Harsh

I love to teach things | YouTube 6.8k subs 👇

[Harsh @devloper_hs]

Vibe coded this solar system planet simulation with near realistic physics, speed controls, orbit disability, and virtual and real scale views with @Kimi_Moonshot (Kimi-K2)

Prompt
"Can you build a simple planetary simulation website. make sure to include everything within a single html file. Keep the planetary controls panel collapsible. Cover all the solar system planets with real details. Keep the planetary controls panel collapsible"

Output (One Shot)👇

[Kimi.ai @Kimi_Moonshot]

🚀 Hello, Kimi K2 Thinking!
The Open-Source Thinking Agent Model is here.

🔹 SOTA on HLE (44.9%) and BrowseComp (60.2%)
🔹 Executes up to 200 – 300 sequential tool calls without human interference
🔹 Excels in reasoning, agentic search, and coding
🔹 256K context window

Built as a thinking agent, K2 Thinking marks our latest efforts in test-time scaling — scaling both thinking tokens and tool-calling turns.

K2 Thinking is now live on https://t.co/YutVbwktG0 in chat mode, with full agentic mode coming soon. It is also accessible via API.

🔌 API is live: https://t.co/EOZkbOwCN4
🔗 Tech blog: https://t.co/n7xxaszqzF
🔗 Weights & code: https://t.co/4ukcXB0iP6

[Harsh @devloper_hs]

@original_ngv @IndiaToday Sir goat ko jarurat nai hoti news walon ko puchne ki, wo to seedhe podcast krte hn 🙂

[Harsh @devloper_hs]

Fr?
Someone show me the light!

[Harsh @devloper_hs]

@rohit4verse Got it.

[Harsh @devloper_hs]

@Harry_The_Nerd @dhruvtwt_ Are you rewamping them , or writing new ones ?

[Harsh @devloper_hs]

@TheVixhal Most companie's tech aren't mature either , they just go with hype not to lose behind , without even considering the foundations.

That's why foundation matters the most. Without it the tech breaks , gets corrupted or hacked.

Common nowadays.

[Harsh @devloper_hs]

Backpropagation is what happens when blame becomes feedback.

[Harsh @devloper_hs]

Building infrastructure and solving real problems is one of the best things one can do in this era, because without infra, agents break down.

Happy to see @contextkingceo & team doing it quite well.

Congrats on the launch, and all the best to @hydra_db for what’s next.

[Nishkarsh @contextkingceo]

Introducing HydraDB.

The graph native context infrastructure for agents. Purpose built to deliver precise context & observability into why agents act the way they do.

We've always believed graphs are the best way to manage AI context, but they've been too expensive to scale or impractical for storing full context. Until now.

@hydra_db combines in memory, NVMe, and object storage into a single graph layer, making context delivery faster, cheaper, and more precise.

We want context delivery to be extremely fast, 1000x cheap, and highly precise. Give your agents a brain.

[Harsh @devloper_hs]

@pandeyragini24 How to make this , happen ?

[Harsh @devloper_hs]

At Mount Kailash 🗻.

[harish.rs @Harish_521]

Where do you see yourself in 5y?

[Harsh @devloper_hs]

@original_ngv He is a ceo for a reason :)

[Harsh @devloper_hs]

@KaiXCreator All we need is nginx to run these now :)

[Harsh @devloper_hs]

@dhruvtwt_ Hm khn hn 😭

[Harsh @devloper_hs]

@helloiamleonie Mailing agent mail to request some funds and reverting back if it got that.

[Harsh @devloper_hs]

@23_dec_2005 The entire hackathon is a joke , but at least we get to learn, implement and see improvement , that's what matters.

[Harsh @devloper_hs]

@mehulmpt taxes are the real king here + govt regulations kills the momentum .

Moving to dubai is also on my bucket list , sf is also not bad , let's see when this manifests.

[Harsh @devloper_hs]

@Kcodess This is the reality , many founders are now opting for core skills and foundation depths.

This is done, so that they can evaluate , how you perform under sudden changes and guess you did quite well , congrats on that.

The same happened with me , my friend and many others :)

[Harsh @devloper_hs]

@VazeKshitij Thank you sir , same to you as well.

[Harsh @devloper_hs]

@kaaaash____ Cursor , Codes and Claude Code

[Harsh @devloper_hs]

@buildwithsid One time setup and maximum output.

Tried my hands on setting up a custom agent swarm that does entire data analysis , engineering , training , validation and submission.

Best part, it keeps on running till it reaches the desired validation score.

Trying it with a dataset :)