LGTM, Ship It: The AI Code Review Problem

This company is merging 60 PRs per developer per week, and everything is a mess. I got a robot.

They are bringing in external dependencies into your projects. Is it worth it? Specific tips on AI and writing UI code. What are local AI models even? We're gonna be talking about all these and more in this Syntax Pollock. That's right. This is our Syntax Pollock episode where you bring the questions, we bring the answers. And if you have questions for us, head on over to syntax.fm and ask a question for us, and we will get to it on a future show. So let's get into it. Yep.

Alright. So the first question today from Deno sea goose. I just listened to episode 1,010.

Wow. It Wes crazy. We're past 1,010, which was a Pollock episode with a lot of AI questions. The problem I'm struggling with is whenever I ask an AI agent to create a component or any large piece of code, it takes me a considerable amount of time to understand what it wrote before I can make any changes.

What advice do you have to help me better understand the code it generates? Yeah. Well, okay. So there's a number of things. Sometimes the AIs, the when they write components, man, they write the craziest, like, rat's nest of function calling a function calling a function and then throw it all in, like, an effect brained kind of way and and really over rely on things like side effects that can make really understanding the code that AI writes a little confusing sometimes because it's quite simply not good, or it is good and it's just different than how you would have written it. One thing that I like personally is having it write terse comments in in line with the code.

Just like, one, it's it's documenting the code. Two, it's it is easier for me to reason about in English than to reason about in, some other way of authoring code that perhaps I'm not used to at that particular time. So if I'm trying to deeply understand the code it's being it's writing after it's written a large component, one, comments. Two, I like to lean on a lot of rules, particularly when I'm working with AI agents myself.

I heavily dictate the style at which they're allowed to write components in.

You know, in the Svelte world, I'm saying do not use effect. Don't even think about it. Side effects are almost never the right decision here. Yeah. But because so many things happen in effects in React, it's always wanting to lean towards that. So, I have a number of strict rules that I give it when I'm prompting. So, therefore, when I go to read the code, it it reads like the code I want to read. Yeah. I think with UI code specifically,

because that often dictates how it looks, I find myself drilling in a little closer with with the UI. Whereas, like, if I'm writing some, like, back end functionality, I have no problem for it to generate several 100 lines or Yeah. Or maybe even a thousand lines of, like, functionality.

You Node, you Scott your database and everything JS is object function.

Yeah. It's a It's ESLint function. Pretty good at, like, actual subjective functionality.

But when it comes to actually building out the UI, I drill in a little closer and spend a lot more time on it and do sort of smaller changes because I think that's why a lot of these apps feel like crap. It's because the UIs on them are just absolutely awful. They technically work Wes. Mhmm. But how you use them, how you interact with them, and and what they look like is often there's a lot to be desired. So my recommendation here is just drill in and use it at a much smaller hell, even go back to using tab completion, if if you must for that type of thing. Other things you can lean on, like Scott says, is is lots of rules in there, example code, prompts over if something should be duplicated or not. I find that that's a really interesting one where Yeah. You have two things that have somewhat overlapping functionality.

And in some cases, you wanna simply just duplicate them, and they're different enough that I just wanna be able to have the code in this component and not have to, like, factor it out into some reasonable thing. In other cases, yeah, you wanna, like, say, okay. Here is a validation step, or here's what we do when you upload an image. Right? That shares enough code that you wanna you wanna bring that out and put it into a util library.

I find the AI is not great at that.

So in that case, you just you should be making smaller changes so that you are more dialed into what is actually happening. Yeah. I mean, on the on the UI tip or on the Sanity tip, you know, one thing that I have in all of my agents' files is is really just that, like, to create shared utilities in one location.

And, like, here's the directory of shared utilities. Always check to make sure a utility doesn't already exist.

Like, here's the place you're looking for utilities.

Do not create one locally. Create it globally. Import it, and you use those global utilities because that way it's easier to check. If it's creating, like, an inline utility function or whatever that it just uses once, it could duplicate that code a 100 times. Yeah. I I think as you do more of the stuff, you'll get to a level of comfort knowing

I'm I'm happy with this type of thing. I'm happy to, like, trust that the code generated is up to my satisfaction. I don't think you have to read every single line and absolutely know what everything is doing, but you should be prompting it in such a way that you you do understand where it is for mission critical stuff. There's certainly other projects where I just I haven't even looked that line of code. Yeah. Yeah. Totally. Exactly.

Yeah. And and I I think it is a lot of it is setting up those standards, those standards of what you wanna see, those deterministic tools, whether that's the linting stuff or whatever. And, like, the the ways that you can have your your code work that just you you don't need to have an eye on it if you have a good set of rules.

This next question is the exact is exactly what you don't want to happen because I love this. Yes. Let's do this. The it's a question from l g t m bro. On my team, a common workflow is an engineer asks an agent to implement a feature or fix a bug, open a PR, address automated review comments if necessary, waits for CI to pass, emerges. In many cases, there is little to no meaningful human review of the actual code. JS this certainly a problem right now JS that the bottleneck is not in generating the Node. The bottleneck is trying to review it and make sure that what is implemented is actually it it not even that it has to be good code, but it actually does what you're doing and not breaking everything and and all that stuff. Right? So it goes on to say, some individual contributors are merging 60 PR releases per cycle, which is about a week. So Jeez. You're merging 60 PRs a week.

Holy hell. Including changes spanning 20 to 70 files and thousands of lines of code. The volume alone makes it difficult to believe that anybody is thoroughly understanding validating these changes. The results appear to be a steadily increasing amount of technical debt, reoccurring bug, architecture inconsistencies, fragile implementations, and code bases that are becoming progressively harder to extend and maintain.

The product is awful. This is kinda interesting because we have these new superpowers, and you're seeing teams do one of two things.

One, feature Sanity.

Any feature you want. Boom.

Let me just copy what you want, paste it into the box, and I will, sit there and watch it it do it. And then the other the other people that is like, I'm looking at some of these apps, and I'm like, why am I using this app for so many Yarn, and there's literally been no features or things that have been added to it yet? Like, you have these new feature canons. Like, what's going on? And what a lot of those companies are doing is they are saying, wow. We have a super superpower.

Let's spend the next year or whatever or six months cleaning up all of the stuff that we have been saying that we want to fix. We've been saying for so many years that I want to be able to get this into Scott. And I have this with a lot of my apps where I'm like, I'd love to implement this feature, but I would love to move to this database first. I'd love to clean up my utilities first. I'd love to to do all of these things. So a lot of people are saying, ah, we got all this technical debt, which we can finally solve.

So they're using AI to refactor all of that type of stuff. So I think, like, we could talk all day about, like, oh, yeah. You should implement, like, features and tools and linting and all of this stuff, but I I think that this problem goes beyond the proper prompts and skills and and and tools that you have. And it's like, well, well, well, turns out all of the stuff that we talked about for 10 about good software engineering still applies.

Right? And and you're just doing it on like, you're just speed running it on on, like, fast mode now Wes it might have might have taken you six years previously to get to a point where your code Bos is a mess and things are duplicated.

The problems you Yarn describing here are not new at all, and the solutions to these problems are not new either at all. You we know how to how to make a good software, and the fact that you guys are just doing it on a speed run is just making it much faster. Man, that's crazy.

60 plus PRs in a release cycle, and that's one week. And some individuals

yeah. At that point, like, the code Bos would be unrecognizable in in several months, which is partially not like, like, some parts of your code base, you just don't understand. It's easy. The AI can explain it to you. But those, like, high level architecture you know, if you're just adding a whole bunch of crap and duplicating it, it's going to be very tricky to sort of unwind a lot of those, especially if you don't have, like, tests and whatnot.

I think one thing that the prompt bros don't realize about this stuff is because, you know, most of them have never maintained old systems or long running systems or these types of things.

That tech debt is really real. And and tech debt is is such a major problem with AI AI now because it's so easy to add code. And, like, AI hates to remove code even when you beg it to remove code. So, like yeah. I just feel like the amount of code that we'd be adding sometimes is going to come back to haunt teams. And, sure, you know, Fable is gonna come out again, and the the the clouds will Yarn, and you'll say, hey. Make it better. And Node make it better.

And and maybe that's the case. But, yeah, in the meantime, I think we're gonna be seeing a lot of software and apps, like, barely held together because that's that's wild to me. Next question here from Sarah Chen. In episode 1,009, you mentioned that nobody knows what a local model is.

Can you do an explainer on what local models are? Well, local models are just, models that you're running on your local machine. I think some people really disconnect that and think, oh, I'm running the software locally.

Therefore, it's local. But it like, when reality, they're going off and doing API calls. That was the whole thing. Like, all All these people are like, just use a local thing. It runs local. They use OpenCode and Deep seek for local. Like, that's not local. It's going to China. Yeah. So it's it's literally running running on your machine. And I will point you to the, CJ did a really incredible video called your guide to local AI. And so instead of just recreating that here because, you know, we could I could just recreate CJ's awesome video on the fly here. Right? I'm just gonna point you to that because it's really super good, and he even uses a piece of bread, to show you how big his computer JS. So that's fine.

Yeah. The the local model stuff is really interesting. We'll we'll keep coming back to it because there's, like, this, like, idealistic thing of, Wes.

I would love to run local models. It's it's private and whatever, and it's on my machine and whatever. And then there's the actual, like, you don't realize, like I again, I'm saying people don't know what local models Yarn, and now I'm gonna say, like, people don't realize how much compute is actually being thrown at this stuff until you try to run something on your own machine, that is of of the the quality that you're expecting from, like, an Opus. I will have

yeah, I'll have a lot more opinions on this once I get a local AI rig and get deep into it because, that's coming for me. It's just I'm waiting for some next round of hardware to be released. It's just yeah. The local AI stuff makes a lot of sense when you're doing purpose built stuff.

When you're doing Sanity detection, when you're trying to detect parts of your face, there are thousands of small models on Hugging Face that you can run in the browser. Look up transformers JS s. We had XenaVaa on the the podcast Wes there JS a lot of really cool stuff that you can run-in real time and, like, that is is good very good in the browser, but they are all dedicated to, specific things, you know, doing speech detection, text to speech, tagging things, categorizing things, vectorizing, all all of that stuff. So the local AI space is not a monolith that does absolutely everything.

It's a specific targeted things that you want to do in your application.

Yeah. And that's not you're gonna get fable at home. It's just not. No. Although, there's a guy there's a guy on Instagram right now, and he's he's saying that he's he's Node Fable on a laptop. And it's, like, it's literally on a TV tray and, like, the psychosis that some of these people have. And he's he's posting screenshots of Claude telling him that he is a 0.1 advanced user, and, like, he'd, like, literally thinks he's he's like, I've only been doing this for thirty days, but we're on to something. We just need a little bit more compute. We're building Fable at home. And, like, he's he's telling Claude, build me Fable, and it's it's pretending to build it. It's got all these graphs and everything. Yeah. Un unbelievable, the the amount of psychosis that goes. And I I feel bad because it's like they don't he's he's never he's never done anything in computers before.

So It's mental illness is really what it is, and that's like It's telling them. It's It's telling them, you're doing a good job. I'm we're building this. Like You're very handsome. You're doing a great job. Everybody likes you. Yeah. It's a bummer. Yeah. Yep. Yeah. It's crazy. I think I mean, we're all we're on the precipice of seeing an endless amount of warp in the world. So you're ready, folks.

Did you ever look into Wes it that that, GPT four o thing?

Is that where the people were mad that their lovers went away?

Yeah. It was something like that. It was like they chained they they discontinued the model, and people, like, were just

losing it. Because they have been talking back and forth with this as, like, a specific friend and a lover. Yeah.

And they found that they can't replicate that with a different model because it has just different temperatures and tones and whatnot.

That's Yeah. That is sad and crazy.

Man, man, that's all I have to say. Yeah. Aaron. That's all I gotta say. Yes, man. I keep seeing people say jiu jitsu is a new superior alternative. Jiu jitsu. Jiu jitsu, not ju Jujutsu. Jujutsu. Jujutsu. Whatever it's called, I'm certainly open to the idea that Git could be be improved. But every time somebody tries to explain their use case, it sounds to me that we are doing something weird with Git that is just not a problem if you use it normally.

I was tired of using so much superglue to keep the spaghetti on my spoon, but this new spoon improves superglue retention by 50%.

What am I missing? Have you guys ever used it? I am in the same spot. I've had people come to me and try to tell me about jiu jitsu, and I'm just like, I don't have those problems.

But, Scott, have you have you looked into this? I've looked into it a bit. You know me. I'm always looking into this type of stuff.

I have not used it because it felt like a a thing I had to spend a lot of time on that I don't necessarily have right now. But the the key bullet points for people who are interested in this thing is that, one, it is fully Git compatible. So it can use Git as a normal dot Git directory.

And even to the point where you can use Wes, you can use Git in a project, I can use JJ in that same project, and there's not gonna be any issues. Right? So it's a personal choice if you want to use this strategy or not. It's not like a full team buy in unless you you, like, want to do it that way. So that part of it is nice. Right? You don't have to, like, really go nuts on it to to get to use it. There's an operation log of all revision logs, including, like, your snapshots, your merging, all commits and everything exist in one giant, operation log. Yep. So at any given point, you can just, like, run an undo command, which

man, having an undo command in git, how many Yeah. Stack Overflow post would that have saved? Right? I still do it. I'm like, shoot. How do I like, I like, I've been using git for, what, probably twenty years or fifteen years, something like that. And I still am like, oh, no. I added something Yeah. That I shouldn't have, and it was not it might get ignored. Now I have to, like,

ask AI to do that. Yeah. And there's a lot of there's a, like, automatic rebasing, which I you Node, there's no staging, so everything is just automatically staged, so you never need to stash anything.

There's no branches.

So there's a concept called bookmarks, that are very similar to branches, but they're like, there's no current bookmark. You can just go back to a bookmark.

And people say that this makes it really easy to go and make revisions or fix things or merging, or or or, conflicts are easier to resolve and things like that.

I'm gonna say this as somebody who has not used it. So if you have used it, leave a comment below and let us know why you think it's better. The one thing that I've noticed when people are talking about JJ is that you hear a lot about from people who are currently using it, have bought in, and absolutely love it. But when I started to dive in a little bit more, I found people being like, honestly, even though it is easier to go make and make revisions or this or that, Not having, like, true branches and and that sort of style of working is too much of a brain shift for me, and I just want like, I don't care if I have to ESLint a rebase as long as I can make my branch, make my changes, or whatever. So I think the next

Git thing is gonna be from like, Cursor last week announced, what was it called? Cursor Origin, which is their GitHub alternative. You know? And, like, I think whatever whoever, like, solves this, like, what is the next Git slash GitHub is going to be building something for agents that can simply just handle the sheer amount of, like, committing every single keystroke or every single thought, you know, so you can roll it back and you can, try six different things. You know? I I think whoever figures that out is going to is going to win this thing. Making something that's, like, a slightly different CLI for humans, I don't think is gonna gonna catch on anymore.

Well, I I did see some arguments saying that JJ was better for agents because of that not needing to branch. Right? It can just create bookmarks and versions of files and stuff really easily.

That's in practice, I haven't gotten to use it that way. That said, Zed, of the, Zed editor did re or Node they didn't release because it's in private beta.

Something called Delta DB, a version control that's built on a single coherent abstraction that transforms your conversations with agents and the work trees they edit into shared artifacts.

So, Delta breaks your work into a stream of fine grain deltas Wes Git captures a snapshot of each commit. Delta captures every operation in between and gives each one a stable identity.

So it seems like Zed is also trying to solve this very problem. So Interesting. Seems like a frontier that a lot of folks are what do people do on a frontier?

I should know this because I live on a frontier. You explore them. They're exploring. Yes. They're digging for gold. I think yeah. That's that's kinda interesting as well because, like, you spend all this time chatting back and forth, and then you lose all of that conversation.

There's often valuable stuff in the, like, why? Why did you do something? So maybe if compute and context and all of this stuff is not an issue, like, why not just store absolutely every single thing that has happened if we need to ever go back and look at it or or use that context in a new conversation? But compute and storage is certainly a problem right now, so we'll see. Yes.

Alright. Next question from Phonics.

As AI enables developers to build features in the minutes that previously took hours or days, how should contractors and freelancers price their work fairly by time ESLint, estimated effort, or value delivered? They should price it the exact same way they've always priced it or should have always priced it. Value delivered.

Pricing, a lot of this Wes something you could've gone back to episode number one of Syntax when we're talking about freelance stuff, and we're talking about, like, time spent or effort.

Like, these aren't what makes for a good quote or anything like that. It can lead you to go off the rails a little bit in spending or or whatever. At the end of the day, I think value derived is the only way to price this kind of work. And, frankly, for me, has always been the only type of way to to price this work.

I I have charged hourly in in my career, but that's almost always where they're like, we will bring you in. What we need is maybe unknown at this point, but we just wanna pay for twenty hours a week of your time, and you're just gonna work on whatever it is. And and those are okay arrangements. Right? Like, that's similar to just having a job. But if you are if somebody is hiring you to do a task and how long you do that task should be irrelevant.

It's it's all about the value delivered. And if you spend a whole bunch of your time learning something, building out a bunch of tools, creating, I don't know, an army of agents that is so special that nobody else can do it as well or as fast as you, that shouldn't be, factored into how how long it takes you. You're right. You just you should be billing based on how much value delivered to the client at the end of the day.

Certainly, like, these things are gonna change.

Like, people aren't going to be paying you $600 to, like, refactor, like, a one page on their website anymore when, like, I think, like, what people are willing to pay for and how much you get paid for doing stuff is is going to be or is already changing. Right? But that that doesn't make make it that you should be charging based on your thing. Maybe tokens.

You know? Like, at what point does your bill include how many tokens you spent on something? Because now people are saying, like, you're not talking about how much time you spent, but you're like, oh, man. I spent $6,000

worth of tokens. Spending money to develop. Yes. Yeah.

Yeah. Weird world.

Emil Babin says, hey, folks. Do you have any idea why WebKitBox Reflect, such a useful and delightful property for easy CSS reflections, doesn't get standardized nor Firefox implements it with a vendor prefix, and this goes on for decades at this point. Yeah. So the WebKit box reflect was something that was added in Safari to basically get the, like, glossy floor effect where you could, like, mirror an image below.

And it was never it never went through a standardization process. It was just added to Safari.

Very likely, a lot of weird stuff gets added to Safari because they need it internally, or they're they're like they have some weird screen in iTunes that needs, that is actually built with HTML and CSS, and they need to be able to replicate that effect.

Why was it never added to any of the browsers? I think that they're hesitant to add stuff that is trendy, and that whole, like, glossy floor reflection was a huge trend that Apple kicked off. And, like, you Node you don't see it anymore. Warp flow on the iPods. It's Yeah. Yeah. Totally. It's not all that useful anymore. And and I think, like, your time is better spent on higher level features that would be able to make that, like HTML and Canvas. Right? Now you can be able to or not now, but once HTML and Canvas is in the all of the browsers, you'd be able to just wrap that component in a Canvas, flip the component, you know, like, paint it, and then just flip it on on its axis at the the very bottom of the x axis, and you can you can replicate that feature rather than saying, like, this cheesy thing that has no knobs.

Like, who needs that? Nobody needs that. It's I don't know. Is it is it delightful? I don't think so.

Yeah. I I mean, like you said, it was a very of a time. And and even though now it, like, feels, like, very useful in regards to, like, the CSS battle, stuff like that, again, very trendy look. I haven't seen a site that I would consider modern use that. No.

Although, like, now I feel like I feel like that's gotta be a syntax challenge. Who can make the most modern ass looking design using WebKitBox reflect? Like, some very beautiful modern UI using WebKitBox

Reflect. I tried to use it in one of our battles because I needed to duplicate I did demon. Wes. But it turned out that it it vertically flipped it. And I Correct. I did that I did that intentionally. Yeah. I didn't I didn't want I didn't want that. For some reason, I just needed to duplicate it to another one, so it wasn't wasn't possible.

Yeah. There's a lot of weird stuff in in CSS like that.

Yes. A lot of fun things.

Alright. Next one here from Moto Moto likes you. Moto Moto? Hey, guys. I work on a medium scale enterprise software product as a front end dev.

We use Angular as our framework of choice, and it's proven itself a sound investment in the age of AI as it's very opinionated, but that comes with a cost. The cost being that Angular developers tend to rely heavily on framework provided libraries, not willing to try many external libraries, which reduce the amount of code we produce and walk the same paths many developers have already walked before us. This is new to me as a former React developer. Pulling in new libraries

was not nearly as contentious. React devs don't bring a library for anything. Yeah?

React devs. Yes. That that that's the bubble. Right? Do you guys have any suggestions for getting the rest of the wider development team on board with using more external libraries, or are they correct and we should use the framework tooling as gospel? Currently, we only use a couple like Lodash and Hammer Wes, but I know there's loads of places we could benefit from it. So this is interesting to me.

Somebody who never made the transition to Angular two, you know, that's when I I Bos out of the Angular ecosystem like many other folks did and then spent a lot of time in the React ecosystem Wes you're right that the emphasis was often on bringing in other libraries. But that isn't always the emphasis. In fact, on level up tutorials, I built just about most components on that site because bringing in libraries has a cost, and it always has had a cost. So, I think you might be looking at this too much with rose colored glasses in terms of just bringing in and getting easy Wes libraries is probably a plus. You're opening yourself up to Wes things. And, again, we talked about in the last episode, the last potluck episode, we talked about the browser standards have evolved, and you can lean on the browser, standards to augment them with JavaScript.

So if the AI is good at writing Angular without bringing in extra libraries, that to me seems like a pro. But that is my perspective, and that that's the perspective of somebody who's often kind of largely chosen to do it themselves for, you know, yeah, I would say a benefit, but other people might argue the opposite. I I think with, yeah, like Scott said, with AI, you don't

necessarily need to bring in libraries. Your utils folder can simply just make implementations of of everything you know. You you certainly have to spend a bit more time making sure that it covers all of your use cases, but it it does exactly that. It just covers exactly what you Node it to do, and you're not adding a whole bunch of extra code for it. And I think a a lot of devs like that because I think that's why Angular has attracted so many non JavaScript devs. You know? Like, you you don't hear about Angular a lot inside of, like, the JavaScript community because we love piecing together everything ourselves.

But for people who are, like, coming from, like, Java or Rails or whatever, like, just give me the way to do it in in absolutely everything.

And I think this question has the answer to that. Currently, we only use a couple, like, Lodash and Hammer JS. Hammer JS is a touch library last Mhmm. Last released seven years ago.

And Lodash, still some use cases here and there, but in many cases, you don't necessarily need Lodash. And and this is what can happen is that you you build the dependency into it, and all of a sudden, you're stuck seven years later maintaining Hammer JS code that that hasn't been touching in forever. So I think I'm going to side with your team on this one. Yeah. Wow. I'm surprised we agreed on that. So okay. Cool. Well, Wes, do you wanna do a sick pick before we get out of here? I Scott one for you. I'm gonna sick pick the headphones I'm wearing here. So these are ten year old Bose QC 30 fives. These are I've had these headphones forever. Ten years. Right? And I have changed out people are always like, oh, yeah. I have those, but, like, I don't use them very much anymore because, a, they're micro USB, which nobody it sucks carrying around a micro USB cable or having one on your desk when, like, it's the only thing. And, b, the, like, headphone cups are shredded. Well, you both of those, I've I've fixed. Right? Like, I have changed these. You can change them to the ones that aren't stupid and obnoxious, but I've changed mine to cow print.

Just you could just find brand new ones, and they feels like a brand new pair of headphones when you put new ones on. The other thing is you can get a charging board from the Bose QC 40 fives online, and you can simply just it's plug and play. There's no soldering or anything. You just take it out. You unclip it. You clip in the new one, and boom. You do have to, like, wallow out the connector hole a little bit. But then if you find, like, a pair of old q c 30 fives online or if you have a pair, if you upgrade the pads, make them USB c, they're still my favorite. I have other headphones that are noise canceling, and none of them Node of them are are as good as the QC 30 fives.

Still a big fan of them.

Nice.

Yeah. That you you have been riding with those Four AC earphones.

Yes. I'm going to sick pick my, new robot, the Ricci Mini.

Hopefully, I'm gonna try to not to drop him while he does a dance here. I'm gonna have him do a dance because I'm holding him by the bass, which I don't I'm not I'm not sure how,

intense that will be. This is explain what this is. This is a oh.

He's going nuts right now. He's doing a little dance.

So this is a robot. It's by, Hugging Face and Pollen. Sorry. He's still making noise. I'll I'll send him down. Are you done? Okay. So this is a a robot from Hugging Face and Pollen, and you have to build it yourself. So every little component of it, you have to build. So it took about, like, two hours to build. I did it with you. All the pieces you snapped together? Yes.

It was a it was a very long instruction booklet here. Yeah. 50 page instruction booklet.

50 steps, I should say, of, yeah. There's several PCBs in here, you know, power, battery, all that stuff, motors, and, there's cameras and microphones. And, because I did it with the kids, I was having them do a lot of the the screwing and stuff, so it took a long time because of that. But it was a ton of fun.

You could see he's got a lot of stuff going on on the inside. He's very expressive.

And the cool thing about this is is, basically, you could think of it JS, like, one, a very expressive Google Home that can move itself, has a camera, and you connect it to any AI provider, and it has, like, a fully open app ecosystem with a ton of great examples. So you can download a ton of stuff for this. Or right now, I wrote an app that connects mine to my Hermes.

So when I'm chatting with this, it's actually using Hermes to get the response. And it has a microphone and speaker, so it will do Oh, okay. Text to speech, speech to text.

I'm still getting into it. Like, I have a ton of fun app ideas. We just arrived last night, and we just finished building at, like, 8PM.

So I haven't had any chance to really play with it beyond just getting it up and running. But what a cool

cool ass little thing here. Alright? It it's pretty sick. So it has Wi Fi, has motors, has cameras, has speakers.

Yes. And it's cute. So you could, like, you could And it's cute. Build

so many things with it. Man, my my mind is just running. The fact that you made it, like, a Google Home that you can program is, like, such a good way to put it. Yes. I think that's what it feels like to me. But, like, what I'm gonna use it for is, like, I'm gonna, like, hold up our kids' spelling list, and then I'm gonna make an app that they can practice spelling with it and use, like, time repetition. Because one thing that our kids get really frustrated at when they're when they're, like, bad at a word, they get really worked up and frustrated about it. And I'm wondering if, like, having this cute ass robot be like, oh, that's all I've tried to remember, like, will will help them, like, be engaged a little bit more than just like, oh, let's try again. You know? So I'm gonna build a spelling app for them. I'm gonna build, like I said, I'm connecting it to Hermes, so that will be fun. And then I'm gonna try to build an app that helps us learn Warhammer 40 k.

So that way, you could, like, look at the battlefield and tell us, like, what to do next and stuff. That is cool. Very stoked. You know what? Like, if anyone listening, like, another cool thing you could do is you could throw

what's the model? I made, like, a, like, a barbell detection couple years ago. Every single time you, like, did a rep that counted, that was using why am I forgetting the name of it? It's like a

TensorFlow? No. Not TensorFlow.

No. MediaPipe. MediaPipe.

It was using MediaPipe. And there's there's three or four models. So you could there's, like, a hand tracking MediaPipe. There's, like, a body tracking one, and it will, like, know where your elbows are.

Perfect example. That thing can run on low power hardware media pipe. You could throw that on the robot and have the robot watch you, and you could figure out like, count the jumping jacks. You know? Like, that would be one little example you could do with this.

That would be cool. So there's two the you there's two versions of this robot that you can buy.

You have to assemble them also if you're not technical and can't you know, you're not comfortable with assembling it. Like, this was not an easy build. There's no soldering, but it wasn't an easy build. Don't get it because it's not plug and play.

And it's definitely, like, more technical, but, like, yeah, you you could in one version of this Wes, there's a built in Raspberry Pi.

This version is not that because those are not out yet. This is the one it does all compute on my laptop.

So it can connect over Wi Fi, and then the compute happens on my machine. So Okay.

Wes. It it communicates, with the laptop to to handle some things and so apps you're running. But the in the Raspberry Pi version of this, it would happen on just all on robot itself. Okay. I see. That that's what I did with my,

Roomba as well. Right? Like, I Yes.

I put no. I didn't, actually. No. I put the I put the code on pnpm ESP 32.

But when I was first doing it, I was communicating wirelessly from my computer and so I could I could do the more intensive stuff. That is sick. I want one of

these fairly to, I'll I'm gonna one, I'm gonna make a a big detailed video on it, on the build, assembling it, all that stuff, talking quite a bit about more about this in the future.

And, I think you'll see a lot for me on this thing because I, I'm very inspired with it. I'll tell you that. So

Sick. Alright. Thanks, everybody, for tuning in. We'll catch you later.

Peace.

Peace.