Daniel Whitenack: Yeah, definitely. I think that there’s a sort of general trend in science where these sorts of AI techniques are being applied, that are making an impact. I’ve mentioned a couple times on the podcast where machine learning started to make an impact on the field that I was studying when I was in grad school… But I think the pattern was similar there, and that’s the sense that we have these experimental observations which are rooted in reality, and we have a little bit of knowledge about how things work, and constance, and other things, and certain laws that shouldn’t be violated, so constraints, essentially…

But the relationships between input and output could be incredibly complicated to write down in terms of equations, or maybe computationally too expensive, so the problems that we are working on, where we know that this atom or molecule has this many particles in it - electrons and neutrons - but to write down the equation and actually make the computations about how all these things work together… It’s actually computationally infeasible to do that, just by the equations that we write down.

[19:39] In this case, with the dark matter, we know some of these constraints, we know the experimental observations, but we’re not able to maybe write down well the statistics that they’re talking about, that govern these things. So kind of plugging a neural network into that gap and having it try to learn some of the features that are important to some input/output, whether that’s input – in this case, input, it sounds like these experimental observations, and output, cosmological constance and other things. Or in the case of atoms and molecules, inputting geometries or numbers of particles, and outputting energies. Putting a neural network into that gap, where things are really hard to model, can make a lot of sense, I think.

Tom Occhino: Yeah, that’s right. So it was declarative UI, right? React sort of pioneered this idea of describe what your application looks like at any point in time. And maybe the era of LLMs, just to connect to the analogy, is “Describe your application at any point in time.” So it’s not just what it looks like now. Now it’s how it behaves. There’s a tweet going around with “Use AI when…” And actually, I think that directive is not that far off. You’ll kind of describe what your application is supposed to do, and if we can ensure that the same predictable behavior and outcomes happen, we’ll be able to – we’re in this sort of era, entering the era of personal software, I think. So we’ll see.

Jerod Santo: How dare you. Number two, a fast and customizable vulnerability scanner based on simple YAML-based DSL. Number three, an abstraction atop all popular frontend testing technologies. Run everything from one place. Number four, a blazingly fast JavaScript framework written in Rust, with AI superpowers. And number five, “Nuclei: Get to the center of your issues.”

Adam Stacoviak: I actually did some math… I couldn’t help it, because I was thinking - you know, when you think about directors, or particular actors, you think about Tom Cruise, for example. I think Top Gun was his all-time – Mavericks, of course; the latest one. It was his top-grossing movie of all time for himself, right? It trumped every other – and sorry, Jerod; I used that word trump. It has exceeded all past records for Tom Cruise as an actor in a movie. And so I was like “Well, what has José done in terms of listenership on the Changelog Podcast Network?” Because you’ve been on Practical AI, talking about different things, you’ve been on this show, you’ve been on now Friends, you’ve had Chris along with you to sort of talk through some things… So collectively, you’re responsible for at least 210,000 listens on our podcast.

Daniel Thompson: [13:54] And it was really important to us that nothing changed. From the outside, it’s still a militantly driven open source community, that now is supercharged with a handful of engineers being paid full-time to do the research, development and maintenance that a massive project like Tauri needs. And that company is called CrabNebula. We chose CrabNebula because we liked the idea of a place where stars are born. A nebula is a star factory, if you will. And we chose crab because - well, Rust, the icon, the avatar, if you will, is Ferris, the little crustacean/rustacean. Now, obviously, you can’t make a pitch to a VC and say “We’re just going to serve as a charity, and donate all of your money to open source.” Things would be nice if they worked that way maybe… But I think that we’ve found the perfect VC to join us on this trip. That’s JJ from OSS Capital, a West Coast-based venture group that only supports early-stage open source projects; commercial open source projects.

And through JJ and through years and years of being around, we sort of collected an all-star regiment of angels who joined us along the way. I could drop all of the names, maybe you can edit them out, or choose the ones you like, but I think of the almost 30 angels that we have, a good dozen that are really relevant are Naval Ravikant from AngelList, Automattic Inc, the company, the investment arm of Automattic from WordPress fame, Guillermo Rauch, the CEO of Vercel, Thomas Dohmke, the CEO of GitHub, Tom Preston-Werner, the original co-founder of GitHub, Paul Copplestone from Supabase, Justin Hoffman, the former SVP of Elastic… If I didn’t say Bob Young, I’ll say his name again, because he’s amazing… Emad Mostaque, who is the CEO of Stability AI, Clement Delangue, who is the co-founder and CEO of Hugging Face, Dave Teare, the founder of 1Password, Adam Wiggins, the co-founder of Heroku, [16:27] Naveen Rudrappa, the founder of NocoDB… Heather Meeker, if you know Heather; she’s not only the general partner of OSS Capital, but wrote the book on OSS licensing… And we also have a couple of people like Cassidy Williams, who’s the CTO of Contenda; you maybe know her as Cassidoo. Also Tejas Kumar… And I think what drove us to work with this number of angels is getting to know your idols, the people that have built open source, the people who are building open source, and people who are poised to build the next open source, who understand the challenges of not only having a product, but also having the machinery, the understanding, and the ability to innovate into products and new products as they come out.

So that happened, and nothing changed at Tauri. I mean, we kept on building Tauri, in line with our foundations, expectations… But behind the scenes, we are working on a few products. I mean, other than investing time in Tauri, we’re also auditing Tauri; that’s one of the things that we love doing actually, is auditing people’s software that they built with Tauri, with Rust. We just completed an audit for a company called Bloop.ai that uses Tauri. We don’t really do custom development. We do, but we don’t.

[18:05] We will pick from clients who wanted to have something done that aligns with our research goals, if that makes sense. We’re not just out there, cutthroat working for half a million dollars because somebody wants to pay us money to build something for them. I mean, we’d consider it, but it has to align with our research goals; things that we want to know, things that we know that the community needs. Those are the kinds of customers that we’ve been looking for, and been finding.

We also recognized from the beginning that shipping apps is hard. It’s really hard. I mean, anybody can build an app, but once you’re done, how do you distribute it? How do you update it? How do you sign it? And for that, we’re building a platform to empower the people to ship their apps super-easy, super-cheap, and in some cases totally discounted for open source. And we know that - this is gonna get a little political; I don’t know if that’s okay… But we know that the incumbents, like Microsoft, and Apple, and Alphabet, and Meta, and ByteDance, they have vested interests… Well, maybe less ByteDance, but definitely the platforms… They have an interest in keeping a hold over a chokehold, over the app signing process. They keep such a chokehold on it that in my opinion, it’s not talked about enough in the supply chain. That final bit of app signing for GUI apps, and in some cases even CLI apps - that’s just like “Oh, yeah, Apple will take care of that for you. We gotcha, we gotcha. Come here, come here. Just give me your 99 euro –” Or Microsoft changing the game suddenly last April saying “Now do you have to get an extended validation dongle, or HSM. But don’t worry, you can use our super-secure, never-been-hacked Azure platform for that.” Google is a little bit less concerned; I think that anything that impacts their ad business is going to be a problem for them, but apps on devices keeps people on devices, keeps them buying devices, so I think that that’s okay for them.

But when I hear from people from the Tauri community that they are having problems signing their apps… Some people don’t even sign them. They just ship their Microsoft apps without a developer signature, and they’re like “People will deal with it.” And there’s no money solution here. There’s no lobbying we can do, but we can, in Europe at least, get involved with the European Commission and its platform policy work, serving as experts and making sure that these changes are respected, that other types of app stores can be on their devices; other third-party apps can now, starting in April, in the European Union, by law, have to be landing on these devices. And I feel very, very deeply about empowering the citizen developers out there who don’t have the 99 Euro, or who are in a so-called third world country where $99 is just a month of food… But they still can build their apps, but they can’t distribute them, and they have no access to these larger markets; I find that compelling in a very sad way. And I think there’s this mixture of good business, good politics, supporting the community is really in the DNA of CrabNebula. So much to the point that one of our first products that we’re going to be bringing out in Q4 - and you can find out about it just by following our socials - is a dev tool, because debugging anything is hard. Debugging Tauri is triple hard. Web has great dev tools, apps not so much, and we want people to be able to connect their app to an analyzer to figure out where things are going wrong, or getting better.

[22:04] And I understand you’re thinking, “Daniel, this is niche. Not many people are using Tauri.” But the great news is that we’re working together with other partners in the Rust ecosystem to define and use emerging standards, so that the work that we’re doing for Tauri people can be used by, at first, others in the Rust ecosystem, and later other ecosystems as they get interested in it.

I think the final thing that I’d like to point out is that - you asked “How is this possible? How can you keep the energy going, the momentum going?” And in open source projects you generally have like three or four models. One, a company sits on top, puts its thumb down… It’s not even a benevolent dictator; it’s like “We’re doing this now this way. We’re calling our project open source, and later on we can rug-pull.” But another one is benevolent dictator, who decides the way the project goes, generally takes all of the funds, and other people will contribute as they have time; people come, people go. And what we have with Tauri though is really quite compelling, because from the core team, five of the people are still around almost five years later. And I think that that’s a testament to the fact that we really enjoy doing it. And yet, I don’t believe that it should be just the goal of one company sitting in Malta to finance an entire open source project. And we do have donors, and that’s amazing… I think where things are tending toward is toward applying for more systemic grants, where we apply for funding from the European Commission, from organizations like NLnet, and potentially even other companies come and recognize the value that they’ve gotten from the community and start giving back. I don’t know, I think that that’s a very long, long play, expecting people that get something for free to give back to a community that they’re not so much involved in.

Adam Stacoviak: Yeah. You can’t get that unless you’re the pro plan user. Yeah, that syncing is – I mean, it’s essentially you’re a team of one, as you’ve described it. You’re a team of one, you’re a power user, you’ve invested in it; you need to be able to move from machine to machine. Or maybe just two. Maybe it’s not three or four or ten, but maybe just two, that you need to keep in sync. And even though certain things in Raycast aren’t present on both machines. For example, I think it’s a quick link, this is what they call them, to certain directories; that directory may not live on that machine, so it just won’t show up as a quick link on that machine. But it’s still in my settings. So when I go back to the machine, that quick link is present on, then I can go and use that quick link.

So it’s really interesting how it helps me become a - what you’re trying to do here, a power user to a native, free feature on the Mac. They’re trying to win over Spotlight users, and then potentially Alfred and LaunchBar users. It’s really about do you desire to be a productive user on a Mac? If the answer is yes, you should use Raycast. And by the way, it’s free, for almost everything. And it’s like a lot of stuff. I was like “Tom, you’re giving so much away for free. Why did you do that?” He’s like “Trust me, Adam, it makes sense. I can’t explain it all, but when you get into the details, you’ll find out why it makes sense.” And so now it does make sense, because the things that I want as a power user, I’m gonna pay them for; like syncing, or the AI features that they have now, too.

Adam Stacoviak: It reminds me of that movie… Tom Cruise, Oblivion-something… Oblivion Horizon… There’s a narrator, I suppose, or this voice in the void that asks this team, Tom Cruise and his co-star, “Are you an effective team?” That’s the question that the AI or something keeps asking them… Like, “Are you an effective team?” And it’s this response of like, “Yes, we’re effective.” And if they’re not an effective team, then the voice helps them be effective. So that’s the question you all are asking yourselves, is like “Are we an effective team, in this movement?”

Chris Benson: Yeah. We’re talking about – think of the movie Top Gun, okay? That kind of dog fight, like pilot Tom Cruise, all that kind of stuff; not the same plane, though. And they train them. And so at the end of it… This was not the Navy, which is what Top Gun is, but the weapon school, which is the equivalent of Top Gun in the United States Air Force, they had a weapon school instructor - this would be like the equivalent of the instructors at Top Gun - go up against the model, and the model demolished the instructor, over and over and over again. And so you’re talking about one of the best fighter pilots in the world, period, getting demolished by the simulated AI that was based on deep reinforcement learning, beating it in the simulator. They were using a real – I believe it was an F-16 simulator. I watched the whole thing live and it was just amazing. So this is some pretty cool technology here, and it’s evolving rapidly.

Chandler McCann: Sure. DataRobot – and I actually found DataRobot through one of those conferences, at Strata in 2016. That’s a funny story… But DataRobot was born from our co-founders, Jeremy Achin and Tom de Godoy, who were early Kagglers, in addition to working in the insurance space… But during their Kaggle days, they realized that they could benefit from automation. They could automate a lot of the tasks that data scientists were doing; not to replace them, but just streamline the workflow.

So following some funding, DataRobot and automated machine learning were born. That was back in 2012. Today, DataRobot is a full, end-to-end enterprise AI platform, and that means really going from the whole value chain of data, from “I have an idea, and a raw, unstructured dataset”, all the way through model building and monitoring a deployment, and even consumption, into automated applications. DataRobot provides a very high level of automation, all the way across the spectrum. So it’s been a lot of fun to be a part of that ride.

Daniel Whitenack: Yeah, definitely. Let’s maybe turn a little bit towards the HPC for AI, and how these worlds are colliding… Because I remember, for example, when I did that internship that I mentioned, the primary applications that I was working with were climate modeling applications. I know people have used these sorts of clusters for quite a while for these climate models… I know also in grad school when I was doing computational chemistry calculations, where you’re basically trying to calculate properties of materials based on what you know about how the physics work for atoms and molecules…

So we were kind of submitting jobs to HPC clusters at that time, in a couple different places around the country… And so I know that there’s been this history of HPC clusters being used for these large-scale (like you said) simulations and scientific computations, and those sorts of jobs. But I also know recently I was talking to – I live in the same town where Purdue University is, and I was talking to one of the data scientists that works for Purdue University, and he was saying that now they have sets of nodes, and they’re continually buying more that are specifically geared towards AI applications. So I know that this is happening, and obviously, you’re working in this area, so could you describe maybe why HPC is relevant for AI, and maybe when would I want an HPC cluster for doing AI, versus maybe just spinning up some stuff in the cloud, and vice-versa?

Daniel Whitenack: Yeah, it’s interesting to come full circle on this… When I was in grad school I was working on computational physics, and Dr. Ghose, as you said, a lot of these brute force techniques will get you to maybe modeling a hundred or a few hundred atoms or molecules… And I remember at the time when I was in school, it was kind of the first time I had seen – at the end of my grad school people started to apply machine learning techniques to figure out the energy functionals and things that we were trying to figure out just from scratch by writing good equations, and instantly out-performed everything that we were doing, and it was kind of a shock to all of us… But it illustrates, as you were saying, Marcus, the power that you can achieve in certain cases by reimagining a problem as an AI or as a machine learning problem.

I was browsing around, as you were talking, on the Xanadu website and a couple others, and I see certain phrases like “Machine learning toolbox for quantum computing powered by TensorFlow” and other things… Is that more on the side of, say, using TensorFlow, using AI to learn more about quantum computing? Or are those things more on the side of creating a quantum computing module for TensorFlow?

Daniel Whitenack: Yeah, definitely. It’s exciting. I’m excited to just learn from everybody on the show. It’s gonna be great. So yeah, I can kind of give a little bit of my back-story. I actually started out in the physics world. I did my undergrad and graduate work in physics, and started out doing academic research in atomic and molecular physics, basically trying to calculate the properties of atoms and molecules from first principle. Basically, what that means is like, from first principles – there’s kind of two camps in this world. There’s the camp that says “Hey, we’re gonna figure out the equations that describe nature, and just using those equations calculate properties about these things.”

Then there’s the other side, which is like “Well, we’re just going to create these incredibly complicated functions, and parameterize all of their various parameters based on experiments, and then they’ll do good at predicting properties.”
[11:57] I was in the camp of using - from first principles - just the equations that describe nature… Which now I see as incredibly ironic, because if you look at what AI is, it’s basically the other camp from which I was in, creating these complicated functions and models and parameterizing them to describe reality… So I find that incredibly ironic.

It was during this time that I was actually exposed to machine learning to begin with, because if you think about that problem, people are trying to come up with all of these complicated models to describe how to calculate the properties of electrons, and eventually someone - a group in Berkeley, they said “Hey, let’s just let the machines do it.” So they started using machine learning methods to actually learn the exact solutions for these functionals to describe the properties of atoms and molecules.

It was kind of sad in one sense, because I was like “Well, that’s basically the way we should be doing it, right?” but at the same time, I’m doing a whole PhD on not doing things that way… So it was just kind of funny in that sense. That was my first exposure to it.

But then I went into industry after Academia… I really didn’t know what jobs were available for a theoretical physicist in the industry; it’s not really that clear. And that was kind of at the beginning of the hype of data science… So pre-AI hype, but data science hype; just the beginning of that.

I found my first job working with this IP firm… Basically, helping lawyers understand math is essentially the job, which is incredibly interesting in certain ways… [laughter]

Juan Pablo Buritica: There is. I think the event will happen in November - the 1st to the 4th November is the dates that we have the venue. We’re opening the CFP right now today. JSConf Colombia right now has three confirmed guests. We’re gonna have Tom Dale (I think some of you know him), we’re gonna have Suz Hinton (some of you know her as well) and we’re gonna have Elba Sánchez, who is a local Colombian rising star. She worked with me at my last company, and she’s been writing a ton of Node, and working with AI and bots and stuff, so we’re really happy to have her.

Then we’re looking for workshoppers and speakers. We pay for travel, we pay for accommodation, we pay for your significant other if you’re a new parent, and we’ll also arrange childcare as well… So it’s a pretty sweet deal. We’re hoping that at some point we will be able to pay speakers, but we’re not there yet; it’s a challenge.

The URL is cfp.jsconf.co. Please apply. I think it ends in June 11th or 12th. It’ll be a blind CFP, and I recuse myself from judging, because I know how some people write… But it’ll be pretty cool to have a broader audience. It’s been super fun to have people come to Colombia, because we’ve been generally isolated from the world, for some fame we’ve gotten ourselves into… But we’ve had some pretty fun times. I think, Alex, you had some good food…?

Nathan Sobo: Yeah. And again, it’s like – I don’t know, if you would have told me that people were forking VS Code left and right because they were so excited to have direct control over an editor UI, and that some of these people were being purportedly valued at billions of dollars in valuation… Yeah. So basically, if you would have just stripped the kind of comparative analysis of all of it when I started and said that was going to happen - sign me up. I love that future, righ? So I don’t know, I always zoom out whenever I’m in doubt.

My goal has always been to kind of build the next generation in tech. Like, really advance the state of the art for the developer experience, period. Full stop. For me, for a long time, once I tried it, we created Electron, this Atom shell, and then we tried it, we gave it a good try… And yeah, could we have done it better in web tech? Probably. But ultimately, we had to leave that piece of it behind to deliver a truly exceptional experience. Otherwise we could never be better than the browser. Like, that put a ceiling on how good the experience can be.

And so for me, I was never going to do it any other way. And so anybody else’s success doesn’t… It’s not really relevant right now. Other than to the degree that it indicates that there’s value where we’re going, I guess. So it’s never really – it doesn’t bother me, I guess. I don’t feel behind or ahead. I feel right where we need to be, because we’re doing what we’re doing with integrity, and we’re going to do it as fast as we freaking can, with integrity. And I think over time, that’s going to prove to be the best approach. But we’ll all see. Or you can see right now. Anybody listening can see right now what they think, you know?

Justin Searls: Yeah. So what I had was I’ve got a blog, justin.searls.co, and it’s just a traditional-looking blog, with a lot of like link posts, and other stuff. I’ve added a whole bunch of kinds of multimedia, especially since starting a podcast. And what I want to do is every time I post something to the blog, I would want it to go on what had been called Twitter, and then there’s Twitter and Mastodon that people are on, and then there’s Twitter and Mastodon and Threads that people are on, and now people actually use BlueSky for something other posting now, so’ve I got to get on BlueSky. And to manually copy-paste the same message or to use a tool like Buffer and like manually click the thing and try to make sure that it looks right would have been so time-consuming. Additionally, now I’ve got four timelines that can each separately derail my train of thought and get in the way of my workflow, or whatever, and I’m going to be way less productive. And there’s a reason why I always had to keep Twitter at arm’s length, because I got addicted to the timeline scroll.

So I built a kind of hodgepodge of Docker containers, this random – somebody has this thing called Feedtoot, but that’s in Python, and I don’t know that, so I’m running in a Docker container and a Synology… So then I made a thing called like Feed to Gram, and Feed to Thread, that would do similar things, and those are just Ruby gems that you can use today. And if you’ve got an Atom feed, it’ll read the feed, and then it’ll syndicate on your behalf if you give it the right API tokens and you jump through the 800 steps that Facebook demands.

[01:43:59.02] The problem with that – it scaled to one really well, but then my wife saw me and the life of luxury of a write-only existence, where I was able to publish all I wanted and be in all these places without having to get sucked into them, she wanted the same thing. And so then now I’ve got doubles of all of these Docker containers running in my Synology. But then unlike me, she can’t log into that and then understand what’s happening in the logs when she’s got like the wrong aspect ratio or whatever in Instagram… And so I built her this platform, this Rails application for her business. Betterwithbecky.com. It’s like a strength training subscription app, but it also has this whole thing called Beckygram in front of it, which is sort of like a blog for people who are Instagram-first. And it looks a lot like Instagram, and it can do video, and it can do photo carousels, and all that stuff.

Well, to get the Beckygrams into Instagram, I had to actually make it part of a real working application. So I had this Rails application now, and it’s got like all of the background jobs and all of the sort of like durability that a real user would need in terms of like error handling, and making sure that they have a way to retry and fix things and remediate if there’s a problem in publishing, and then like a link back to the post, or whatever it is.

And now I’d gone through all that work, and I’m starting to get – I’m finally done with Becky’s app, I’m starting to get people who are like right into my website, or the podcast, and be like “Man, how do you do all that stuff again? I really want to be able to do that, too. Would you release like a Hugo template?” But that would just create more problems for me, because then I would be supporting a copy-pasted project file… That’s not the solution. So what I decided to do is I’m going to build a dead simple, this does one thing, and it solves exactly my problem, and if it solves your problem too, you’re welcome to subscribe to it… A little app that I call Posse Party, so you can run with my Posse.

It will basically be a place where you sign up, you give it a feed URL, probably Atom, RSS, JSON feed when I’m all said and done… And then you add your social accounts. Now, it’s working for me in production, for me, and I’ve got the big four Twitter-like things; I probably do Instagram, LinkedIn, Facebook… And all it does is read the Atom feed and a little bit like a JSON sidecar payload that is also posted on my blog as part of the feed. And that little sidecar payload says “For each of these posts, this is how I want to present this, this is the format string I want to use”, and any customizations on a per-platform basis. Like maybe for BlueSky I want it to look this way, and for Twitter I want it to look this way.

And then it’ll of course be platform-dependent. So there’s certain things - on Instagram, it’s going to look a little bit different. And BlueSky’s got different rules about embeds, and all that. So all the app really does is read a feed, turn it into these normal database rows, and then a job runs that will syndicate those to all of the different services by reading the configuration that you supply, as well whatever defaults you configure in the app. That’s it. That’s the whole thing. And if that has value to you… Jerod and I, before the show, were talking about our information diets these days, and we’ve both reverted to RSS… But I’m not going to get my mom to use RSS, or most of my friends in my life. They’re in one or two of the social apps, and they’re casual at this point, looking at it… I just want them to be able to click my face and see my stuff. Even if the downranks me for posting via the API, even if the links get the fact that you’re linking outside to a third party, the platform doesn’t like it, as long as my stuff is literally there, I’m a lot happier, because then I don’t have to tell everyone about everything in my life.

And since getting started with the Posse Party stuff, I’ve noticed that more and more people in my life – like, I had a guy detailing my car a couple of weeks ago, and he’s like “Oh man, I really loved that clip that you did about whether or not to have children and your decision-making process.” I’m like “How’d you see that?” and he’s like “Oh, I saw it on X.” I was like “Great.” I haven’t logged into X in like three years.

[01:48:04.23] So I don’t know, that’s the theory. Now, I don’t know what the market is. I kind of don’t care, because as long as I want this, I’m happy to build it… But if it can help people, encourage people be the thing that gets people to sign up, and make their own website, and own their own content, or buy their own domain and like centralize themselves as kind of the king of their castle, and only treat these platforms at arm’s length, as those are just “newspaper stands to throw my stuff on” - there’s always the canonical URL - then that just supports the open web, and it really promotes people as thinking of themselves, less as just like me, like a take maker or just operating in the Twitter stew, and instead being somebody who’s developing a voice and curating an audience. And if I can help people along that trajectory, it seems like it’d be putting some good in the world.