RobbieTheWagner 
Adam Argyle In today's modern technological world, developers face many difficult questions, artificial intelligence, autonomous agents, JavaScript frameworks, even CSS, with so much happening so quickly, how can a developer possibly keep up? Simple. Just listen to the world's most important AI podcast. Two experts, countless opinions, zero hesitation. The internet is complicated. These men are not. Each week they review the latest developments in artificial intelligence and modern web development. This is Whiskey, Web, and WhatNot. The world's most important AI podcast. Hey, what's up, everybody? Welcome to Whiskey Web and WhatNot with your host, Robbie the Wagner and Adam Thomas, our guy, the nerd. I'm here y'all. What's up?
Yeah, we are here live from React Miami, podcasting from the clerk booth. Thanks so much to clerk for your generous sponsorship to have us here. And who's our first guest today, Adam? Oh, you might know them as executive director of the React Foundation, former React Lead meta, React compiler champion, relauncher of React, conf creator of React Impact Signals, mentor, leader, speaker, photographer, strategic growth advisor. Overall, Kat, that keeps the pipelines clean, so tools and devs can flow and grow. Now working at expo is chief developer evangelist, but still continues to help react and react native remain open, sustainable and community led. This is Seth Webster, y'all.
Yeah, I need to hire you. How's it going, Seth? It's going great. Man, I'm super excited to be here at React Miami. It's my first time both to react Miami and to Miami. So it's been amazing so far. This event is blown my mind in multiple ways, even in the first few hours. Yeah, it makes two of us, man. Yeah, it's a great event. They do a great job. Miami's always fun. Yeah, a little loud for me. Like every event is like, what if the music was super loud? You know, you get to take it up every I get out to the last conference. Yeah, that's true. Sometimes volume is your only choice. Yeah. All right, so we have a giant list of hot takes. I don't know if we're going to get to all of them, but I want to start with one that I feel like you'll have opinions on. All right, is React a framework?
I thought I was done with this conversation. You know, I think, I think, you know, just depends on where you draw the fault lines, man. Like to me, React is meant to be tiny. Yeah, it's meant to be a library. But over time, as you know, the needs of users and developers have been surfaced more and more stuff has been added to react. And I think there's a growing feeling among the react. The folks that are actually still working on react is is actually, you know, getting back to that react core and react native core such that such that the library. The library is very thin and that it's easier to bolt things on and it's easier to use these scape hatches and so forth. So I think, I think generally speaking, and it's done what every piece of software throughout time and memorial has done, which is to grow as the needs grow.
And I think it's a good time. It has been a good time for a while to look back at scaling that back to the core and putting more things out into the community for community management and that's something foundations looking hard at. One of the projects I'm most excited about this year is actually working on. I have a bunch of companies that have stepped up to sponsor stable API work for React Native, which basically means that we're going to have this API that platforms can target. And no longer be victims of every time we do a react native release and it's going to three months catching up, they'll be able to target a stable API and we're going to send for that API such that you can depend on it as long as you're targeting a specific version.
So like stuff like that's like long overdue and gets us back to being a library and less of a framework, but I would say we've drifted pretty, pretty squarely into the framework space over the years. Yeah, so it depends, you would say. Yeah, I mean, they, you know, I guess I'm a little bit of a politician. So, you know, they got it. It was at the very end, but that's what I think I think the right thing is to be a library. Yes, yeah. Can you move your mic just a little further this way, like like tilt. Yeah, cool. Because if you're facing, I'm just trying to get as much sound as we can anyway.
That's good. Sorry. Yeah, you're good. Yeah. What about what do you think about React server components? Was it a good idea? I mean, yes, I was championing that from the inside of the original code name was Flight and it was worked on by Lauren Tan and I think Dan Abramov worked on it. And the original, the original goal was the same as it is today, which is to give React the full stack story. You know, the, you know, clear indications after many years of developing spas was that there was no like canonical way to do data fetching and to do secure data fetching close to the server. You're always having to wire something up. You're having to do API keys and bear tokens and all this stuff to do relatively simple things that Ruby on Rails developers were like, just do this.
You know, it's so easy. And so I think, you know, the in spirit, it was definitely a good idea. One thing that was, you know, super hard in the development of it is that meta has such bespoke web infrastructure. We could not use server components at meta. And so we had to find an external partner to take on the work of like integrating it into some kind of full stack. And that's where Verself stepped in and it did an amazing job with the first version. Of course, being a being a friend and framework, we didn't have, we don't have still probably the muscles to understand like the full stack and like the security implications.
And so now you're seeing because of that that we have for the first time in React's history, we're vulnerable to exploits, which we weren't before. And so I think generally it's a good idea. I think having that story and being able to say that we can operate anywhere from the server to the client and also like, you know, removed likelihood of leaking secret keys to the web, I think is an important step forward for sure. I don't know that the implementation has reached, excuse me, I don't know that the current like implementation of server components has reached its max yet. I don't know where necessary or like server components prime, but I do think that the continued development there is going to only make things better.
The biggest thing that I think is confusing for people now is like things like caching and so forth. It's like there's not a really clear and Verself trying some things, but there's not really a clear like all encompassing story for caching that that really works for everybody. And so I think you're going to see evolution there, but I do think it was a really good idea. And I just think it's as even though it's been out for a couple of years, it's still early days because it hasn't had that many iterations. Yeah, I do think on the one hand, I think it's nice to be able to use the same things on the client and the server. On the other hand, I like the separation of like, I know for sure which one I'm operating on because it's a totally different language or something for server stuff.
Yeah, I can see arguments for both ways, but yeah, I think you know, when you think about it from like the artisan or craftsperson perspective, I think I definitely agree with you that it's like nice to have the ability to turn and tweak all the knobs. I think where we found that server components has been such a huge hit is a companies that you know are looking to move really fast and want to hire like more of a single archetype of developer and not have people not have to have not having to have the conversation about are you a full stack developer. Right, are you are you do you know reacts good that's enough and I think that's that's a big hiring advantage for companies because the react market is very strong.
It's one of the strongest markets in all of engineering and so I think you know that's where companies let lean and that for me means that we have more companies available to sponsor react and support react and so I want to see that virtue of cycle working. Yeah, yeah, that's definitely been so I'm an Ember guy. Oh my gosh, I was an Ember guy before react nice nice Lauren Tan was as well. Yeah, I'm familiar. Yeah, but yeah, that's a lot of places I've worked, you know, it's like we're chugging along they're like it has to be react now because you can't hire Ember developers and I'm like I hear you but you can also teach any developer.
If you're a good developer, you can learn any front-end framework so it's like I don't know how do you feel about that argument. I think I think it depends again sort of coming back to the idea of like are you a software engineer. That's a crafts person and you're really caring about spending your lifetime minutes in the code or are you a builder or are you someone a company is hired to build something and the two equations are very different. Right, if you're if you're building a table like right here, you decide to be a carpenter, you want to do woodworking. If you're building that table, you're going to obsess over every corner, every edge, you know, every nook and cranny, every joint and so forth.
But if someone if you go to, you know, restoration hardware bedbath and beyond to buy a table or something, I don't think they have tables at bedbath and be out. If you went to front-end framework, you're showing my age. So if you go to these companies, like you're not going to have the same attention to detail on this table you're buying as the table that you're building. And I think this is how companies feel about software development. I mean, I was reading a series of tweets on the bad place today that was like nobody cares the code is messy. And ultimately nobody cares what the code is messing. That's where the crafts people come in. And so I think there's still room for crafts in software, but I think that companies, expectations of you as a developer have changed.
And what they want is speed, they want efficiency, and they want a delightful outcome. And for me, that's where frameworks like Ember and React and others, you know, I think of rising tide lifts all boats are so so powerful because they provide primitive tools. Because they provide primitives that AI can use to lead to a delightful outcome, where I think if AI were forced to live in the completely imperative space, as we've seen, like you see this all the time, it's just not as good. There's lots of edge cases, there's lots of bugs, and that's where frameworks are so important, even for AI. React has an advantage because in the training set, we have the preponderance of code.
And so the models understand React better than I think they build terrible react. I don't know if you've looked at the code. Oh, yeah. I mean, it's not limited to react. They build terrible code in general. True. But it's still that question of like does it matter if it's done, right? And that's how business is thinking about it. And so I think we have to as a discipline, you know, as a center of excellence in engineering, we have to accept that there is that trade off. And you have to decide when you need to lean into the code and craft, and when you need to lean into the get it done. Yeah, Adam and I have talked about this some kind of like making an analogy to like a car, an engine of like, okay, so if you vibe code your engine, right?
And then you take it to a mechanic, see me like, what? I don't know how to. So I think that's the only time that it matters if the code is messy or not is like, if you're going to need someone to go fix it later, is it readable at all? But maybe we're getting to an age where that doesn't matter. Like if AI gets good enough to where it's building stuff and you're sure it works, maybe you never care. And it'll fix its own problems. Is there a cliff will fall off where it won't be able to fix it? I think about it a little bit differently. This is also how I think about the why would you choose in this age where it's super easy to just ask Claude to convert your iOS app to Android?
Like why would you choose a cross platform framework? No, that's true. It's the same question to me actually, which is that it doesn't, the economics of tokens right now are unbelievably unsustainable. Right, they're so heavily subsidized by the big enterprise companies paying open AI and anthropic. I think this is more true of anthropic than open AI, but you have these big enterprise contracts that are subsidizing our tokens in order that we can afford them. At some point, the economics is going to have to be worked out and I think tokens are either going to be driven to commodity prices by open models, which is to be transparent what I hope happens.
But these tokenomics are still going to be a factor for a while. So if I'm having to pay for three or four or five times the number of tokens to get cross platform, or if I'm having to pay for many more tokens in order to achieve like, you know, perfectly crafted outcome or something, then I'm going to be spending way more money on my product than my competitor who chose to do something that was already codified like reactor, ember or angular or whatever this very well-worn practices and we know how to do things. I was talking about in my talk this morning, my brief talk was like, we're in this phase of like refiguring everything out.
So all we had is all figured out and buttoned up, right? And now we're like, wait, how do we do this now? Because we're not actually writing the code. How do we cure it? And so I think it's going to be about that you can't abdicate your responsibility. The code the agent writes is your code. And so if it's not writing the code, it should be writing for long-term maintainability and to reduce later maintenance tokens cost, that's on you. So you have to be thinking about like the stuff to just like we've always done is engineering. This hasn't changed. Think about the future. Who's going to pick this up later, even if it's an agent.
You want that agent to be able to pick it up and understand and work with it quickly because that's much less expensive. Yeah, that's true. I think about it too. Like we've got factories which make things out of plastic. They've got molds, right? Everything's pre-made, fast to make, cheap to make. You know, it's got the fast and cheap side down, but you got a crafter who's kind of got the opposite thing going on, right? It's slow, but the quality is super high. You can get something bespoke that perfectly fits your scenario. And I think we're just going to have those tensions forever components. I've seen so many times where you pull them off the shelf and they seem to just be exactly what you wanted.
And then a year in, you're like, we want to modify this one thing a bit. You're like, oh crap, it's a black box to us though. We bought it off the shelf. We got to go off the rails now and build our own thing. And so you've got this pendulum that just won't stop swinging and everyone wants the pendulum to stop except until you want something bespoke. They both need to exist just like they do building homes. You got people building homes that are cookie cutter. You got crafters still. And so there's just tension between factory creation of something and bespoke and I kind of like it. Yeah, I need both. I want to go to a local market and get something handmade, but I also want to go pick up something fast from the store so they can both coexist.
I think that tension drives innovation. I mean, I think it drives us to innovate because you know, you get started, you get in, you get your product moving and then you have a need that isn't served by your component. So you have to invent, right? And so I think that tension is incredibly healthy. That's why I love the fact that there's multiple competing reactive matter frameworks. At some point, the remix guys are like, you know, React is not for us right now. React is no longer for us. We're going to we're going to fork off and do this thing. And they're doing really interesting things that all of us can learn from. So, you know, to me, I've never been somebody who thinks that that innovate or the competition is like this thing to be resisted or to be to be squashed.
That's something that to me celebrate, like people are out there inventing and creating and building. That's going to be something I'm always excited about. Hell yeah, I love that. Yeah. Okay. Is HTML programming language? You're trying to get me in trouble. Sure. Why not? Okay. There's CSS. I mean, sorry to all the designers. I don't think it's as much of a programming language as it is, you know, a declarative manifest, but we have things like variables and they have some ability to do functions and stuff now. Yeah. It can do calculations. It's it's it's blurrier than it used to be. I mean, I don't particularly use CSS that way. You know, I'm definitely on the tailwind bag and bandwagon for the most part.
Hell yeah. What do you get called yesterday fail wind? Yeah, I never know. Not happy about tailwind yesterday. I didn't I did not hear the I did not hear the confuffle. Yeah, there was, um, were you at the party thing last night? Yeah, I was there briefly. Yeah. Okay. Yeah. So when they did the thing on stage, something about tailwind and everyone was just like, like, no tailwind, like, like, learns CSS. Like, okay, it's it is CSS. But all right. Yeah. You know, I mean, like, I've been I've been a big advocate for semantic approaches to programming and declarative approaches to programming. And I think that the, um, I think folks that are hanging on to like being able to tell the computer what to do.
Um, I think you're missing out. Like, it's just much better to describe state and the things you want to see when a particular state is present. I feel like that's what you get with with tailwind. You know, it's very semantic and declarative. And I can read it. Yeah. The year later and still understand what it's doing. Yeah. I don't have to hunt through six different files to figure it out. It's right there. Uh, to me, it's, I'm sorry to ever, I'm sorry to the tailwind team for people calling it fail. Yeah. Yeah. Um, all right. So you mentioned a few things that you wanted to chat about. And I think this is a good one. Um, do you think when you frameworks at all is, is AI getting rid of the need for any of this?
Uh, I think I kind of already answered this accidentally, forgetting that we had this question on the list, but I do think frameworks will play a vital part of even the agentic programming world simply because it provides a well thought out and well tested primitives for AI to use. If AI is writing everything from scratch, you're going to end up with a slightly different version of everything every, every, every single time because it's not to turn on deterministic. And so wouldn't we rather know that the boundaries are the same so that we can plug and play pieces. Right. So if you're writing using AI to generate your team's design system and then that design system needs to be plugged into a different application is gets developed later because you have developed it with a framework that provides contracts and the way you expose those things to other.
Applications, you're going to be able to drop that in in music. If it were slightly different than you're always going to have to AI be like AI retrofit this thing and then, thus, you're spending more tokens and it's the same thing for cross platform and so forth. The fact that you have people like, you know, I'll shout out shout out to my expo team, obsessing about like the smallest details of how an API works on, you know, the, you know, to adopt. iOS is new glass views or something right like they've obsessed about those details so that the animations are fluid and buttery and smooth and the refractions are great and all that stuff right people have obsessed about that.
So the agent doesn't have to and so the agent who's quickly by just dropping something in and now you're not waiting for five minutes while code cloud comes up with something you're waiting for five seconds or 15 seconds. So I think frameworks still play an incredibly important part and I would rather see agents using and this is what you're generally seeing that the agents are smart enough that the models are smart enough to use well, you know, well to find libraries or well, well respected libraries. I don't know if they're doing it based on forks or star count or what but they're picking they're definitely picking good libraries for the most part.
You do have to educate it and that's where I think the community has such a huge role to play and like, you know, creating the MCP of how, how you think right and creating creating the skills of how you think about those things and to help people get the same for similar results. But ultimately, I think frameworks are here to stay and going to play a vital role for as far as I can see into the future, which I spend a lot of time thinking pretty far in the future. Yeah, it's almost like the framework that was made to optimize all of us to keep us all within the same kind of confounds and bounce and APIs so that things were predictable as you move forward.
Those also play into what LLM's want. They need to know that there's rails. They need to know what's a copycat. They need to follow the framework. No, but you know, like I'm in on rails like you're in a cart, you know, I'm just giving you shit. It was great. I was about to go Donkey Kong on you. Remember that level you're on rails. But yeah, all those things you need to provide those patterns and it's either going to discover them from your your mess that you've made as a code base or it's going to try to pull them from its memory. But if you're there providing them really explicitly and a framework can help do that even more, then there's success on that horizon naturally.
So the same things that empowered your team are now empowering your agents. I think that's important to continue on. 100% agree. I think, you know, the thing I've learned about using Claude and others is that they do much better just like human beings. They do just way better with constraints. You know, if you just say, build me a website, what are you going to get out of Claude? But if you give it all of the different things that you want to see, the stack it should use, the best practices it should apply, it does way better job. If you heard that you're not supposed to be mean to your LLM though.
No, I personally have found they responded really well to like harsh criticism. Yeah, you and Wes Boss saying there's a lot of people out there cousin and then you'd see in the Claude code leak that it tracks all of your cuss words at it. They're like monitoring how angry people are. Yeah, I've not seen that, but I would I would imagine I'm on a list somewhere. Yeah, I do like the Claude now. I mean, so I usually use open code, but I've been using Claude code because I got the like free open source maintainer one or whatever. I was going to mention it all the time. I didn't even get that. I'm just important, I guess. No, I think I was one of the first people like within minutes of them posting it.
I applied. So I think that was what it's news. But now it asks you in your Claude session. I don't know if it's based on how long you've been talking to it or like, how's your session going? Like rated one to four or whatever. So if I'm having a bad time, I'm like one, you know, like terrible. I think I think that's super smart that they put that in there, but it's always I have I have answered. I've given a lie so many times because they get me just the right wrong time where I actually hit like trying to type 20 and I give them a heads going fine. Yeah, when I really am not happy. And then other times I've given it like a really bad rating when it's kicking ass.
You're confusing. Yes, I'm very confusing. Claude does not understand me. Yeah. So I do encourage. Reverb back to like the the previous point kind of for a minute. Do you think I feel like as a developer or as an AI consuming JavaScript error messages and things like I feel like we do a really bad job of telling you what went wrong. It's like, hey, something broke and you're like, cool. Let me manually step through everything to find where it actually broke. Do you think that there will be a push to improve a lot of that so that like because if you're using rust, like LLMs can just build you whatever you want in rust because it tells you exactly what went wrong and how and it can just go to town.
Like if we could get that with like various JavaScript things, I feel like that could be really helpful. Yeah, I think it's kind of ironic that the impetus to improve error messages is going to be the machines are going to be reading them. I think people have really tried. I don't know. Over the years, I mean, the reacting has gone through so many different permutations of how to provide error messages. And for a while, there were some like actual technical limitations that prevented us from like getting deep enough into the stack of what happened to give you a better error message. But a lot of that's been fixed. I mean, I was watching Aurora's talk today and the couple of the error messages that came up when she was like in the middle of a compile or something were so much better than they used to be.
And I think for cells led the way on some really good improvements to error messaging. But I absolutely do think that it's incumbent upon all of us. Again, like we should all have a mindset of how do we make? I hope that we all agree that AI programming with AI is something that amplifies our ability to deliver. And so we want to all be focused on like making our tools and so forth really good for AI's as much as we want to focus on making them good for humans. And I think the balance is going to shift such that eventually we might care more about making tools work well for AI than we do about people because people are using a completely different user interface than they used to.
They're not reading the error messages like the user or their copy pasting them in an ideal world. You have an integrated system, an IDE kind of from from the bottom to the top that is watching the error messages for you. And what I always do is I always tell Claude or whoever I'm using spawn these tasks in the background so you can monitor the logs. I don't want to do copy paste. And so I think what we want is systems that do all of those kinds of things by default. And in order for those to be successful, we're going to have to be really good as two builders in making the agents completely understand what it is that's wrong.
And you're right about last. That's just a wonderful experience using an agent to build for us. And I think it's going to be a great experience. I think it's going to be a great experience. I think it's going to be a great experience. I think it's going to be a great experience. I think it's going to be a great experience. I think it's going to be a great experience. I think it's going to be a great experience. I think it's going to be a great experience. I think it's going to be a great experience. I think it's going to be a great experience. I think it's going to be a great experience. I think it's going to be a great experience.
Yesterday, I think somebody opened source the same thing for iOS simulator builds. So you can now control, automate iOS simulator builds and get the error messages back from that. Wow. And that has React Dev Tools built in, which is based on Chrome MCP. So I'm imagining you can probably automate that with the Chrome MCP too. Don't take my word for it. All right. So what are some of the goals for the React Foundation? Yeah. Let's just stop with that. I answered that one first, I guess. I think the big one for this year is making sure that we transition out of being a single company, steward into a multi company foundation.
Successfully. You know, it's a lot of voices. The founding partners all have different perspectives on thing. They've been great partners so far. But they all have different perspectives and different priorities. And so making sure that we're able to successfully take care of everybody. It's going to be a challenge for a framework that's never done that. I talk a lot when I talk about the story of the React Foundation. For those that don't know, it took over four years to get it done, which is a long time. And that's like a web standard. Twice that. Twice that for web standard. We were first looking at it and starting to like write, you know, put pen to paper.
What we recognized pretty quickly was like the React team itself was not. Ed Metta was not ready to operate like a foundation. We were so used to Metta's tools. We were so used to being able to move at a breakneck speed and only like release what we had done. We've been using for two years. We release it two years later. And so there are a lot of different like points of dissonance when they came to being able to operate as a foundation. And so we pulled back from doing the foundation for a couple years and said let's fix. Let's fix everything internally first such that our interface boundary is better.
That happened. Things like, you know, React Native went from releasing like on a very ad hoc schedule and not very often to doing six reliable releases a year. React core is like one solid release a year. And so and where it's like before you could be like, I, you know, I don't know when React is going to do a release. But you know, at least it's super backwards compatible, right? So like there's there's different equities people are bouncing. So that's kind of coming together. So the big goals this year is kind of to come back to your question really about like stability. Stabilizing React in a foundation.
For the first time supporting financially maintainers events around the world communities. We want to help build communities where they're struggling to get those started because of resources. So things like for me, Asia is a big market that I want to expand to. We've never done a good job in Asia of localizing our documentation or reaching out to them. And then there's something like 10 million developers in China alone that are not counted in our surveys that are not among the people we target with things. So I really was just there in China and they are so excited about React and like we want to be involved.
We want to help with stable API, which is the big project I'm excited about for React Native this year. That will empower more out of tree platforms, which I want to get rid of that term. All platforms are platforms that they want to be part of react. They're welcome. But we're going to basically rebuild the API layer such that the React Native things sit atop that as opposed to being forks. Every platform is a fork that's going to take a while. We're going to start on that this year such that the operating systems in China can use React just as easily as macOS, iOS, Android, and Windows. So there's a big ones.
But the things I'm most excited about are actually starting to give back to the community because for too many years people have given so much blood sweat and tears to this project. And other than like accolades and appreciation, they haven't gotten a lot out of it. And so I really want to start turning that around. I want to see some new communities spring up. Those are the big ones for this year. Hopefully we can get React Comp again this year. The only reason that it's a hope and not a promise is that we want to make sure we have an event that's worth everybody attending. And so I'm in the process now.
We're in the process now of curating what the program would be and making sure that it's up to the standards. The high standard that we set for React Comp. So yeah, those are the good ones. Nice. Yeah. So going back to your point a second ago about giving back to open source and stuff. Like do you think in general contributing to open source is helpful for your career? Like let me clarify, like directly contributes to getting hired somewhere versus is just like learning. Like you definitely learn from it. Do you think it like anyone goes, oh, I know that guy from open source. Let me hire him or like.
I mean, I can think of dozens of examples of what you just described actually happening. Yeah. I think it would be pretty heavily biased if I just relied on those memories. So like anything else, yes, but it depends. It depends on your ability to network your ability to make sure that you're seen in that community. I think that there are a lot of unsung heroes in open source that are brilliant engineers. I have given just as much of themselves as other people, but nobody knows their name. Yeah. Nobody knows what they've done. And that's because they're not maybe into self promotion or someone else hasn't come along and sponsored them to say like,
look at the great work this person is doing. Right. Everybody else know. And I think that you know, it's it's just like any other skill. You have to know how to make sure that your work is appreciated. Yeah. And there are too many people are focused on getting the work done and not being recognized. And it's sad in some ways that like you have to be promoting yourself. But if we're taking the lens of does it help me get hired? I think it does not by itself get you let you help you get hired in the last someone sponsors you. Or you are an advocate for yourself. And the not to open a whole can of worms that I'm going to get in trouble for.
But I think the sort of newest generation of software developers that I've seen are not great itself promotion. They're not a great itself advocacy. Right. And need sponsorship more than ever. And so I would call upon those folks to develop that skill. But moreover, I would call on the engineers who are a little further in their career to be a sponsor. Be an advocate because it would be a shame to lose the kind of amazing engineering into the shadows that we're seeing from some of these folks because they didn't have someone. You know, vocalizing for them that they're doing good work. So advocacy and all is always been an important part of every career for every person always needed somebody to give them an opportunity.
And this is the opportunity we have to advocate for people. Yeah, yeah, it's always been a problem of like, you know, there's every every few years somebody blows up like the fake or JS thing was a big thing. Or like, you know, somebody's like, hey, all these companies are using my thing for free and giving me no money. And I've spent a ton of time on it. It's like, yeah, there's got to be a fix for that. But there's I want to fix that problem. And I also want to fix the problem of, you know, AI code Providence. I mean, like the the elements are spitting out full copies of open source libraries rather than pulling them off the shelf.
And I think we have to fix that. I think we also have to fix. For example, I mean, if you think about it from reacts perspective and I'm not somebody who spends time wasting my energy on being frustrated. But if you think about, if you go and you ask cloud or any of these systems pretty much to build you any commerce website, what technology will they use? With that sentence, they'll use react. Yeah. And they'll use probably next. You used to be tailwind, but they also, they don't necessarily love tailwind as much anymore. I hope that's not true. But, you know, react isn't seeing anything from open AI or or anthropic.
And no one is so rampant theft of everything. Exactly. And so even, you know, even the big ones like us are struggling with the same problem as the small ones. And I think that as an industry, we need to stand together and figure out how to solve this problem. I don't have any good solutions for you. Like the things I'm thinking about are, as we talked about earlier, like cloud and codex and other code agents. They don't write very good react code. Can we provide the e-vows, can we provide the post training data or pre training data to help generate code that is performance? It's not using refs for state tracking because it doesn't like our life cycle and so forth.
So, you know, there's opportunities there, I think maybe to partner with the AI companies in order to develop, help them develop products that are better for their customers. And then it's a win-win, right? So I'm looking, trying to look at it from different angles and see, you know, how can we all benefit from this? You know, it does feel like a gold rush, right? It feels like this vibrant, this energy is just percolating everywhere. It feels like a really good time. But it's also important that we keep our minds and keep ourselves focused on the outcomes that affect human beings. And this is what I said earlier too.
It's like, you can look at the efficiency gains you get from your engineers at your company using AI as an opportunity to save money. I think the smart money is on using the opportunity and the efficiency that you're saving to do more of the things you've been unable to do throughout the years because you have to make hard decisions of prioritization. You still have to be good about prioritizing, but I think this opens the door to like, explore those business vectors that you've not been prioritizing because you were scared. Like, now you can do it. Now you can afford it. You have 10 times as much money as you did before because you're engineers with 10 times more valuable.
Yeah. So, laying people off is not the way to save money. It's build more, like Adam and I talked about this a lot of like, you probably have a lot of SaaS products you use, right? Make a whole team, have them build that. That'll save you some money. Like firing people is such a short-sighted way of saving money. Yeah. I agree. I think also the thing you just said, it's something like I worry about a little bit which is like, how will I reshape the SaaS market? There's so many brilliant tools out there that... Yeah, like clerk. Check it out. Like, I'm a clerk, I'm a clerk customer. Yeah. But I have, I have also rolled my own off.
I've been in a lot of customer, a lot of zero customers as well. And I just like so, you know, not wanting to go through the experience of setting up off, setting up off again. I'm like, I'm just going to do magic links and that's it because it's easy. But then you guys decided to add billing and I was like, now I'm like, now I'm like interested again, right? Yeah. But I have a few apps running with clerk and a great job. I mean, what a great product clerk is. I wonder if we can learn from the music industry here where they kind of went off the wall where if you play eight seconds of a song, you got to open ease to the artists.
And there's an amount of that that feels pretty silly. But at the same time, at least they're giving attribution and paying back for using somebody else's stuff. And I've been sent DMs already, which people said, hey, I invoked this LLM to make a design. And I literally saw your name show up in the results of either the styles that were written or my name was in the HTML that was producing stuff like that. And I'm like, I don't think I want pennies to start trickling towards me or whatever. But, you know, CSS tricks did the same thing. Chris Coyers, like I built 10 years of CSS content for LLMs to strip it and steal it
in a couple of days and give me nothing. So we got the music industry that's like got trolls to pop up over all these bridges and they go, yeah, you used a fraction of our song. You know, you must pay BMG or whatever it is. I don't want to do that. But I wonder if there is something that we can learn from them as being sort of too far in that pendulum swing. Maybe AI can do something else. And yeah, instead of taking the money that you're saving and you're enhancing and pocketing it as a big company, giving some of that back to what you clearly are borrowing. I don't know. Yeah, I mean, I think it's definitely something I want to work on figuring out.
I'm in a good position to do so at the foundation. So for context, those that don't know, Linux, Linux Foundation is the parent company of React Foundation now. And one of the big efforts they've started this year is the Agente AI Foundation, which is looking at way more than this, but looking at all of these types of things. You know, if you take a step back and you think about those of us who've spent a long career in the developer tooling space, we are excited about helping other people build things. And so I don't need the pennies to trickle into me because I am the executive director of the React Foundation.
What I want is for the developers who built the code that Claude was trained on to be rewarded for the energy and effort and time and lifetime minutes they spent on that. Because my whole North Star was helping other people bring their ideas to life, in this case, helping other people bring their ideas to life that help other people bring their ideas to life, that, like I said, that chain of providence has to be preserved, such that people can be at the bare minimum recognized for their contributions. And in a better world rewarded for those contributions. You know, rewarding record recognition, go hand in hand.
And I think that, you know, I think we can all agree, I don't know if this is true, but I think we can all agree that from helping people bring the ideas into the world perspective, AI is a huge unlock. And as somebody who's spent a career helping people do that, I'm so inspired and excited about this arrow. But it comes with these trade-offs that I think we have to reckon with. Otherwise, what will we be building upon 10 years from now? You know, it won't be novel, it won't be innovative, it won't be exciting. It'll be, again, back to what I talked about in my talk this morning, regurgitation of the same ideas over and over.
Yeah. I could sit in a different, in a form, but it's still just all, you know, shitty ice cream. So I really want to see us, I really want to see us figure this problem out. I call that the mainstream just getting wider, where you've got a mainstream, and it's popular, but as we regurgitate more and more mainstream, the mainstream just gets wider, and the indies and the edges get smaller and smaller. That's true. And I think that's dangerous. The same way that you don't want pure factories, or you don't want pure craft. There's always a healthy middle ground here, and we've got to find some balance.
So I'm happy to hear you say that, man. I do think there's a lot of catching up we have to do as an industry. Like, you know, Adam made a nice marketing page for us the other day. We have a couple of prompts done better than, like, in my opinion, our website. So I was like, huh, like, when that's that good, and there's still websites that suck, it's like, that we need to just raise everybody up to that good level first. Let's work on that. Then we can figure out what's next. I agree. You know, I think we've talked a lot about this at Expo too. It's like, what's going to be the differentiating factor in the future?
Or now the future is now the future is here. And I think it still comes down to like taste, you know, knowing what you want to create and what you want to be different about that. And I think that that's going to be a big one. I think longevity and stability are going to be things that are going to be highly prized and valued, you know, sticking around. It's going to be harder than it used to be. Like, entire market verticals are going to get saturated with players. And at that point, like, how do you differentiate it? If every hotel has the same or like, really similar or like, really awesome digital experiences,
how will they differentiate themselves from one another? You know, remember when we got the phone based hotel room key, and like, only a couple to it and it's helping or whatever that has this, it's awesome. But if every place has that, you're not going to feel like, oh, that's so cool anymore, right? It's just like, why find an airplane the first time you ever used it? You're like, oh, I got to just shift the diff from 30,000 feet. That's so cool. And now you're grumpy anytime it doesn't work, right? You're not amazed by it anymore. You're just grumpy. And so we're going to get to that same place, I think, with the saturation in every niche,
which I'm hopeful about because I think what it does is it pushes us back into the analog world of differentiation. It puts us, pushes us to reach different heights of differentiation. The weird web, right? Yeah, going back to Photoshop slices and throwing it all on the floor. Please God, no. That is, I don't want to slice any corners. But I think there's a really interesting shift coming because we're going to democratize so much more of this development that tastes and differentiation is going to be even more huge. And I'm hopeful that it actually pushes us back to some degree into the analog world
where I think differentiation matters much more. Yeah. Yeah, I hadn't considered, like, the sole way I decide is this resort dope. I'm like, is the website really good? If you put a lot of money into the website, then I know you probably put a lot of money into your experience there. So if everyone gets that, then yeah, you're right. It's like, how do I know if it's a good one or not? Exactly. You're all going to know by trying. Yeah. There is high correlation between a well thought out website where the content is there. It's beautiful and the experience that you get somewhere. So I don't think that's ever bit me really where I've gone to a beautiful website.
And then I showed up to the brick and mortar store and I'm like, what the? I feel like that's a movie in the making. I think I should write that. Yeah. All right. We are about at time here. Is there anything that we miss? Do you want to mention plug, etc? If you have any interest for everybody out there, if you have any interest in participating in the activities at the foundation, definitely let us know. There's a ton of great stuff, you know, reach out to me. You can, there's a contact form on the website. That generally goes to me. So reach out. I'd love to hear from you. There's a lot of activities that we need help with on the website.
I mean, on the, at the foundation, the last thing I'd say is like, I really want to encourage folks to consider this moment. Like a call to action to start building. I don't think we need to have 70 million of the same things out there. But getting really good at commanding agents and commanding your fleet of agents and learning how to parallelize your work and learning how to work on five or six different things in the same day at the same time is not a skill. A lot of engineers have who were much more used to going deep and focusing. So start practicing this because this is where the industry is going.
And it's, you know, the way I think about it right now is there's never been a better time to be a builder. And so I'd love to see, you know, many more ideas, you know, leverage expo, leverage these framers, leverage and acts, leverage, you know, tan stack, whatever your, whatever your favorite is. To get your idea out there and get it into hands because you're not going to learn so much more by actually shipping something. Especially if it's a project of passion for yourself, then you will, uh, by reading blogs, sitting there reading blogs about air. Or just get your hands dirty, get in there, make all the mistakes we're all making and help us figure out this next air of software development.
Hell yeah. Love that. Yeah. I got one final request though. All right. Will you say whiskey, web and whatnot in your dope ass voice? You're listening to whiskey, web and whatnot and react by me. Yeah. Nice. Is it all right? Thanks guys. It was really great to be here. Oh my gosh. Yeah. Thanks for coming on. Thanks everyone for listening. If you liked it, please subscribe, leave some comments and reviews. We appreciate it. And we will catch you in, well, I would say next time, but in a few minutes because we'll be back on. Yeah. With some other, some other requests. Yes, indeed. All right. This is great.
Congratulations, citizen. You are now prepared to face the challenges of the modern technological world. With the proper tools, a steady hand and a reasonably configured terminal, any developer can orchestrate AI agents with confidence. You'll do just fine. Thank you for listening to whiskey, web and whatnot, the world's most important AI podcast. For additional resources including official merchandise, visit whiskey.fund. And remember to subscribe, rate and review so this vital information may reach others because the future of technology depends on it. You