Bun Got Bought by Anthropic  ## Metadata - Author: [[YouTube]] - Full Title: Bun Got Bought by Anthropic - Category: #articles - Document Tags: [[advocacy]] [[metrics]] [[monitoring]] - Summary: Bun got bought by Anthropic (yes really) - URL: https://www.youtube.com/watch?v=Yj9677a3y6c It's no secret that I've been a huge fanof Bun for a while. Way back in 2022, Iwas Jared's first interview talkingabout it. And the ride we've been onsince has been crazy. I'm a fan, a user,a builder, and an investor, all workingwith Bun wherever it makes sense to. AndI'm really, really lucky to considerJared a good friend. So, knowing thatthere is bias to account for here,especially when you realize that theywere just acquired by a company I'm notnecessarily known for being the biggestfan of. Yes, Bun is joining Anthropic. This seems just absolutely insane. Andas a result, I've seen even more insanetakes about it. From, oh, they're beingaqua hired, Bun's dead, to, "Of coursethey did that. They're going to build acloud environment to compete withVercel for Cloud Code." And yeah, allof this is insane. Just as insane as metechnically having equity in Anthropicnow, which is kind of hilarious that Ihave equity in Anthropic and not OpenAI. All of that aside, I have thoughts here.I also have a lot of inside info, so Ihave to dance around it to the best ofmy ability. But, uh, this is a weird topicbecause I am uniquely qualified both asA friend and investor of Jared and Bun,but also as somebody who's deep in theweb world, deep in the building worldfor full stack stuff around JavaScript,and also really deep on the direction ofthis AI stuff and where Enthropic is trying to go. We have a lot to talk about here, guys, and I can't wait to do it,but I didn't get a payout for this yet.So, let's hear a quick word from today's sponsor. There are a few things that hurt developer velocity morethan slow builds. If a user reports a problemand you go and fix it in 10 minutes, but you have to wait two hoursfor the build to complete just to realize there's a small mistakeand then go through the whole loop again, there goes all of your agility.You're throwing away so much opportunity to move fast and ship fast.And those blocks will keep your engineers from staying motivated on those tasks.If you're looking at a huge pile of tickets that have been building up over time,there's a good chance your build time is a part of why that's happening.That's why I look at this number and just can't fathom why anyone isn't already usingDepot. Seriously, Post Hog's actualProduction builds went from 153 minutesto 4 minutes. I know it sounds too goodto be true, but let's be real. A lot ofthis is because GitHub doesn't know whatthey're doing. But the other side is that these guys are so locked in andfocused on performance. Depot isobsessed with squeezing every ounce ofpotential improvement in your GitHubactions. Honestly, if this costs 10times more than GitHub actions, I'dprobably still be a happy customerbecause of the amount of time it'ssaving my teams. But it ends up beingcheaper, which is just insane. Sincemigrating from default GitHub actions,our Docker image build times haveliterally been cut in half. Dev isincredible for reducing our build timesby 10x from 8 minutes to 20 seconds. Depot's awesome. Got my build in pushfrom 30 minutes to 4, and it will beabout 12 seconds when it's cached.Moving to Depot takes 5 minutes. So, ifyour builds are longer than that, yournext build will cut the cost difference.It's so worth it. Give them a shot nowat soyave.link/depo. Where to even start with this one? Thereare two articles as well as all theThings people are saying and then, ofcourse, my own speculation that we'll bedoing near the end. So, if you want tohear my thoughts on why they're doingthis and where it's going, stick to theend for that. Oopic acquires Bun asClaude Code reaches the 1 billionmilestone. Pretty crazy revenue forClaude Code, but also they're basicallygiving discounts for Claude by usingClaude Code. So, I get it. Claude is theworld's smartest and most capable AImodel for developers, startups, andenterprises. Kind of crazy that thefirst sentence is nothing about Bun atall. In fact, you have to go quite a bitbefore Bun is mentioned. I heard someonesay that the amount of mentions ofthings in here is weird. Claude ismentioned 25 times. Claude Code ismentioned 11 times. And Bun is mentioned18 times. Fun. Cloud Code represents a new eraof agentic coding, fundamentallychanging how teams build software. InNovember, Cloud Code achieved a significant milestone. Just six monthsafter becoming available, it reached abillion in run rate revenue. Today,we're announcing that Anthropicsacquiring Bun, a breakthrough JavaScriptRuntime to further accelerate CloudCode. This is the key piece here. We'lltalk more about the importance of thisin a bit. Bun's redefining speed andperformance for modern softwareengineering and development. Founded byJared in 2021, Bun is dramaticallyfaster than the leading competition. Asan all-in-one toolkit combining runtime,package manager, bundler, andtestrunner, it's become essential infrafor AI software engineering, helpingdevs build and test applications atunprecedented velocity. Bun has improvedthe JS and TypeScript developerexperience by optimizing forreliability, speed, and delight. Forthose using Cloud Code, this acquisitionmeans faster performance, improvedstability, and new capabilities. Together, we'll keep making Bun the bestJS runtime for all developers whilebuilding even better workflows intoCloud Code. Here's a whole section ofglazing Cloud Code. We've been a closepartner of Bun for many months. If youdidn't know, Cloud Code bundles Bunalongside it when you install it as theruntime for Cloud Code. Ourcollaboration has been central to theThe rapid execution of the Cloud Code team directly drove the recent launch ofCloud Code's native installer. We knowthe bun team is building from the samevantage point that we do at Anthropic,with a focus on rethinking the developerexperience and building innovative anduseful products. I did notice this whenI last installed Claude that there was abunch of callouts for native in it. Ithought they might have done the unnecessaryRust rewrite that a handful of theseproducts have done that didn’t need it. No, that’s not what they did. They didthis to create a single binary that you caninstall using bun. It’s simpler, morestable, and doesn’t require Node. Werecommend this as the defaultinstallation method for all Cloud Codeusers going forward, with brew or curlor on Windows with IRM. Makes sense. This bundles a single binary that iswrapped with bun so that it can executeand manage its own runtime and it won’trequire you to have the right versionof Node installed. I know a lot ofcompanies get really annoyed aboutsetting up Cloud Code and these thingsbecause they want a single binary theycan approve and distribute internally. But if you also have to have Nodedistributed internally with the rightversion and everything, then the ITteams get all confused with it. I thinkthat's stupid. Your IT team should becompetent if they're controlling whattech your developers can use. But a lotof companies aren't competent, and assuch, this type of single approvablebinary is necessary. I've heard frommany that this is the reason why Codeexgot rewritten in Rust, which is theworst possible reason to rewritesomething in Rust. I much prefer themethod that Anthropic took here. Theyeven got the CPO of Anthropic to leavesome comments. Bun represents exactlythe kind of technical excellence we wantto bring into Anthropic. Jared and histeam rethought the entire JS tool chainfrom first principles while remainingfocused on real use cases. I kind of disagree here. They didn't rethink thetool chain so much as rewrite it so thatit's fully compatible. Something likeDeno was the rethinking of the JS toolchain, which failed, and now they'retrying to do compatibility as well. Bun is acompatibility play. Bringing the Bunteam into Anthropic means we can build. The infra to compound that momentum andkeep pace with the exponential growth inAI adoption. The decision to acquire Bunis in line with our strategicdisciplined approach to acquisitions. Wewill continue to pursue opportunitiesthat bolster our technical excellence,reinforce our strengths as the leader inenterprise AI, and, most importantly,align with our principles and mission. And one last important piece becausepeople are worried about Bun goingprivate, which makes sense because cloudcode is the only one of these majoragent coding tools that is still closedsource. Bun will remain open source andMIT licensed, and we will continue toinvest in making it the runtime bundler,package manager, and test runner ofchoice for JavaScript and TypeScriptdevs. I already see people in chat beingskeptical of that. I'll say it myself. If they were to screw with the opensource nature of Bun, I will be theloudest person complaining about that byfar. They will never hear the end of itfrom me. I would argue that this is nowthe time to do the opposite. They need totriple down on open source. They need toopen source cloud code alongside this. There is no world in which they stopsupporting Bun as a result of this. Infact, it's the opposite. Enthropic betsso hard on Bun that Bun not gettingprofitable and then dying is a huge riskfor them. We'll talk about all of thatin just a bit. First, I want to readthis from Bun's perspective. Bun isjoining Anthropic. Anthropic is bettingon Bun as the infrastructure poweringcloud code, the cloud agent SDK, andfuture AI coding products and tools. So,what doesn't change? I really like thisis at the start. Bun stays open sourceand MIT license. It continues to beextremely actively maintained. Same team is still working on Bun. Pretty mucheverybody at the Bun company is comingthrough for this acquisition. And theone person who isn't that I know ofbecause they were still in the processof being hired, they got a reallygenerous payout. So like Jared wentreally hard to make sure everybody atBun and Oven won here. And that'ssomething that matters a lot to me. Mybiggest concern with Bun has always beencan Jared run a company that makes moneyand get the win back to his team? I wasconcerned he might not have it in him. But this is the perfect path. Jared's abuilder. He wants to build. He doesn'twant to be an executive. He doesn't wantto run a company. He doesn't want todeal with profitability and all of thesethings. Someone who's talked with him abunch about it, interviewed him androasted him privately and publicly overhis lack of vision for a way for them tomake money. Like, this is the perfectending for that. This is letting him gofocus on what he's good at, get his teamwhat they deserve for the work they'vedone, and everything gets to accelerateas a result. Also, of course, Bun isstill public on GitHub. They're notgoing to change that. And Bun's road map will continue to focus on highperformance JS tooling, node compatreplacing node as the default serversideruntime for JavaScript. The one thing Ido expect to change is that we're goingto see less of these like Reddusconnectors or like the MySQL pluginbuilt into Bun. I would expect less ofthat type of thing and more of the likelocked in focus on making Bun the bestruntime for node applications. Cloudcode ships as a Bun executable tomillions of users. If Bun breaks, cloudCode breaks. Anthropic has directincentives to keep Bun excellent. Yes, perfect. So, what is changing? Wewill help make coding tools like CloudCode and the Agent SDK faster andsmaller. We get a close first look atwhat's around the corner for AI codingtools and making Bun better for it. Andmost importantly, Bun will ship faster.Now that Jared's not distracted byrunning a company, dealing withinvestors, and all his other things, andcan just focus on shipping, he's goingto ship hard. My only concern is thatAnthropic has enough interesting stuffgoing on that he might get nerd snipedby something else. But now that Bun hasthe resources it needs internally, theycan keep moving. There's a fun little like how Bun started story in here. I'lllink it in the description if you wantto read the whole thing. I'll breezethrough it really quick though. Five yearsago, Jared was building a voxel game inthe browser, kind of like Minecraft.Cobbase got large. Iteration cycles took45 seconds to test if changes worked. Hewas using Next.js and the hot serverreloads were taking almost a minute. Hegot frustrated, so he started porting ESBuilds JSX and TypeScript transpilerfrom Go to Zigg all the way back in2021. Three weeks later, I had a somewhatworking JSX and TypeScript transpiler. I spent much of the first year in acramped apartment in Oakland coding andtweeting about Bun. In order to get theserver-side rendering to work, you needa runtime, and all the runtimes need tointerpret and JIT the code. So, hestarted working on top of WebKit to makehis own runtime. Bun 0.1 was released inJuly. This is when I interviewed him, bythe way. Looking at that in retrospectis wild. There are 20K stars on GitHub. Whathe doesn’t mention is that he also had a YCcompany that was still active but inpivot hell, and he pivoted it to Bun. So technically, Bun is a YC company, but it’ssimilar to how T3 Chat’s a YC companywhere Ping was a YC company. Ping didnot go great. And then T3 Chat happenedunder the same company and went waybetter than expected. They raised a biground from Kleiner Perkins. He took asmall salary. He always underpaidhimself. Believe it or not, Jared untiltoday was broke. So, it’s going to befun to see how he handles having money.My suspicion is nothing will change. They raised again, this time a $19 million series A. This was in September 2023, after the 1.0 launch. 1.1 added Windows, and 1.2 made big changes fornode compatibility, added built-in Postgres and S3.This is when X and Midjourney started to use them. It's my understanding that X is using them mostly for package management, not theruntime. Midjourney, I believe, is using the runtime as well. This is part of what makes Bun complex. It's a bundler,a runtime, an alternative tool set, and a standard library for a bunch ofthings. It's a lot of stuff in one. It seems like the majority of people using Bun are primarily using it for theinstaller and bundler. Anthropic is one of the few companies using it for the runtime, which I think is a really safeand good thing because them using it for the runtime means that the acquisitionisn't going to be the end of the runtime. Like if Twitter or X purchased Bun, they would be all focused on thebundler and the package manager side, probably not the runtime side. Anthropic, one of the few companies that uses most of Bun and can benefit from all of it. Then 1.3 dropped. I actuallyI did an interview with Jared in person ata small 1.3 launch event in the cityjust a few weeks ago, and I grilled himon not having a plan to make money, andhe really didn’t at the time. It seemslike it was just a few days after thatthat things started to reallyaccelerate. AI started to get good inlate 2024. AI coding tools went fromcool demo to actually useful. I knowthat Jared has been using Cloud Code andClaude models for doing stuff in Bun fora while now. He used to very excitedlysay that for the first time, Jared’s notthe one committing the most code to Bun. It’s now Claude. And since Bun is asingle-file executable, you don’t haveto do a whole system install with abunch of dependencies. You can just bundle Bunitself. It’s easy to bundle with the CLItools. You can compile any JS projectinto a self-contained binary that runsanywhere. Even if the user doesn’t haveBun or Node installed, it works withnative add-ons. It has fast startup, andit’s easy to distribute. That’s whyCloud Code, Factory AI, Open Code, andtons of other things are all built withBun. And the result of this was thatJared got obsessed with Cloud Code. GitHub username with the most merge PRsand Bun's repo is the Claude Codebot. Wehave it set up in our internal Discordand we use it to help fix bugs. It opensup PRs with tests that fail in earliersystem-installed versions of Bun beforethe fix, and it passes in the fix debug Bunor build. It's doing the full process of writinga test, using the new version, changing ituntil it works, and responding to reviewcomments. It does the whole thing. Thisfeels a few months ahead of where thingsare going; certainly not years. And hereis the key piece: today, Bun makes 0 inrevenue. If you watch my most recent Bunvideo, the last third of it is meroasting Jared for this and trying todescribe a path for them to make money. This has clearly been stressing himas well as the team at Bun. I can'ttell you how many private conversationspeople at Bun have worried aboutwhat the long-term plan is for thecompany. One of the most commonquestions I get is about sustainability.Questions like: how does Bun becomea business? If I bet my work project orcompany stock on Bun, will it still bearound in 5 or 10 years? These are key.This is one of the big things I was nervousAbout Bun, I don’t know whereit's going to be because it wasn'tmaking money, and you need money tosurvive. I know Nex.js is still going tobe here in five years, even if it's in arough state because it's a huge part ofVercel's business. They have a lot ofmoney, and they want to keep making money. They need Nex to succeed. I don’t knowthat about Bun. The DA has always beenwill eventually build a cloud hostingproduct integrated with Bun's runtimeand bundler. Same with Deno, and that didnot work for them. Getting this right isreally hard and would take Jared's fullfocus, which means he wouldn’t beworking on Bun, the actual open sourceproject as much anymore, which he doesnot want to do. Jared just wants tofocus on the thing he cares about. Andmaking money is not the thing he caresabout. As someone who’s a good friend ofhis, there are few people who are lessmotivated by money than him. The onlymonetary motivation I’ve seen from himis getting his team paid. The money wasa way for him to keep working on thisthing he cares about, not a thing thathe wanted to have a whole bunch of inhis bank account. Now his team is paid. Now his burden is cleared and he can golock in and build. Forcing ourselvesdown the prescribed path felt wrong whenAI coding tools are getting this goodthis fast. He also just didn’t want todo it. I love this. The walk. Thispart's very funny for reasons we'll getto in a minute. We've been prioritizingissues from the Cloud Code team forseveral months now. It'd be a lot easierif Cloud Coco's open source. Justsaying. I have so many ideas all thetime and it's really fun. Many of theseideas also help other AI codingproducts. Look at that. Dax getting alink in an anthropic post. Me every timeI need something in Bun fixed for opencode and it's the video of just themsaying Jared, Jared over and over again. A few weeks ago I went on a four-hour walkwith Boris from the Cloud Code team. Wetalked about Bun. We talked about whereAI code is going. We talked about whatit would look like for Bun's team tojoin Enthropic. The link here is veryfunny. Oh [ \_\_ ] no. What happened October28th? This is the only day that we knowof where Jared wasn’t writing commits.He has one of the craziest contributiongraphs we've ever seen. Jared, all things happening on October28th, there were no contributions. Dax tweeted, veryconcerned about this. That was becausehe was on this walk with Boris from theCloud Code team.It's very funny that Dax got two tweet linksin this article. We talked about what itwould look like for Bun's team to joinAnthropic. Then we did that about threemore times over the next few weeks. ThenI did that with many of the competitors. I think Anthropic is going to win. Yeah,he did his due diligence. I'm sure hetalked to everybody from Codeex toCurser. I don’t know. He didn’t tell mewho else he talked to. All he told mewas how excited he was to go toAnthropic. He really seems excited aboutworking with these guys in particular,the Cloudco team. And I even have somegood friends on the Cloudco team who areawesome. Shout out to Daisy. She's thereason I’m not a lot harsher onClaude Code. She clearly really caresand I know there are a lot of peoplefighting internally to make it opensource. Just to reinforce the point,Anthropic, we can bridge thisgap. You guys are making a lot ofprogress right now. Of good things, and this is a huge winfor the whole ecosystem. As long as youdon't screw over Bun and the Bun teamand people using it in any meaningfulway, we're going to be a lot better. But if you, on top of that, were to opensource cloud code and start contributingmore to the open source JavaScriptecosystem,Slate's cleared. We're good. I would bea very proud investor and shareholder inAnthropic at that point. So, uh, to thehandful of people on Cloud Code that arestill convinced it's your secret sauceand open sourcing it would bedisastrous, you're wrong. Just do it.Betting on Anthropic sounds like amore interesting path to be in thecenter of things, to work alongside theteam building the best AI codingproduct. Jared likes chasing what he'sexcited about. This is ADHD winning, andit's winning hard here. Jared's the typeof person who was half an hour to anhour late to his own launch eventbecause he was coding. This is all a bitcrazy. At the time of writing, Bun'smonthly downloads grew 25% over the lastmonth, passing 7.2 million monthly downloads. To figure out monetization. We didn'thave to join Anthropic. They had fouryears to figure this all out and they'restill doing it. That said, it does getharder every month to figure out how tomake money. Every product I've everlaunched started with a paid tierbecause we don't want to have this risk.Shout out to the people who do. It'sscary. I don't want more users to meanwe go out of business faster. Personally, this is how I build and thecompanies I invest in tend to alreadyhave revenue. It's scary. But if you'resmart enough and know what you're doing,I'll throw some money in, which is whatI did with Jared and Bun. And here weare. Instead of putting the users andthe community through Bun, the VC-backedstartup tries to figure outmonetization. Thanks to Anthropic, wecan skip that chapter entirely and justfocus on building the best JS tooling. When people ask, will Bun still bearound in 5 or 10 years? Answering with,we raised $26 million isn't a greatanswer. Those investors eventually needa return or you have to raise more moneyuntil you can get that return. There's abigger question behind that. What doesWhat does software engineering even look like in 2 to 3 years? Yeah, we're now at the pointwhere the smartest engineers I know aresaying this. Two years ago, the dumbestengineers I knew were saying, "AI is going to replace everyone." Lastyear, everybody was pushing back on it,like, "No, this stuff's not that good. You can't build real software with it."This year, the smartest people I knoware like, "Oh, things arechanging." AI code tools are gettingreally good, really fast, and they'reusing Buns, single file executables toship CLIs and agents that runeverywhere. If most new code is going tobe written, tested, and deployed byagents, then the runtime and toolingaround the code are way more important.There's a lot more code overall thatneeds to be written and tested a lotfaster. And humans are more detachedfrom every individual line. So theenvironment it runs in has to be fastand predictable. Bun started with afocus on making developers faster. AI codingtools do a similar thing. It's a naturalfit. And that's why they're joiningAnthropic. Anthropic is investing in Bunas the infrastructure for cloud code and cloud agents. SDK and future AI coding products. Thisis why people are thinking that this isa cloud bet. I'm sure if they do makemore cloud plays, there's a good chanceBun will be used for that. But thisisn't what I thought initially. Thisisn't Anthropic buying Bun to make themcreate the internal Anthropic cloud. Thisis them investing heavily in the toolsthat work. Because, as I've said manytimes, the code quality at Anthropic isnot great. Just go use the cloud websiteif you don't believe me. Like, it's justthe quality of the code beingshipped there does not even come closeto the quality of the models and researchthat they're doing. They almost feel atodds with each other. The quality ofunderstanding of the JS ecosystem andtooling at Anthropic has more than doubled through this acquisition. Theirjob is to make Bun the best place tobuild, run, and test AI-driven softwarewhile continuing to be a great generalpurpose JS runtime bundler, packagemanager, and test runner. Being part ofAnthropic gives Bun long-term stability,a home, as well as resources, so peoplecan safely bet their stock on Bun, afront-row seat to where AI coding toolsAre headed so we can shape Bun aroundthe future instead of guessing from theoutside, as well as more firepower. Theycan actually hire more and not worryabout their run rate dwindling and theirtime until they are out of money happeningfaster. It’s really scary to hire whenyou have a certain amount of runway. When your company only has enough moneyfor three years and then you hire andnow you have two and a half years, youhire again, you have one year. It’sterrifying. And of course, same keypoints from the top. It’s open source.It will still be built in public. Thesame team is working on it, and they’restill obsessed with making JST faster toinstall, build, run, and test. And ofcourse, if you’re skeptical, make sureyou’re subscribed to my channel. Onlyhalf of you guys are. It’s just a littlebutton at the bottom of the screen youcan click. I’ll be covering this. Ifanything changes going forward, I willbe the loudest annoying voicecomplaining if Anthropic screws this upat all. I genuinely don’t think theywill. Again, as the person who is mostskeptical of Anthropic most often, Ithink this is a good alignment and pullsAnthropic is moving in the direction I want to seethem going in. This feels good to me. I'm not just saying that because I’m nowtechnically an investor. That one'sweird. I'm saying this because I reallybelieve in the BUN team and Itrust the people who see what I see in them. So, subscribeif you want to make sure that I can holdthem accountable. Quick FAQ, and thenI'll dive into my own skepticism here.Is Bun still open source? Yes. Willit still be on GitHub? Yes. Do they stillcare about node compatibility? Yes. Isthe same team working on Bun? Yes. And now they don't have to worry aboutmaking money somehow. What does this dofor Bun's roadmap? Bun's team will workmore closely with the cloud code team.It will probably look similar to therelationship between Google and V8,Safari and JS Core, Mozilla Firefox andSpider Monkey, but with moreindependence to prioritize the widevariety of ways people and companiesboth use Bun today. Let's do some quickmemes.We finally can see the back of the Bunlogo. Quality meme. Good work, Terminal.Dax quoting, "How funny is this?"The 4-hour walk thing being the day thathappened, which is just hilarious. But also Jared's— this is my favorite.People would always ask, "How is bunsustainable? If I bet my company's tuckon bun, will it still be around in a fewyears?" We didn't have a great answer tothis question until today. Yeah, theydid it. And I still see a lot ofskepticism in chat. Used to love bun.It's going to [\_\_]. I can't trust them.I'm so worried. All of these peoplesaying things like this. Again, I am theperson who loves to do this more thananybody. It's like my whole thing to[\_\_] on anthropic. But the harsh realityis that there's a lot of goodengineering going on that makes thismake sense. We need to go back in time alittle bit to when OpenAI dropped 04Mini, because I was excited about theircodec CLI in here in particular, but Ihad a particular statement I made aboutit. I want to find ink.js. Ink, ifyou're not familiar, is a React rendererfor the terminal, similar to how ReactNative lets you render on iOS orAndroid, or React 3 fiber lets yourender in 3JS and canvas. Ink.js letsyou render in the terminal. Very cool. Stuff means that you can use all theReact patterns you’re used to. Thatsaid, I’ve seen a couple of people move offof Ink because it gets rough at scale.Just maintaining it well with theperformance can be a real challenge. TheWrangler CLI by Cloudflare used to beusing Ink. Then they moved to Go. Thenthey moved back to TypeScript, but nomore Ink. JJS, so it’ll be more performantand maintainable. So yeah, I’ve seen alot of back and forth with this. And Ihave a bad feeling if they haveperformance issues, they’re just goingto leave JS. Then realize they werewrong and go back to it. So, uh, to theOpenAI people watching this, to Fluid inparticular, don’t move off JS. If youhave problems, talk to me. We’ll get youguys off Ink if it isn’t the right thingfor what you’re doing. But if it doeswork, great. Keep using it. It’s a yeah,they moved to Rust a week after I saidthat, almost like I called it. Yeah. So, this has been a problem with alot of these. They move out of theJavaScript ecosystem. So the other biglab that could afford them, OpenAI,won’t do it because they moved off of JSbecause they wanted the fastest, mostThe powerful, best possible thing. What'swrong with Rust? It's impossible tocontribute to unless you're like afull-time Rust person. Iteration speedgoes down a ton. The flexibility of thetool set goes down a ton, too. If itturns out you need to make a change,good luck. Have fun. There's a lot ofthings that are making the developmentof codecs a lot slower. They're puttinga lot of engineering effort and moneyand time into it, but it's still iterating way slower than Claude is. Andeven then, like Open Code is, and all ofthese smaller teams and companies thatare building on top of JS. There's a lotof ways they could have made Codexfaster and a single binary, which werethe main goals that didn't involverewriting in Rust. Yeah. Yeah. This iswhy that happened. Their strategy wasthis bold rewrite that slows down thewhole team. Anthropic's strategy was toacquire the dependency that helps makethings way faster and safer withoutrisking the whole business. Even Rustfans can admit this. I love Rust andwish I could use it in production, butit just takes too long to write. Yes. Also, TypeScript and Python are almostIt's never slow enough that you need C orRust for speed. TypeScript and Pythonare just as good for security. Yes, andon top of that, since the thing that isslow in these apps isn’t, hopefully isn’tthe code, it’s the LLMs. If you’realready having to wait 13 minutes for aresponse from one of these LLMsgenerating all this stuff, the amount ofexecution cost that you’re eatingusing JavaScript versus Rust is very, verylittle, like almost nothing. This is abig thing I’m talking about a lot withJavaScript; its backends went from making somesense to making a lot of sense when theaverage request time went up so much. Because if I’m observing 500 requests ata time, all doing LLM generations whereI’m streaming down tokens from a thirdparty API, having the JavaScriptconcurrency model around the event loopmakes that work really, really well. Because most of the time your JavaScript is justwaiting for the next packet to come inso it can forward it down to the user.JavaScript is great for this. Setting up goodenough multi-threading systems in otherlanguages to handle that type ofconcurrent workload is not fun at all.Go can kind of do it. None of the otherMajor languages handle that specifictype of concurrency through requestloads particularly well. JavaScript doesthis beautifully. You're going to bebuilding your own concurrency and eventmodeling solution if you're not using JSor Go. Why not just use them? Like, yeah,it's insane to me. So that's anotherhuge part. Anthropic's the biggestcompany by far that is going all in onbun like this if they want to make suretheir dependency survives, because likeyou know the classic XKCD dependencychart where you have all moderninfrastructure and then a project fromsome random person in Nebraska. Think ofthis as everything Anthropic is building. The guy who doesn't know how to make moneyis building bun, VC-backed, a huge, huge risk forAnthropic. But now that this dependencyis owned by Anthropic and they havecompetitors that are using that samedependency, there is some incentive forthem to lock it down and not help withthings that help those competitors. Thefact that Cloud Code itself is closedsource really makes it feel likesomething they would do. And I, again, Iimplore y'all on the Cloud Code team,Boris, Cat, Jared, everyone else. Involved in these decisions, there's nogood reason to keep cloud code closedsource. It doesn't make sense. I'veheard arguments for everything fromsecret sauce to maintaining the repo, andthe community around it is difficult.You now have one of the best open sourcemaintaining groups ever in this webecosystem working for you guys. There'sno reason not to do this. It will makepeople so much less scared about thisBun acquisition. If you want to reduceall of the FUD, the fear, and theconcern people are having in my chatright now, make the good faith play. Open-source the cloud code. I don't know whatyou're hiding or what you're scared of.You need to just do it. I thinkthat's all I have on this one. I'mexcited for the future of Bun and thefuture of cloud code and all the thingsthat are going to happen here. The onlymajor risk I perceive is Jared gettingdistracted by something else and nolonger focusing on Bun as much now thathe's going to be surrounded by coolthings going on. There's a decent chanceof that happening, but I am sure thatBun will be in very good hands now thatit is finally funded properly. Honestly, I’m just relieved that I no longerhave to talk about whether or not Bunwill ever make money or how it will behere in the future. We have our answer:Anthropic. One of the best things thatever happened to software developmentfrom AI is this: the fact that AI moneyis allowing Bun to be maintained andexist long term. That’s a very goodthing, and I’m excited for the futurefor the BUN team. I’m also moreexcited than ever for Anthropic. And no,not just because I’m now technically aninvestor, but because this is a really good faith play. I trust Jared’sjudgment. If Jared talked to all theteams and this is the one he wants towork with, I trust him. I’m askingyou guys, too, as well. I really hopethat my trust doesn’t get burned bythese guys. But Jared’s never burned mebefore, and I trust him to do this right.He did his team right as well. Andthat’s something that’s really importantto me. You want to watch me hold theseguys accountable? Make sure yousubscribe. Until next time, peace, nerds., Bun Got Bought by Anthropic, YouTube, 01kbjn3wr2ackpryxny5anm4za.