Scale, and not an ad, but they are sponsoring us very soon.
Uh, and you know I love them.
Uh, you can go get a free count today.
They're free forever.
That's their that's how they are.
And so if you're a home labber or somebody has sent you with a hundred or less devices, Jerry.
You gotta have one.
Basically free forever.
So you can home lab to your heart's content to 100, and they're happy to have you.
Happy to have you.
All right.
So you should use it.
Quick side quest here.
Yeah.
In what calendar year do you suppose you, Adams Takoviak, are gonna hit that?
You're gonna have 101 devices on your tail.
Like, when are you gonna when are you personally gonna hit that?
We're talking like you're you're getting close?
We're talking like a decade, never.
What do you think?
Let me log in real quick.
Devices are proliferating on our networks, you know, sometimes before we even know it.
But doesn't mean we're necessarily running tail scale on them.
Yeah, tell us what you got.
What are you working with?
Let me tell you uh how many are here?
54.
Oh.
And that's only because there's no cleanup.
Okay.
You could probably get that down into the 40s.
Yeah.
Um let's see.
I haven't announced it yet, and I'm not sure when I'll actually open source it, but I'm building a self-hosted GitHub runner.
Because I was just tired of how slow GitHub runners are.
And it's called.
Uh yeah, they do.
Yeah, they they do.
It's it's uh well, they may change the ball.
They may change the the playing tool that front there.
Yeah.
Uh but for now, you can build an app and connect it to GitHub and build your own.
Now, I'm trying to do this, and so the reason why I have so many is because it's in test mode uh and it's got a ton, it uses tail scale for the networking to connect to things.
And so because of that, it's got a lot of test machines that are actually not active.
So these are all dead ones part of my test.
So truly around 20 devices across my network, but in actuality around fifty four per the count right now.
And a lot of that's just crupped from so what's the calendar year you're thinking?
Like you're you think you ever hit that at your at your home office slash home lab?
No, I think honestly, I'll probably I'll probably will hit it within this next year.
Okay, but at that point, it's it's propping up either an open source project that has a business attached to it or could have a business attached to it.
So I think at that scale, if those were truly sustained, true devices on the tail net, uh that's not a home lab anymore.
I mean, I think I don't know, maybe actually home fraud.
A home lab with a hundred devices?
Somebody out there's calling foul.
They're like, Adam, I got ninety-nine, man.
I got to do it.
Somebody's out there with like in the world home lab, I'm not really sure.
Yeah.
I'm not sure I will personally ever have a hundred.
Now, if I had new projects I'm building, and each one's self-hosted in my home lab and they're self-serving, that's a lot of software, man.
I don't know.
That's tough.
You want to talk more about this.
It's just impossible essentially to hit that number.
You want to talk more about this runner?
I mean, I'm intrigued.
So this is like basically self-host your GitHub runners.
Right?
Yeah.
That's so let's see what I can share now.
So there is a a uh I guess newer to me, probably less newer to others out there, is a new thing called Incas.
It's new to me again.
That's why I'm saying new.
Incas.
Incas is I'm on their website, uh Linuxcontainers.org slash Incas.
Incas is a next generation system container application container and virtual machine manager.
Uh some history behind this, which I'm loosely uh aware of, is a canonicals LXD.
It was Canonical's LXD.
And the the the fellow behind this, his name is Stephanie or Stefan.
I'm sorry, not Stephanie, Stefan.
Uh short named just a Steph, I believe.
And I could be wrong on that front, but I'm trying to get them on the pod uh in the near future.
But this is the underpinning.
So when I looked at Incas, and because it's uh it supports like any Linux, ZFS, copy on write, you know, it's gonna be really easy to build on top of Incas.
And so I'm I'm building Turk right now because I needed to build some let's just say some stuff on GitHub through runners, and they're super slow.
And we've known this because we've been sponsored by Depot and sponsored by namespace or users of namespace.
And this is well known that GitHub runners are uh just slow.
They're not purpose built uh runner platforms to run faster.
And so the runner part of it, so you have GitHub Actions inside of your project, you obviously have.github or something like that, uh, and you have slash workflows.
And inside of those workflows are things will run on GitHub Actions.
And so when you push your repo, those live there, and they can run on schedules, you can hit them, they could do your releases, you can tag a release, it'll automate that stuff.
And uh, and so then that instantiates GitHub Actions, which is essentially a logic platform, and if then else builds Linux, a bunch of crazy stuff.
You could run a runner that just checks things, kind of like Cron in a way, too.
Uh, but then it it it uh it dumps that off to their own actions runners, which as I mentioned, they're kind of slow.
But you can also do self-hosted runners.
Okay.
And so if you just want to build your own simple self-hosted runner, you can do that today.
But the problem is that I've found, and the reason why I went this route was because they give you temporary keys, they expire in 90 days, they make it painful.
They basically make it painful to self-host your own runner, and then you also have to do it per repository.
You can't do it like I'm Adam Stack or I'm uh you know, slash the change log and put it at the org or user level.
You have to do it one by one and the same painfulness on the key expiration one by one.
And I'm like, well, how did Home D uh not Home Depot?
Where'd that come from?
How well depot, how did Depot do it?
And so I sort of look into this, and it's an app.
You build a GitHub app.
Gotcha.
And so I'm like, okay.
Uh and my goal with Turk, it's called Turk, because our friends back in Amazon, they used to have Mechanical Turk, right?
I was like, man, the coolest thing for this.
Now I'm gonna reveal this now.
If you go there, you're gonna see nothing because this is like literally brand new.
Turk.run.
Okay, that's the coolest domain ever.
That's cool.
Turk.run is what it's gonna be.
I'm still not sure on the licensing, but I'm thinking about building this uh into something that could be commercial.
I'm just not sure yet exactly where it'll land.
But the licensing battle on open source is a struggle because if I give this too permissive of a license, it's easy for all these existing runner platforms just to use it.
So I'm I'm torn there.
I could use some guidance and some advising on that front.
But I'm thinking, gosh, man, don't hit me.
Don't don't hate me for saying this, man.
What?
I'm thinking some version of uh what do they call that?
I'm not even sure what it's called.
What it when you have open code but not open source.
Open core.
Oh, open uh business source, perhaps business source, but source available.
Source available.
That that infamous uh phrase we've said a few times.
So I'm thinking source available.
And the the call we had with uh the SDK folk, that was what really gave me a lot of insight into this world of like okay, you can be source available, respect the world of open source, but still build a business on top of it.
So I'm just not sure yet.
I'm just not sure yet.
At this point, it's a fun project I'm building internally, yeah, that has some long legs.
And uh I'm using it in prod.
It's actually right now building DNS hole.
So that was actually the most recent win I had was let me actually dog food this thing.
And so DNS Hole is a private repo right now, uh on the DNS hole user.
So if you go there, getup.com slash DNS hole, uh, there is a an org there, but nothing for you to see yet because it's coming soon.
Um, and the biggest issue for me was being able to build this thing, not in a VM locally and a couple other things to kind of get it over this finish line and time.
And time, really.
Sure.
Uh and so Turk is what I built to do that.
Because I was just like, I want to self-host runners because GitHub GitHub Actions runners are just so slow.
Went down that rabbit hole, built Turk, and now I'm actually building DNS Hole with Turk.
And so I'm just so close to like two cool things out there happen.
Well, that's exciting.
I'm sure you'll let all of us know when these things go out there, right?
I mean, ASAP.
Why not?
So soon, so soon.
I actually thought about a uh an image registry as well, because uh another project I have has to have a custom Linux image.
And this threw a curb bought me because I was like, I was not expecting to have to build a registry for this thing.
Uh, but a a Turk image registry is the next thing.
I just started to pin the spec on uh on the Turk image registry.
Way uh way off on left, but yeah, coming soon, Turk.run for now.
Well, just know that I'm building it.
And if you are interested in the if you believe that self-hosted is the future, which is why I'm which is why I feel so strong about building it.
Because I'm like, this can solve our friends' problems, this could solve some enterprises' problems, and I want to build it right to solve my problems, but also think of it in a way that's reusable by the world.
And so I'm trying not to make it an internal project, which is kind of easy in this world, and just say, okay, it'll be internal.
I won't do anything with it.
But I really want to build it in a way that's usable by me.
I want to prove it works for me, which was DNS hole being built with it, and then being able to take it to the to the masses.
So that's the next step.
That's cool.
Very cool.
Should we move to our next story?
There's more self-hosted.
I think we'll gloss over a couple.
You got so DH.
This is not in the list, but he's talked about you know moving away from the cloud for a while.
Uh Hetzner's prices are up 30 to 40 percent.
I think this is largely predicated on the fact that RAM and CPU and storage.
Mostly RAM is just astronomical.
Jared, I think I told you this recently.
I bought four sticks of RAM for like 200 bucks recently.
Uh four.
Two of those is a thousand.
The four and dim problem.
Back to that, yeah.
So you spent 40,000 bucks on them?
I didn't spend two grand.
No.
No, I spent 200 bucks on four.
Okay.
Those four now are worth two grand on the market.
Good time.
My gosh.
Like, I think the last day the prices were low.
I think.
It's like I was surprised.
I'm like, I just bought RAM.
What is it?
This can't be true.
Troll those ones.
Craigslist, man.
That's a good that's a good markup.
I got I got self-hosting to do, bro.
I can't do that.
Those things are too valuable to me.
Now I'm like, oh MG, I'm so glad I bought them when I did.
Yeah.
Um yeah, and I think nanoclaw was the other one I'm I wanted to put in the list too.
Is gotcha.
Uh nanoclaw move from Apple Containers, which I'm not that familiar with.
Yeah.
But I'm about to be because of Turk.
Uh okay.
And then because part of Turk is gonna be built on built on Mac as well, not just Linux.
Interesting.
But actually, to automate Apple containers, which is cool.
So I did download Nano Claw and I used it.
I think maybe I told you that we were talking about Apple containers, you and I, weren't we?
Yeah.
Recently.
Yeah.
And so that introduced me to Apple Containers, which is uh Apple's first party project.
It doesn't ship with Mac OS.
You'd think it would just be in there, but it's not.
There's like it doesn't?
No.
Bummer.
It's like a container project on GitHub.
And you go out and like click releases and download a DMG.
It's very manual and weird.
Ugh.
Maybe it's just an experiment and they're not gonna like want to support it long term.
I don't know.
You never know what Apple's up to.
But it's basically, you know, like Docker for Mac without Docker.
It's like an Apple version of what Docker does for Linux, uh, for Mac.
Has very similar even command line uh UI and stuff.
And you know, it's fine.
And I used it because nanoclaw was built on top of it.
Because the author of NanoClaw wanted to do that, and I wanted to try something out real quickly on my laptop.
It was cool.
I got nanoclaw up and running.
The problem with it was he only supports WhatsApp.
Oh.
And he also doesn't take pull requests.
It's kind of an interesting project.
Like the way he runs it, I thought it was fascinating at first, but then I thought it seems actually impractical, which is that you do not add features to nanoclaw.
You add skills and you have your agents build features on your personal cloneslash fork of nanoclaw.
Yeah.
And so it's basically like if you want the feature, you can add a skill that like teaches will teach all of these things to use it, but he's not going to accept any PRs that says, you know, let's support Discord, let's support whatever Slack.
And I thought that's an interesting model.
And then I thought that'd be very fragmented and very difficult to actually stay up to date with nanoclaw if you have tons of modifications.
Like everybody has tons of custom features that are never going to be merged upstream.
So I don't know.
Seems kind of impractical, but this is the same way that uh OpenClaw went too.
This is uh Peter Steinberg.
I was actually, he said this on the 21st on X.
He said if you were a service, I think it's also because he's getting inundated with like a lot of inbound, right?
Yeah.
Uh side note, Peter, if you're listening to this podcast, respond to my DMs, man.
Okay.
Now we were actually it's funny because just before like the the uptick and the major craze, like there was a lot of uh inertia around Claude Bot at the time.
Now OpenClaw and all the names it has been.
There was names even before ClaudeBot.
Uh we were talking and trying to coordinate timing.
And then he got hit with that DMCA from Anthropic, which I think was just a little side tangent.
What a bonehead move, right?
How do you not see that this was an opportunity?
And then you slap the guy with a rename and a DMCA.
Yeah.
You know, like that was just like how did you lose the trees, you know?
You know, like you're basically bringing a bunch of attention to a thing that otherwise had not that much anyways.
Yeah, absolutely.
And not only was it a cool project, but somebody who had a following.
Yeah.
Like you don't do that.
That's just not very fastest growing repository in GitHub history.
Yes.
That's well deserved, too.
Buddy Buddy with, you know, saying he's a genius.
He's a genius.
And now your competitor, Open AI, employs him.
That's right.
So I feel like Anthropic did drop that ball.
Oh gosh, man, they botched that one hardcore.
That's a such a sad moment.
But what he says here, and this alludes to what you're saying about Nanobot.
He says, uh, if you are a service, make a skill or plug-in.
There you go.
Not gonna add random features or skills to core for visibility.
Yeah.
So I think this is the same kind of stance, but I think the nanoclaw guy is like way more uh hardcore on it.
Like his is way smaller already, right?
He's still not gonna add anything, so it kind of makes it where it's like I'm gonna go use either the real thing or something that's closer to it.
Um, and so I've I'm planning on not continuing with nanoclaw.
I don't use WhatsApp, you know?
It's just not my ball of tricks.
So his name is Gabrielle Cohen, to my knowledge.
Got a DM out to Gabriel as well, nano block creator.
So Gabriel, come on the pod and share your thoughts on nanoclaw if you have them.
Nanoclaw nanoclaw, sorry.
Change that name.
Nanoclaw to be super clear, nanoclaw dot dev.
In one sense, it's like good for maintainers, you know, open source maintainers, because they're just like, you know what?
You know, it's kind of the old uh Ben Johnson, open open source, not open contribution.
The instigator of all this stuff, I think.
Open side.
Trailblazer, wasn't he?
He was, man.
It's kind of that.
It's like, yeah, we're not taking PRs, but you can build a skill and you can just fork it and have your agents build whatever you want to.
And I get it.
I think that's a it's an approach.
I think you have a hard time fostering a community around that, but maybe I'm wrong.
Yeah, I know that open claw has not had a hard time fostering a massive community.
Or did I say the right one?
Yeah, I did.
And they have, I think, two orders of magnitude more source code than nanoclaw.
It's like 400,000 versus 4,000 lines.
I don't know.
So different approaches, different results.
Yeah.
Different approach than results.
I I think that uh you know, as a maintainer of an open source project in the world of agents, you are being inundated with pull requests.
Yeah.
You know, the side note, even on pull requests, and what's your stance on code review and pull requests, Jared?
I know you've traditionally just not been a fan necessarily, but like, do you think it's a dead thing?
Do you think code review is dead uh in the sense of the word we've known it?
Do you think pull requests are a version of dead?
Well, I think the the old phrase the future is here, it's just not evenly distributed is so true right now in so many different ways.
And I think code review is gonna continue to live uh in different places for very long time.
Even our good friend Amel was like death on no code review.
And we started talking about just like basically vibe coding and you know, merging, vibe merging stuff.
I mean, sh in her domain, that was like anathema, basically.
I think that'll be the case in a lot of places, um, especially high-stakes software for a long time, but I do think the way the wind is blowing, and I did give her, I think, my own experience of change, and I've had a similar experience with self-driving cars, where it's like at first you're like super nervous and you're criticizing everything they do, and then like a couple weeks later, you're just like, Yeah, I watch it, but it makes the right decisions like most of the time, and then a few months later, you're like, why do I have to even be watching the road?
Because this thing is darn near flawless.
And I do watch the road, but it's so good that I'm just like, I'm not checking on it as often as I as I would.
Now there are moments when you do, of course, and like high-stakes traffic, you know, or dangerous scenarios, yeah.
Weird one-offs where you're like, Yeah, here comes an emergency truck, and etc.
Now I'm starting to pay more attention.
But I think that is the trend line for code review from agentix coders.
Um, unless the the advancements in their ability to write software stops right where it is.
Like if it stops right here, code review is gonna exist forever.
But if it continues to advance, I think even as a longtime software engineer, you just stop watching as closely, and you're thinking at a higher level, and I think eventually, if it continues to advance, we all do that, and in that sense, code review dies.
That's my kind of dual-sided take.
Maybe a cop out because I I answered both ways, but that's what I'm thinking.
Yeah, I think it's gonna evolve.
I think it's gonna be code review in the sense like you push a PR, someone, something reviews it there, is gonna become an evolution.
I think we'll care more about code quality than code review, which is really the purpose of code review.
Yeah.
Is what what changes?
What's the implications of this change?
Is it massive?
And so in the sense of human review, massive change, massive PRs have been largely frowned upon, where smaller changes are easier.
That's where you get the you know LGTMs on a massive PR versus a small one, which gets scrutinized to the N degree, right?
Oh my bike shed it.
Uh because it's so much easier to reason about, right?
Yeah.
And so I think you have this idea of reasoning about the code change, and you care about that.
I think we'll always care about some version of code quality, but that's gonna change.
It's gonna be about who.
It's the same thing with shift left.
It's like, we sure shift left, but now it's just like more on the developer.
How much further left can we actually go with the shift left?
Same idea there.
I think you'll have more to reason about to think about, and you'll always care about some version of code quality.
But the purpose of caring about code quality is is the intent met?
Does it pass tests?
Does it actually work in you know, in development environments?
Does it pass the smoke test where we spin it up 50 times and battle test it?
And that's the point of like where you can do Turk type things where you can just spin up a ton of runners because they're more free in a self-hosted world.
So like you wouldn't spin up 50 runners on GitHub to test because like one, uh potentially super expensive.
Uh two, my gosh, super slow.
Okay.
But in the world of Turk, where you can like self-host and infinitely test on you know, easily clinical VMs or containers, system containers, like that's a world you live in.
So you can you can ensure code quality at a much higher degree.
And I think AI will continue to evolve this and we'll care about what has been code review.
I think will translate into some version of code quality.
And I'm actually wondering if the next big thing on this front isn't some version of a code quality gateway.
You see this happening in AI where you have AI gateways where you want to gateway the AI.
This happened networking, right?
You have a network gateway or something like that where you can log into something.
So like open router is a good example that you're talking about, right?
Exactly.
Something that lets you access more the network, you know, it's a is a gateway.
I think I have more of that in the code quality, and what has traditionally been code review is gonna be more in the real time.
I want to know just in time, not before I push and hope and pray.
But you know, I just built this feature.
Does this meet my intent?
How does it work with others?
Like, I want to see it on my machine, not push to you know, some place and and in hope and pray.
I think things will change on that front there.
I'm not sure exactly how, but I'm thinking code quality over code review.
That's where I'm leaning.
Two more things on our list.
They're both kind of deep.
Okay, and I'm not sure how deep you went with them because I shared this list with you moments before the show.
I was just gonna ride your coattails, which is you know okay.
Uh dangerous, but I do like it live a little bit dangerously.
It could be dangerous.
Okay, so I'm gonna I'm gonna loosely mention uh Boris Tane.
So Boris Tane's gonna come on the pod soon.
Uh the invite has been accepted, the plan has not been made, so there's not a date on the calendar yet.
But one thing he said recently was the software development lifecycle is dead.
You will often hear the term, Jared.
If you haven't, this is new to me.
I learned it recently, SDLC.
I'm like, what is SDLC?
Oh, software development lifecycle.
So only if you sell SDLC type products do you know SDLC acronyms, right?
Right.
Because you don't want to keep saying software development life cycle.
It's a lot.
Right.
It's a lot.
Yeah.
So I get it.
I get the acronym, but for a while there was like, okay, what does SDLC mean again?
Okay, it kind of makes sense.
But he proposes, I should say purports that uh the SDLC as you've learned it is a relic.
He shares, let's see if we could throw some video up on this, but he shares a classic software development life cycle where we're taught requirements, system design, implementation, testing, code review, hello, deployment, monitoring, and recycle back to requirements again.
Like this has been a known loop.
And he goes on to suggest how the uh every stage is collapsing, how AI is really collapsing a lot of this.
And it's not because AI is eating it, it's because more of what happened in the SDLC is happening more at once and in real time versus in different disparate places.
And I think that's why I really care about just in time code review as well.
And it's not really code reviews, just more like, dude, is my code okay?
You know, you know, like I want to I actually want to software call that dude is my code okay uh com.
And then that'd be kind of cool.
I want that, right?
Uh, because I don't really want code review, but he kind of goes on through this.
And I think that kind of leans into the fact that we just talked about code review and code quality.
Yeah.
That's gonna that's gonna be the thing.
You know, it's gonna be more this is getting compressed, more this is getting uh replaced and compartmentalized in a way that is just like it's just collapsing.
A lot of this is collapsing.
And I I kind of mourn in a way the old way, but I don't, but I do.
Like I really I was watching how it's made recently, Jared, and there was this fella, and he was making uh um, let's see, like you can make tortillas and stuff like that, and you steam them and you want them in this thing, you want them to be steamed.
Uh, or maybe it's rice.
Uh it's this basket.
And so he takes these four bamboo trees, and you watch him go from four bamboo trees to many baskets.
But it was not, it was not like minutes, it was hours, potentially even days of labor to go from multiple bamboo trees down to you know, baskets.
Right.
And I think that's like serious craftsmanship to to handmake those things.
And that's what we did in the old days, which was like basically last week in software.
Yeah, you know, is we used to make software by hand, we used to hand craft some of us are still doing it, some of and some of us are are still doing it, but I think that world is so changing, and then you know what Boris Tane is suggesting here is collapsing and changing, is is part of that, and so like you have said before, the the future is here just not evenly distributed.
I think maybe that's the case for most folks.
Maybe the SDLC is not dead for everyone, maybe it's just dying or changing or evolving.
The word dead is just thrown around so easily, right?
Well, it's sensational, right?
Sure.
I mean my attention, right?
Yeah, you want to get attention.
You don't say that things are changing, you just say, like, this is dead.
That's right.
And that's how you get attention.
So uh yeah, I mean, I've always looked at the SDLC with disdain myself, you know.
Um, probably because I was forced to learn it in college, and anything you're forced to learn in school is lame, right?
De facto lame.
So S DLC was not new to you.
This is a a uh textbook stuff.
Like they teach it took you in the textbooks, you know.
See, um, and then you come out of school and you're like, I'm not doing that.
That looks rigorous, you know.
You know, Gerhard would do it, he would love it, he loves rigor.
Uh, the rest of us in this room, not quite as much.
Um, it's hard to love rigor as much as Gerhard.
The guy loves it.
So I always didn't like the SDLC.
Of course, it gets kind of associated with waterfall, but you I mean, there are all these steps, and I always have felt like it was like the way that it was described to me and the way he's showing it as this like step-by-step process.
You're assuming each of these is like a massive undertaking.
I always felt like it should be smaller, tighter, and loopier.
And he's basically saying it is it's getting that way.
Um, so in a sense, I think it was always should be less uh serialized and things happen at the same time, and less people doing more things together at the same time.
And um, now we're just taking the people out, you know.
Like that's the big difference now.
Is like it's you and an agent, and you're not waiting on design review or whatever it is.
You're not waiting on DevOps team, which as our as our friend Ellie Hustable pointed out, as soon as you have a DevOps team, it's not like DevOps has failed, basically, because the whole point was to you know integrate the dev with the ops and have them all be together.
So, anyways, right waiting on some other org, waiting on some other person, you know, waiting for your code to compile like the old comic where we're outside in the hallway.
Sword fighting, sword fighting, yeah.
Waiting for my code's compiling.
Okay, cool.
Exactly.
Yeah.
Carrying easy.
I understand where your horse around, it totally makes sense.
You could do nothing else.
Right.
Like, why are you guys not working?
It's like, well, we're waiting for system design, you know.
We shouldn't be waiting for these things.
And I certainly think that it's been collapsed for a while, and it's just getting tighter and tighter.
Well, friends, this episode is brought to you by my good friends over at Notion.
We use Notion here at Changelog.
I love it.
It is so helpful.
It keeps me on track.
Notion agents, Notion AI, Notion, all the things.
Here's the thing.
You probably spent weeks building out your team's knowledge base.
You probably wrote the onboarding docs by hand, the run books, the architecture decisions, the how we deploy guide.
Of course you did.
And it's all Notion, organized, searchable, beautiful.
And your team still DMs you.
Hey, quick question.
Except it's never quick.
It's not just one question, it's 10 questions.
Every interruption costs you 23 minutes of context switching.
Your productivity is going down, and your carefully documented knowledge base collects dust while your DMs are on fire and stack up.
Well, Notion is an AI powered connected workspace for teams.
It brings all your notes, all your docs, all your projects into one space that just works seamless, flexible, powerful, and it's actually fun to use.
With AI built right in.
You spend less time switching between tools and more time creating great work.
And now with Notion's new custom agents, the busy work that used to take hours or actually never get done at all, it now runs itself.
Custom agents automate time consuming repetitive workflows directly within your Notion workspace where a team already works, where all your knowledge base already lives.
Here's what makes them different from most AI because most AI still waits for you to prompt it.
But custom agents run on schedules and triggers.
One person sets it up, putting a workflow for repetitive tasks like weekly reports or triaging feedback, and it becomes a shared resource for the whole team.
Think of custom agents like AI teammates with a specific job.
Whether it's a status update agent, a QA agent, a task routing agent, or a custom agent you've just built for your workspace.
The busy work is now handled.
And don't just take my word for it.
Notion is used by over 50% of Fortune 500 companies.
And some of the fastest growing companies like OpenAI, Ramp, Vercell, they all use Notion AI every single day to help their teams stay ahead.
So here's the next step.
Try custom agents now at Notion.comslash Changelog.
That's all lowercase letters, notion.com slash changelog to try custom agents today, right now.
And when you use our link, you know you're supporting the show.
And we love that.
So Notion.com slash changelog.
Who was it that uh taught us about cues?
His name is is uh is missing on my brain in this very moment.
Dear friend, he loves process.
What else was he talking about?
Been on the pod before.
Gosh, man.
Oh, anytime you're queuing, like you're not efficient, kind of a thing.
Yeah.
Yes.
I want to say Barrett Hubert, but it was not Barrett Hubert.
It's in that same wheelhouse.
Lucas De Costa, I think, wasn't it?
Yes, Lucas DaCosta could call him.
Yeah.
Yeah, Lucas.
He would talk to, I mean, that was really where I was.
I mean, I knew that.
So it's funny whenever you know something, but then you don't know it until somebody gives it a name and defines it for you.
Like, you know what?
I've experienced that problem 17,000 times.
Okay, now I know it.
So I feel like I learned it there, but I experienced it for a long time.
But just this cue process, like the when you have a queue and a backed up cue, that's an inefficient system.
That's obviously not a good thing, right?
If you can solve the queuing problem, then you solve a lot of the bottleneck.
And you know, one thing that uh that Boris suggests here is that he, you know, going back to CodeView in a way, or I guess directly, not in a way completely, he says, um, you're gonna like this man.
Give me one moment.
Where is it at?
Sure.
Code review.
Give it up.
Okay.
Think about it.
An AI agent generates 50.
Sorry, at a zero, 500 PRs a day.
Your team can review maybe 10 review queue backs up.
This isn't a bottleneck worth optimizing, it's a fake bottleneck, one that only exists because we're forcing a human ritual onto a machine workflow.
And I believe that's true.
Like in the SDLC as you've learned it, and as we've all experienced it over our years, what is the point of the SDLC to ship software?
Yeah, right.
And we've ritualized and held sacred these steps along the way because we've had to.
And that's been our comfortability.
So as humans, we have this comfortability gate.
Well, we do this because that's what gives me confidence and comfortability to go to the next step.
But if that's not required anymore, or in these scenarios, if you're still doing it by hand, then obviously this SDLC as it has been is the way.
But if you're moving to this new AI agentic world, it's just not.
It's just not.
We'll link it up in the show notes, but uh Boris Tane goes into a lot of detail there, worth checking out.
A couple of good comments I could pull out briefly before we switch context here.
Uh, because man, the 17 comments and they're pretty high quality.
So yeah, definitely dig into those comments.
Uh one Kevin said, I'll read the comments, man.
Comment section sometimes is the best stuff.
I meant that in the I do read the comments, but you're generally like, hey, don't read the comments.
Uh so one Kevinson says a really high quality article, blah, blah, blah.
I generally agree everything you've described here, but now I'm wondering what's left of the software engineered job.
So as everything gets collapsed down, it's a really good question, of course.
I think that's what on a lot of our minds.
And Boris's response, I thought was pretty good.
He says everything that isn't typing code, typing on a keyboard was never the hard part.
The job shifts from person who writes code to person who knows what to build and why.
Taste, judgment, knowing when the agent is confidently wrong.
Yeah.
Nothing new there.
We've been saying that kind of stuff, but I think it's just I think that's on point.
Yeah.
It's it seems.
I mean, even Scott Hansman comes to mind in this scenario here.
Like when he says person who writes code, and in particular, typing on a keyboard.
One thing he told us in that show we did probably eight years ago when I say Jared.
Wow.
Was he said he had so many keystrokes left in his life?
And that's that stood out to me really as a very wise thing to think about.
And he he judges.
I think he had a even a program that that's a good idea.
There's a website you can go to.
Yeah.
I knew he liked it in a way.
Like it was so it was important enough to not just say it on a podcast, but to actually put something behind it to um to quantify it for himself.
I think about that.
Like I you know, I would love to, while I'm a craftsman at heart or craftsperson at heart, I want to be able to make the best thing.
I have taste, I have judgment.
I think, I think, Jared, I like to know when the agent is confidently wrong.
Yeah.
You know, that's questionable these days.
I'm like, what the what did you say?
Right.
What did you just say to me?
That really is the hard one, right?
Because you need to know more than it knows and you don't in order to don't.
Yeah.
And a lot of times you're like, well, I'm up, I'm in over my head.
Um the tooling that I've been building, it's all pretty simple.
Now I've been building that Mac OS deal where it's like, I don't know, Elika Swift, I don't know the UI kit, Apple, you know, uh it's on SDK APIs.
And so I'm just like, I'm swimming in deep waters when it comes to like being like, nah, that's not the right way to do this, because I literally have no clue what's the right way to do it.
Now, when I write node-based tools and Go-based tools, and I'm looking at some JavaScript and it's kind of streaming by while you know Claude's doing his thing, I can be like, Yeah, this is looking pretty gnarly.
I bet we could rewrite this and I'll I'll talk to it.
But yeah, how do you know when you don't know?
That's kind of scary.
You know, that is scary, but then I think it's gonna get so good that even when you don't know, at some point it catches up to fix itself.
At least that's what the Thomas side.
That's the hope is it gets better from here.
I mean, it's gotten way better in the last 12 to 18 months.
So tell me if this is your flow with Swift uh and it in an Xcode.
Are you just uh command Ring and uh getting the next build?
Or are you actually doing the less stuff in Xbox?
Less than that.
I'm not even running X code.
In fact, when I first started, Claude told me to go ahead and launch S code and here's how you compile and run, which is like, yeah, command R, hit the build button.
And I said, I don't want to run X code if I don't have to.
Do I have to run X code?
Can't you just write some sort of a build script that does it for me?
And then it was like, yes, you're absolutely right.
I can write a build script.
And so I had to install Xcode, but I never have to launch X code.
And so I just tell it to I just tell it to rebuild for me.
And if I want to run the script by myself, I can.
I can't just so all I do is double click on the app bundle, and then I close it, I double click on the next one after it's rebuilt it for me.
Yeah.
That's the same version of a command R, but not really.
I guess command R just gets you a command R versus a close and open, close open.
It's the same repetitive task.
You're still an API in some way.
Yeah, yeah.
Uh, but but I guess in my case, I'm the QA RAM consumed because X could probably requires more RAM to be open just so I can command R.
Hog, man.
I think that thing lacks a RAM.
Well, the only experience I've had recently with with Swift was actually with my Remarkable, is I started to build a file browser essentially for it.
Because like this thing is kind of slow, if I'm being honest.
It's a Linux and it's not the best CPU, but it's it's good for what it's intended to be.
But navigating it, it's a touch interface, it's kind of latent a little bit.
Uh and so the API I wrote against it, uh, a Go API against it, is super fast.
Super fast.
And then I built an app you, it's a lot a lot like Finder to navigate it, rename it, reorganize.
But I've learned that it can't be a real-time system.
I have to like create a version of the truth over here in the Mac app that I then synchronize back to the Remarkable.
And then the Remarkable has to do a refresh because of the way um the Linux itself, or at least the the UI uh persists new um new rights to the database and memories to the to the disk.
So it's gotta like do a little do a little restart there.
But uh in the weeds a little bit on on uh command R and ice code and Swift app development.
I say, you know what?
Listen, if you're if you're out there and you're like, man, I got an idea for for this or that, go and try it.
Yeah, explore new world.
Tant to you know, touch all the things if you can.
You know, you're in a candy store, it's all for you.
Eat it, enjoy it, don't get fat, just enjoy the candy.
Yeah.
Here is a quote from Wes McKinney, who wrote a piece of software, I think he might have vibe coded it, who knows, called Message Vault that I've been playing with.
Um, in his post, The Mythical Agent Month, which was the the title of the last changelog news, he says the coding is easier now and honestly more fun, and I can spend my time thinking about what to build rather than wrestling with the tools and systems around the engineering process.
I found that to be absolutely the case with me.
Just to, you know, give my amen to what you just said and to back it up.
It's just fun to work at this level and just touch touch all the things, try all the things, you know, build something, see if it sticks.
If not, throw it out and build something new, something else.
Touch all the things that you should touch, okay?
Respect terms of service.
Yes.
If that's your life, if not, live dangerously.
Right.
Consent, you know, you need consent for sure.
Yeah, you you do need consent in the world and the word of touch for sure.
Let's not don't don't clip just touch all the things.
Only clip it in context.
Clip the disclaimer.
Clip it in claim.
One thing we didn't mention, and I'm just gonna mention it here on the out, that you should check out, and then I want you to also check out a Nish.
I'm not sure how you say his last name, Acharya, maybe.
He is from uh A16Z and Dreeson Horwitz.
And there was a you may have seen this, Jared, and tell me if you did, but there was a post that circulated called the 2028 Global Intelligence Crisis.
Does this ring a bell to you at all?
No.
Tell me about it.
Well, uh, it is a thought exercise in financial history from the future.
And so it actually writes itself.
The very first the the dated line is February 22nd, 2026.
That's scratched out, and it's June 30th, 2028.
Okay.
It's largely, you know, bearish, not bullish and negative.
But it's it's they say it's grounded on some research, and it's essentially the consequences of abundant intelligence.
And it paints this picture of a of uh just a downside that's coming when it comes to what automation does.
And I'm just not sure.
I'm just not sure.
Yeah.
Uh that's all I'm gonna say because it's it really is that deep.
It's a whole podcast worthy uh topic thing.
But one thing that a Nish shared was a different way that this could go down.
I'm only got both because I don't want anybody out there reading this thing uh and getting all negative and sad and you know going into a hole when we say go touch all the things, and then touch only things turns into all this nasty stuff that this uh seemingly well-researched article from the future purports.
So I would read Anisha's uh prose on X, which we'll link to in tandem with this, and try your best to come out not crying like a little baby.
Because it is pretty, it's pretty, it's pretty wild.
I'll read the first line.
Okay.
The unemployment rate printed 10.2% this morning, a point three percent upside surprise.
The market sold off 2% of the number on the number, bringing the cumulative downdraw in the SP to 38% from its October 2026 highs.
It just goes in.
It's one of those kind of like doom and gloom.
But they even say in the preface that they're not trying to be doom and gloom, but it is a doom and gloom kind of post.
They say what follows is a scenario, not a prediction.
And it's based to some degree from what I understand on uh on research.
So we'll see.
We'll see.
All right.
But don't go and read that.
Get sad.
Go read that and then go read Anisha's post on X and uh and just see a different side.
There's always two different sides.
Don't go read it like I did and get super sad for a day or two, crawl into a hole, and I didn't.
I just I was pretty bummed out for like 45 minutes.
I was I was extant, I was existential for a minute.
I was like, gosh, babe, I'm having so much fun.
Right.
Tell my tell my wife this.
Gosh, babe.
That's that's that's usually how I I sigh in life is I'm usually telling my wife some sort of story from the from the battlefield, so to speak.
Gosh, babe.
This high or this low or this, you know, this this uh possible future.
Uh and I was like, she's like, what's wrong?
I'm like some big consequences out there coming.
I was like, I don't really know how to take this.
So I was a little down for a little bit, but then I read Anisha's uh post and I was like, all right, I'm right back up again.
Let's see the bright side here.
So we'll see.
There you go.
We'll see.
We don't know.
We don't know.
That's why it's interesting.
And uh read both.
Yeah, read both.
Read both.
Well, friend, it has been uh the absolute pleasure of uh of a big part of my life to spend time with you on the weekly, on the daily, yeah.
In the trenches of this uh this war room called software development SDLC, as you may say.
Oh, yeah, living the life cycle, you know, living the life cycle, so to speak, you know.
Um yeah, I guess uh yeah, yeah, man.
What did uh what did Lloyd Christmas say?
Goodbyes are hard, man.
What he said to uh to uh Holly Hunters to the person he just met a few minutes for it prior.
I hate saying goodbyes, yes, something like that.
Yeah, yeah, goodbyes are hard.
Um we can always add a for now uh on there that lessens the blow.
Uh not for good, but for now.
And um, yeah, it's been a blast.
Appreciate the many years of potting together and doing doing life together, you know, yeah from afar, from afar, and we met on the internet, and we've uh you know built a life for each of our families for a long time, and I'm excited to see where where the change law goes from here.
You know, you were changelogging before I got here, and you're the changelogging after I leave, and yeah, um I'm excited for what's next for me.
I have zero plans for those who are interested in my future.
I don't have one.
Uh I'm happy about that.
So I've not locked into anything.
I don't have ideas.
I am gonna take a break.
I definitely need to step away.
I've been I've been uh going full bore for for a while now and excited for what's next.
Open to ideas.
Don't have any of my own.
I have a few things I like to build, but you know, who knows?
I've had the thoughts of going taking a jobby job, you know, going working for somebody else.
It's been since 2011 was the last time I had a real job.
So I'm not sure if I could ever actually do that.
My wife thinks probably I can't.
Like I'm just at this point, unemployable.
Could be true.
Um don't know, but happy to see uh where where life takes me next.
So yeah, it's been a blast.
I think the only thing left to say is buy friends, right?
We we always finish with buy friends.
Well, I was gonna leave a little surprise for our plus plus folks.
I feel like we owe them a little bonus.
So you're in that crowd.
Oh, good idea.
Just get a little more excited.
I'm not sure what's there.
I'm not sure.
I don't have a plan either.
It's just an extended take, potentially, or extended.
Yes.
Maybe a couple more details that are for our pertinent uh really close folks who we really appreciate who've supported us over the years.
Sounds good.
Uh plus plus content is always good.
So if you're on that, if you're on that uh that dial, then stay tuned.
And if not, changelog.com slash plus plus is how you become that person.
So with that, friends, bye friends.
Bye, friends.
Well, friends, it is a finale, but it's not over.
That means this show continues.
Yeah, nothing changes really.
Jared's stepping away.
He's retiring.
All the things, and we'll miss him.
But the show must go on.
The Changelog Remains.
Changelow.com is here to stay.
And if you haven't yet, join us in Zulup chat, changelog.com slash community.
It's free to join.
Everyone is welcome.
And I mentioned the bonus.
If you want the bonus, you gotta be a plus plus subscriber.
Changelaw.com slash plus plus.
It's better.
You know what?
It's better because bonus content rolls, getting closer to the metal rules.
And of course, you roll.
Changelaw.com slash plus plus.
That's it.
The show's done.
We'll see you real soon.
Alright, I don't have a plan, but I figured a little bonus for the plus plus folks would be cool.
Let's do it.
I have no idea what to say there, except um, yeah, the I would say the the treadmill of of creating content is probably pertinent, right?
Like yeah, yeah, yeah.
It's easy to look at it's easy to look at this.
Uh and maybe even you know in the comment section in Zulup.
Uh which I'm very happy that people are sad, I suppose.
I don't know how to say that.
I'm very happy that uh there's a feedback loop, I would say, right?
Like it's one thing to put something out in the world on this media treadmill for so long.
And I think where we get most of our feedback loop has largely been in Slack, in Zulup, and at conferences.
Yeah.
Sure, there's DMs, there's people coming on the show, there's people who say, I've listened to your show for years, and maybe that's the polite thing to say, maybe it's just something you say, but then somebody who actually gives you that hug or that handshake in person, or someone who says, Man, I'm I'm I'm really just thankful because you got me through some hard times when I navigated my my career in software, and you were just entertainment in my ears.
I think there's just something to be said about the the media treadmill and the absolute chore it is to show up weekly for so long, yeah, and uh and not be biased or negative, which I think we've both been a little cynical and negative in the situation.
It's hard not to be when you've been through so much, you know.
Yeah, yeah.
Uh I mean, I used to describe it internally, and you can edit this part out if you think it's too negative, is what I say.
We've like created this, we've created our little jail cell, and now we've got to spruce it up and make it a nice place and live in it, you know.
Yeah, which is the creator of Joe.
You know, me and my uh in my most cynical, just kind of and we but we made it a nice place, you know.
We we it's it's a very nice one.
Um that's just how it feels over time because of just the constant repetition of like a new week, a new set of shows.
You know, you know it as well as I do, if not better, that it's just nonstop.
And um sometimes that's good and sometimes that's bad, and then over the course of years you get good at it.
I mean, I feel like we've been pretty good at it, um, to where I don't think about it all that often.
And um you do in a certain extent, especially with podcasts, feel like you're just putting something out into a void and hoping that people like it.
And I would say that, especially in the last couple of years, we've had some, I think we've made some real connections with some friends.
A lot of our changelog plus plus people are those friends, so thank you all.