Aryan

2024 wrapped 🚀🚀🚀 - Shipped projects with an amazing team, supporting 15,000+ users in production. - Made $25,000 from hackathons (started with $0), working less than a month total to free up time for other pursuits. - Built my own AI agents with local LLMs, changed how I work. - Took 10 flights in 2 months, still missed the Goa trip. - Old MacBook died, downgraded from Pro to Air. - Became an @AtriumAcademy alum. - Learned to bake zero-sugar, high-protein brownies.

@MayankSaraswal @ChaiCodeHQ @Hiteshdotcom @nirudhuuu @piyushgarg_dev @yntpdotme Well deserved, Mayank👏

Secured 2nd rank in the Buildathon 🚀 and received these amazing swags 😊 Big thank you to @ChaiCodeHQ @Hiteshdotcom @nirudhuuu @piyushgarg_dev @yntpdotme @aryankumar877

@adymitruk @TomLibelt @naval I'm with you on that fs, it's just a matter of when

That's not a big problem that can't be generalized. AI will do a better job than humans. Think of Event Sourcing and the many custom models it can provide. AI can figure out the index tree in no time and make it work across clusters with paxos, raft, casper, whatever. It not an if, it's a when.

AI coding agents can now deliver one-shot custom apps straight to your phone. It’s the beginning of the end for the iPhone’s dominance.

@adymitruk @TomLibelt @naval Tho I'm with you on that they'll improve, they aren't there yet tho for my use cases

@adymitruk @TomLibelt @naval For that matter real time sync between thousands of users, not just 2 sometimes

@adymitruk @TomLibelt @naval that's literally the most simple visualizer one can make, and hence you proved my point

@aryankumar877 @TomLibelt @naval Wrong. I made a better optionstrat in 5 mins.

@hiteshdhayall For sure hitesh, no worries and awesome 👏

@aryankumar877 Ohh . Sorry sir, thoda mis understanding hogya tha . But still looking forward to learn a lot from you

2024 wrapped 🚀🚀🚀 - Shipped projects with an amazing team, supporting 15,000+ users in production. - Made $25,000 from hackathons (started with $0), working less than a month total to free up time for other pursuits. - Built my own AI agents with local LLMs, changed how I work. - Took 10 flights in 2 months, still missed the Goa trip. - Old MacBook died, downgraded from Pro to Air. - Became an @AtriumAcademy alum. - Learned to bake zero-sugar, high-protein brownies.

@adymitruk @TomLibelt @naval Todo apps for sure yes, but for most real world utility apps agents are just writing bugged out code that doesn't work

@TomLibelt @naval Many people make apps in minutes. It's over.

@hiteshdhayall Hey hitesh, I work at chaicode, not a student. Happy to connect with you too.

@aryankumar877 Hey bro.. Let's connect Also in the same COHORT26 And I also liked the kinda stuff u post

We saw some of the most interesting submissions to date, demonstrating the immense interest for decentralized governance powering startups on Solana. Honorable mentions 👏 - @JustSparkdotFun - RealmsSmith by @Digitalmonarch - ClawdDAO by @aryankumar877

@theo Why did you hop back if it's so awful 😔

Had to hop back into Claude Code for some UI changes. I have no idea how y'all deal with CLIs for working with agents. This is awful compared to a real GUI.

@yoavtzfati @theo zcode by @Zai_org is built on tauri

@theo Same/Better than tauri?

“Why did you use Electron for T3 Code?” Because I evaluated every option, and Electron had the best performance across all platforms for most things.

@addore_ @complex_maths @Rasmic @theo you both know t3code uses the same codex cli underneath the js wrapper right?

@complex_maths @Rasmic @theo He's not comparing codex (the cli) with t3 code( the desktop app)... He's comparing codex (the electron based desktop app) with t3 code... They're both electron

t3code is more performant than codex in alpha and that’s just crazy 🥲😂

@ChaiCodeHQ @kunalsh2403 congrats @kunalsh2403 !

And we have a winner! 🏆 The JavaScript Quiz Bounty goes to @kunalsh2403 Congrats, Kunal for winning ! 🎉

3/ 🥈2nd place — ClawdDAO ~5% voter turnout in Solana DAOs. ClawdDAO's answer: AI agents that internalize your governance values, vote through Realms on your behalf, and publish their reasoning to Tapestry so you can audit every decision. Built by @aryankumar877 🔗 Clawddao.aryankumar.dev

@ZixuanLi_ it's already crashed

GLM-5 traffic is surging while Claude is down. I hope our service doesn't crash and burn in the next few hours.

@ashoKumar89 @naval but sometimes u just gotta check each line of code, the models just can't do somethings beyond their limits yet

@naval Coding is expanding from “write logic” to: - Orchestrate models - Define evaluation loops - Build guardrails - Design feedback systems Execution cost is falling. Judgment cost is rising.

The cost of code is coming down, so we will consume more of it. The productivity of coders is going up, so they will become more valuable. Coding now includes training and driving models.

@srtenginamath @ChaiCodeHQ @Hiteshdotcom @nirudhuuu ❤️

Masterji app’s new UI is 🔥😍 Clean, simple, and intuitive. Great work @ChaiCodeHQ Team @Hiteshdotcom @nirudhuuu

@ankgogawat @ChaiCodeHQ @Hiteshdotcom @yntpdotme @nirudhuuu ❤️

Wow look at this masterpiece 😍 new UI of masterji @ChaiCodeHQ @Hiteshdotcom @yntpdotme @nirudhuuu

@mehulmpt Your biggest mistake is trusting the benchmarks.

Gemini 3.1 Pro beats Opus 4.6, which was released like 10 days back, at half the price. The bloodbath in AI world is crazy