menu close

Dax's Vision Pro Review, Cloud Pricing, Rebuilding Plex, and Moving to an AI Utopia

/ 01:23:47/E79
Adam:

Look at us talking about things. I guess Chris can always cut stuff out. That's what having an editor is all about. Leave it up to you, Chris. Save these people from us.

Dax:

Whenever we talk about Chris will probably cut this out. I feel like he leaves it in. And I can't tell if it's a prank. There's something deeper going on?

Adam:

No. My no. No. No. No.

Adam:

I don't care what I look like. No. The my current mental, emotional, psychological, physical, all of them. All the states. We had some, like, production issues.

Adam:

And do you know, like, when everything in your week is, like, planned out Mhmm. And then, like, suddenly, you're doing none of it? Yeah. You you know, those type yeah. It's like that.

Adam:

I'm also I'm very hungry. I've I've done, like, 36 hour fast. I've done I've done 48 hour fasts. Like, I've done a lot of fasting in my life. Well, not in my life, in the last year because of jujitsu, for, like, cutting weight and stuff.

Adam:

For some reason, I am, like, shaking, and I've only just not had breakfast. I skipped breakfast, and I'm I'm, like, jittery. It's it's actually kinda wild. My hand is shaking.

Dax:

Man, there's always something

Adam:

What's that mean?

Dax:

I feel like every time you get on, there's just, like, something. Oh, sorry. Can you just be normal? Why can't you just be normal, Adam?

Adam:

Can't you just be normal?

Dax:

I'm just kidding. The production issue thing is funny because, I feel like there's been a few moments in my life where, like, something goes wrong and it always tends to, like, stack up on other stuff that's, like, kind of already going wrong. And I always have this, like, 60 second period where I just feel all the stress.

Adam:

Mhmm.

Dax:

And and, like, it almost feels like my vision is tunneling. I can just, like, I just, like, let it all in, and I feel it. And then it fades, and then it's gonna just kinda go and get the work done. Every time this has happened, it's like something like that has happened. It's funny how, like, dramatic it ends up turning out.

Dax:

Like, it's in a way, it feels not serious because we're just, like, you know, pushing some buttons on a computer. But it's like we haven't experienced, like, we're in combat or something. It's just so, like, over the top.

Adam:

Yeah. It really is. Yeah. The amount of, like, fight or flight that you go into

Dax:

Yeah.

Adam:

Over people not being able to see, your website or not being able to do certain things on your website, in our case.

Dax:

Is it all fixed?

Adam:

Yeah. No. It was all fixed yesterday, 2 days ago. It's been 2 days ago. It it's just there's been a lot of fallout of, like, making changes to prevent it from happening again, That kind of thing.

Adam:

Mhmm. So just all the stuff I was working on, that I was thinking I was working on, not working on. And then you just feel, like, everything feels so scattered, like, when you have a disruption like that. Like, my, I don't know, just the things open on my computer. I'm, like, closing stuff.

Adam:

I'm, like, blah. There's just a lot.

Dax:

Yeah. It's, do you ever think about how that stuff only happens when you're making changes? And if you never made any changes, it would never happen. And then you imagine there are some companies out there where their product just never changes, and they never do anything with it because they're, like, past the point of doing that. And it must be so nice.

Adam:

Is software rot a thing though? The if you just, like, wrote some code and then leave it running for a long time, not in Rust, let's say in JavaScript, at some point, will it just stop working?

Dax:

Yeah. I think so because it's memory leaks are extremely common and prior to moving to serverless stuff, like, a lot of my time was spent, like, not a lot, but, you know, figuring out a memory leak was not uncommon. And I think we've all had the people that have, you know, pre serverless days, if you've shipped stuff, You've all had that experience of, like, that chart where, like, the memory just, like, goes up over the course of 8 hours and then it reboots and it, like, drops. And then you finally fix it and it's flat and it feels amazing. But, like, with serverless, it's just it's just like it's rebooting your system, like, every every couple minutes.

Dax:

So Yeah. You can kinda ignore all those problems.

Adam:

I will say I built a serverless app. One of the first apps that I built just kinda like using all these different new, what are very common paradigms for us now. It was 3 or 4 years ago for, like, a local company, and they use it every day. It's like they'll just say it's like a health care thing, and they use it in their office to manage, like, their patient experience. They use it every day.

Adam:

There have been no issues. I mean, like, this is not a thing that I experienced early in my career. If I built something for somebody, I had to constantly

Dax:

Babysit.

Adam:

Fix that yeah. Babysit it. And maybe part of it, I was I was still young. I was a newer engineer. But I don't think I've gotten that much better, and this app has just been rock solid.

Adam:

Like, I don't have to think about it. There there's this, like, once a year. It's the daylight savings time. Apparently, some the library I was using, has like a like a 1 hour gap where there's like a bug during daylight savings time. And that's it.

Adam:

That's the only issue, and I still haven't fixed it. I should probably fix that. I'm so sorry if you're listening.

Dax:

Yeah. I've had a similar experience. The first thing I ever built, in this extreme way it's so crazy. I can just, like, go to it at any time and it still all works. And that's such a novel experience if, you know, coming from the way things were done before.

Adam:

Yeah.

Dax:

I think the other kind of rot that exists though is this is annoying and I'll just ignore it. Because, eventually, every single one of your dependencies has, like, some critical security thing in it. And so you do have to, like, upgrade those. And I think with that comes some risk of something breaking. But, yeah, like, the fewer dependencies you use, like, you know, the better.

Adam:

Yeah. I guess, like, software rot is a term that was coined long ago, I'm sure. And even back when, like well, was it? I don't know. Was it when there was box software when you were, like, buying DVDs with software on them?

Adam:

Or is it in the web?

Dax:

Like physical rot where, like, the the medium would

Adam:

be great? No. Not not physical. No. I just I I I think, like, the term I wasn't even talking about, like, necessarily applications that you deploy serverlessly.

Dax:

You mean, like, literal code?

Adam:

Just like yeah. Code. Like, over time, stuff doesn't work as well. Let's let's

Dax:

look at what software rot is.

Adam:

Yeah. What is software rot? Maybe I should just know what I'm talking about first.

Dax:

Slow duration of software quality over time, that was a little bit different. Not a physical phenomenon, so I'm gonna actually decay. Suffers from a lack of being responsive and updated with respect changing environment in which it resides. Yeah. Okay.

Dax:

That makes sense. Like, if the environment changes and the software is not updated, then it feels like it's rotting.

Adam:

So is this like the environment being like the operating system? Like, if it's native software like, the operating system's evolving, so your software has to continue evolving with it. That's why you have to keep paying Apple $99 a year. So if you wanna pay it

Dax:

to the Or opens just like things like, like touch screens came out, and now some people are trying to use your software on touch screens and it, like, just doesn't work. Or their expectations change or or all of that. So, yeah, I think we all dream both as someone making software, you can just, like, finish it and put it out there, And someone, like, consuming software that you can just, like, implement it once, and it never changes. But, yeah, just the world is way too dynamic.

Adam:

So is that why everybody does subscriptions? Is that why it's not one time payments anymore? Because it is an ongoing effort, and there's no such thing as just being done with the software product.

Dax:

There's a few things to this. Right? So, everyone's like, oh, things used to not be subscriptions. No. It didn't.

Dax:

Does no one remember how we all paid for Photoshop? I mean, we all we all stole it. We all pirated it, I think. But if you if you did pay for it, Photoshop 7 would come out. You pay $700 for it.

Dax:

And he would use it for a year and a half. And then they're, like, oh, Photoshop 8 is out. Give me us another $700. That's just a subscription with, like, a yearly contract. It's it's the exact same thing.

Dax:

The reality is is, like, how are you the only way to really do this is if you look at the fundamental model. You're saying every single customer you acquire will only ever have one transaction event with you. So your only way of growing is continually finding brand new customers, which companies are already trying to do. And on top of, like, subscribe like, existing customer resubscribing. So that model just doesn't make any sense.

Dax:

Just like companies already struggle to do both. Even by doing both, they still struggle. You're gonna, like, cut off half the strategy and expect it to work. I don't know. I do I do know that tailwind, people talk about I was

Adam:

gonna bring tailwind.

Dax:

Yeah. Yeah. Because they they do go down with that model. But,

Adam:

yeah, it's

Dax:

a good 3 or 4 person team. I think that makes a lot of sense for them. And there is to be fair, there is the reality of, there's the world is less static than we you imagine. Like, it feels like there's, like, a pool of 2,000,000 people, and you need to, like, try to get to all of them. But, really, what it actually looks like is there's a pool of 2,000,000 people.

Dax:

Some of them are aging out, and a bunch of people are aging in. So it's like, you do have a shot at, like, continuously finding customers. But, yeah. I think for products that require ongoing maintenance, but also, like, stuff that's hosted this is another thing that I always think about. I don't know if you ever thought about this, but, you know how, like, you have to most SaaS software, like, stores your data forever?

Dax:

Does it? Well, like, sort of. Like, if you build like, let's say you build a b to b SaaS product. Most of them aren't, like, oh, yeah. After, like, 2 years, we're gonna expire your data out and we're gonna archive it.

Dax:

It's, like, pretty much there. And, like, if I paid them do you see how, like, over time that data could get so large that it, like, exceeds what I'm paying them per month?

Adam:

Oh, yeah. Mhmm.

Dax:

Yeah. And I do wonder if there are some, like, more mature, bigger b to b companies that have, like, had to institute a policy of, like

Adam:

We're gonna purge your stuff.

Dax:

Yeah. Or any inactive records we're gonna, like, archive. If you think about something like Slack, just pure let's say it's a small company that's using it. Let's say 10 people. They're messaging in there every single day.

Dax:

Slack's bill to store those messages increases every single month. Yeah. But my bill for how much I'm paying to Slack is fixed. So their margins, like, get worse over time. The ratios are insane.

Dax:

Like, I'm sure, like, the cost for them is, like, a quarter of a percent or something, but, it's still like a weird it's something that, like, doesn't line up there.

Adam:

Yeah. Is it, like are they banking on storage costs just get cheaper over time too?

Dax:

Like That's true. Storage costs have gotten a lot cheaper over time as well.

Adam:

The Google work spaces just increased the pricing. I I got, like, 15 emails.

Dax:

They did?

Adam:

Yeah. So, like, my I have a Google Workspace just for mine. I don't know why even, but I for one of my email addresses.

Dax:

Yeah.

Adam:

And they just they just bucked the price. I don't remember how much. I didn't even look at the email or any of the emails. But it wasn't probably a lot. Just, that's notable.

Dax:

So I think that model of, like, the pay once thing, it works if you're just shipping a delivered product. I think that makes sense for Tailwind because they're, like, trying to

Adam:

ship you. Tailwind UI is it makes sense. It's not like, there's not, like, this ongoing infrastructure requirement. I guess, like, there's a website and you download the stuff.

Dax:

But, you know, basically free.

Adam:

It it seems less complicated. Yeah. Basically free in terms of infrastructure.

Dax:

Yeah. And then for for the one like, the DHH stuff where he's trying to sell it once, really, that's that's just a marketing scheme. But to take it seriously, if his company was actually growing, like, he they would have to release, like, a new version that you have to pay for again. The lifetime subscription thing, buy this for life, it it just doesn't work. Like, I did this with Plex, like, years ago.

Dax:

Like, 10 years ago, Plex offered, like, a buy it for life thing and I did.

Adam:

Yeah. I bought it for life this year.

Dax:

Oh, they they still do it?

Adam:

Yeah. They still do it.

Dax:

Okay. Cool. Yeah. So, like, I guess it it does kinda work, but it's just extra challenging.

Adam:

Yeah. Yeah. No. It's I I I mean, I think we've covered all my thoughts on it. I don't know.

Adam:

The one thing, I didn't really read the big page that DHH published. Like, what was the idea? That they're gonna make a bunch of things that are you buy it once or once it's a product?

Dax:

They self host it. So I think self hosting is kind of a requirement for this because then then there's no ongoing infrastructure cost for Yeah. The creator. Yep. The thing though is, like, they don't I'm super into the self hosting thing.

Dax:

Like, SST is all about self hosting stuff that other people are paying a service for. Make it, like, really easy to self host, you know. Like, I I don't wanna self host like a rails app that, again, is gonna have a memory leak that needed to have monitoring on. Like, I don't that's not that's not easy self hosting for me.

Adam:

Yep. Yeah. Because you you guys have gone down that route, right, with the SSC console making you're trying to make that self hostable?

Dax:

Yeah. I mean, we're gonna have to. Like, every b to b company like this, especially the lower level you are, like, the closer you are to infrastructure, big enterprise companies will not use a hosted version. It just doesn't Yeah. It doesn't work.

Dax:

But I think what's nice is your incentive should be if I'm running a hosted version, I want it to be really easy for me to host it. I don't want a lot of operational overhead. I want to be able to, like, create copies of my environment easily. If you do all those for yourself, it, like, also aligns with making it easy for other people to self host. Because then they'll also have low operations and you'll be able to, like, submit a copy in their environment.

Dax:

So Yeah. Yeah. I'm seeing a few more projects in this space. I don't know if we talked about this already, the Cloudflare Analytics 1.

Adam:

I don't think so. I I think you've mentioned it to me, but I don't think we really talked about it much.

Dax:

Yeah. I forgot the name. I keep forgetting the name of it, but, it's from someone that works at Sentry. I think his name is Ben. And he just built a very bare bones analytics product.

Dax:

I think it's bare bones today. Like, it's it's new. But it does like what Google Analytics, the basics for Google Analytics, page view tracking, etcetera. But it's just built on top of Cloudflare's analytics engine. So you just deploy it into your Cloudflare account.

Dax:

It's basically free because the Cloudflare are free tier and, it's no operational overhead because the analytics engine is serverless.

Adam:

Yeah. That's pretty awesome.

Dax:

So that's great. Like, I think we're actually considering using that, for our own tracking. I'm seeing a few more things like that.

Adam:

How much is yeah. How much is Ion gonna make SSD SSD, like, a Cloudflare? I mean, like, how much are you guys gonna kinda start defaulting to Cloudflare in some of your constructs? Or is that gonna be a thing? Is it still gonna be largely AWS focused or no?

Dax:

Yeah. So we we, like, went back and forth on, like, the right model for this. So there's a few different ways we could do this. Right? So if you take one of our constructs, like, AstroSight construct, right, to move an AstroSight, we could update that that thing to give you, like, options inside the component to be like, oh, I want the CDN in Cloudflare and I want, like, the server functions to run-in Lambda.

Dax:

You can, like, give you some complex configuration there. Yep. We realized that that was, like, a bad route. That was just gonna, like, create, like, a really leaky abstraction, and, like, it was gonna be hard to really customize that correctly. So we just have completely separate namespaces now.

Dax:

There's, like, an AWS namespace with all of our AWS constructs. There's a Cloudflare namespace, which today we just have, like, a worker construct. But we'll bring, like, the AWS the Astro site over to it. We'll bring Next. Js over.

Dax:

We'll bring everything to it.

Adam:

Gotcha. Okay.

Dax:

And that lets us just say, like, okay. We're we looked at Cloudflare, and here's, like, our best possible take on deploying Astro there. And they're, like, completely separate, and the communities that know AWS and are using Astro can, like, own theirs. And communities that, you know, know Cloudflare and using Astro, they can they can own theirs too. So we're keeping it completely separate.

Dax:

The first release is is, like, really still AWS focused, but, yeah, we wanna grow our library of Cloudflare components as well.

Adam:

Cool. I have a a completely, new topic

Dax:

Yeah.

Adam:

To jump into. I wanna know what you think about, like, framework devs versus app devs. So this is, like, a topic that comes up every once in a while. Like, oh, well, you're just a library author. You don't actually build things.

Adam:

So what do you know about what's best for me? Have you ever heard that? Like I mean, that that comes up. I'm not saying you. Like, you do both.

Adam:

I mean, you guys, like, building the console is a great, way to, like, dog food your own technology. But do you think, like, the creators of React, the creators of whatever UI library, like, do you think that they lack context? Like, are all the people debating the internals of the latest React? Are they all just, like, too detached from the real world, I guess, is my question.

Dax:

Yeah. I would say I think this is a real dynamic that I think I've been seeing my whole career. There's, like, this concept of a hands off architect that a lot of companies I don't know if that's as common anymore, but at least when I was first starting, that was very common to have, like, this architect role. And it often didn't work out well for the exact same dynamics. Like, they theorize these things, and then on the ground, it just doesn't really feel right.

Dax:

Yeah. You see this in the AWS space a lot. Like, the, like, de facto way to do AWS, the correct way to do quote, unquote correct way to do AWS things doesn't really, like, feel good for the day to day developer doing it. They they're kind of like, they make sense in theory.

Adam:

So I

Dax:

think I've seen this dynamic a lot. On the flip side though, I've also seen individuals that are very like, when you talk to them and, like, there are people that are framework authors, you can talk to them about anything and they know exactly, like, what the experience of it is like. And so I think it just comes down to the individual. So a few people that I personally have had this experience with, Ryan Carniato.

Adam:

Mhmm.

Dax:

It's funny because you would think he's, like, the godfather of, like, front end framework stuff. Yeah. You would think that he's the most attached.

Adam:

Epitome of a library author.

Dax:

Yeah. Exactly.

Adam:

In the sense that he knows all the libraries.

Dax:

Yeah. He knows, like, all of these. He knows all of the work. Yeah. Yeah.

Dax:

But, he is an excellent back end engineer. You know, we all know him for the, like, the front end world stuff. I've talked to him about so many specific practical real world things, and he's very good at understanding, like, here's, like, the thing we're pitching with this framework, but here's how it'll it'll actually get used. It's he knows that it's not gonna be as slick as what the way people talk about it. He talks a lot about, he actually coined the term for me.

Dax:

He calls it, DOA projects, dead on arrival projects. Because there's a whole class of projects in the real world where you just build it once and then

Adam:

you or

Dax:

touch it again. And he's one of the few people in the framework space that I've heard, like, directly say that sometimes frameworks target those, and

Adam:

Mhmm.

Dax:

It doesn't make sense for non DOA projects. Whereas I've seen other framework authors, like, show these really simple, slick things and be like, it works for everything. Another person is Ryan Florence. I think he also is very, very good about understanding practically what happens on the ground. And, I mean, now that he's at Shopify, a big part of his role is to literally talk to actual companies that are using this stuff and, like, figuring out what they commonly do wrong, what their problems are, etcetera.

Dax:

So I think he tends to also be very, very good at balancing both. So, yeah, I think with, like, everything, it just comes down to, like, the individual and how Yeah. Talented they

Adam:

are at random and

Dax:

all this stuff.

Adam:

Do do you think, like, just talking with people who who build with your stuff is good enough? Like, does that give enough context?

Dax:

No. I think that's, like, the bare minimum that you can do. I think the best people are people that, like, are actively trying to build something, and they realize there's an opportunity for tooling that helps them build it. But the opportunity for tooling is like like a detour for, like, the actual thing they wanna do. Yeah.

Dax:

That's obviously so I think people that fit in that, they build up that stuff. I think the another issue is this is a scale issue. If you're directly working on React, like, React is so big that the reality is is you could have a very nice, very great career just being someone that focuses on, like, the meta React stuff as opposed to, like, down on the ground using React stuff. So I think when something gets so big, like, you inevitably have people that are incentivized to kind of stay in a certain zone.

Adam:

Yeah. Could you explain I don't know if I really understood what you meant. You mean, like, there's enough React is such a huge, ecosystem that there's room for there to be people whose whole career is just, like, in the minutiae? Is that what you're saying?

Dax:

Yeah. So I guess the the counter I can provide is if I just wanted to spend all day working on Solid JS and Solid JS internals and, like, thinking about high level abstracted stuff, no one on the planet is gonna pay me to do that. There's no way I can do that full time. Yeah. But that's not the case with React.

Dax:

There's, you know, like, Vercel hires people to do exactly that. There's opportunity for people to do exactly that. And I think what I think the the big shift in React was previously, every feature that was released to the public was already in production in Facebook. It was being used by teams to build stuff and hardened, and then it got, you know, shown to the world. I'm sure there's maybe exceptions to that.

Dax:

I don't know. I might not know the details, but, generally, that was the case.

Adam:

Mhmm.

Dax:

And with RSCs, Facebook's not using that. Nobody's using it. Nobody it's like it's the inverse. Like, we're being pitched it to try it out as opposed to being told, hey, we're already using this. If it's helpful for you, you can use it too.

Dax:

So that that that's, like, the shift, I think, and I think we can kinda feel that.

Adam:

Yep. Do you ever just get a random taste in your mouth? Like, I just I just taste pumpkin pie right now. I haven't had pumpkin pie in, like, a year, but I taste it. Is that like I'm having a stroke?

Adam:

What does that mean?

Dax:

I don't really know. I feel like your hands are shaking. You're having weird taste in your mouth. Okay. Your your emotions are impacted.

Dax:

I don't know. Like I said earlier, just try to be normal. Okay?

Adam:

I'll try. You know what? I'll try. I've been a little, extra. I feel like I'm the high maintenance one of us between the 2 of us

Dax:

by law. A 100%. That's not debatable.

Adam:

A 100%. Speaking of Vercel, so there's this runaway bill, in the in the news.

Dax:

Oh, yeah.

Adam:

And by in the news, I mean, on Twitter. And this is the thing that it drives me crazy about, like, cloud providers. Like, why can't I I've said this a lot. On AWS, why can't I just say, here is a dollar amount. And if you hit that dollar amount or I just want a free account.

Adam:

Like, here you give me the free tier. Well, if I use anything that costs anything, shut it down. Yeah. I know it's a hard problem, but it seems like that should just be baked in to all of these platforms. And I know Vercel being a middleman and just reselling AWS, they kinda have the same problem AWS has, and it's like this case by case bureaucratic decision about whether we refund customers, if it was their fault, if it was a exploit, if it was a vulnerability on our side, like, all those things.

Adam:

Sorry. Go ahead.

Dax:

Yeah. So just like broadly, I find it annoying when because I'm sure people expected me when this came up to code dunk on Vercel, but I actually find these situations really annoying. I think that, one, they they're always resolved. They never actually end up being a problem.

Adam:

Right.

Dax:

But it just provides so much opportunity for, like, a ton of misinformation about all this stuff to get out. And it's just, like, bad for the whole ecosystem that people see these, like, shock situations. And they see the $23,000 whatever and they, like, never see the resolution. And it just continues to spread, like, all this confusion around this stuff. So I don't like like, I just hate it when this happens.

Adam:

Yeah. I mean, to your point, like, AWS, there's so many people who won't use AWS just because they're afraid of a runaway bill that they've read on Reddit. And if

Dax:

you and if you look at the last couple I mean, all the last ones that I can remember when because this happens every couple months. Right? Someone posts something like this. I don't remember an AWS one in a while. It's, like, pretty much have been Vercel ones.

Dax:

So my thoughts on this are, I think it's a hard problem, not technically, but it's like a product issue. So the question is, like, what headline are you trying to avoid? If you're trying to avoid, oh, no. I just got a $23,000 bill, You're you make it so when you sign up, there's, like, a limit that you have to turn off. Right?

Dax:

Yeah. And most people won't even know about that. Like, they won't look into the right setting page. They don't even know the limits there. But when this happens, they, you know, they they don't get this runaway bill issue.

Dax:

Yeah. But then you have this other headline, which is I spent 2 years working on my product. It finally went viral and Vercel shut it down.

Adam:

Yeah. That's that's worse. And you

Dax:

see this with, like, payment providers and stuff all the time. Like, because some, like, spike in payment activity looks like fraud and they shut it down and, like, lose a bunch of sales. So you have both these headlines that are possible and whatever default you pick, default cap on, default cap off, determines which one you wanna deal with. Yeah. So I understand why they opt for the 3rd option, which is we'll just refund people when this stuff happens.

Adam:

But I guess, like, that's default. In this case, there is no ability to say I want a hard spin limit. And that's what bothers me. It's like, why on AWS? Why on Vercel?

Adam:

Why on all these platforms? I know why on Vercel because they can't I mean, AWS, you can't do it. So Woah. It's like they're upstream. I guess they still could.

Dax:

Yes. They can because they they have the ability to, like, freeze a project. So because remember, they have their own routing layer before it hits the Lambda function, so they can always Right. Like, shut it down for it. But but remember what AWS is doing now.

Dax:

They, when you create a new account, your function concurrency is, like, capped at 10.

Adam:

Is that every new account? Is that a thing?

Dax:

It's been happening way more frequently. It happened to us, and, I've been seeing it a lot more. So, like, that's now they're switching to Optium for the default where, like, you go to production, you're like, what is happening? Everything's broken. But the the limit is so low that you'll probably see it very quickly

Adam:

Yeah.

Dax:

As opposed to, like, you know, seeing it, like, 6 months later.

Adam:

Yeah. I guess that that just spun off a whole new set of thoughts. Like, you guys use SCS in the console, don't you, for email? Yeah. What's the process of getting actual production usage of SCS?

Adam:

SCS? I've never it's been too much of a hurdle. I've never tried.

Dax:

Do you know what I do? I've done this so many times.

Adam:

Do you DM a buddy and just, like, they no. Okay.

Dax:

No. I do something even funnier. Okay. So people that aren't aware, when you are using AWS SES, which is a service to send emails, by default, you can only send the emails that you've, like, registered. It's just for, like, dev use.

Adam:

Yeah.

Dax:

If you wanna send emails generally, there's a process. You have to open up a support ticket. I've done this a few times where I've realized I'm very certain the other end is a bot. So I will always provide what they ask and they always respond with the same information. Respond the same message being like, thank you.

Dax:

And they ask for more information. But it's already information I put in the first message. Yeah. So I can just copy paste the first message back, and then they unlock it. And I've saved this message now, and I just repaste it every time.

Dax:

And Okay.

Adam:

I need it. I need the message. Or does it have, like

Dax:

No. No. Yeah. I could probably just give it to you. The the Yeah.

Dax:

Part of me wanted to be like, we just have a component in SST that just does this. It just turns on the CSB. It's it's also funny because the initial limits are so hardcore. They're, like, you can only send a 100 emails per day to to, like, whatever. And then you go through this and they're like, okay.

Dax:

You could send, like, 4,000,000,000 emails every hour. It's like, okay. I don't need that much Yeah. Of a jump.

Adam:

That's funny. Wow. I I want this message. Yeah. Or just build the construct.

Adam:

That's so funny. That that's the power of what you guys could do with SST. Like, you really can cover a lot of rough edges on these cloud providers by just doing weird stuff like that.

Dax:

Yeah. So yeah. I mean, I think, yes, Vercel should add I mean, they already have something. It triggers a webhook and it's on you to, like, handle it and and freeze the project. They could make that one click.

Dax:

I don't know why they haven't, but, like, it's tricky. It's not gonna be I think everyone thinks it's so straight I think that a lot of people externally, it seems so straightforward that the fact they're not doing it seems, suspicious or intentional. Nobody enjoyed the last couple days, you know. Nobody on their team enjoyed having to deal with this. So, and they almost always have to refund.

Dax:

I think their bigger problem is, you know who's not refunding? AWS is not refunding them.

Adam:

So yeah. So Vercel is just out that that money. Like, they just refunded their customer, but they paid the bill, paid to be as well.

Dax:

Yeah. I mean A

Adam:

5th of it or whatever. Whatever it is.

Dax:

But still, you know, there are

Adam:

a couple

Dax:

of but it's probably, like, I think I think I did the math. I think it was, like, $2,000 or something.

Adam:

Okay. So

Dax:

Plus the time of all the people investigating and all that.

Adam:

Yeah. I guess Vercel can't then go to AWS and say this was an accident from our downstream users.

Dax:

Yeah. They can't.

Adam:

Could you please refund?

Dax:

Yeah. The more I think about Did

Adam:

you say they can or they can't?

Dax:

No. They cannot. I I I guess cannot. Yeah. I don't think that would work.

Adam:

Because they look bad or or just because they don't have a case? They're like a big customer. They pay their bills every month. It's different than, like, a hobby person coming in and being like, I did not mean to spend $30,000. Yeah.

Adam:

That that AWS works with you, but this is a case where probably not.

Dax:

Like, again, everyone has a shared responsibility model. AWS has a shared responsibility with Vercel, not with Vercel's customers. So I can see why

Adam:

That makes sense. Yeah.

Dax:

They couldn't do that. The thing I keep thinking about is what what is AWS, and what are these cloud providers? What they did was they built, like, crazy multi tenant system. Right? You can Mhmm.

Dax:

Get resources. Your resources are scoped to you. The configuration is scoped to you. The billing is scoped to you. Like, everything is just your stuff.

Dax:

And they put in the work to, like, create all of this separation so that even though it's one data center, you feel like it's your own private data center that nobody else is in. Yeah. Then you have companies like the middleman companies that go and act like they're one tenant, but they're actually representing, like, you know, thousands of companies that are using them. So they're stuck rebuilding all the same stuff that AWS did. They're stuck, like, rebuilding a multiten system on top of this.

Dax:

And there's such weird problems. Right? If you look at so I was thinking, like, oh, why do because, AWS has WAF, right, the web allocation firewall to prevent I don't know how effective it is, like but I'm sure it's very pretty effective for preventing, like, DDoS and, like like, this malicious traffic. I was like, oh, why can't they just turn that on? But then I was like, oh, but if Purcell turned it on, that would turn it on for all of their customers with the same rules for all of their customers.

Dax:

And if, like, one customer is targeted with, like you just don't have you lose that granularity. So they're stuck rebuilding that in their layer.

Adam:

Yep. And

Dax:

a lot of their development time goes into these types of problems. And, like, you as an end user, you don't care about those. Those are, like, issues that internal issues that they have, internal architecture that they need to solve. You wouldn't need to solve those if you were just using AWS directly. So, yeah, I can't just I just keep discovering new ways to understand, like, what a tough business it is.

Adam:

Yeah. What like, how do quotas work for Vercel? If they're just like if all their customers are really just in, like, Vercel accounts AWS accounts, I'm sure they have more than one account, but, like, they have to deal with these quotas that potentially as their business grows, if they have more customers using more of their cloud services, they're gonna butt up against more and more AWS quotas where you can only have so many resources deployed. And they've gotta just continually, I guess, keep that provisioned higher than what they need.

Dax:

Well, AJ was telling me that when he worked at serverless framework, they had a similar problem because they had, like, a hosted thing. It's kinda similar. They they actually would just load balance across multiple AWS accounts. Not, like, literally load balance, but they would, like, shard their customers across many AWS accounts. Yeah.

Dax:

So that's I wonder if they do something similar.

Adam:

Yeah. That makes sense.

Dax:

Again, just, like, weird problems that like, you imagine that when you pay a company like this, it's because they're doing the work that you would have to do. Yeah. But they spend a lot of time doing work that you would never have to do because you're not trying to manage a 1000000 deployments.

Adam:

Yep. Yeah. The WAF thing is interesting. And just being, like, DDoS ed on Vercel. As someone who is very involved with a consumer facing website that gets a lot of eyeballs on it.

Adam:

Like, we deal with DDoS attempts or just constantly. I mean, we're constantly getting hit with these, like, bots that are trying to either scrape all our data or whatever, like, just big bursts of activity that's not really stuff that's desirable. We don't want that activity. Like, Like, we have to deal with that stuff, but we have the full suite of AWS tools at our disposal.

Dax:

Mhmm.

Adam:

If you're a Vercel customer, what like, I've seen the example where a very notable, tech influencer got DDoSed, and he just, like, DMs Lee Rob, and they take care of it. But what do you do if your Joe doesn't have a relationship with Lee Rob on Vercel, you're a customer, and you get DDoSed? You're just what? Like, what happens?

Dax:

It's funny because you just drop down into a system that has proper multi tenant support. So everyone goes and signs up for Cloudflare, which, again, is a fully complete multi tenant system. They give you a firewall that's scoped to you, and they put it in front of their site. And now they can control the rules that they need.

Adam:

So people so if I'm a customer of Vercel and I'm actually concerned about these things, I just put Cloudflare in front of my Vercel account.

Dax:

That's what everyone is told to do, and that's what everyone does, does, and it effectively fixes the problem. But it's, like, it's it's so awkward. Like, there's there's the Yeah. Like, now you're routing through Cloudflare to get into the AWS to Vercel's AWS network to go into their Kubernetes layer to then go back out to Cloudflare Worker.

Adam:

Yeah. I I hate the idea of, like, extra hops.

Dax:

Uh-huh.

Adam:

The the stuff like that just drives me crazy. And I don't know, maybe it's just DNS, and it's not, like, a big deal. Or it's not just DNS, is it?

Dax:

No. It's not DNS. It's a 100%.

Adam:

It's your CDN in front of another CDN.

Dax:

Yeah. Yeah. I don't love it. Yeah. I I wonder if they will end up moving off of Lambda because okay.

Dax:

So if you look at some of the things they've done, they have, like, their own, like, bubble environment thing. Like, when you, like, deploy a Vercel serverless function, it, like, kinda lives in this, like, bubble. There's, like, certain stuff that's injected into the global. There's certain, like, APIs for even Yeah. Receiving the request.

Dax:

Like, they had a post yesterday, and now you use a standard, like, web request response stuff. So you're not exposed to any of the Lambda specifics at all?

Adam:

Yeah. You're saying, like, there's a Lambda function wrapper, sort of.

Dax:

Yeah. Exactly.

Adam:

There's a bunch of Vercel code that lives in Lambda, and it shields you and kind of, like, abstracts away Lambda Yeah. To from your function. Okay.

Dax:

Yeah. So what's nice about that is they could just take your stuff and move it somewhere else to their own compute wherever. And now they, like, you know, it's not running a land anymore and you probably wouldn't even notice. And they could probably improve a bunch of some of these issues, like, some performance issues. They have more control.

Dax:

But it's a weird place to be because I don't really believe that you can offer a serverless offering unless you're at AWS scale. I don't think I I think you literally make negative money until you tip over to being, like, AWS scale. It's it's hard to, like, keep your machines busy otherwise.

Adam:

Yeah. What are you saying? You're saying it wouldn't work for Vercel?

Dax:

I want I see why it makes sense for them to, like, move to different forms of compute that they, like, directly manage more. Like, if they have their own servers, like, physical servers and physical data centers.

Adam:

Oh, I thought you're saying, like, Cloud Fire workers or something instead of Lambda? No. Oh, I see.

Dax:

I meant, like, just

Adam:

Going lower.

Dax:

Going lower. Yeah. Which is not something they generally do but I can see why that kind of has to happen because, their margins don't, like they're just so expensive and it has to be because it's they're building on top of another product. Yeah. If you look at, Jay pointed this out to me.

Dax:

If you look at, the companies that invest in that's right. The VCs that invest in, like, Vercel and some kind of some other companies, like, that are similar to them. You'll see a bunch of common VC names, but you'll never see Sequoia. And that's because Sequoia specifically believes that you cannot build a profitable business on top of someone else's cloud. That's why they're never in any of these deals.

Dax:

So that is, like, a real question mark and a real problem. So you kinda have to move to your own stuff to make it sustainable. But, I don't know, you have to be at a crazy scale before you can do that. So, yeah, I don't really know. I can see them doing it, but I can also see the challenges.

Adam:

Are they not already at that crazy scale, Dax? You don't think you don't think they're already there? I'm just kidding.

Dax:

I mean, I don't

Adam:

think Cloudflare

Dax:

is there. I think Cloudflare has a crazy profitable other business that lets them bet on this direction.

Adam:

Yeah. But that's something sometime I'd love for you to give me, like, the the history of Cloudflare. Because I don't know if I understand, Like, it started just as a CDN, and then they've moved into all this other stuff. But you don't have to do it right now.

Dax:

They don't really start as a CDN. They actually, I need to look up more on this myself. I'm not that educated on it, but they they started as the problem we wanna solve is security, like a DDoS issue. How do we solve it? And the way to solve it for them was to, like, integrate into all these crazy places in the world, like, all these crazy places where Internet traffic flows through.

Dax:

And they did that work, and nobody else in the world, not even AWS, is as tightly integrated as they are. Like, at the ISP level, they're they have, like, deals with every single ISP. It's, like, crazy, crazy embedded. And I realized now that we're here, we can do all kinds of other stuff. I think it just all all flowed from that.

Dax:

I'm not super aware of the details, but, yeah, it's a cool I feel like that that's one of those, like, 100 year businesses. Like, you'll never really see their what they're gonna become for, like, a 100 years or so. It's a long term play.

Adam:

I have another dumb question. Yeah. I just saw you tweet something about Supabase. And I realized I see people tweet stuff about Supabase, and I don't know what Supabase is. What is Supabase?

Adam:

It's like not Firebase. It's like open source Firebase, but what's Firebase?

Dax:

Okay. So

Adam:

Is that even right? Did I even get that right?

Dax:

I think yeah. Supabase and Firebase are very similar. I think what Supabase initially set out to end still is, tackle that back end as a service market where you really just want to build a front end. You want to write data to a generic back end. It has stuff like auth permissions, like, all that stuff built in.

Dax:

And they were like, we were gonna do, like, the best possible version of that. And I think they did achieve that. They built the best possible version of that concept, did it really well, built a lot of really cool technology as a side effect of that. And when I saw them, you know, first come out and started to get popular, I was like, oh, they're gonna win this, like, market. It's not the biggest market in the world.

Dax:

Again, like those DOA projects or, like, agencies that do, like, 8 projects a year. Perfect for them. I was like, okay. They're just gonna dominate that or, like, you know, win a large percentage of it. And my tweet was I found it fascinating that they broke they, like, went so specific and so deep with this very specific market, but people just use them for generic Postgres hosting now.

Dax:

Like, they don't even care about the full back end of the service thing. Like, as a side effect of them building that, they also built a great Postgres experience. Because it's a full back end as a service offering, it has, like, great tools to, like, go look into your database and see what's going on. And it's all nice even if you're just using it just for the post cross functionality. Yeah.

Dax:

And, yeah, I think that I think that was, like, a very unique trajectory. I didn't really see that I didn't predict that happening at all.

Adam:

Yeah. That's just making me think of all the things I use. No. I can't think of a single example right now because I haven't eaten. But there's so many things where it's like you use one sliver of it Mhmm.

Adam:

And you kinda feel bad. You just have this nagging feeling that, like, you should be taking advantage of more of this stuff that it offers, But you just need this one part, and you just use that part, and there's nothing stopping you from just using that part.

Dax:

That's great when it doesn't feel the rest of the product is getting in your way.

Adam:

Yeah. Sometimes yeah. Sometimes it's it it's very much in the way.

Dax:

Yeah. It's a they're they're they're trying to get you to use all of it. And you don't want to, and then that's very annoying. I think Supervase does a good job not doing that.

Adam:

Plex is a good example of that. Like, I use Plex to direct play media that I have stored on a server, but they wanna shove all this other Plex stuff on me. Like, no, thank you. I signed up for Plex Pass. Can I get rid of all that something?

Dax:

It's so hard to get rid of it, and they're always adding these stuff that you then you have to, like, go and turn off in the settings too. Yeah. Do you see my tweet about that? Like, that's one of my, like, dream projects.

Adam:

Oh, yeah. You you wanna build Plex?

Dax:

Yeah. Plex is written in Python.

Adam:

Oh, really?

Dax:

I think so. I think a large amount of

Adam:

Which part? Like, the the server?

Dax:

Yeah. I think it is. Maybe I might be wrong on this. But I think I can remember seeing Python files in in in the folders. I'm digging through.

Dax:

It's so slow. Source?

Adam:

Wait. Oh, no. You dug through the

Dax:

Just like the config folder file. Okay. So there's, like, some Python files end up there. Yeah. It's so slow.

Dax:

Like, what I don't know if you've you've experienced this, but, like, why, like, why is there any lag going between any screens in Plex? Like, it's storing, you know, even big libraries, 2,000 records. 2,000 records, 2,000 movies that you have, whatever it is. That's like you can just fit all that in memory literally, like, the whole database.

Adam:

Yeah. And and every time I open like, every time I start up my theater, it is slow again. Like, why didn't it at least get fast after the first load? Yeah. Why didn't it store all that stuff for the next time?

Dax:

Yeah. So there there's that, which yeah. Someone just built, like, a local first architecture on this. It would just blow it out of the water. It would feel really great to use.

Dax:

It would feel like a native app, which it should because you're it's running on, like, dedicated devices.

Adam:

Installed like a native app. Yeah.

Dax:

Yeah. So that's 1. 2, this is this is, like, maybe over the top, but I always thought the transcoder of Plex should be decoupled so it can run on several servers if you wanted to. So you can have, like, a cluster.

Adam:

Yeah. Let's just have, like, a dedicated transcoding machine.

Dax:

Yeah. Exactly. That's, like, separate or something. So it's kind of decoupling those.

Adam:

I

Dax:

think the whole thing should be written in Go. Single binary you install, like, just keep it really stupid simple, instead of, like, this complex application. But yeah. Like, ultimately, all it is is, like, it's in a way, it's just just like OBS. OBS is a big complicated thing, but all it's doing is running FFmpeg commands at the end of the day.

Dax:

It is a UI for FFmpeg, and Plex is kind of also that. I'd love to just, like, build a true open source version of it that doesn't have all this bloat, that's not trying to become a Netflix, that's not trying to, like, shovel this stuff in your face. It'd be fantastic.

Adam:

That would be great. I'll use it. I'll be your first customer. That reminds me too. I saw somebody tweet about, Fabrice Belard.

Adam:

Did I say his name right?

Dax:

I have no idea what you did.

Adam:

Of FFmpeg? Okay. Well, it's someone from France. Yes. For, it's a French name.

Adam:

I knew that. I'm cultured from the Ozarks. Fabrice Bellard created FFmpeg. Have you read his Wikipedia? Because, I don't know how I never heard the name.

Adam:

This the his body of work is kind of insane.

Dax:

I think I've heard

Adam:

And still going.

Dax:

I think I've seen this pop up before. I don't remember the details, but, like, what is it?

Adam:

So well, I'm on his Wikipedia now. Obviously, FFmpeg is nuts. And I didn't know there's, like I mean, I'm sure there's a lot of contributors to FFmpeg, but that's impressive. In 2014, he proposed BPG images as a replacement for JPEG. I guess I don't know if that actually took off.

Adam:

In July 9th or 2019, released QuickJS, which is a small embeddable JavaScript engine.

Dax:

He works on QuickJS?

Adam:

What's that? Created it. And QuickJS, that's how this came up because, what is it? The new LLRT runtime is based

Dax:

on LLRT. And, it's funny. I I met a guy last week, here in Miami, and he built this really cool serverless type thing, as a very cool product. And he's trying to sort of copy around it, and he also uses Quick. Js and, like, randomly came up twice

Adam:

in a

Dax:

little while.

Adam:

Oh, really? Yeah. Yeah. So the the we're kinda spanning all these weird different things. Right?

Adam:

Like, the the he built a JavaScript engine, but also FFmpeg and also an image format. But then in 2021, this artificial, neural network based data compressor, NNCP I'm just reading this straight off the Wikipedia. But took place 1st place out of 100 in the large text compression benchmark. It uses his own artificial neural network library, libinc, which is publicly available. So a large text compression benchmark competition, he won it with this neural network based text compressor, which blows my mind.

Adam:

And then in 23, last year he released TS Zip, which sounds like a derivative of that work.

Dax:

Oh, it's a

Adam:

lossless text compressor using LLMs. Yeah. Just like AI, media, and then a JavaScript runtime. It's just such a spread of things he's done. Like, this guy's gotta be brilliant.

Adam:

Do

Dax:

you see the commonality with all of these? No. He's very good at making things small.

Adam:

Oh, interesting.

Dax:

Compression stuff. It's all, like, how

Adam:

can we

Dax:

make this small? Which is very difficult, obviously. But, like, I I mean, that's, like, the common thread. Like, and it's spanning all these different places, which is impressive. I mean, q did he invent qem qemu also?

Dax:

Because that's used a ton.

Adam:

Oh, yeah. Wow. I didn't even read the little opening. They I could have summed this up a lot better.

Dax:

And that's, like, a really fast, emulator.

Adam:

Don't know how I never heard his name. I never thought, like, who made FFmpeg? FFmpeg is one of the most impressive pieces of software that's ever been written. Like, you can do literally anything with media with FFmpeg.

Dax:

I know. It's amazing.

Adam:

You can, like, build a whole career on just using and being good at FFmpeg. You You know what I mean?

Dax:

You know what? One of the first unlocks for me with chat gpt was, like, chat gpt plus FFmpeg, and I was like, wow. I can do everything.

Adam:

Like, it can create FFmpeg commands. Yeah. Because the amount of it's like its own programming language.

Dax:

Yeah. That's cool. Do. Super impressive. Yeah.

Dax:

There's just some people like that that, are just everywhere and prolific and

Adam:

That yes. So you just said there's people like that. Now I gotta pull up this. Somebody sent me this yesterday. Well, not just somebody, a friend, Roger, who listens to the podcast.

Adam:

I know that because he said something yesterday from our podcast. Mhmm. He sent me this, individual, Johnny Kim, his Wikipedia article. This is just crazy. So he was an American or is an American US Navy Lieutenant Commander.

Adam:

He was a SEAL.

Dax:

Oh, he's a he's an astronaut, doctor?

Adam:

Yeah. Also an astronaut and a doctor. While in was it basic training? What was it? While he's, like, going through his Navy SEAL training or something, he gets his doctorate, his medical what what is it?

Dax:

I gotta read this. Yeah.

Adam:

Gets his medical doctorate and gets accepted into the astronaut program and then becomes an astronaut. Like, what in the world? Yeah. What's wrong with us?

Dax:

He's he's Korean. Right? Everyone I remember when when Yeah. When I first heard about this, the joke everyone was making was like, oh, great. Like, every Asian parent has, like, someone to compare to what impossible to compare you to.

Adam:

Yeah. Oh, I'm comparing myself. It's like, what have I done in life? I can't hardly have a podcast and get programming done at the same time. This guy was an astronaut and a Navy SEAL.

Adam:

What?

Dax:

It's nuts. Has he killed people in space?

Adam:

I mean

Dax:

And resuscitated them because he's a doctor?

Adam:

He served in the Iraq War, it looks like. He got a silver star. Wow. There's just, like there's people like that. That's what Roger said.

Adam:

There's people like that.

Dax:

Chad is saying I bet he doesn't know Rust. Someone also asked about Jellyfin. So Jellyfin is the open source version of Plex, but it sucks. I think it sucks. I don't think it's all really good.

Adam:

There's also Embi. Embi sucks too. Right?

Dax:

He's been around for a long time. Yeah.

Adam:

I love that, like, that all the hundreds and thousands of hours of work that probably hundreds of people have put into Jellyfin and MB, and we're like, yeah, it sucks. I barely use it. I installed it on my Synology NAS, and I'm like, MB sucks.

Dax:

Yeah. But, I mean, you know, you gotta get people to that moment quick or you suck.

Adam:

Yeah. That's right. It's it's a marketing thing as much as anything. I mean, it's a packaging a packaging thing.

Dax:

Yep. The other thing I forgot to bring up, we never talked about, is I tried the Apple Vision Pro.

Adam:

Oh, Apple Vision Pro. I went to ask you this first thing.

Dax:

Yeah.

Adam:

This was the first thing I wanted to talk to you about. Tell me, do I need to buy it? Is it amazing?

Dax:

So backstory, I have a friend here in Miami, and I went over to his apartment for the first time.

Adam:

Is this your rich friend?

Dax:

You know, I was He

Adam:

said everybody has to have a rich friend to go to all the rich parties.

Dax:

No. It's not it's not it's not that category of rich. But I did after hanging out with him, I was talking to Liz and I was like, oh, this is the urban version of Adam. Okay? And I'll explain why.

Dax:

So, he, like, he he, like, sold a company a couple years ago, and he's, like, working on a new one now. He's doing pretty well. Apartment is amazing. Super nice. Really well done.

Dax:

We go into his room, and he just has every single gadget, like, every single thing. And he has a he has a gaming PC. Okay? You've seen the size of, like, a typical desktop. This is the biggest PC I've seen in my life.

Dax:

It was literally, like, this tall. It was, like, it was on the ground. If you hadn't mounted it yet, it was, like I got my ribs. It was super tall. I was, like, what I was, like, oh, does it have, like, 4 GPUs in it?

Dax:

He was, like, it was just a normal PC. It has 1 GPU in it, 1 CPU in it. But for some reason, it's giant and, like, crazy water cold and stuff. And he's gonna, like, mount it in the wall, with, like, this crazy, like, built in wall stuff.

Adam:

I like this guy.

Dax:

He's got this massive painting, but oh, another thing that I loved, he had a bunch of art in his house. And for me, my whole life, I've realized if the bigger the painting is, the more I like it. And every single thing he has is these giant, giant paintings. And it sometimes looks better when it's big. You know?

Adam:

Yeah. No. I am big on the train of the bigger the thing, the better it is. Casey's always like, just because the apple is bigger doesn't make it better. Like, no, it is.

Adam:

It's better. Any True American. Fruit, any yeah. If it's bigger, it's probably better.

Dax:

Continue. Exactly.

Adam:

So he's

Dax:

again, so he he reminds me a lot of you because he just, like, has all this stuff, but

Adam:

it's, like, the urban group is just

Dax:

in an apartment.

Adam:

Because he's not from the Ozarks. Yeah. So if I move to the city, could I be the urban version of Adam? When you said he's the urban version of Adam, I was imagining, you know, tall, rugged, handsome, well spoken, those kind of things. But, no, it's the consumerism.

Dax:

Well, he also kinda looks like you too. Like, similar definitely similar category of person physically.

Adam:

But But most of the thing that reminds you of me is that

Dax:

he has a lot of stuff. Yes. He's had a lot of stuff. He's he seems, like, very particular and, like, you know, there's a lot of similarities. Mhmm.

Dax:

So he has, like, every single headset ever. Like, the Quest 2, the Quest 3, he's got, like, those be real glasses. He got he he has these are the ones that, like, they 3 d print them to your face so that the light doesn't leak in.

Adam:

Oh, yeah.

Dax:

And he obviously got the Vision Pro. And I tried it and I will say I it was as someone that has like, I have the Valve Index, like, I'm not new to VR or anything. Yeah. It was, like, a pretty crazy experience. In hindsight though, I think it was crazy because I've never the thing I was new to was the mixed reality, like, AR stuff.

Dax:

I'd only ever had, like, full VR experiences.

Adam:

Camera?

Dax:

Yeah. So the immediate what's funny was the thing that blew my mind the most is when I put it on and I looked at my hands and I was, like, moving my hands around and I was, like, touching my hands, there was, like it just looked like real life. And it was so hard for me to remember that I'm looking through a screen. But the pass through is it So it's that good. Yeah.

Dax:

Yeah. I mean, but, like, technically, it's not perfect. Like, it's a little grainy and, like, the light the color is, like, a little bit dull and it, like, changes depending if it's highlight or low light. Yeah. So, of course, it's not, like, equal to reality.

Dax:

But as someone that never experienced that, I was, like, this is very good. Like, I I really don't feel like I'm inside something. I feel like I'm just out, like, looking at people and and I can see them. And then, obviously, if you have a believable base version of reality when you start to, like, put fake stuff in there, it's like a really crazy experience. Like, he he showed me this thing, some dinosaur demo where, like, this portal opens up in front of you and a dinosaur comes out and you can interact with it and it's, like, in the room with you.

Dax:

It's really, really convincing. It's really impressive. It was really definitely a magical experience. It felt like

Adam:

Was it a meat eating dinosaur?

Dax:

Or Yes. It was. That. Well, it was funny because I was, like, touching I was, like, trying to touch it, and it snapped in my hand. And I, like, instinctively, like, pulled my hand back, and then everyone laughed at me.

Dax:

But that was so on the pure entertainment side, that was really impressive. When I dialed the thing to make it you know, how you can make it fully immersed so you can't see the real world.

Adam:

Yeah. Yeah. Like, be on the moon or whatever.

Dax:

So none of that was impressive to me because I've seen that a lot of times in VR. I don't think that experience is particularly different.

Adam:

Mhmm.

Dax:

The other thing is the controls are insane. I thought it would take me a little bit of time to get used to that.

Adam:

Like the pinching the fingers

Dax:

looking to where you wanna click. It is so fast. Like, I got the hang of it in 30 seconds and there's no interface faster than looking where you wanna do and click. There's so there's a little bit of precision issues, but I think I would just get used to it. Yeah.

Dax:

But compared to, like, all the stuff I set up on my keyboard keyboard for, like, navigating my computer quickly, it's just not gonna be, like, looking somewhere and know have the computer, like, knowing that that's what you're focused on.

Adam:

Yeah.

Dax:

Super fast. So I was like, this tech would be great in my normal desktop where if I just look somewhere and they actually do make stuff that that does that for, like, accessibility issues. But, controls are amazing. And the last thing that I wanted to try was, okay. Can this replace my monitor?

Dax:

Like, is this something that I can work on? Yeah. There's some, like, novel things with it where, so you can mirror your Mac. So you have one floating window that is your Mac. And it's really cool because you can, like, pull it up really close to your face and you can, like, it's a different experience than using a monitor.

Dax:

So I think it's hard to compare it one to one where you just have something that's replacing the rectangle that is your monitor. Mhmm. Because you can do all kind of stuff with it that you can't. You can, like, angle it really easily. You can, like, put it below you and kinda look like this while you're typing.

Dax:

But that's very interesting. The other interesting part of it is any, like, like, auxiliary apps that you have like Slack or Discord or, like, you're watching TV in the background. You can pull those out in separate windows because you can just have those as iPad apps. So you have your main app that's your Mac mirror thing, and you can just have, like, your email up here and your Slack over here and whatever else you need. You can just look at it.

Dax:

It. So you kind of that infinite spacing, I think that does work out. I don't think you can work in it all day.

Adam:

Like, too heavy or eye strain or both?

Dax:

I think it's a eye strain. At the end of the day, every single headset has this problem where you are looking at an image that's right in front of your eyes, and your eyes aren't changing focus as much as they would. Even looking at a flat screen, like, looking at different parts of it, like, your muscles are your eye muscles are moving.

Adam:

Yeah. How how does it work? Like, I know the thing is just like an inch from your fit your eyeball, but you're seeing the camera pass through at objects that are far away. Does your does your eye still not have to change focus to see those things?

Dax:

It does, and I feel my eyes doing that, but there's something that's not the same.

Adam:

And you could feel your iris dilating? Are you kidding me?

Dax:

No. Not that. It's like, like, when I was trying to, like They

Adam:

had some superpower.

Dax:

When I was trying to, like, look at the right things so I can click on it, I could feel my eyes, like, focusing differently on the thing. Yeah. Yeah. But I that's something that's just different when you're looking at, like, a static object. Yeah.

Dax:

Everyone says the eye strain is an issue. It doesn't matter who you are. Like, it just

Adam:

it just Gonna be a thing.

Dax:

Affects you. And the the comfort on the head, like, it's still, like, kind of front heavy. So I think, hypothetically, if it wasn't heavy, I think it would actually be better for your posture. If you can, like, lean back and, like, look up here. And this is, like, pretty good posture, you know?

Dax:

As opposed to, like, punching down.

Adam:

Like, as opposed to kind of, like, slumping down on your laptop. Yeah. That makes sense. Yeah. I this is so hard for me because, like, I've bought every Apple product ever.

Adam:

I've never had a headset. Like, I've never done VR, so I am new to that. Mhmm. And this is the first time I feel like I'm just kinda, like, being cautious. Because I just know I've bought so many stupid devices that I've used for a week and then they set, and I just know this is gonna be one of them for some reason.

Adam:

Not that it's a stupid device. I think it's actually this cool moment where I think we're we're maybe heading toward, better and better versions of this, and I think this is the beginning of it. But I just I do I know I'm not gonna use it. Like, I know I'm gonna use it a little, and then I'm not gonna use it again. And that Yeah.

Adam:

That's caused me to pause. Am I am I shaking my consumerism just a little bit? Just a touch.

Dax:

Yeah. It seems like it. So have you ever tried anything like this?

Adam:

I've the only VR experience I've ever had was the the the jump mania place or whatever it's called. It's like a trampoline place. We take our kids, and they have a VR thing where you throw snowballs at each other. That's it. I've done it with my son.

Dax:

This is, like, a very cool experience if you've never tried any of this stuff really. Yeah. The the downside is it's not like a gaming device. So you can't really have like, the craziest VR experiences are games. Yeah.

Dax:

And the Mac world just doesn't do that. So for gaming, it's not gonna be this device. But it is like like I said, it is like really magical. It's impractical because you're not gonna buy 4 of them for your whole family.

Adam:

Right.

Dax:

If you're not not considering buying 1. So, like, when are you gonna use it really? Like, is this gonna be kind of a solo device? But you can kinda, like, pull up cool stuff and that you can, like, share it and then each person tries it and whatever. But it's like a pure entertainment device.

Dax:

I don't think you're really gonna it's gonna be part of your daily workflow.

Adam:

Not gonna be like a replace my monitor and just work on this thing.

Dax:

Yeah.

Adam:

Yeah. I'd love to, like, like you said, dinosaurs. My son, my 8 year old, loves dinosaurs. Yeah.

Dax:

He would love it.

Adam:

Well, we watched Jurassic Park. It was probably too soon, but I I I just forever, I've wanted him to get to watch Jurassic Park, just, like, experience real dinosaurs. Forgot it's like a horror movie Mhmm. And it's very dark and rainy, and there's a giant T Rex about to eat some kids. And it was it was too much.

Adam:

We didn't make it all the way through it, and he hasn't hardly slept since. I mean, he has, but, like, every night before bed, it's like there's there's this new instilled fear in him. And when I thought about it, it's the first movie he's really seen. Like, his his poor little soul was completely untarnished from the outside world. He doesn't go to public school.

Adam:

Like, he has not experienced any negative or, like, evil things in this world. And then I set him down, and we watched Jurassic Park on my giant screen Yeah. With the crazy audio and the rumbling, like, subs and all that. It was it was too much. So if if the VR experience were like a, you know, triceratops or like a pleon eating No.

Dax:

Okay. The VR experience is not scary. Not at all. Like

Adam:

Okay. Well, if it's a t rex in the room, it might be a bit much for him. I don't know.

Dax:

I mean okay. Yeah. I'm sure they have a a a an herbivore. Is that what they're called? Yeah.

Adam:

Yeah. That's an herbivore. Yeah. Yeah. That's what we're looking for.

Adam:

The ones that the meat eaters eat.

Dax:

Yeah. So exactly.

Adam:

The big meat suit dinosaurs that just walk around and get eaten.

Dax:

Yeah. So I think I think your kids would find it, like, really fascinating. And I think there's, like again, the library of stuff, that's another weird thing. You can tell all the people that have apps for it out already didn't have a Vision Pro when they made the app.

Adam:

Oh. You could tell Apple

Dax:

didn't give them one. It was all in the simulator because it's compared to the native Apple ones, like, they're really, like, janky and not worth it.

Adam:

Yeah. Oh, another cool thing

Dax:

that I didn't realize was okay. So you know how the the controls are? You look at stuff and you click and you can, like you've seen in all the demos? Yeah. But you can also pull any window close to you and right in front of you.

Dax:

And you can touch it and, like, use it like an iPad.

Adam:

Okay.

Dax:

So you can, like, you know, scroll through stuff like it's a normal iPad app.

Adam:

Yeah. So a lot

Dax:

of those little details, like, really polished and, like, the Apple apps are very well done. But, again, like, it's a pure entertainment device that you kinda have to share with the people around you. So

Adam:

So I have a couple more, takes on this whole situation, or just thoughts. The one, I saw Zuck's response video Mhmm. And that was the first time I had seen, what is it, Quest?

Dax:

Yeah. The Quest. 3?

Adam:

Let's be real here. I don't care how much it costs. Okay. It's $500. It looks ridiculous.

Adam:

Like, if everybody is, like, trying to compete with Apple on hardware, just, like, don't be so bad at it. Like, what did what are those three little ovals on the front? It just looks hideous. Like, I can actually imagine a smaller version of this, the VisionPRO, people wearing it around unironically, and people getting used to people wearing them around Mhmm. Unironically.

Adam:

I can imagine a future, future, like, where we we wear these things, and it's not super weird. Maybe that's a dystopian future people don't want. I don't know. Time will tell. But I can remember it being weird when people had AirPods in, and it was kind of like and then it's just normal.

Adam:

And, like Yeah. The I I think I can imagine it with what Apple has built, like, the aesthetics of it. It looks like a thing that it's not a huge leap to imagine. But, like, that Quest, like, I guess it's a different thing. It's a gaming device, like you're saying.

Dax:

You never you would always do it at home. Like, in private.

Adam:

At home alone.

Dax:

It's a

Adam:

good thing. Because it's

Dax:

And they also have the Meta Ray Ban stuff. Right?

Adam:

Oh, right. Yeah. I have those. Again, devices I buy that I don't use. I wore them on the podcast once.

Adam:

I think it's the last time I wore them. Yeah. To me, it is exciting to think this is the 1st generation, and it could be the iPhone moment where this totally changes computing for the next 10 years. I know it's a long ways from that right now, like, the it's not there, but that that part of it makes me wanna get it even if I don't use it just to, like, experience it in these early days. I had the first iPhone, kind of, like, same same reasons that I would wanna have it.

Adam:

But, yeah, I just I don't think I'll use it, And that is a problem.

Dax:

Yeah. I think the thing that was interesting about Zuckerberg's thing was, the price gap is so massive. Right? It's like the $500 device and the $36100 device. Yeah.

Dax:

And I honestly can see someone getting something in the middle that's kinda compromises on everything a little bit better. Because the the final device is missing some of the stuff that Apple that the Vision Pro some stuff that makes the Vision Pro feel magical, like the eye tracking and the hand tracking and all of that. If you double the price of the Quest, I'm sure you could get those stuff in. And the quality of that is just, is is software. So if you put in the effort.

Dax:

Yeah. You can probably get close to that. And that really like, all that tracking stuff is a big part of what makes us feel seamless. So if you can get that at, like, a $1,000 price point, definitely more reasonable. But, again, still the the solo entertainment device is just a hard sell.

Dax:

Like, we don't really have anything else in that category. Like, what entertainment device do you have that's, like, a one person experience today?

Adam:

Yeah. I don't have one person entertainment experiences, Dax.

Dax:

Yeah. Because I

Adam:

have kids.

Dax:

But, But, like, even if you don't have kids, like, you it's like I don't think it's the widest market.

Adam:

Oh, you're saying, like, it's not a great category. Oh, I gotcha.

Dax:

Yeah. Yeah. Yeah. It's like, how, like, do you have anything that's in that shape of, like, it's an entertainment device that can only be used 1 person at a

Adam:

time? No. Yeah. I guess not.

Dax:

That's very expensive.

Adam:

The the idea that, like, anybody's gonna compete with Apple on this. It's it's like there's just people like me who will always I'll just always buy the Apple thing because I know it makes me feel a certain way. Not because, like, I have it and it's a status thing. It makes me feel a certain way when I hold it and when I look at it and when I touch it. And, like, looking at the Quest is just the epitome of, like, they don't care about the details, and I care about the details.

Adam:

I'm not buying the experience. I don't care what the software experience is like. I don't care if I don't use it. It's how it makes me feel, the actual product. And that's why I'll always buy Apple products.

Adam:

And I don't care. I'm not a fanboy. I don't, like, have some allegiance to Apple. Like, I just I like the way they care about every single detail, because I care about every single detail. And I know some people don't care, but I care.

Dax:

Yeah. I think that but the thing to remember though is there's some categories that Apple just is not in. And one of those is gaming, and a big use case for this is gaming.

Adam:

Do people people game on iPhone. Right? I guess that's all third party.

Dax:

I think a big use case for these headsets is, like, really high end gaming experiences.

Adam:

And

Dax:

that's just not the space that Apple is in. So

Adam:

Yeah. I'm also not interested in that space. I'm interested in us, like, having this. Maybe it's contacts. Maybe that's what it's gonna have to eventually be.

Adam:

I don't know how they're ever gonna get tech into contacts, but just like an augmented experience in everyday life, like, having digital stuff available. I don't know. Maybe that's just maybe we don't need it. I don't know. I'm talking myself out of it.

Adam:

Like, maybe that's too much. I'm

Dax:

looking for I I think, for me, the incremental step is something that replaces my monitor. I think that's, like, the first thing. Something that is, like, a no brainer. This is better than having a monitor on your desk. Know what like, we're gonna look around the world, and then there's not gonna be monitors on people's desks anymore.

Dax:

There's just gonna be a little device that's sitting there for them to pick up to work on.

Adam:

Yeah. But that's the the eye strain question. Like, is that, is that human biology that we just can't change, and there's no way there's ever gonna be a device right in front of our face that doesn't cause eye problems. That's the question, I guess.

Dax:

Yeah. I mean, the other angle is that there's no work to do anymore because AI does it. So Oh,

Adam:

that's true. Yeah. We're heading toward an an AI, what's the word? Not dystopia. An AI, what's the good version of dystopia?

Dax:

Utopia.

Adam:

Utopia. That's the one. We're just headed toward this, like, world where nobody does anything because the AI just makes all our food and does all our work.

Dax:

I, I laughed really hard the other day because, you see Ken Wheeler replying to Sam Altman being like, you're gonna turn us into fucking batteries, aren't you?

Adam:

That's awesome. What what was the I didn't see that tweet. What was the Sam Altman 7 trillion thing? I don't see the news. I just see the reaction to the news on Twitter, and sometimes I miss the context.

Dax:

I found this thing so annoying. I, like, I even tweeted about it and, like, deleted it right away because I'm, like, I'm just not gonna engage in this topic. So there is this figure going around that Sam Altman is trying to raise $7,000,000,000,000 and everyone just ran with it. Every, like, media outlet, every, like, whatever just talking about.

Adam:

Oh, he didn't say it? It's like a rumor?

Dax:

It's floating around from some talks, some presentation he gave, like, some private presentation that was related to fundraising. Yeah. He's obviously not raising $7,000,000,000,000. Like,

Adam:

what is $7,000,000,000,000? Can anyone even comprehend $7,000,000,000,000?

Dax:

If he if he was raising $7,000,000,000,000 today, that's valuing his company today at, like, let's say, like, $50,000,000,000,000 or so modestly.

Adam:

Well, I guess we do have $1,000,000,000,000 in the the public company, scale now. Like, Apple is a $1,000,000,000,000 company. Yeah.

Dax:

So if but if you're investing in that today, that means you're hoping that they become a quadrillion dollar company one day to make it. Right. Obviously, this makes no sense. There's, like this is not what was going on. I think what's going on was he was talking about the impact that AI stuff will have and the whole industry that it'll spawn and, like, all the stuff that needs to be in place as a society for that.

Dax:

And all those collectively over, like, 10 years will require, like, 1,000,000,000,000 of dollars in investment. That's my guess to, like, put that $7,000,000,000,000 thing in context. But I was I kept imagining, like, imagine you're Sam Ullman and you're giving this for this presentation to whatever And some information leaks, and the next day, you just see everyone being, like, he's raising $7,000,000,000,000. Like, we could use that to solve world hunger. Like, this is so fucked up.

Dax:

You must think that the whole world is stupid. Like, you are already building It

Adam:

really is. You're building

Dax:

this thing that's like

Adam:

The whole world is stupid, Dex. That's the thing.

Dax:

But, like, it's not great that someone that's building, something that's better than humans potentially is having this experience where the whole world is so clearly stupid, and people are so, so dumb.

Adam:

Yeah. Is it a journalism problem?

Dax:

I just, like I know some people were just joking about it. But

Adam:

Did did the public run with it? Like, did it just spread on Twitter? Or did some journalist report that you know what I'm saying?

Dax:

Like No. There was plenty of very serious Substack articles being, like, we could do better things with this money.

Adam:

Just have, like, any common sense. What I don't understand how anyone jumps there. I thought he said he wanted to raise $7,000,000,000,000. I'm like, why would he say that? That's dumb.

Adam:

Clearly, he doesn't wanna raise $7,000,000,000,000. But, no, it's just this thing where people are dumb. People are dumb. It's just it's a sad reality.

Dax:

Yeah. But, like, the problem is I'm okay with a dumb person being dumb. What's annoying is someone being dumb, but then taking the role of, like, into an intellectual that's gonna, like, write an article about

Adam:

Yes. That's what I'm saying. Like Like, take they're they're, like, dumb, but they're in positions of, like, they should know better. You know? Like, if you're gonna have a Substack and write your your newsletter, you should know better.

Adam:

Like, just know anything. It's like when I watch a movie, we just watch something the other night where like, do they have anybody consulting on, like, facts about technology? The I know it's hard for, like, being a developer and watching a movie that has a software developer in it or whatever. And, like, I'm sure Doctor is still the same when they watch Doctor stuff. But, like, why can't they just have one person in the room that's like, actually, if you put that in the movie, you look like an idiot.

Adam:

Yeah. Just like just so you know.

Dax:

Yeah.

Adam:

And maybe they do. And maybe they just do it anyway because they're, like, whatever. Nobody knows the difference.

Dax:

Well, there's the the reverse examples of this have been really great. So mister Robot is a TV show, and it's all about, like, a hacker. And it's a really surrealist, very good show. And all this stuff in there is almost, like, overly accurate. Like, they they really had people, like, you know, that knew what they were talking about.

Dax:

And it's, of course, it's dramatized and, like, he's able to do this stuff that in practice you wouldn't be able to. But, technically, it's not totally out of the question. Like, it's stuff that, like, you know, makes sense. Like, it could make sense. Silicon Valley also, I think, captures a lot of specific details.

Dax:

And I always wonder, like, who are these people that were able to describe, like, deep Silicon Valley type things that can also, like, turn that into, like, a great joke? I was, like, wondering what that process is like. Like, I don't think it's the same person that, like, deeply understands what these companies are like that's also, like, turning it into a joke.

Adam:

It's more of a relationship between the creative mind and the person who has the information.

Dax:

Yeah. Yeah. So Mike Judge used to work in software. Yeah. So I I do know the creator who had some background in it.

Dax:

But, yeah. It's, it's impressive when when media does that.

Adam:

Get it right?

Dax:

Going back for a second to people being dumb and AI, did you try out Google Gemini, the advanced one?

Adam:

No. So it is

Dax:

a it's like a it's chat gbt. It's a competitor to chat gbt.

Adam:

Didn't they do Bard?

Dax:

They renamed Bard to Gemini, and that's like their My god.

Adam:

I hate Google so much. They're just so dumb.

Dax:

I know.

Adam:

Google. The they just got rid of Google Podcast, by the way. So if you used to listen to us on Google Podcast, sorry. Google sucks. You're gonna have to listen somewhere else.

Adam:

Sorry. Continue. Gemini. So it's good?

Dax:

So they they've had a few different versions of it, and they've released their advanced version. Okay? And it's just like ChatJakBT. I tried it casually. Frank has been using it a little bit more directly, and it's good.

Dax:

It's as good as chat gbt in a lot of ways. They did some things better. Like, some of the UI is a little bit more rich. It's not just, like, raw text. Like, there's there's more interesting stuff that's in there.

Dax:

I don't know if in net that's, like, objectively good or bad. But worst case, I think it's a reasonable equivalent. Like, you could reasonably use either and have

Adam:

Like, gpt 4?

Dax:

It's like Yeah. The best experience you have on chat gpt is very similar to this.

Adam:

Okay.

Dax:

You go and try to use it, and they charge you $20 a month to sign up. Okay? And this seems

Adam:

like it makes sense. Gemini dot com? I'm going right now.

Dax:

Is that I don't know where it

Adam:

is. No. That's turn your money into crypto assets.

Dax:

Yeah. That's funny. You wanna turn your money into crypto?

Adam:

Googledot com.

Dax:

So it costs $20 a month, which Mhmm. They clearly just copied Dash EBT's pricing model. Yeah. Right. Good.

Dax:

And Dash EBT just made that number up as well. Yeah. You are a company the size of Google that has 1,000,000,000 and 1,000,000,000 and 1,000,000,000 of dollars in profit. You have this competing company that's an existential threat to you. You've built a product that's as good.

Dax:

Why are you charging for it? Literally, if you made it free oh, by the way, it can also it also has access to your emails and all Google Workspace stuff. So, like, that's like a nice integration.

Adam:

That's awesome. Yeah.

Dax:

If they made it free, literally overnight, why would anyone use chat gbt? They could just put a huge, like, obstacle in front of OpenAI. Yeah. That's a great point. They charge $20, so we should charge $20.

Dax:

It just makes no sense to me.

Adam:

That doesn't make sense. Or at least, like, why not do it cheaper? Like, yeah. Why why not compete on

Dax:

Oh, it's 2 months free. I didn't know that. Okay. That helps.

Adam:

2 months free?

Dax:

But still

Adam:

It is I mean, it's a huge threat. Right? Like

Dax:

Yeah.

Adam:

I I I do feel like it has to affect Internet search over time. Right? The LLMs and their ability to surface information.

Dax:

Yep. And, like, that's they're gonna existential threat to your company. Why are you charging this made up number for it, which I here's another funny thing. I was talking about this the other day. Someone making fun of themselves was like, I thought I was super smart, and I wasn't gonna pay chat gbd $20 a month.

Dax:

I was just gonna, like, use their API and just pay for exactly what I use. So they, like, canceled their chat. Gifd subscription and switched to using it directly to the API. And they were, like, oh, my bill ended up being higher. What that means is a point of our number is entirely made up and doesn't make sense.

Dax:

And, like, for a lot of people, OpenAI is probably losing money on anyway. So, like, copying that number makes no sense. You might as well just make it 0.

Adam:

Well, now I wanna play with it. I think I remember the announcement, the Gemini announcement when it's like the multimodal thing. It's like the first LLM that's been it's, like, trained on all these different types of not just text, but images.

Dax:

Well, that video

Adam:

is fake.

Dax:

It was completely fake. I don't know if you saw it. Oh, was it? Yeah. It was totally fake.

Dax:

And it was I'm, like, so embarrassed from them they did that.

Adam:

Why yeah. Why would they make a fake video if it's a real product? Or maybe it wasn't real yet? I don't know.

Dax:

It's not as good as it is in the video. It's actually just like chat gbt.

Adam:

Oh, okay. So the video was like vaporware, what it's gonna be like someday maybe.

Dax:

Yeah. Exactly. The, much like Ion, the, the thing that I'm annoyed by though is, you know, how back in the day there was just Netflix and you wanted to watch something? You would just open up Netflix. That was a one streaming service.

Dax:

If you wanted to have an Internet video experience, that's what you opened up. Now there's, like, 500 and you're, like, where is this thing on? And it's, like, a mess. Yeah. I really don't want that to happen with these LLMs.

Dax:

Like, I don't wanna be like, okay. I'm gonna go talk to ChatTBT. Oh, ChatTBT didn't give me a good answer. So after a minute Yeah. If I wanna

Adam:

watch this football game, I gotta sign up for Paramount Plus. What? Well, I guess the didn't the Chad GPT, didn't Sam Altman, didn't he say there was, like, this future of multiple, like, specific GPTs? I mean, they they you can create custom GPTs now. Isn't the idea that there will be didn't he even say there won't be one to rule them all?

Dax:

Yeah. But, like, you don't wanna have to tag them. So currently, BinChat GPT, there are there are, like, specialized GPTs, but you have to tag them. So you have to, like, know that I'm doing this thing, and I know that there's this specific GPT that like, the SST GPT exists and it can write SST code. I'm gonna that's not a good experience.

Dax:

You need to automatically select it and you don't want it to be forwarded.

Adam:

Can we just have a GPT that selects our GPTs for us?

Dax:

Exactly.

Adam:

Just GPTs all the way down. What is a GPT? It's like an AWS account. What is this layer of abstraction?

Dax:

Nobody knows. Nobody knows.

Adam:

So I

Dax:

was just asking how much you weigh. Complete change of topic.

Adam:

I saw that. Yeah. Did we just say that on the podcast? My weight is a fluctuating it's it's not a static number. A couple months ago, I was a £188.

Adam:

And then around Christmas, I think I was 208, and that's £20, increase. Now I'm, right now, at 198, so I'm £10 down from then. I'm I'm all over the place, and it depends on what I'm competing at, which Saturday I'm competing in the 195 division. So I will be exactly £195 in 2 days. I gotta lose £3 in 2 days.

Adam:

No big deal.

Dax:

Is that hard? I I have no context for it.

Adam:

You would think so, but it's actually not. It's really not. You just don't eat.

Dax:

Water. Oh, just don't eat.

Adam:

Well, there's just so much food in your body any given day. So if, like, you can basically lose £2, £3. I mean, depending on how big you are. I'm a 200 pound man. Like, for my wife who's a 100 and no.

Adam:

Never mind. My wife is smaller. I don't know why. Because I think the thing you don't do is, like, say other people's weight. But for somebody who's half my size, you know, maybe it's different.

Adam:

But for me, I can lose 2, £3 pretty fast. Water weight and, yeah, just eating less. That that's more than you asked for, trash. I'm sorry.

Dax:

Is anyone this is a funny thought, from my party at React Miami. It'd be funny if anyone else also did jujitsu, and then you can do a little, like, fight in my backyard.

Adam:

Will roll? Yeah. Well, Melky does.

Dax:

Oh, that's true. You wanna find

Adam:

him? Is trash coming?

Dax:

No. Trash refuses to go to Miami. Miami

Adam:

trash? Oh, that sucks.

Dax:

Every time I ask him about it, he's just like, it would be bad for me. And I just like I don't really know what what he's talking about. This is a wholesome place. So

Adam:

Okay. So me and Melky, let's do, like, a pay per view. Let's, like, let's have a featured, jujitsu fight night thing, I don't know, in Miami. That'd be fun. I've put on £2 of lean mass, by the way.

Adam:

In in a month in a month of training with my personal trainer, Sam. They scanned me at the beginning, and they just scanned me a couple days ago. £2 lean mass. So, this is probably the last time I get to compete in the 1 95 division, because I'm just gonna be huge. That's where I'm headed.

Adam:

I'm headed towards, you know, months from now. Who knows? £210. What what's

Dax:

the weight class called again? Like, the high highest one? Is it, like, super heavyweight or something?

Adam:

Super heavy yeah. Something. It's, like, 225 plus.

Dax:

So my friend who also does jujitsu, he's like he he doesn't weigh a lot. And what's the the super light ones call it? Like, featherweight or something? Featherweight. Yeah.

Dax:

Featherweight or something. It's just so, like, demeaning or emasculating.

Adam:

It really is. Yeah. I I debated, like, do I just wanna be thin and wiry and compete in the 195 forever? But it's, like, there's the absolute division, so you can compete with people in all weights. And it's like, in the real world scenario, I wanna be able to compete with anybody no matter how big they are.

Adam:

And the smaller you are, the less chance you could really compete with the big, big people. So I don't wanna be 225, but I do wanna kinda be we've spent way too long time about my weight. What are we doing? I need to go eat. It's like an we're an hour and a half into this podcast.

Adam:

Yeah.

Dax:

You need to not eat because you're trying to lose £2.

Adam:

Yeah. No. I still have to eat that. You gotta

Dax:

have a couple of nuts.

Adam:

Because I'm shaking. I'm gonna have a couple nuts. That's probably yeah. Probably what I'll

Dax:

eat. Yeah. You guys have low blood sugar right now, that's why you're shaking.

Adam:

No. I think so. I have blood sugar issues. I'm pretty sure.

Dax:

To the adrenaline?

Adam:

Not like diabetic. But anyway, let's stop talking about whatever we're talking about because this is a long one. Do I say that a lot? I feel like that's an

Dax:

hour now. 42. That's the this is the longest one maybe we've ever done.

Adam:

Maybe. Wait. 22? Not 42.

Dax:

42. 42. 42.

Adam:

An hour and 42?

Dax:

No. That's what I'm guessing my twitch stream is on for you.

Adam:

It's probably

Dax:

I was late.

Adam:

Yeah. It's less. Hour and 25. Anyway, now we're talking about that too much. Goodbye, I guess.

Adam:

What are we we're 80 episodes in. I still don't know what to say to you when we're done. Just like, see you?

Dax:

See you.

Adam:

See you next time.

Dax:

Bye bye.

Adam:

Okay.

Creators and Guests

Adam Elmore
Host
Adam Elmore
AWS DevTools Hero and co-founder @statmuse. Husband. Father. Brother. Sister?? Pet?!?
X Website Link
Dax Raad
Host
Dax Raad
building @SST_dev and @withbumi
X Website Link
Chris Enns
Editor
Chris Enns
I edit podcasts so you don't have to.
X LinkedIn Mastodon Threads Instagram Bluesky YouTube Website Link
Dax's Vision Pro Review, Cloud Pricing, Rebuilding Plex, and Moving to an AI Utopia

headphones Listen Anywhere

More Options »
Broadcast by