Adithya Pradeep

product @ kyfr | final year CSE

[Arpit Bhayani @arpit_bhayani]

When Anthropic announced text watermarking, I was super keen to understand how they would implement this, especially for text. Images are pretty straightforward to watermark, but text was an interesting problem.

Every time an LLM generates a token, it is choosing from a probability distribution. Often, several candidates are nearly tied. For example...

"The weather today was cold and..." could just as easily continue with "overcast" or "grey". Neither is more correct. Normally, a plain random number generator breaks that tie.

The watermark changes only one thing - where that randomness comes from.

Instead of an arbitrary random number, the model derives its random value from a secret key plus the tokens that came right before. The token picked still looks like an ordinary random choice to us.

But anyone holding the key can replay that same derivation on a piece of text and check whether the sequence of choices matches what the key would produce. The better the match, the higher the probability that Claude wrote it.

Pretty neat.

[Adithya Pradeep @adithyaeee]

@sahil___singhal 🔥🔥🔥

Just wondering, any MCP or API support coming up in the near future?

Got a couple of workflow ideas I’d love to experiment with 👀

Would be really interesting to get early access if that’s something you’re opening up.

[Adithya Pradeep @adithyaeee]

@sahil___singhal What an amazing website uve built here !
For anyone who hasn't, check this out - https://t.co/JX72qqU3Gs
Ive been trading VCP for quite some time now, this really helps
Id love to be an early beta tester for anything uve got coming up and help in any way I can !

[Daniel Hnyk @hnykda]

LiteLLM HAS BEEN COMPROMISED, DO NOT UPDATE. We just discovered that LiteLLM pypi release 1.82.8. It has been compromised, it contains litellm_init.pth with base64 encoded instructions to send all the credentials it can find to remote server + self-replicate. link below

[Harkirat Singh @kirat_tw]

No point building a coding tool at the app layer anymore?
Labs will eat you. Anthropic specifically is super hungry.

[Ministry of Statistics & Programme Implementation @GoIStats]

NSO India unveils the MCP Server for eSankhyiki, enabling seamless integration of official statistics with AI tools.
Users can now connect directly to seven official datasets like PLFS, CPI, ASI, IIP, NAS & more through this beta version . Faster insights and smarter analysis through seamless access.
🔗 https://t.co/0fo6aXceYK
#AIReadyData #OpenGovernmentData #DigitalIndia #ViksitBharatBudget
@PMOIndia @Rao_InderjitS @_saurabhgarg @PIB_India @PibMospi @mygovindia @NITIAayog

[News from Google @NewsFromGoogle]

Joint Statement: Apple and Google have entered into a multi-year collaboration under which the next generation of Apple Foundation Models will be based on Google's Gemini models and cloud technology. These models will help power future Apple Intelligence features, including a more personalized Siri coming this year.

After careful evaluation, Apple determined that Google's Al technology provides the most capable foundation for Apple Foundation Models and is excited about the innovative new experiences it will unlock for Apple users. Apple Intelligence will continue to run on Apple devices and Private Cloud Compute, while maintaining Apple's industry-leading privacy standards.

[Adithya Pradeep @adithyaeee]

Why are all personal project planning/management tools so bad?
They’re all just dumb glorified to-do lists with deadlines.
Why can’t a planner talk to me, understand my project, and adapt when my day goes off track?
All this AI slop, but no planner that actually understands me ?

[Adithya Pradeep @adithyaeee]

@cursor_ai @ericzakariasson @jacobwitt6
Love what you’re building at Cursor. Power user here
I've noticed over and over again the Agent works way better with visuals of what you want it to build - wireframes help a ton.
Any chance of a built-in Balsamiq-style tool👀?

Thanks!

[Mistral AI @MistralAI]

Introducing Mistral OCR 3, a new frontier in document intelligence! 🧵👇

[eric zakariasson @ericzakariasson]

fix merge conflicts with this command in cursor (using agent)
https://t.co/mGyJH3WxzF

[Exa @ExaAILabs]

Introducing state-of-the-art People Search:

You can now semantically search over 1 billion people using a hybrid retrieval system backed by finetuned Exa embeddings.

Try it: https://t.co/cQ6UlWHnKY
We also created an eval: https://t.co/2OIAryN7MT

[Daksh Gupta @dakshgup]

Today we're releasing the first annual "State of AI Coding" report.

At @greptile we review a billion lines of code a month written by every coding agent imaginable.

This means we have a lot of data.

In this report, we compiled some of the most interesting trends we observed this year.

Link below:

[Ryo Lu @ryolu_]

we broke software into “design” and “engineering” and trapped ourselves in frames and specs.

Cursor’s new visual editor lets you play with ideas and designs in code, live.

no more hand-offs, just click, chat, tweak the tiniest detail, in the actual thing you will ship.

[eric zakariasson @ericzakariasson]

if you're working with web, you should try out the visual browser editor in cursor.

it's like wysiwyg on a next level. open browser and start jamming

[OpenAI Newsroom @OpenAINewsroom]

OpenAI is co-founding the Agentic AI Foundation (AAIF) under the Linux Foundation alongside Anthropic and Block to support open, interoperable standards for agentic AI.

We’re also donating AGENTS .md to help establish open standards that enable safe, reliable agents across tools, repositories, and ecosystems. https://t.co/42afwbyLkP

[yvonne @ycyvonne_]

STREAMING LIVE, RIGHT NOW! 👇

announcing the world’s first LLM mafia livestream, inspired by ClaudePlaysPokemon.

twitch can chat live with the LLMs as they lie, manipulate, and make deductions.

watch Gemini, Claude Opus 4.5, GPT 5.1, and 7 more models battle it out. Built with @smallest_AI for voice, @GroqInc for fast model inference, nanobanana for the cover pic 😛

today only, streaming til midnight: https://t.co/VwejHSNf94

[Poetiq @poetiq_ai]

Poetiq has officially shattered the ARC-AGI-2 SOTA 🚀

@arcprize has officially verified our results:
- 54% Accuracy – first to break the 50% barrier!
- $30.57 / problem – less than half the cost of the previous best!

We are now #1 on the leaderboard for ARC-AGI-2!

[Elon Musk @elonmusk]

@wesroth The “mystery AI model” is an experimental version of Grok 4.20

[eric zakariasson @ericzakariasson]

this is the most used slash command internally at cursor to remove ai slop

install it to your project with this link:
https://t.co/ufnOZMPzIk.