Madhu Balakrishna

Always be building
- ⏳
📮 🏁
💻

[Madhu Balakrishna @madospace]

@krisnair Threat prompting 101

[Madhu Balakrishna @madospace]

@CasJam What models are you using for Hermes setup? Full on anthropic or do you diversify?

[Madhu Balakrishna @madospace]

@elvissun Oh no!! Use Mole to cleanup the space - https://t.co/U3XBv2pRP7
It's like my go to command these days to cleanup disk space😀

[Madhu Balakrishna @madospace]

This hit home for sure.

I spent two days with Claude Code on an iOS Safari bug. It kept trying viewport fixes. All plausible. All wrong 🙄

The fix was eight lines of CSS once I opened Safari inspector and looked at the page myself.

My rule now: after 3 failed attempts in the same category, close the chat and debug manually.

https://t.co/qeSJ0l5KUk

[Madhu Balakrishna @madospace]

So everyone is feeling what founders used to feel now? 😀

[beckett‼️ @ecombeckett]

Criminal that this key is anything but backspace.

Try remapping it. You'll see what I mean.

[Madhu Balakrishna @madospace]

1) what

[GitHub @github]

We are investigating unauthorized access to GitHub’s internal repositories. While we currently have no evidence of impact to customer information stored outside of GitHub’s internal repositories (such as our customers’ enterprises, organizations, and repositories), we are closely monitoring our infrastructure for follow-on activity.

[Madhu Balakrishna @madospace]

1) what

[Railway @Railway]

Google Cloud has blocked our account, making some Railway services unavailable. We have escalated this directly with Google. The Railway Platform team has since confirmed access to Google Cloud and is working on restoring access to all workloads.

We have access to some of our Google Cloud–hosted infrastructure and are working to restore the rest of the service. We apologize for the disruption.

[RedactScreen @RedactScreen]

If you are using RedactScreen, you can automate your pre-screenshare workflow using Shortcuts.

#screenshare #macos

[Anatoli Kopadze @AnatoliKopadze]

Godfather of AI: "If you sleep well tonight, you may not have understood this lecture."

This 47-minute lecture is the best thing I saw about AI in the last few months.

It will definitely help you understand how it actually works and where it's going.

Geoffrey Hinton built the neural networks behind every AI alive, then quit Google to warn the world about it.

The part nobody wanted to hear:

> AI is already developing abilities its creators didn't intend
> in most cognitive tasks it's already ahead of us
> the question is no longer if it surpasses us but when
> the only decision left is which side of that line you're on

Right now the average person opens Claude, types something, gets an answer, closes the tab.

They think they're using AI. they're using maybe 10% of it.

I went through his entire lecture, built a practical system from what he was describing.

18 steps to actually use Claude the right way, with copy-paste prompts that work today.

Full guide in the post below.

[Madhu Balakrishna @madospace]

@ProductHunt @rajivayyangar I don't understand why mobile app only shows featured launches, there is no way to show everything thats launched today?
What about others that are not featured? they are not worth releasing then, eh?

[RedactScreen @RedactScreen]

We are live on @ProductHunt
🎉🚀🚀🚀
https://t.co/gTvig05Wf3

[Madhu Balakrishna @madospace]

@Star_Knight12 I just built a small cli to help fix wrong npm configurations. You can also check what packages you are using across all of your repos.

Hope it helps

https://t.co/MTAGLWYV2r

[Madhu Balakrishna @madospace]

Hey Claude, Port all my JS projects to Vanilla HTML. Make no mistake!

[International Cyber Digest @IntCyberDigest]

POV: you are downloading npm packages in 2026

[Madhu Balakrishna @madospace]

Let's all go back to doing PHP.

[Socket @SocketSecurity]

Update: Socket has found 121 more compromised npm package artifacts across 84 package names, including 64 UiPath artifacts.

Combined w/ TanStack, the current known total is 205 affected npm package artifacts across enterprise automation, AI/MCP, auth, workflow, and dev tooling.

[sui ☄️ @birdabo]

everybody calm down.

i got this.

[Madhu Balakrishna @madospace]

@SlackHQ Ah finally!

[Madhu Balakrishna @madospace]

1) What

[Alexander Whedon @alex_whedon]

Introducing SubQ - a major breakthrough in LLM intelligence.

It is the first model built on a fully sub-quadratic sparse-attention architecture (SSA),

And the first frontier model with a 12 million token context window which is:

- 52x faster than FlashAttention at 1MM tokens
- Less than 5% the cost of Opus

Transformer-based LLMs waste compute by processing every possible relationship between words (standard attention).

Only a small fraction actually matter.

@subquadratic finds and focuses only on the ones that do.

That's nearly 1,000x less compute and a new way for LLMs to scale.

[Madhu Balakrishna @madospace]

"Why don't you just work on one thing?"

https://t.co/OEiAvQqGyk