Nikhil Khandelwal

[NASA @NASA]

The wet dress rehearsal for the Artemis II mission around the Moon began at 8:13pm ET on Jan. 31 and will continue tomorrow with @NASAKennedy teams fueling the rocket.

The results from this test will help inform teams about a possible launch date. https://t.co/8wBtyGUeh6

[U.S. Embassy India @USAndIndia]

We welcome India’s new #SHANTIBill, a step towards a stronger energy security partnership and peaceful civil nuclear cooperation. The United States stands ready to undertake joint innovation and R&D in the energy sector.

[Nikhil Khandelwal @nikhil71]

poetry! ❤️

[Jeff Bezos @JeffBezos]

Good overview of the landing. We nominally target a few hundred feet away from Jacklyn to avoid a severe impact if engines fail to start or start slowly. We’ll incrementally reduce that conservatism over time. We are all excited and grateful for yesterday. Amazing performance by the team! Gradatim Ferociter.

[Jeff Bezos @JeffBezos]

Good overview of the landing. We nominally target a few hundred feet away from Jacklyn to avoid a severe impact if engines fail to start or start slowly. We’ll incrementally reduce that conservatism over time. We are all excited and grateful for yesterday. Amazing performance by the team! Gradatim Ferociter.

[Eric Vallieres @EricVallieres84]

Just casually welding itself to the barge, nbd

[VishalSid | CW: Jojos @VishalSid_1]

INDIAN BROS WE ARE SOO BACK
THIS LOOKS WAY TOO GOOD😭😭🔥

WE WILL BE SEATED🗣️🔥🔥
#BahubaliTheEternalWar

[Nikhil Khandelwal @nikhil71]

just a fad. 🥲

[OpenAI @OpenAI]

We've trained an unsupervised language model that can generate coherent paragraphs and perform rudimentary reading comprehension, machine translation, question answering, and summarization — all without task-specific training: https://t.co/sY30aQM7hU

[Andrew Mayne @AndrewMayne]

@yacinelearning @OpenAI I worked at OpenAI for four years. Started a consulting firm. But now also serve as host for the official OpenAI podcast.

And I still publish two novels a year.

[Nayan @0xNayan]

when you finish rebuilding @karpathy nanochat only to remember your actual job for the foreseeable future is still gonna be building agents that are just OpenAI calls in a for loop

[murat 🍥 @mayfer]

🔥 Today we’re excited to announce new funding for `grep` (at a $1.3B valuation) to continue building the foundation of agent observability and text search infrastructure.

grep began as a humble UNIX utility in 1973. Since then, it’s evolved—through recursive innovation and the rise of ripgrep—into a core platform for developers, sysadmins, and agents. Our tools now power engineering and AI teams across @OpenAI, @Anthropic, @Meta, @Cloudflare, @Replit, @NASA, and thousands more.

Over the decades we’ve iterated from grep to `egrep` to `ripgrep`. Our goal has always been to figure out what intelligent agents of the future need to see, filter, and extract—and then build the tools that make that possible.

While our journey is still just beginning, we also want to take a moment to reflect on how the space (and our role in it) has evolved. You can read our reflections and details on this funding milestone here:
https://t.co/um8zFkOqDI

We also share more about the funding that will power our future there. Thank you to @IVP, @Benchmark, @Sequoia, @CapitalG, and the open-source community for their belief in the enduring power of regex.

What excites us most today is what’s next:
grep 5.0 with AI-assisted pattern synthesis
ripgrep Cloud, bringing distributed search to agent clusters
pgrepGPT, an agent-native process discovery layer
And new no-code integrations for autonomous observability pipelines

We’re in the midst of a transformation in computation itself. grep and ripgrep will remain at the core—helping humans and agents alike find what matters, faster.

[Nikhil Khandelwal @nikhil71]

@vijaythirumalai would probably value them in billions if they can hire a few UX Designers and their apps stop looking like a sarkari app from early 2010's.

[Narendra Modi @narendramodi]

Deeply appreciate and fully reciprocate President Trump's sentiments and positive assessment of our ties.

India and the US have a very positive and forward-looking Comprehensive and Global Strategic Partnership.

@realDonaldTrump

@POTUS

[Nikhil Khandelwal @nikhil71]

पृथिव्यां यानि रत्नानि गुणवन्ति गुणान्विते
तान्याप्नुहि त्वं कल्याणि सुखिनी शरदां शतम्

O accomplished one! O treasure of all virtues! May all the riches of the world belong to you. May you be happy for 100 autumns.

[Aditya Raj Kaul @AdityaRajKaul]

Bharat Mata Ki Jai! #OperationSindoor 🇮🇳

[MrNeRF @janusch_patas]

[SIGGRAPH '25] TeGA: Texture Space Gaussian Avatars for High-Resolution Dynamic Head Modeling

Note: On the left that's a 3DGS rendering!

Contributions:
1. We propose a simple approach for rigging 3D Gaussians within the continuous tangent space of 3DMM face models, allowing Gaussians to move freely across mesh triangles.

2. We propose a novel CNN-based deformation model that is agnostic to the number of 3D Gaussians, naturally enabling adaptively densification of the representation to improve detail where most needed, with expression-dependent shading.

3. We show significant improvements over baseline SOTA methods and demonstrate the ability to render even extreme close-up images at high quality.

[Sarvam @SarvamAI]

Meet the all-new Bulbul 🚀

Natural, familiar speech in 11 Indian languages, with authentic accents that sound just like India.

▪️Not robotic. Not rehearsed. Just real.
▪️Lightning-fast performance
▪️Custom voices for your brand

From lower latency and India-first pricing to wider language, Bulbul sets a new benchmark for Speech AI in India.

If you’re building for India, say it with Bulbul!

Watch it in action ↓

[Nikhil Khandelwal @nikhil71]

just do it, Rohan 🙂

[Rohan Pandey @khoomeik]

oxford indologists rly tryna ragebait me into solving OCR for handwritten manuscripts 💀💀💀

[GDP @bookwormengr]

Don't feel the AGI with O3 🧵
----------------------------

Follows wrong logic, lands close to approximately right answer despite wrong logic.

This is like student memorising answer for the popular question in exam and landing in the ballpark of the answer.

A better model, but not certainly a path to AGI...
(no model till date has answered this properly yet).

Prompt:
"6ND is a common formula to estimate flops for training & fine tuning of tranformer models. What would be the formula for flop estimation for fine tuning with LORA? Assume LORA adapters add negligible parameters.

In 6ND formula, N is the number of parameters and D is the number of tokens in data. This formula assumes most computation happens in FFN and most parameters are also there.

Please answer in English"

@teortaxesTex @natolambert @teknium @dwarkesh_sp

[Transluce @TransluceAI]

We tested a pre-release version of o3 and found that it frequently fabricates actions it never took, and then elaborately justifies these actions when confronted.

We were surprised, so we dug deeper 🔎🧵(1/)

https://t.co/IdBboD7NsP

[Ravi Vijayakumar @thehiphopswami]

A career peak for myself! Led physical implementation of this chip from concept to silicon debug.

https://t.co/vfMhRfpSxS

[Rohan Pandey @khoomeik]

ghiblification growth in india is explosive rn but i am kindly begging my indian brethren to use this magical technology to explore, develop, and revive aesthetics that befit the Indian nation in the 21st century instead of co-opting a [cute] japanese cartoon